Hi fidelity SSB bandwith [Archive] - Page 2

K3VR

03-09-2007, 07:50 PM

That's an interesting article Mark. The authors say "High-quality speech systems need to cover the frequency range of about 80 Hz (for especially deep male voices) to about 10 kHz (for best reproduction of consonants, which are crucial to intelligibility)." The article reinforces the fact that intelligibility improves with bandwidth. They also say speech needs to be about 12 dB above a broadband (40 Hz~20,000 Hz) noise level in order to achive an 80% speech recognition rate.

Here's another article that seems to indicate that bass notes (below the typical 300 Hz cutoff of older SSB amateur radios and telephone systems) improve the functional signal-to-noise ratio (SNR) by 10-15 dB for speech recognition in noise.

I've found this to be true also. Increased upper bandwidths seem to help with general intelligibility, up to about 7 kHz. After that, the returns are less noticeable (in my system, with my ears). I also notice that bass notes down to about 100 Hz help me recognize and discern speech from noise. This research has tremendous ramifications for those with and without hearing problems who also happen to be amateurs!

Title: Unintelligible low-frequency sound enhances simulated cochlear-implant speech recognition in noise

Author(s): Chang JE (Chang, Janice E.), Bai JY (Bai, John Y.), Zeng FG (Zeng, Fan-Gang)

Source: IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING 53 (12): 2598-2601 Part 2, DEC 2006

Abstract: "Speech can be recognized by multiple acoustic cues in both frequency and time domains. These acoustic cues are often thought to be redundant. One example is the low-frequency sound component below 300 Hz, which is not even transmitted by the majority of communication devices including telephones. Here, we showed that this low-frequency sound component, although unintelligible when presented alone, could improve the functional signal-to-noise ratio (SNR) by 10-15 dB for speech recognition in noise. We argue that this low-frequency enhancement effect cannot be due to linear addition of intelligibility between low- and high-frequency components or an increase in the physical SNR. We suggest a brain-based mechanism that uses the voice pitch cue in the low-frequency sound to first segregate the target voice from the competing voice and then to group appropriate temporal envelope cues in the target voice for robust speech recognition under realistic listening situations."

.

ab0wr

03-09-2007, 08:19 PM

Quote[/b] (n5rfx @ Mar. 09 2007,07:52)]Quote[/b] (ab0wr @ Mar. 06 2007,18:43)]I *STILL* have not seen a single person address why fire annunciator system requirements today specify at least 6-8khz bandwidths if 2.1khz bandwidths are "plenty good enough".
I couldn't find the frequency response specified in the NEMA Fire Service Annunciator Standard, but in doing a Google search it appears that most RFP's #and regulations require a 200 to 4000 Hz frequency response. #So it may be stretching it a bit to say that fire annunciator system requirements today specify at least 6-8khz bandwidths, but I get your point, higher frequencies may allow for more intelligible audio.

73,

Mark N5RFX
Section 4-3.1.5 of NFPA 72, National Fire Alarm Code®, now requires that "emergency voice/alarm communications systems shall be capable of the reproduction of prerecorded, synthesized, or live (for example, microphone, telephone handset, and radio) messages with intelligibility"

The requirements in NFPA 72 requirements equate to an 80% word-comprehension rate and a 95% sentence-comprehension rate.

For single syllables in typical noise, 3.3 KHz bandwidth yields an syllable accuracy of 75 percent, as opposed to over 95 percent with 7 KHz bandwidth.

Word accuracy in a reverberant space will go from about 50% with a bandwidth of 4khz to about 80 percent when the available bandwidth is 8 KHz.

The cumulative articulation index for 2 word groups is about 75% at 3khz bandwidth and is somewhere between 90% and 95% for 6khz bandwidth. If you use 10 word groups, the index is almost zero at 3khz and about 65% for 6khz. You don't hit 90% for 10 word groups till you hit around 9khz.

The articulation index is a measure of the intelligibility of speech in the presence of noise. The usual noise referency is 65db if I remember correctly.

http://www.audiologyinfo.com/ai/aitut3.htm gives one understandable way of estimating the AI.

This is a very complex subject with all kinds of information on the internet. It's taken me over two years to even become marginally conversant in the jargon let alone understand the intricasies.

I got into this as a result of my hearing loss. My unaided hearing rolls off around 4.5khz, typical for older people. I was having a distinct problem understanding people in the middle school where I work. I simply could not believe the difference even the least expensive programmable digital hearing aids have made in my hearing. They are fixed program units set up to amplify from around 3khz to 8khz (about the upper limit for this unit). From understanding people to appreciating the high school orchestra to listening to the FM Stereo they have made a huge difference in my life. The word recognition tests we ran as a baseline before getting the hearing aids and after the hearing aids graphically show how much difference the upper frequencies make in intelligibility. I remember in one test of 15 words spoken by my wife I got 14 wrong. In a similar test after being fitted and adjusted I got 14 right out of 15. So much for the excuse of not hearing my wife tell me something! She now knows it is truly *selective* hearing loss when I ignore her!

Those who say that 2.1khz is the be all-end all for radio communication have just never experienced the difference first-hand.

tim ab0wr

wd0ct

03-09-2007, 08:42 PM

"We suggest a brain-based mechanism that uses the voice pitch cue in the low-frequency sound to first segregate the target voice from the competing voice and then to group appropriate temporal envelope cues in the target voice for robust speech recognition under realistic listening situations."

Very cool. Now if I could just tilt and swilvel my ears like a fox I could totally tune out qrmers!

ky5u

03-09-2007, 09:05 PM

Good info Tim. Just remember that the NFPA codes are written by the various committees. The NFPA fire service codes requiring full turnouts and SCBA were written by a committee consisting of Lyon MFG, Globe MFG, MSA, Scott. There was exactly ONE firefighter on the committee. Lyon and Globe makes fire clothes, MSA and Scott make SCBAs. With all the mandated "safety" the number of firefighter deaths has remained unchanged in proportion to the number of calls. Know why?

Because the #1 cause of FF deaths is heart attack and the #2 cause is traffic related. None of the self serving NFPA codes fixed those things. You want to put a dent in FF deaths? Mandate every FF over 40 must have a stress test every year and a physical fitness program. Unions would never allow it. Mandate that when FFs are out at an accident, traffic is stopped in all directions within 300 feet of the scene. The public would never accept that.

So I know I digress from your point, but my guess is that those NFPA regs were written by a committee of PA Sales and Alarm System sales people and designed to require 90% of current alarm systems to need changeout.

wd0ct

03-09-2007, 09:18 PM

"So I know I digress from your point, but my guess is that those NFPA regs were written by a committee of PA Sales and Alarm System sales people and designed to require 90% of current alarm systems to need changeout."

Nothing like mandating more money for the company. Those guys must take lessons from K street.

Back to intelligibility, on another hammie forum they were talking about things only New Yorkers know. One was that no one ever understands any subway announcement.

n5rfx

03-09-2007, 10:23 PM

Quote[/b] (ab0wr @ Mar. 08 2007,14:19)]Quote[/b] (n5rfx @ Mar. 09 2007,07:52)]Quote[/b] (ab0wr @ Mar. 06 2007,18:43)]I *STILL* have not seen a single person address why fire annunciator system requirements today specify at least 6-8khz bandwidths if 2.1khz bandwidths are "plenty good enough".
I couldn't find the frequency response specified in the NEMA Fire Service Annunciator Standard, but in doing a Google search it appears that most RFP's and regulations require a 200 to 4000 Hz frequency response. So it may be stretching it a bit to say that fire annunciator system requirements today specify at least 6-8khz bandwidths, but I get your point, higher frequencies may allow for more intelligible audio.

73,

Mark N5RFX
Section 4-3.1.5 of NFPA 72, National Fire Alarm Code®, now requires that "emergency voice/alarm communications systems shall be capable of the reproduction of prerecorded, synthesized, or live (for example, microphone, telephone handset, and radio) messages with intelligibility"

The requirements in NFPA 72 requirements equate to an 80% word-comprehension rate and a 95% sentence-comprehension rate.
Tim,

Thanks for the clarification. I was mistaken when I thought there was some fire annunciator standard that specified at least 6-8khz bandwidths fire annunciator systems.

I agree the subject is very complex and subjective. One of the most difficult books I have attempted to get through is "Speech and Hearing in Communications" by Harvey Fletcher. I bought a copy of this book when you and I were debating whether digital and analog signals could co-exist. Man what a book.

Working in the Land Mobile Sector at Motorola was a great lesson in how subjective folks are in evaluating audio quality. I can't count how many times we would replace a public safety system and the first complaints would always be about the audio quality. The standard was typically that the dispatchers didn't sound the same as they used to, so it was hard to tell who was who.

Heavily compressed digital audio is also very interesting. It is very hard to distinguish one voice from another.

73,

Mark N5RFX

WB2WIK

03-09-2007, 11:07 PM

Quote[/b] (n5rfx @ Mar. 09 2007,15:23)]Working in the Land Mobile Sector at Motorola was a great lesson in how subjective folks are in evaluating audio quality. I can't count how many times we would replace a public safety system and the first complaints would always be about the audio quality. The standard was typically that the dispatchers didn't sound the same as they used to, so it was hard to tell who was who.
Dispatchers usually sound like crap because their sound reflects the way they operate, which is terribly.

If they used boom headsets and learned how to speak properly, they'd sound fine.

WA0LYK

03-10-2007, 12:06 AM

Quote[/b] (WB2WIK @ Mar. 09 2007,16:07)]Quote[/b] (n5rfx @ Mar. 09 2007,15:23)]Working in the Land Mobile Sector at Motorola was a great lesson in how subjective folks are in evaluating audio quality. I can't count how many times we would replace a public safety system and the first complaints would always be about the audio quality. The standard was typically that the dispatchers didn't sound the same as they used to, so it was hard to tell who was who.
Dispatchers usually sound like crap because their sound reflects the way they operate, which is terribly.

If they used boom headsets and learned how to speak properly, they'd sound fine.
Certainly wider frequency response isn't the whole story. Good diction, good modulation of your voice, and a myriad of other things such as mic response and phase variations in audio amps come into play. That is why this is a good area for experimentation. The entertainment industry alone pays a lot of folks for their expertise in audio systems.

Who knows, by participating in this line of research in amateur radio, we may be training the next Alexander Graham Bell whose main interest was working with the deaf. I guess that's one reason I'm always loath to denounce any area of interest with hams. There may have to be limits when infringing on others right to use the bands and one must be open to valid criticism when poor signals are being put on the air but in the end, experimentation is a good thing.

Jim
WA0LYK

K3VR

03-10-2007, 01:51 AM

I've spent the last few days reading national and international standards and journal articles about bandwidth and intelligibility. Here's a thumbnail sketch: Apparently, in the days of old, the telephone company settled on approximately 300-3,200 Hz due to cost and design characteristics more than anything. Microphones and speakers weren't capable of much more bandwidth when the phone system was designed.

Art Collins used the telephone company's figures to research and design his SSB filters. The intelligibility Articulation Score was fixed at about 65-70% for those rating the telephone system using the old Articulation Index. Collins engineering notes said, "Practice has shown that a transmitted bandwidth of approximately 3500 cps is satisfactory for communications circuits. Anything greater than this just uses more of our spectrum and produces little or no additional intelligibility."

(Note: In fact, modern research indicates an additional 25% intelligibility is realized with increased bandwidth.)

[Articulation Index - A measure of the intelligibility of voice signals, expressed as a percentage of speech units that are understood by the listener when heard out of context. Note: The articulation index is affected by noise, interference, and distortion.

Articulation Score (AS) - A subjective measure of the intelligibility of a voice system in terms of the percentage of words correctly understood over a channel perturbed by interference. Note: Articulation scores have been experimentally obtained as functions of varying word content, bandwidth, audio signal-to-noise ratio and the experience of the talkers and listeners involved.]

With the advent of SSB, the game was on, and the stage was set for Art Collins to compete for military dollars against my old friend Leo Meyerson. According to an article in QST by Gil McElroy, VE3PKD:

"General Curtis LeMay became aware of the successes of amateur SSB work, and in 1956 undertook two flights, one to Okinawa and the other to Greenland, during which SSB was put to the test using Amateur Radio gear and hams themselves. Two of the hams invited to operate on those flights were Art Collins, WØCXX, of Collins Radio, and Leo Meyerson, WØGFQ, of World Radio Labs. SSB far outperformed the conventional AM communications systems then in use by the military. In 1957, it was formally adopted by SAC for use in its (then) new B-52 bombers, the same year that General Francis Butch Griswold, KØDWC, of SAC would give the keynote address on the subject at the ARRL National Convention in Chicago. Sideband had won the day."

Ironically, The new standard for the FAA and NAS in their Next Generation Communications (NEXCOM) radios will be VDL modes 2 and 3. In 15 years the plan is to move all NAS/FAA comms to VDL-3, e.g., VHF and UHF transmissions utilizing 25 kHz AM DSB and/or 8.33 kHz AM DSB modulation schemes that will share analog voice with data streams transmitted by low earth orbiting satellites. By the time they actually implement this plan, it's widely predicted that it will be hopelessly obsolete. A new hybrid plan is currently being developed using Sirius and XM radio satellites as models/case studies.

Meanwhile, back in July 1961, the pages of QST went on to report that, "General Curtis LeMay, then K3JUY/K4RFA, announced his nomination as chief of staff of the US Air Force. During the mid 50s, Gen LeMay converted Strategic Air Command communications from AM to SSB, based on his ham experience with the newer mode. Sideband was what the Doctor ordered and ARRL effectively spread the word.

In 1955, Collins all but abandoned production of AM gear and threw its considerable resources behind development of sideband gear, having prepared the way with a series of full-page Engineering Notes that appeared in QST in late 1954. In May of 1957 Collins would make history with the launch of the KWM-1 transceiver, the first mobile transceiver, the advertisement in QST read, and the first to offer SSB. A review of the rig in the April 1958 issue would be positively glowing:

By April 1948, QST Technical Editor, George Grammer said, It may not be too much of an exaggera-
tion to say that our present-day phone methods will be just as obsolete, a few years from now, as spark was a few years after c.w. got its start. Old-fashioned phone will eventually be something that can be tolerated only where there is plenty of room for it."

You can still hear people on the air today using the same bandwidth rationale as an excuse to deliberately QRM others. As previously discussed, switching to PSK31 or CW during crowded band conditions would yield well over 6,000 more frequencies and opportunities to discuss gall bladder surgery, so there's no excuse for deliberate interference on the basis of bandwidth, but I digress.

AM enthusiasts of the day derided SSB as sounding like Donald Duck and some accused the ARRL of "shoving SSB down their throats," as well as selling hams out to the government's military industrial complex in exchange for the enhanced revenue that would pour into the ARRL's coffers as a result of hyping the new technology. They thought it was about advertising dollars in QST... Sound familiar?

The current national and international standards for intelligibility are based on the early phone company research that led to the Articulation Index and later, to the US military's Speech Transmission Index, which expanded on Ma Bell's earlier research.

Basically, these two indices ask people who speak the same brand of English (Midwest American, or what-have you) to rate their ability to understand the spoken word and groups of words under varying conditions.

The military and telecommunications industries have since expanded on that research to include things like the effects of reverberation, distortion, propagation, reflection, broadband noise, and distractors (like ambient noise, interference, and other people talking).

The Journal of the Acoustical Society of America has almost 2,800 articles touching on bandwidth and intelligibility. Our military's MIL-STD-188, and the American National Standards Institute’s approved procedure (ANSI S3.2-1989, “Method for Measuring the Intelligibility of Speech Over Communication Systems) are being re-written to take the latest research into account.

Fire Annunciators and Classroom speech systems are widely expected to adopt the new wider bandwidth standards to ensure the safety of the public. As far as I can tell from reading a few dozen articles, most current American researchers agree that the optimum communications bandwidth is between 125 Hz and 8,000 Hz. These assumptions are based on numerous experiments where the science has been replicated over and over again. There also seems to be general and specific agreement among American safety, and telecommunications experts that something like 125 Hz~8,000 Hz is the optimal bandwidth for truly intelligible communications.

Medical research has concluded that both the AI and the STI have significant shortcomings. They're concerned with people with hearing deficiencies and those who suffer from upper register hearing loss due to aging. A hybrid index, designated mSTI, which takes the best features from each procedure, has been described and demonstrated to be the best alternative presently available.

Meanwhile, the accepted European protocol is just as wide, with the Europeans having adopted an 8.33 kHz bandwidth as the future protocol for air traffic control and 50~7,000 Hz as voice transmission standards. The G.722.1 standard specifies operation at 24 kbps and 32 kbps for wideband (50 Hz – 7 kHz) and is currently defined in the form of a fixed-point arithmetic implementation.

Conclusion

The opportunities for bandwidth experimentation are limitless, and acoustic research is a vital tool that's on the bleeding edge of aerospace, communications, defense, and medical industries today. As amateur operators, despite the proven benefits of wider bandwidths, we still have to contend with band crowding on certain bands at certain times.

It pains me to say this, but unless we're involved in handling emergency traffic, all of our conversations are superfluous. What I'm getting at is: Unless we have a declared emergency, my (insert mode) conversation is no more important, and has no more precedence than the next guy's (insert mode) conversation. #

I'm not advocating wide modes for everyone all the time, or for anyone at any time, but none of us has the right to QRM or chastise anyone because we don't like the mode he or she chooses. People who use AM or ESSB have a right to pursue their interests without interference or name calling, just like CW ops, contesters, PSK31 afficionados, and SSTV picture traders. If bandwidth is an issue at any given time, there are lots of alternatives, including turning the radio off and watching TV.

The bottom line is this: We need to respect each other. We need to follow the rules concerning interference and good engineering practice, and above all we need to develop compassion for those who have different tastes in modes of operation. Anyhow, that's my opinion; I've got my flame retardant suit on, so have at it, and as always, YMMV.

By the way, sorry for taking up so much BANDWIDTH!

http://www.qrz.com/iB_html/non-cgi/emoticons/biggrin.gif

QST Article on the Advent of SSB (http://www2.arrl.org/qst/2003/01/McElroy.pdf)

NEXCOM (http://www.arinc.com/aeec/general_session/gs_reports/2002/presentations/13_vdl_3/nexcom.pdf)

ITU Standard G.722.1 (http://www.polycom.com/partners/0,,pw-1234,FF.html#What%20is%20G.722.1)

Telecom Dictionary Reference (http://www.faxswitch.com/Definitions/telecom_dictionary_a.html)

FAA E-2958 (http://scholar.google.com/scholar?hl=en&lr=&safe=off&q=cache:Ejp26A6qQq4J:www.asu.faa.gov/faaco/attachments/RPDE_J-4_SRD_10.pdf+Appendix+A+of+NFPA+72,+1999+intelligi bility)

Alternate Satellite Hybrid Link System (http://spacecom.grc.nasa.gov/icnsconf/docs/2003/03_A2/A2-04-Farneth.pdf)

Collins Engineering Notes (http://www.wa3key.com/qst/qst5410.jpg)

KA4DPO

03-10-2007, 02:39 AM

By golly you're right Brian. We need to all get along and respect one another.

And now everyone can show their respect for amateur radio by supporting the ARRL bandwidth proposal. If you're a legal operator this will have absolutely no effect on you. It's only purpose is to stop the scofflaws, freebanders, and bootlegers from QRMing legal QSOs.

So how about it everyone? Help put a stop to bootleggers, support the ARRL bandwidth proposal.

http://www.qrz.com/iB_html/non-cgi/emoticons/wink.gif

K3VR

03-10-2007, 02:42 AM

Hey, thanks John. What is the ARRL Bandwidth proposal anyhow? I've been looking at MIL-STD gobbledygook for 3 days and haven't had a chance to look into that yet. What's the ARRL asking for and why?

Brian

w4wtf

03-10-2007, 03:06 AM

Quote[/b] (ab0wr @ Mar. 09 2007,06:15)]Quote[/b] (w4wtf @ Mar. 08 2007,20:21)]Quote[/b] (ab0wr @ Mar. 07 2007,06:10)]You know, I find it just amusing as all get out that not a single person on here has anything to refute what the people that design fire annunciation systems figured out a decade ago and are beginnng to implement nationwide.

All I seem to be seeing is "Well I can understand it fine!"

Everyone seems to want to defend the old systems designed with narrow bandwidths based on 1940's telephone company research that resulted in an economic decision about required bandwidth as much as an intelligibility decision.

If you want another example go look up where the design standards for student classrooms are going - especially the ones set up for distance learning. You'll find that it is no longer acceptable to just use a 2.1khz long distance telephone line to bring in the audio. Many system requirements specify at least an 8khz bandwidth #and some as high as 12khz.

There is a REASON for this folks. It is called *INTELLIGIBIITY*.

You can stick your head in the sand if you want and say "2.1khz is all anyone needs" but the research over the past decade shows that when intelligibility is the *key* you probably don't even know that you are missing anything.

tim ab0wr
Sure, a wider signal is going to be more intelligible, but in both those cases there is really no drawback or disadvantage, other than a sligthly higher cost, to having such a wide frequency response.

They designers of these systems are going for one thing and one thing only, maximum intelligibility, and they have nothing they lose by going for it.

We do however, for every wide signal we lose bandwidth, and we have a limted amount of bandwith available. So the question becomes do we want to trade that intelligibility we gain for bandwith and possible interference?

Where do we reach the point of diminishing returns?

I imagine of the fire alarm designers were faced with a situation where wider bandwith meant allowing fewer alarms per structure, would they deceide that bandwith or number of "stations" was more important?
Just out of curiousity, did you read the entire thread?

Did you see anybody advocating that everybody use essb all the time everywhere?

I just can't seem to find that stated anywhere.

I'd appreciate it if you could provide the quotes that say that.

tim ab0wr
Of course nobody said that. Nor did I claim they did.

But if ESSB is being advocated as better for emergency comms (as I have seen on here) and other stuff because of greater intelligibility, we need to make sure we find that balance between the two. There is no free lunch... so how much bandwith do we wish to trade for this enhanced intelligibility, and when do we reach the point of diminishing returns?

n5rfx

03-10-2007, 03:13 AM

Thanks Brian for the links

Quote[/b] ]Fire Annunciators and Classroom speech systems are widely expected to adopt the new wider bandwidth standards to ensure the safety of the public. As far as I can tell from reading a few dozen articles, most current American researchers agree that the optimum communications bandwidth is between 125 Hz and 8,000 Hz.
This is certainly consistent with what Tim AB0WR has been saying.

Quote[/b] ]selling hams out to the government's military industrial complex in exchange for the enhanced revenue that would pour into the ARRL's coffers as a result of hyping the new technology. They thought it was about advertising dollars in QST
Classic! Now its the Homeland Security part of the federal government. There seems to be a trend http://www.qrz.com/iB_html/non-cgi/emoticons/smile.gif

73,

Mark N5RFX

KA4DPO

03-10-2007, 03:13 AM

Quote[/b] (K3VR @ Mar. 09 2007,21:42)]Hey, thanks John. What is the ARRL Bandwidth proposal anyhow? I've been looking at MIL-STD gobbledygook for 3 days and haven't had a chance to look into that yet. What's the ARRL asking for and why?

Brian
You wouldn't like it. I was being a smart %$$. You have to admit that was a better spin job than a democratic candidate promising to lower taxes.

ARRL has put a proposal on the table that would limit SSB signals to a 3 KHZ bandwidth at 26 db down. That means the -6 db bandwidth would be about 2.3 KHZ. There are a lot of rigs out there right now that wouldn't pass. I think the hidden agenda is to kill pactor, winlink, and other wide band digital modes.

I actually disagree with the proposal because it imposes limits rather than allowing the amateur community to exercise discretion based on conditions. I do support a band plan that would segregate wideband digital modes into a discreet window. Some of the worst interference on the bands is from automatic digital stations that pop up over ongoing PSK and CW transmissions.

n5rfx

03-10-2007, 03:20 AM

Quote[/b] (w4wtf @ Mar. 08 2007,21:06)]so how much bandwith do we wish to trade for this enhanced intelligibility, and when do we reach the point of diminishing returns?
I must confess that I am a 300 to 3000 Hz guy, but as long as we don't interfere with each other, I don't see a problem with sometimes stretching out. I am going to have to listen around the phone bands and see if I can hear some ESSB and see how it goes. I hardly go into the phone bands exept to copy SSTV, Digtal Voice and listen to the AM guys.

73,

Mark N5RFX

ke4pjw

03-10-2007, 03:38 AM

I am a little curious about intelligibility gains with the change in spectral power density. I would imagine that as you occupy more bandwidth, your effective power output would drop. Probably not enough of a change to matter. I'm not good at math http://www.qrz.com/iB_html/non-cgi/emoticons/smile.gif

Updated: effective power, meaning spectral power residing in the normal 300-3K bandpass.

K3VR

03-10-2007, 04:27 AM

John, I just read Chris Imlay's notice of an ex-parte conversation in February of this year, between the FCC and ARRL regarding RM-11306. The ham community was against that petition by a margin of 8:1, if I recall correctly.

The ARRL has amended their requests, but they're still asking for automatic RTTY and DATA stations to have access on all bands where RTTY and DATA are allowed now -- plus, they're asking for bandwidth limitations again. They want a maximum of 3kHz on ten meters through UHF with exceptions for AM DSB and FM.

Hmmmm. The FCC just told us that wider bandwidths are part of legitimate experimentation. So what's the ARRL up to now with respect to 10 meters and up -- Why limit bandwidth on 10 and VHF/UHF?

PROPOSED RULE CHANGE:

"97.221 Automatically controlled stations transmitting RTTY or data emissions.

"*****

"( c ) A station transmitting a RTTY or data emission may be automatically controlled on any other frequency authorized for such emissions provided that the station is responding to interrogation by a station under local or remote control.

"( 1 ) (Deleted)

"( 2 ) (Deleted)"

Latest Comments in Opposition to ARRL Proposal (http://gullfoss2.fcc.gov/cgi-bin/websql/prod/ecfs/comsrch_v2.hts)

ARRL's Notice of Ex-Parte Conversation with Amended Appendix (http://wa3vjb.amham.com/pics/11306-cont.pdf)

k5xit

03-10-2007, 12:03 PM

Quote[/b] (KA4DPO @ Mar. 09 2007,20:13)]Quote[/b] (K3VR @ Mar. 09 2007,21:42)]Hey, thanks John. What is the ARRL Bandwidth proposal anyhow? I've been looking at MIL-STD gobbledygook for 3 days and haven't had a chance to look into that yet. What's the ARRL asking for and why?

Brian
You wouldn't like it. #I was being a smart %$$. #You have to admit that was a better spin job than a democratic candidate promising to lower taxes.

ARRL has put a proposal on the table that would limit SSB signals to a 3 KHZ bandwidth at 26 db down. #That means the -6 db bandwidth would be about 2.3 KHZ. #There are a lot of rigs out there right now that wouldn't pass. #I think the hidden agenda is to kill pactor, #winlink, and other wide band digital modes.

I actually disagree with the proposal because it imposes limits rather than allowing the amateur community to exercise discretion based on conditions. #I do support a band plan that would segregate wideband digital modes into a discreet window. #Some of the worst interference on the bands is from automatic digital stations that pop up over ongoing PSK and CW transmissions.
I read another post from a fellow ranting about ARRL's hidden support of winlink etal. I guess I will just have to read the league news more closely.

ky5u

03-10-2007, 01:59 PM

Quote[/b] (K3VR @ Mar. 09 2007,21:27)]John, I just read Chris Imlay's notice of an ex-parte conversation in February of this year, between the FCC and ARRL regarding RM-11306. The ham community was against that petition by a margin of 8:1, if I recall correctly.

The ARRL has amended their requests, but they're still asking for automatic RTTY and DATA stations to have access on all bands where RTTY and DATA are allowed now -- plus, they're asking for bandwidth limitations again. They want a maximum of 3kHz on ten meters through UHF with exceptions for AM DSB and FM.

Hmmmm. The FCC just told us that wider bandwidths are part of legitimate experimentation. So what's the ARRL up to now with respect to 10 meters and up -- Why limit bandwidth on 10 and VHF/UHF?

When we talk about wider bandwidth, let's distinguish between compatible HUMAN controlled modes like SSB and ESSB controlled by a person and wideband data which is incompatible with narrowband modes used by people. Add to that the AUTOMATIC stations.

Last year the FCC wanted to turn wideband data loose in the voice bands which would have hurt ESSB, SSB, AM, SSTV, etc. Myself and others worked for an 8 to 1 margain in comments AGAINST this. That failed so now they're trying in the CW/RTTY/Data subbands to let wideband data run amok. What I need, Brian, is for you phone guys to support US now. Please comment AGAINST the Inmay letter and repay the favor we did to keep this out of the voice bands. I also request when you speak of wideband as being ok, that you specifically identify voice and not data based on compatibility. Thanks!!

W3MIV

03-10-2007, 02:07 PM

Quote[/b] (AG4YO @ Mar. 10 2007,08:59)]What I need, Brian, is for you phone guys to support US now. #Please comment AGAINST the Inmay letter and repay the favor we did to keep this out of the voice bands. #I also request when you speak of wideband as being ok, that you specifically identify voice and not data based on compatibility. #Thanks!!
Agree. Get to it, Brian.

http://www.qrz.com/iB_html/non-cgi/emoticons/wink.gif

KA4DPO

03-10-2007, 02:45 PM

My opinion as stated before is that the operator of a voice mode station has a measure of control over the bandwidth of the signal.

However you may feel about ESSB or AM, the bandwidth of such signals should be left to the discretion of the operator taking conditions into account. I assume a responsible operator would narrow the signal bandwidth to accommodate crowded band conditions when necessary. Those who don't will hear about it from fellow amateurs and will be branded as LIDs.

Also, allowing automatic digipeaters in the CW and data sub bands has the potential to wreak havoc. Already a lot of "Maritime Mobile" stations use these amateur systems for E-mail to avoid paying commercial carriers. The proposal put forth by the ARRL is technically and practically unacceptable. The use of these systems by boaters is on the very thin edge of the law dealing with pecuniary interest.

n5rfx

03-10-2007, 11:26 PM

Quote[/b] (ke4pjw @ Mar. 08 2007,21:38)]I am a little curious about intelligibility gains with the change in spectral power density. I would imagine that as you occupy more bandwidth, your effective power output would drop. Probably not enough of a change to matter. I'm not good at math http://www.qrz.com/iB_html/non-cgi/emoticons/smile.gif

Updated: effective power, meaning spectral power residing in the normal 300-3K bandpass.
Your question prompted me to look at the recoding posted earlier in this thread.
Audio Demonstration (http://www.ve7kfm.com/bw.mp3)
I broke this recording up into it 9 bandwidth demonstrations. What I hoped to do was to provide statistics that would answer the spectral power density question. I ran into some problems. When looking at each bandwidth, the noise level did not change with the bandwidth, which raised a flag. When I analyzed this further, it sounds to me like there is a ground loop hum, and this hum is not affected by the filtering. The main audio level varied by a little more than 4 dB from the 10K sample, to the 2.4K sample. This is to be expected since the same audio was not used for each bandwidth. This was accounted for in my calculations.
By calculation there should be a 6 dB increase in spectral density between the 10K sample and the 2.4K sample. This is true because the majority of the power lies in the 2.4 KHz bandwidth of the last sample. 10LOG(BW1/BW2) gives us this expectation. When I checked the samples and compensated for varying audio level, the increase in spectral density from the 10K sample to the 2.4K sample was 6.2 dB.

Since I had the samples already broken up, I wanted to look at the signal to noise ratio. Unfortunately the hum really prevents us from seeing the true SNR changes because of bandwidth and the audio samples were different for each bandwidth. We should have seen a 6 dB increase in signal to noise ratio between the 10K sample and the 2.4 K sample. Again most of the energy lies in that 2.4K bandwidth. I did not see this because the hum is constant throughout the samples and is not affected by filtering This type of event would occur when the noise distribution is not Gaussian, and is concentrated in a certain part of the bandwidth, or is being added after the filtering, which I believe is the case with these samples. Since the noise is not Gaussian, I did not see a change in SNR between the 10K and the 2.4K samples. The SNR stayed slightly below 25 dB for each sample when corrected for the varying audio levels.

Had the noise been Gaussian I would have seen an overall gain of 6 dB between the 10K and the 2.4K samples, with the non-Gaussian noise the gain was 0 dB.

To determine the effectiveness of widening the bandwidth, we would have to have a curve that shows the articulation index vs. signal to noise ratio for each bandwidth. We have an example of that here (http://www.meyersound.com/support/papers/speech/mf.htm). The main loss is going to come from the increase in noise because of the increase in bandwidth. As long as the decrease in signal to noise ratio is justified by an increase in the articulation index, then there is an overall gain. This is a very subjective and the results are going to vary from individual to individual. The wider bandwidth audio is easier for me to listen to for sure, but I have to wonder how this is affected by Gaussian noise? Here are some examples. The 10K sample has 11 dB SNR and the noise is Gaussian. The 2.4K sample has 17 dB SNR due to the reduction in bandwidth.
10K Sample (http://home.comcast.net/~mdmiller7/noise/10K_noise.mp3)
2.4K Sample (http://home.comcast.net/~mdmiller7/noise/2_4K_noise.mp3)

73,

Mark N5RFX

K3VR

03-10-2007, 11:50 PM

Mark,

It's not a ground loop, it's the fan on the SDR you're hearing in the background. Obnoxious isn't it? I chose not to use a noise gate for the recording because the gate opens and closes between syllables, and that adds an artificial quality to the audio. Luckily, my two boys didn't come trip-trapping up the stairs while I was recording! Anyway, carry on, it's interesting to see how many avenues of scientific inquiry are being explored in this old-fashioned "legacy mode" technology of ours!

n5rfx

03-11-2007, 12:20 AM

Quote[/b] (K3VR @ Mar. 09 2007,17:50)]Mark,

It's not a ground loop, it's the fan on the SDR.
Ok, it was the fan eh? I guess the bandwidth of that noise was wide enough to be present in the entire 10K. Are you using a Delta 44? That is what I am using. I am thinking hard about the SDR1000.

Anyway, I added 2 more files, one with 5dB signal to noise ratio on the 10K side, and the equivalent 11 dB on the 2.4K side. I think that if intelligibility is the goal, then preempahsis on the TX side and deemphasis on the RX side would go a long way. I will have to play with that too.
10K 5dB SNR (http://home.comcast.net/~mdmiller7/noise/10K_5dB.mp3)
2.4K 11 dB SNR (http://home.comcast.net/~mdmiller7/noise/2_4K_11dB.mp3)
73,
Mark N5RFX

N6WK

03-11-2007, 04:12 AM

Quote[/b] (wd0ct @ Mar. 09 2007,19:15)]Tnx ep. I was using kenwood 480, mixer, and condenser mic. It won't do essb. Jim was on his sdx, heil mic, and ART audio processer at well plus 3khz.

Hopefully things may be a bit better tonight as my 870 may show up today.

Now watch the band go in the toilet if the new rig shows up!

edit: next time, jump in.
Steve, You will LOVE the 870. Very Nice radio.
Congrats on getting one.

73,
'Gordon.

N6WK

03-11-2007, 04:20 AM

Quote[/b] (n5rfx @ Mar. 11 2007,01:20)]Quote[/b] (K3VR @ Mar. 09 2007,17:50)]Mark,

It's not a ground loop, it's the fan on the SDR.
Ok, it was the fan eh? I guess the bandwidth of that noise was wide enough to be present in the entire 10K. Are you using a Delta 44? That is what I am using. I am thinking hard about the SDR1000.

Anyway, I added 2 more files, one with 5dB signal to noise ratio on the 10K side, and the equivalent 11 dB on the 2.4K side. I think that if intelligibility is the goal, then preempahsis on the TX side and deemphasis on the RX side would go a long way. I will have to play with that too.
10K 5dB SNR (http://home.comcast.net/~mdmiller7/noise/10K_5dB.mp3)
2.4K 11 dB SNR (http://home.comcast.net/~mdmiller7/noise/2_4K_11dB.mp3)
73,
Mark N5RFX
In had a fellow email me today and offered me a SDR1000 in trade for a rig I have here. I declined. I just don't have enough info on the SDR1000 yet. I still like playing with knobs.. hi hi..
What can you tell us about it? If anything .

Thanks,
Gordon

K3VR

03-11-2007, 04:59 AM

The fan on my SDR is pretty noisy. A lot of the fan noise is due to the back pressure created due to inadequate intake opening volume. #Using a Dremel tool, some people have cut out an octagon using the outside of the existing vent holes as a guide. #The increased opening seems to make a MAJOR difference in the fan noise. It's just a black box so it's not like cutting into an IC-7800 or anything.

I like the flexibility built into the transmit and recieve filters. I have an Icom ProII that I never use anymore. The Flex is very quiet on the low bands. Great for SWLing. I wish it had knobs. I can't think of a more advanced radio that offers good SSB and AM performance that also contains the flexibility or has the same customer support.

I should probably post a review on eHam but I'm trying to formulate my comments on the ARRL bandwidth petition right now. I spent the day at my 3 year olds birthday party today so my brain is fried. See you guys tomorrow.

EDIT: Forgot to mention -- Yes Mark, I'm using the Delta 44. Also, some guys have put in a temp regulated fan that hardly moves in receive, that of course moves more in transmit. The beauty of the SDR is that so many people are working to improve the code it's sometimes updated and improved 2 to 3 times a day. Naturally, you just import the changes as they happen, downloading them using subversion.

SDR REVIEWS (http://www.eham.net/reviews/detail/4108)

ab0wr

03-11-2007, 10:15 AM

Quote[/b] (ke4pjw @ Mar. 09 2007,20:38)]I am a little curious about intelligibility gains with the change in spectral power density. I would imagine that as you occupy more bandwidth, your effective power output would drop. Probably not enough of a change to matter. I'm not good at math http://www.qrz.com/iB_html/non-cgi/emoticons/smile.gif

Updated: effective power, meaning spectral power residing in the normal 300-3K bandpass.
The spectral power density shouldn't change because of your amplifier, at least the way I understand the math.

In the systems we use, your amp is limited by it's peak power capability, not it's average power output over a total spectrum.

The spectral power density is determined by the audio characteristics, not the average power output of your radio.

E.g. If you voice consisted of three tones, one at 1000hz (0dbm), one at 2000hz(+5dbm), and one at 4000hz (-20dbm) your amplifier would put out a signal at say 3.901khz, 3.902khz, and 3.904khz with exactly the same power relationships. Your power output would be greatest at 3.902khz and it would be limited by the peak power capability of your amplifer. The effective spectral power density in any frequency band, at least in relative terms, is then determined by the speech characteristics themselves.

Of course if you filtered out the 2khz audio tone (or used compression to knock it down to 0dbm) you could raise the "effective" spectral power density put out by the amplifier in the other bandwidths but, again, that is because you have changed the audio characteristics, not because of any change in the amplifier capabilities.

Mark, n5rfx:

Yes, pre-emphasis IS a help. It's early in the morning and I haven't had a chance to get to the shack in the basement yet but I believe my copy of SSB Systems and Circuits book by Sabin and Schoenke has exactly the curves you are looking for. If I remember correctly they show the increase in articulation index acheived by pre-emphasis. I'll look later today and see if I can find them.

tim ab0wr

ab0wr

03-11-2007, 12:25 PM

n5rfx:Quote[/b] ]To determine the effectiveness of widening the bandwidth, we would have to have a curve that shows the articulation index vs. signal to noise ratio for each bandwidth. #We have an example of that #here. #The main loss is going to come from the increase in noise because of the increase in bandwidth. #As long as the decrease in signal to noise ratio is justified by an increase in the articulation index, then there is an overall gain. #This is a very subjective and the results are going to vary from individual to individual. #The wider bandwidth audio is easier for me to listen to for sure, but I have to wonder how this is affected by Gaussian noise? #Here are some examples. #The 10K sample has 11 dB SNR and the noise is Gaussian. #The 2.4K sample has 17 dB SNR due to the reduction in bandwidth.

According to Sabin and Schoenike in SSB Systems and Circuits you get results like the following:

no pre-emphasis, 10db S/N ratio, Gaussian noise:

2500hz speech # - #AI = 40%
4000hz speech # - #AI ~ 47-50% (graph not easy to interpolate)
6000hz speech # - #AI ~ 52-53%

6db/octave pre-emphasis, 10db S/N, Gaussian noise

2500hz speech # - #AI ~ 48%
4000hz speech # - #AI ~ 62%
6000hz spech # # - #AI ~ 72%

As you can see, the AI increases in both cases as the bandwidth is expanded. The biggest jump seems to be for 2500hz to 4000hz but it is significant out to 6khz.

The real impact here is by the pre-emphasis. The AI goes up by a factor of 1.4 for 6khz speech and about a factor of 1.2 for 2.5khz/4khz speech.

At 0db S/N, the AI varies from about 20% to 28% from 2.5khz to 6khz un-emphasized speech and between 26% and 40% for 2.5khz to 6khz emphasiszed speech.

The AI for 2500hz pre-emphasized speech remains very close to 2/3 the value of pre-emphasized 6khz speech regardless of the SNR. For un-emphasized speech the difference is more like 3/4 of the value.

I think the point is that pre-emphasis is an area that could pay big dividends in intelligibility regardless of bandwidth used. This is one reason behind the popularity of D-104 mic's with their original crystal elements and the Heil HC-4/HC-5 mic elements. They provide some built-in pre-emphasis - but not nearly 6db/octave (at least as I remember).

Of course, this is going to result in what some would consider "audio distortion". This amount of pre-emphasis is certainly going to impact how people sound.

It is a fertile area for research but one that few people even seem to know exists. Where is the ARRL in pushing this type of experimentation? Because of the mixture of digital/SSB it would seem that huge amounts of dividends could be gained from advances in intelligibility of SSB as compared to development of new digital modes.

Yet you see nary a peep out of the ARRL Technology Advancement people on this subject. Is it a "glamour" issue? If the ARRL is in it for the "glamour" instead of the practical advancement of the hobby then they are not being true to their charter.

I recently read a paper concerning the ability of humans to distinguish words in a noisy environment based on "positioning', i.e. interaural cues or the very small time difference between when sound hits one ear vs the other. While I haven't even begun to digest the paper it would seem that, since Gaussian noise is non-directional, if spatial cues could be provided for a speaker location vs the noise being heard, the human brain could more easily distinguish the speech from noise. The paper would seem to indicate that this spatial positioning using interaural timing is very dependent on sound around 500hz.

This leads to some interesting experimentation possibilities. If you were to develop an ISB implementation for a stereo signal with the ISB containing only 300-700hz, your increase in bandwidth would be minimal but the increase in intelligibility by using the stereo sound as a positioning aid might be significant.

Again, a wide open area for experimentation that the ARRL just seems to be blissfully unaware of.

I wonder why that is?

tim ab0wr

KA4DPO

03-11-2007, 02:53 PM

It appears as if you need a very large increase in signal bandwidth to realize a significant gain. From your data doubling the bandwidth produces less than 3db improvement. I don't know what the sensitivity is so I can't really say how significant the result is but it doesn't look like much on inspection.

K3VR

03-11-2007, 03:56 PM

And now a word from Martin F. Jue, K5FLU President and Founder MFJ Enterprises, Inc.

"As I got older, my high frequency hearing loss was destroying my ham radio for me... "I know I'm not the only ham who can't understand all the speech in a QSO caused by high frequency hearing loss. I developed a solution that I want to share with my fellow hams.

I almost gave up my ham radio hobby. I have been a passionate ham radio operator for over 40 years ever since I was a teenager. I loved every minute of it. Still do, but I almost had to give it up. As I grew older (I'm 56 now) I found myself asking "What did you say?" so often it got downright embarrassing. I can hear pretty good most of the time. I just can't always understand what people are saying and my left ear is weaker than my right ear. It got to where I was having trouble carrying on QSOs. I could hear, but I just couldn't quite make out all the words. My hearing problem almost put a stop to my lifelong hobby. There was no way I was going to give up ham radio...

Research showed me what to do. I searched the literature and spoke to hearing and speech experts.
According to their research on intelligibility of speech in hearing English words:

1.The frequencies important for speech intelligibility are the consonant sounds from 500 to 4000 Hz. They contribute 83% of word intelligibility. Frequencies from 500 to 1000 Hz contributes 35% of word intelligibility and 35% of sound energy. Frequencies from 1000 to 4000 Hz contributes 48% of intelligibility but has only 4% of sound energy!

2. In contrast, frequencies from 125 to 500 Hz contributes 55% of sound energy but only 4% to word intelligibility. In other words, nearly half the speech intelligibility is contained in 1000 to 4000 Hz frequency range with only 4% of the speech sound energy. On the other hand, the low frequencies 125 to 500 Hz have most of the speech energy but contribute very little to intelligibility."

The Speech Intelligibility Enhancer (Or Receiver EQ in Colloquial Terms) (http://www.mfjenterprises.com/man/pdf/MFJ-616.pdf)

http://www.mfjenterprises.com/pictures/MFJ-616t.jpg

K3VR

03-11-2007, 04:06 PM

Forgot to add -- current research indicates 75% intelligibility out to about 3,000 Hz, correct?

The missing 25% of intelligibility is contained in the frequencies between 3,000 Hz and 8,000 Hz. That's why the new European bandwidth standard for speech in air comms etc, etc, is 8.3 kHz.

Cutoff everything above 3,000 Hz and lose 25% intelligibility. Add distortion, reverb, S/N below 12 dB, poor microphone technique, high ambient noise, (amplifier or fan noise), and speech compression, and you lose X (%) more intelligibility.

This thread, while rocky at times, is leading to some significant understanding about how physio-psychoacoutics applies to amateur radio. Congratulations guys; well done.

n5rfx

03-11-2007, 04:16 PM

Tim,

Thanks. I have the book you are referencing. I wish there were tabular data for those charts. I looks like you were using the F2 noise charts to do the comparison.
Quote[/b] ]The biggest jump seems to be for 2500hz to 4000hz but it is significant out to 6khz.
The real impact here is by the pre-emphasis. The AI goes up by a factor of 1.4 for 6khz speech and about a factor of 1.2 for 2.5khz/4khz speech.
I agree, here is another way of looking at it:

Without Preemphasis:

Going from 2500 to 4000 Hz gives an AI increase of 1 dB and requires a SNR increase of 2 dB

Going from 4000 to 6000 gives an AI increase of .25 dB and requires a SNR increase of 1.76

Going from 2500 to 6000 gives an AI increase of 1.22 dB and requires a SNR increase of 3.8 dB

With Preemphasis

Going from 2500 to 4000 Hz gives an AI increase of 1.1 dB and requires a SNR increase of 2 dB

Going from 4000 to 6000 gives an AI increase of .65 dB and requires a SNR increase of 1.76 dB

Going from 2500 to 6000 gives an AI increase of 1.77 dB and requires a SNR increase of 3.8 dB

No Preemphasis to Preemphasis

Going from 2500 with no preemphasis to 4000 with preemphasis give an AI increase of 2 dB which is equal to the SNR increase required.

Going from 2500 with no preemphasis to 6000 with preemphasis give an AI increase of 2.56 dB and a SNR increase of 3.8 dB
Quote[/b] ]I think the point is that pre-emphasis is an area that could pay big dividends in intelligibility regardless of bandwidth used. This is one reason behind the popularity of D-104 mic's with their original crystal elements and the Heil HC-4/HC-5 mic elements. They provide some built-in pre-emphasis - but not nearly 6db/octave (at least as I remember).
Of course, this is going to result in what some would consider "audio distortion". This amount of pre-emphasis is certainly going to impact how people sound.
I agree. I don't think the benefit of ESSB is intelligibility as much as it is nice sounding audio. Most folks won't go with the Heil HC4 for normal ragchewing because of its harsh sound, but it will get you noticed.

Very interesting and thanks for making me crack open the Sabin and Schoenike book.

73,

Mark N5RFX

n5rfx

03-11-2007, 04:29 PM

Quote[/b] (K3VR @ Mar. 10 2007,09:56)]Frequencies from 1000 to 4000 Hz contributes 48% of intelligibility but has only 4% of sound energy!
I think this could cause a perceived QRM problem with ESSB enthusiast. Since there is so little energy in the 1000 to 4000Hz range, it might be hard for others to realize that the very low signal they are hearing is actually an essential part of the ESSB QSO.

On the other hand those non-ESSB'rs that run across this part of the ESSB spectrum, might hear it as splatter because there does not seem to be any intelligence associated with the signal, when in fact, it is a very important part of the QSO spectrum.

I don't think that anyone will have a problem with the MFJ Speech Intelligibility Enhancer, since it is for a receiver, the objections will be expanding the audio bandwidth of the transmitted signal.

73,

Mark N5RFX

K3VR

03-11-2007, 04:32 PM

Mark, if you've read some of the articles posted, you know that research indicates that the missing 25% of intelligibility lies above 3,000 Hz.

We also (now) realize that when you add distortion, reverb, S/N below 12 dB, poor microphone technique, high ambient noise (amplifier or fan noise), atmospheric noise, and speech compression, you will lose X (%) more intelligibility (an unknown quantity at this time).

For guys with hearing loss, 25%++ is more than just "nice sounding audio," it's the difference between understanding and giving up the hobby. Agreed?

Real world example: My Grandfather was born in 1907. He was a lifelong ham and Engineer who loved to build and experiment with ham radio. He had to give it up due to hearing loss. If we'd had the appropriate adaptive technologies at the time, he might have been able to enjoy the hobby until his death at the age of 88. He was very depressed at the prospect of selling his S-Line and taking down his monobanders, I can tell you that for sure.

Brian

n5rfx

03-11-2007, 04:45 PM

Quote[/b] (K3VR @ Mar. 10 2007,10:06)]That's why the new European bandwidth standard for speech in air comms etc, etc, is 8.3 kHz.
Here is what we have to remember. The main reason for this bandwidth standard mentioned may be intelligibility and not spectral efficiency on shared spectrum. ATC communications are very safety oriented and this may be the overriding reason for the increase in bandwidth.

Amateur radio operators are obligated to share spectrum. Perhaps there needs to be a SI (safety index) created.

73,

Mark N5RFX

n5rfx

03-11-2007, 04:48 PM

Quote[/b] (K3VR @ Mar. 10 2007,10:32)]For guys with hearing loss, 25%++ is more than just "nice sounding audio," it's the difference between understanding and giving up the hobby. Agreed?
Brian,

I agree, but what percentage of the ESSB community uses pre-emphasis? That is where you get the intelligibility bang for the buck.

73,

Mark N5RFX

KA4DPO

03-11-2007, 05:18 PM

I recall reading Martin Jue's writup on his loss of hearing and his speech enhancing appliance that increases the highs on a standard signal (not an ESSB signal). I guess if you need it then you can buy it as an appliance.

As an evaluator for the Army I have conducted a great many studies and tests of intercom and radio systems that have to work in high (and I mean high) noise environments. Maximum intelligebility is critical, you can get killed if you miss a word or command.

First and formost, it's very difficult to measure intelligebility since all measurements no mater how scientific, are subjective. This creates large standard deviation in a test population sample.

All of the research I have seen for intercoms that have to work inside armor vehicles, aircraft, and watercraft suggest an audio response between 300 and 2890 hz. This seems to be the window for maximum intelligebility in very high noise environments.

I can tell you that the bulk of this data comes from the Army Aeromedical Research Lab and much of the actual measurement data came from independent lab (contract) sources including some from the equipment contractors.

Since they want us to buy their equipment they make it as good as possible for the intended purpose. We don't ever take their data at face value and use many independent government and private sources to either validate or refute their claims.

I have read numerous ( hundreds ) of reports on inteligebility under these kinds of conditions. The only thing I can say is that the typical users are primarily male but some females between the ages of 19 and 27 years old. Their ears are better than mine.

All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself.

The only unfortunate reality is that you can quote your data sources, I can't, mine are not available for public disclosure. I am however, after 20 years of study convinced of their validity.

K3VR

03-11-2007, 05:53 PM

Mark, that's another good field (pre-emphasis) of study for ham radio. Lossless compression techniques and intelligibility algorithms are built into the 8.3 kHz standard being fielded for digital voice right now -- eventually there will be a trickle-down effect into ham radio. Apparently, for now we have to rely on analog techniques. What would be really neat would be to see digital speech recorded in 15 kHz at the transmitter end, sent over the air in 500 Hz packets, and have it be "unpackaged" at the receiver end, and reconstructed back to its original 15 kHz size.

John, do you know when the studies you're discussing were written, or any of the authors? All of the studies should be referenced in the MIL-STD documents. None of the human factors studies are classified. I'm going to the library this afternoon, so if you can remember any of the studies you read, that would help me in trying to locate them. MIL-STD 188 and the ANSI are being re-written right now to (try) to keep up to date with current research.

Also, imagine adding data to the voice stream as in VDL-3. Think of audio with a sub-sonic data component (below the threshold of hearing) so you could read text simultaneously as voice is being transmitted. The data could be automatically encoded using voice recognition algorithms. Closed captioned ham radio... That would open a whole new world of ham radio for hearing impaired people.

WA0LYK

03-11-2007, 06:18 PM

Quote[/b] (KA4DPO @ Mar. 11 2007,10:18)]I recall reading Martin Jue's writup on his loss of hearing and his speech enhancing appliance that increases the highs on a standard signal (not an ESSB signal). I guess if you need it then you can buy it as an appliance.

As an evaluator for the Army I have conducted a great many studies and tests of intercom and radio systems that have to work in high (and I mean high) noise environments. Maximum intelligebility is critical, you can get killed if you miss a word or command.

First and formost, it's very difficult to measure intelligebility since all measurements no mater how scientific, are subjective. This creates large standard deviation in a test population sample.

All of the research I have seen for intercoms that have to work inside armor vehicles, aircraft, and watercraft suggest an audio response between 300 and 2890 hz. This seems to be the window for maximum intelligebility in very high noise environments.

I can tell you that the bulk of this data comes from the Army Aeromedical Research Lab and much of the actual measurement data came from independent lab (contract) sources including some from the equipment contractors.

Since they want us to buy their equipment they make it as good as possible for the intended purpose. We don't ever take their data at face value and use many independent government and private sources to either validate or refute their claims.

I have read numerous ( hundreds ) of reports on inteligebility under these kinds of conditions. The only thing I can say is that the typical users are primarily male but some females between the ages of 19 and 27 years old. Their ears are better than mine.

All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself.

The only unfortunate reality is that you can quote your data sources, I can't, mine are not available for public disclosure. I am however, after 20 years of study convinced of their validity.
The only thing I can mention is that the measurement of intelligibilty in a high noise ambient environment may be different that intelligibility on a high noise radio channel (qrn, qrm, etc.). I don't know this, but it seems logical.

I know this is an area ripe for research and for that reason I don't believe it is appropriate to place artificial bandwidth restrictions on folks that want to experiment. On the other hand, if the only reason for having a wide signal on a crowded band is to crow about how good your signal sounds versus everyones elses then that is pretty selfish and does infringe on other's ability to use valuable spectrum.

It has been mentioned that folks should reduce their bandwidth on a crowded band and only experiment when there is free spectrum available. I agree with this wholeheartedly. We are talking experimentation here and as such it can be delayed until the time is appropriate.

Again, my two cents.

Jim
WA0LYK

KA4DPO

03-11-2007, 06:56 PM

Quote[/b] (K3VR @ Mar. 11 2007,12:53)]Mark, that's another good field (pre-emphasis) of study for ham radio. Lossless compression techniques and intelligibility algorithms are built into the 8.3 kHz standard being fielded for digital voice right now -- eventually there will be a trickle-down effect into ham radio. Apparently, for now we have to rely on analog techniques. What would be really neat would be to see digital speech recorded in 15 kHz at the transmitter end, sent over the air in 500 Hz packets, and have it be "unpackaged" at the receiver end, and reconstructed back to its original 15 kHz size.

John, do you know when the studies you're discussing were written, or any of the authors? All of the studies should be referenced in the MIL-STD documents. None of the human factors studies are classified. I'm going to the library this afternoon, so if you can remember any of the studies you read, that would help me in trying to locate them. MIL-STD 188 and the ANSI are being re-written right now to (try) to keep up to date with current research. #

Also, imagine adding data to the voice stream as in VDL-3. Think of audio with a sub-sonic data component (below the threshold of hearing) so you could read text simultaneously as voice is being transmitted. The data could be automatically encoded using voice recognition algorithms. Closed captioned ham radio... That would open a whole new world of ham radio for hearing impaired people.
Brian, I just completed a study for the Vice Chief of Staff Army dealing with acoustic comparison between several types of combat helmet and headphone/microphone/intercom systems and combinations. #

The study was about a one year effort and was completed a few months ago. #The details and results of the study are classified per G3 Army. #

I have also done similar studies for the Modular Integrated Communications Helmet for SOF use, again the details of the study are classified. #Over the years I have done numerous other studies for acoustic enhancement devices for SOF and also radio and intercom systems for a wide variety of SOF applications. #Some of the older tests may be published but SOCOM doesn't advertise test results for obvious reasons.

Since The war began the official position of the Joint Chiefs is that the capabilities and limitations for any and all military systems or commercial systems tested for, and as they pertain to military use are classified.

n5rfx

03-11-2007, 07:16 PM

Quote[/b] (K3VR @ Mar. 10 2007,11:53)]Lossless compression techniques and intelligibility algorithms are built into the 8.3 kHz standard being fielded for digital voice right now -- eventually there will be a trickle-down effect into ham radio. Apparently, for now we have to rely on analog techniques. What would be really neat would be to see digital speech recorded in 15 kHz at the transmitter end, sent over the air in 500 Hz packets, and have it be "unpackaged" at the receiver end, and reconstructed back to its original 15 kHz size.
Brian,

This is interesting. When I ask digital voice enthusiasts, what is the motivation for using digital voice, the answer revolves around intelligibility and low noise.. I think that these two go hand in hand. The question is what value do we apply to intelligibility? 2.4 KHz analog voice is very spectrally efficient, and provides the best overall voice performance when taking many factors into consideration. Perhaps intelligibility needs greater weight in the evaluation, and spectral/power efficiency less weight in some circumstances. This can also apply to non-voice digital communications.

73,

Mark N5RFX

K3VR

03-11-2007, 07:38 PM

I'm rethinking this whole "crowded environment" rationale because I think it's based on an extremely transparent fallacy.

On the 75 meter phone band, NSSB enthusiasts use to claim a standard 3 kHz or so bandwidth for a total of 132 possible channels. RM-10740 asked for a limit on SSB and AM bandwidth. The FCC shot it out of the water. Recently, the FCC also disregarded the League's bandwidth petition which asked for a total bandwidth of 3.5 kHz for SSB and 9 kHz for AM.

The FCC could have easily incorporated the League's petition in their R&O, FCC-06149, which refarmed the bands, but they failed to include the ARRL's request in their Order.

PSK31 and CW enthusuasts will tell you that partitioning 80 meters for PSK31 and/or CW would provide space for more than 6,000 channels. So, if you look at the issue from their perspective, the guys "wasting" 3 kHz for NSSB fall into the band hog category too.

The fact is, unless there's a declared emergency or someone needs to order food supplies to the Amazon or the Arctic circle, all of our conversations are superfluous -- just talk for the sake of talk. And that's great. I wouldn't have ham radio be any other way. I enjoy chit chatting about cars and women as much as the next guy.

This brings me to my thesis (which will undoubtedly cause the use of bad grammar, virtual expletives, and multiple exclamation points). Just because people can use 3 kHz or less, doesn't mean they're correct or justified when they say "others are being selfish." Others aren't in fact being selfish, they're simply using legal bandwidths in order to pursue their enjoyment of the hobby. The 3 kHz guys could use narrower bandwidths but they enjoy using NSSB. That's fine. I'm not going to pressure them to use CW or PSK31 and no one else should either.

There are quite a few valid options if AM or ESSB is inhabiting your preferred watering hole:

1. Wait til the AM or EESB stations clear
2. QSY
3. Use a digital mode
4. Use another band
5. Use the internet
6. Use the telephone
7. Etc. - only imagination limits your options!

To pressure others or to make them feel guilty about their choice of mode is not only rude, it constitutes bandwidth discrimination and elitism © Brian Crow, 2007, all rights reserved.

Bottom Line: Mode discrimination is illegal, and promoting NSSB unity is often used to promote animosity against people who would prefer to listen to speech with greater fidelity.

Cain't we all jest git along?

.

KA4DPO

03-11-2007, 09:54 PM

I don't have any problem with it at all and welcome any experimentation as long as you don't QRM me when I'm 4 Kc down from you trying get a contact with rare DX.

Most of the guys who run ESSB have been around for a while and know how to act. #I have a lot more problems with some Bozo that tunes up on me in the middle of a QSO.

By the way, did you find anything at the library Brian??

ab0wr

03-12-2007, 12:36 AM

Quote[/b] (KA4DPO @ Mar. 11 2007,11:56)]Quote[/b] (K3VR @ Mar. 11 2007,12:53)]Mark, that's another good field (pre-emphasis) of study for ham radio. Lossless compression techniques and intelligibility algorithms are built into the 8.3 kHz standard being fielded for digital voice right now -- eventually there will be a trickle-down effect into ham radio. Apparently, for now we have to rely on analog techniques. What would be really neat would be to see digital speech recorded in 15 kHz at the transmitter end, sent over the air in 500 Hz packets, and have it be "unpackaged" at the receiver end, and reconstructed back to its original 15 kHz size.

John, do you know when the studies you're discussing were written, or any of the authors? All of the studies should be referenced in the MIL-STD documents. None of the human factors studies are classified. I'm going to the library this afternoon, so if you can remember any of the studies you read, that would help me in trying to locate them. MIL-STD 188 and the ANSI are being re-written right now to (try) to keep up to date with current research. #

Also, imagine adding data to the voice stream as in VDL-3. Think of audio with a sub-sonic data component (below the threshold of hearing) so you could read text simultaneously as voice is being transmitted. The data could be automatically encoded using voice recognition algorithms. Closed captioned ham radio... That would open a whole new world of ham radio for hearing impaired people.
Brian, I just completed a study for the Vice Chief of Staff Army dealing with acoustic comparison between several types of combat helmet and headphone/microphone/intercom systems and combinations. #

The study was about a one year effort and was completed a few months ago. #The details and results of the study are classified per G3 Army. #

I have also done similar studies for the Modular Integrated Communications Helmet for SOF use, again the details of the study are classified. #Over the years I have done numerous other studies for acoustic enhancement devices for SOF and also radio and intercom systems for a wide variety of SOF applications. #Some of the older tests may be published but SOCOM doesn't advertise test results for obvious reasons.

Since The war began the official position of the Joint Chiefs is that the capabilities and limitations for any and all military systems or commercial systems tested for, and as they pertain to military use are classified.
A couple of notes:

-Most of the studies we quote are based on open air hearing.
-It sounds like your studies are based on results using devices that inject the sound directly into the ear.
-It sounds like your studies are based on results using close-contact mic's if not direct contact mic's instead of "open air" mic's used in most ham stations.

If the testing environments are not the same, entirely different results will be reached. If the testing schemes are not the same entirely different schemes will be reached.

The method used in evaluating these systems should not be classified even if the results are. Do you use articulation index, speech recognition threshold, speech discrimination tests, speech intelligibility index, speech transmission index, rapid speech transmission index, or something else entirely?

The research I look at are what I hope will help me understand what happens when using my rig on the ham bands. It is not very likely that I am going to use mic/earphone systems designed for the military. It is quite likely that I will be using systems based on open air hearing, i.e. microphone and speaker, that most hams depend on.

We need to make sure we are comparing apples and oranges when involved in discussions like this.

tim ab0wr

k4kyv

03-12-2007, 12:47 AM

You may find this interesting. George claimed he got his data from experiments reported by Bell Telephone Labs right after WW2, although I have never seen anything written on the topic but what George wrote. This is particularly in reference to his statement that the human voice scans across a spectrum of frequencies at a certain rate, and that it is the rate of scan, not the frequency distribution of the voice spectrum, that determines the apparent pitch of a human voice.

http://amfone.net/Tech/EAM.pdf

WA0LYK

03-12-2007, 01:47 AM

Quote[/b] (K3VR @ Mar. 11 2007,12:38)]I'm rethinking this whole "crowded environment" rationale because I think it's based on an extremely transparent fallacy.

On the 75 meter phone band, NSSB enthusiasts use to claim a standard 3 kHz or so bandwidth for a total of 132 possible channels. RM-10740 asked for a limit on SSB and AM bandwidth. The FCC shot it out of the water. Recently, the FCC also disregarded the League's bandwidth petition which asked for a total bandwidth of 3.5 kHz for SSB and 9 kHz for AM.

The FCC could have easily incorporated the League's petition in their R&O, FCC-06149, which refarmed the bands, but they failed to include the ARRL's request in their Order.

PSK31 and CW enthusuasts will tell you that partitioning 80 meters for PSK31 and/or CW would provide space for more than 6,000 channels. So, if you look at the issue from their perspective, the guys "wasting" 3 kHz for NSSB fall into the band hog category too.

The fact is, unless there's a declared emergency or someone needs to order food supplies to the Amazon or the Arctic circle, all of our conversations are superfluous -- just talk for the sake of talk. And that's great. I wouldn't have ham radio be any other way. I enjoy chit chatting about cars and women as much as the next guy.

This brings me to my thesis (which will undoubtedly cause the use of bad grammar, virtual expletives, and multiple exclamation points). Just because people can use 3 kHz or less, doesn't mean they're correct or justified when they say "others are being selfish." Others aren't in fact being selfish, they're simply using legal bandwidths in order to pursue their enjoyment of the hobby. The 3 kHz guys could use narrower bandwidths but they enjoy using NSSB. That's fine. I'm not going to pressure them to use CW or PSK31 and no one else should either.

There are quite a few valid options if AM or ESSB is inhabiting your preferred watering hole:

1. Wait til the AM or EESB stations clear
2. QSY
3. Use a digital mode
4. Use another band
5. Use the internet
6. Use the telephone
7. Etc. - only imagination limits your options!

To pressure others or to make them feel guilty about their choice of mode is not only rude, it constitutes bandwidth discrimination and elitism © Brian Crow, 2007, all rights reserved.

Bottom Line: Mode discrimination is illegal, and promoting NSSB unity is often used to promote animosity against people who would prefer to listen to speech with greater fidelity.

Cain't we all jest git along?

.
Your logic is a little flawed. First, time and tradition have shown that in most cases ~3 kHz is satisfactory for most of the 'superfluous' conversations that take place. Additionally, your argument does not take into account the fact that we do have shared frequencies. Part of the sharing is to respect the right of others to use the bandwidth available. This means when the bands are crowded, unless there is an overriding need to use a wide bandwidth, then it is perhaps it would be best to reduce your bandwidth and let someone operate also. That doesn't mean you have to, nor does it mean that others shouldn't respect your right to use a wide bandwidth signal if you so wish. However, what goes around comes around. If you're using the last 6 kHz of a band for one conversation, when there could be two conversations, then one could say you're not really sharing.

When sharing, no one may get everything they want. That's life and is part of the 'all getting along' that you propose. Remember, just because you have a right doesn't mean it is appropriate to excercise that right when it infringes upon others. I can exercise my right to drive at only 45 mph on a crowded interstate, however, I'd better wear a bullet proof vest if I make a habit of it!

Jim
WA0LYK

WA0LYK

03-12-2007, 02:25 AM

Quote[/b] (k4kyv @ Mar. 11 2007,17:47)]You may find this interesting. George claimed he got his data from experiments reported by Bell Telephone Labs right after WW2, although I have never seen anything written on the topic but what George wrote. This is particularly in reference to his statement that the human voice scans across a spectrum of frequencies at a certain rate, and that it is the rate of scan, not the frequency distribution of the voice spectrum, that determines the apparent pitch of a human voice.

http://amfone.net/Tech/EAM.pdf
I read this but I'll be honest, I don't know what his definition of 'scan rate' means. I sincerely doubt that a women's voice has the same frequencies as a mans (in general) and that some kind of scan rate varies the percieved pitch. I'm not a doctor but unless my recollection is incorrect, I was under the impression that, in general, a women's vocal cords were smaller in diameter and shorter which results in resonant frequencies/harmonics that are higher in pitch than a man's.

Jim
WA0LYK

KA4DPO

03-12-2007, 02:31 AM

Quote[/b] (ab0wr @ Mar. 11 2007,19:36)]Quote[/b] (KA4DPO @ Mar. 11 2007,11:56)]Quote[/b] (K3VR @ Mar. 11 2007,12:53)]Mark, that's another good field (pre-emphasis) of study for ham radio. Lossless compression techniques and intelligibility algorithms are built into the 8.3 kHz standard being fielded for digital voice right now -- eventually there will be a trickle-down effect into ham radio. Apparently, for now we have to rely on analog techniques. What would be really neat would be to see digital speech recorded in 15 kHz at the transmitter end, sent over the air in 500 Hz packets, and have it be "unpackaged" at the receiver end, and reconstructed back to its original 15 kHz size.

John, do you know when the studies you're discussing were written, or any of the authors? All of the studies should be referenced in the MIL-STD documents. None of the human factors studies are classified. I'm going to the library this afternoon, so if you can remember any of the studies you read, that would help me in trying to locate them. MIL-STD 188 and the ANSI are being re-written right now to (try) to keep up to date with current research. #

Also, imagine adding data to the voice stream as in VDL-3. Think of audio with a sub-sonic data component (below the threshold of hearing) so you could read text simultaneously as voice is being transmitted. The data could be automatically encoded using voice recognition algorithms. Closed captioned ham radio... That would open a whole new world of ham radio for hearing impaired people.
Brian, I just completed a study for the Vice Chief of Staff Army dealing with acoustic comparison between several types of combat helmet and headphone/microphone/intercom systems and combinations. #

The study was about a one year effort and was completed a few months ago. #The details and results of the study are classified per G3 Army. #

I have also done similar studies for the Modular Integrated Communications Helmet for SOF use, again the details of the study are classified. #Over the years I have done numerous other studies for acoustic enhancement devices for SOF and also radio and intercom systems for a wide variety of SOF applications. #Some of the older tests may be published but SOCOM doesn't advertise test results for obvious reasons.

Since The war began the official position of the Joint Chiefs is that the capabilities and limitations for any and all military systems or commercial systems tested for, and as they pertain to military use are classified.
A couple of notes:

-Most of the studies we quote are based on open air hearing.
-It sounds like your studies are based on results using devices that inject the sound directly into the ear.
-It sounds like your studies are based on results using close-contact mic's if not direct contact mic's instead of "open air" mic's used in most ham stations.

If the testing environments are not the same, entirely different results will be reached. If the testing schemes are not the same entirely different schemes will be reached.

The method used in evaluating these systems should not be classified even if the results are. Do you use articulation index, speech recognition threshold, speech discrimination tests, speech intelligibility index, speech transmission index, rapid speech transmission index, or something else entirely?

The research I look at are what I hope will help me understand what happens when using my rig on the ham bands. It is not very likely that I am going to use mic/earphone systems designed for the military. It is quite likely that I will be using systems based on open air hearing, i.e. microphone and speaker, that most hams depend on.

We need to make sure we are comparing apples and oranges when involved in discussions like this.

tim ab0wr
Tim, you should read more thoroughly. If you had read a few posts back you would have read:

"All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself. "

So I don't try to compare apples and oranges. By the way, why are you so defensive about a hobby thing?

K3VR

03-12-2007, 05:59 AM

Warning Will Robinson -- My tongue's pressed firmly agin my cheek, and I'm fixin' to play Devil's advocate for awhile...

Quote[/b] ]First, time and tradition have shown

That the world is flat...
That man will never walk on the moon...
That women don't have the mental capacity to vote...
That blood letting lets out the bad bile...

Look, I already said we need to be good neighbors, and we need to share, but I'm not prepared to say NSSB should take precedence over ESSB or AM, or any other mode. Do the math. If there's an ESSB round table of 12 guys using 6 kHz, that's decidedly more efficient than a group of 4 guys on NSSB using the same 6 kHz. Likewise, if there are 20 guys on AM using 9 kHz, that's a far more efficient use of space than 6 guys using the same 9 kHz for NSSB.

Do you want to be the guy with the pocket protector and the calculator counting heads and kilohertz while attempting to regulate the efficiency of bandwidth useage on 80 meters? I don't. Neither does the FCC.

Let's extend the 'scarcity of resources' argument to the real world:

Oil reserves are scarce, so let's pass a law outlawing SUVs and Sports Cars. Better yet, let's outlaw NASCAR where all they ever do is burn gas and turn left anyhow. And these plane trips to Disney Land, how much fuel does it take to visit Mickey anyway? Time to outlaw vacations!

By the way, land is scarce too, so let's pass a law that says a family of 4 can only own a maximum of one acre. While we're at it, water's also damn scarce, and those stinky beef cattle emit methane like nobody's business. They also consume 31 gallons of water and 20 pounds of grain a day. Let's outlaw consumption of steak! Let's outlaw lawn watering too; it's just a bunch of stupid grass and it contributes nothing to the greater good!

And we have enough babies, so let's limit families to one child, and if that means we abort most of the girl babies because we like boys better... then we must live in Communist China, because that's exactly how they run their country!

In this country, we try to tolerate differences. And we develop new technologies to deal with problems. We don't need to advocate mode discrimination and promote a bunch of unwarranted animosity just so we can preserve our right to discuss our last oil change using NSSB. If it gets too crowded, we'll tell the FCC we need more room, and we can start by reclaiming 11 meters!

No, I damn sure don't want to be the one to tell a big group of AM guys they can't enjoy their gear because some NSSB guys in Texas want to talk about Britney Spears new haircut.

Bottom Line: The FCC did the right thing with the latest Omnibus R&O. They decided to protect users of AM and ESSB by law and then they expanded the phone bands so people can use the modes they enjoy without undue harassment from the 'moral' majority.

.

wa3vjb

03-12-2007, 11:43 AM

Quote[/b] ]Last year the FCC wanted to turn wideband data loose in the voice bands which would have hurt ESSB, SSB, AM, SSTV, etc. Myself and others worked for an 8 to 1 margain in comments AGAINST this.

Geez, Charlie, I thought someone would have nailed you on this real quick.

It was NOT our friends at the FCC who are threatening this proposal. It is yer man Rinaldo at the ARRL, and yes, the comments are overwhelmingly OPPOSED to the idea.

In fact, when it comes to the FCC's position on this, the agency decided 30 years ago that the greater Amateur community had made it clear it did not wish to have the existing, successful system of coordination scuttled.

The FCC's proposal (with Docket 20777) failed on public comment from Amateurs who said NO to its 1976 version of a segregation-by-bandwidth scheme. I was part of the campaign against the proposal, and have photocopies of the comments, deliberations, and conclusions by the FCC.

Despite that clear ruling, the League's Paul Rinaldo (who is plenty old enough to remember it) more recently has tried to push it again.

The comments filed this time with the FCC weighed even more heavily against the threat than in 1976.

But the ARRL, again demonstrating its insulation from the greater Amateur community, has not taken no for an answer and in February had an unpublished meeting with FCC officials to attempt some artificial respiration for its discredited plan.

Any questions ?

Paul/VJB

ab0wr

03-12-2007, 11:47 AM

Quote[/b] (n5rfx @ Mar. 11 2007,09:16)]Tim,

Thanks. #I have the book you are referencing. #I wish there were tabular data for those charts. #I looks like you were using the F2 noise charts to do the comparison.
Quote[/b] ]The biggest jump seems to be for 2500hz to 4000hz but it is significant out to 6khz.
The real impact here is by the pre-emphasis. The AI goes up by a factor of 1.4 for 6khz speech and about a factor of 1.2 for 2.5khz/4khz speech.
I agree, here is another way of looking at it:

Without Preemphasis:

Going from 2500 to 4000 Hz gives an AI increase of 1 dB and requires a SNR increase of 2 dB

Going from 4000 to 6000 gives an AI increase of .25 dB and requires a SNR increase of 1.76

Going from 2500 to 6000 gives an AI increase of 1.22 dB and requires a SNR increase of 3.8 dB

With Preemphasis

Going from 2500 to 4000 Hz gives an AI increase of 1.1 dB and requires a SNR increase of 2 dB

Going from 4000 to 6000 gives an AI increase of .65 dB and requires a SNR increase of 1.76 dB

Going from 2500 to 6000 gives an AI increase of 1.77 dB and requires a SNR increase of 3.8 dB

No Preemphasis to Preemphasis

Going from 2500 with no preemphasis to 4000 with preemphasis give an AI increase of 2 dB which is equal to the SNR increase required.

Going from 2500 with no preemphasis to 6000 with preemphasis give an AI increase of 2.56 dB and a SNR increase of 3.8 dB
Quote[/b] ]I think the point is that pre-emphasis is an area that could pay big dividends in intelligibility regardless of bandwidth used. This is one reason behind the popularity of D-104 mic's with their original crystal elements and the Heil HC-4/HC-5 mic elements. They provide some built-in pre-emphasis - but not nearly 6db/octave (at least as I remember).
Of course, this is going to result in what some would consider "audio distortion". This amount of pre-emphasis is certainly going to impact how people sound.
I agree. #I don't think the benefit of ESSB is intelligibility as much as it is nice sounding audio. #Most folks won't go with the Heil HC4 for normal ragchewing because of its harsh sound, but it will get you noticed.

Very interesting and thanks for making me crack open the Sabin and Schoenike book.

73,

Mark N5RFX
Mark,

I was using Fig 2.2 and Fig 2.4, Intelligibility of normal speech in white noise with a 200hz lower cutoff freq versus signal-to-noise ratio. One without preemphasis and one with.

How are you converting the AI index to a db figure?

I'm also interested in how you are coming up with the fact that an increase in AI from wider bandwidth *requires* an increase in signal-to-noise ratio. According the the graphs above, wider bandwidths give higher AI's for *all* signal-to-noise ratios. You don't have to increase signal-to-noise ratio to gain in AI if you use a wider bandwidth.

Wider bandwidths give you a higher noise figure for sure. But that is really only a penalty if the wider bandwidth does NOT also carry more intelligence which can be discerned.

We also need to be careful in interpreting Sabin and Schoenike's results. I was probably not clear enough about this. They use a specific model of how frequencies contribute to intelligibility so that the AI can be directly calculated.

Their model is that intelligibility is defined by a weighting function given by:

W(f) = (5 x 10**-4)exp[ (-4.2 x 10**4)(f) ]

You can then integrate this from Fa to Fb to get the total AI.

I've never researched their references to see if this is a valid weighting function. I have always assumed that it is since they used peer reviewed references and I've never seen their method questioned.

In any case, integrating this function will give increased intelligibility (i.e. discernable intelligence transmitted) over wider bandwidths even though the contribution goes down based on an exponential function.

tim ab0wr

ab0wr

03-12-2007, 12:05 PM

ka4dpo:Quote[/b] ]Tim, you should read more thoroughly. #If you had read a few posts back you would have read:

"All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself. "

So I don't try to compare apples and oranges. #By the way, why are you so defensive about a hobby thing?

WAIT A MINUTE!

You say that the data may not translate well to hobby interests and then turn around and try to say that it speaks for itself?

That's an argumentative fallacy called Amphiboly. You are presenting two valid interpretations with one being carefully substituted for the other while hoping the other person won't notice.

If your data doesn't translate well to hobby interests then it CAN NOT also speak for itself. You *are* comparing apples and oranges if the data doesn't translate well.

I don't consider myself defensive about anything. I do consider myself as being very careful in how I weigh facts and data to come to a judgement about an issue.

To me, this is more than a hobby thing. If you had read the thread carefully, you would have seen what my underlying interest in this is - my hearing loss as I get older. Specifically how the impacts of the bandwidths and gain equations chosen for my hearing aids affect my ability to understand those around me. This translates directly to my ability to understand those I talk to on the radio and I have become very interested in this subject.

Oh, BTW, I notice you didn't specify your testing methodology. Again, that should not be classified. I am very interested in what test method you used.

tim ab0wr

wa3vjb

03-12-2007, 12:10 PM

Quote[/b] ]use the modes they enjoy without undue harassment from the 'moral' majority.

Hardly a majority my friend.

The push comes from one guy at the ARRL who has rounded up his rubber stamp brigade to file on behalf of a group with a vested interest in obtaining federal grant money for "homeland security" communications.

The majority of us refuse to subscribe.

N5LRZ

03-12-2007, 12:43 PM

It is completely irrelevant how one sounds be it good or bad.

What IS releveant is what does the rules and regs STATE.

All you as operators have to do is to OBEY the rules and regs as defined by and instituted by the FCC.

Bottom line IF they say you will not sound like a commercial broadcast station then you had better NOT sound like one if it means violation of regs.

And the feds HAVE come down on amateurs running enhanced signals that have voilated tecnical specs of the FCC.

Valid Experimentation does NOT mean violation of the rules and regs.

N5LRZ

wa3vjb

03-12-2007, 12:45 PM

Quote[/b] ]It is completely irrelevant how one sounds be it good or bad. What IS releveant is what does the rules and regs STATE. All you as operators have to do is to OBEY the rules and regs as defined by and instituted by the FCC
Raymond,

Glad you checked in again on the issue you often get exercised about. You've implied that there is a potential rules violation automatically accompanying certain levels of audio quality you may encounter on the air.

Not true.
Please review Part 97.

You will find that the prevailing rules concern only technical purity of emission and a requirement to minimize the chance for interference. These simple and effective constraints have served the Amateur community very well for the 70+ years phone communications have been around.

Any further suggestions and judgments you've made in your posting are your latest effort to blow smoke. I may cough, but I won't blink.

Paul/VJB

wd0ct

03-12-2007, 01:09 PM

"And the feds HAVE come down on amateurs running enhanced signals that have voilated tecnical specs of the FCC."

More smoke. Name those that were came down on by fcc.

If you use those letters sent to several essb ops several years ago to prove your point, then we will all know you are full of it.

Those letters were not about any violations.

ab0wr

03-12-2007, 01:10 PM

n5lrz:Quote[/b] ]And the feds HAVE come down on amateurs running enhanced signals that have voilated tecnical specs of the FCC.

Just like they HAVE come down on amateurs running non-enhanced signals that have violated technical specs of the FCC.

The operative phrase here is "have violated technical specs" and not "running enhanced signals".

I can find no technical specs from the FCC that state specific bandwidth's to be used on any amplitude modulated signal in the voice part of the amateur spectrum. Can you?

tim ab0wr

KA4DPO

03-12-2007, 02:02 PM

Quote[/b] (ab0wr @ Mar. 12 2007,07:05)]ka4dpo:Quote[/b] ]Tim, you should read more thoroughly. #If you had read a few posts back you would have read:

"All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself. "

So I don't try to compare apples and oranges. #By the way, why are you so defensive about a hobby thing?

WAIT A MINUTE!

You say that the data may not translate well to hobby interests and then turn around and try to say that it speaks for itself?

That's an argumentative fallacy called Amphiboly. You are presenting two valid interpretations with one being carefully substituted for the other while hoping the other person won't notice.

If your data doesn't translate well to hobby interests then it CAN NOT also speak for itself. You *are* comparing apples and oranges if the data doesn't translate well.

I don't consider myself defensive about anything. I do consider myself as being very careful in how I weigh facts and data to come to a judgement about an issue.

To me, this is more than a hobby thing. If you had read the thread carefully, you would have seen what my underlying interest in this is - my hearing loss as I get older. Specifically how the impacts of the bandwidths and gain equations chosen for my hearing aids affect my ability to understand those around me. This translates directly to my ability to understand those I talk to on the radio and I have become very interested in this subject.

Oh, BTW, I notice you didn't specify your testing methodology. Again, that should not be classified. I am very interested in what test method you used.

tim ab0wr
Tim, The data is classified, trust me. #I don't care about extended speech and your blah blah etc and its effects on ham radio. #As I get older I turn up the volume. #Unless someone’s signal is ripped up I never have trouble understanding anyone on the air.

By the way, your transmitting a 6 KC wide signal won't make the other guy more understandable. #Get an MFJ speech enhancer.

You said:

(" I don't consider myself defensive about anything. I do consider myself as being very careful in how I weigh facts and data to come to a judgement about an issue.")

Oh, thats right, you're an expert. #I forgot.

You said:

(" Oh, BTW, I notice you didn't specify your testing methodology. Again, that should not be classified. I am very interested in what test method you used.")

Yeah, must have been an oversight on my part. #I'm sure the Army Staff would just love for me to tell you and can't wait for your expert analysis. #Our team of physicians and PHDs at UVA, Duke University, GW University Medical School, Walter Reed AMC, And USAARL often make a lot of stupid #mistakes in our analysis of critical issues. #Please help us out OK? #What I and my team do professionally is for real, we don't play and we don't accept hyperbole. #

If you don't think these systems, the test methodology and the results, should be classified feel free to call the Army Chief of Staffs office and tell them about it. Please be sure and let me know when you plan to call I want to sell tickets. #

I won't bother to argue with you any further since it appears to be pointless. #If you want to pursue your Audio experimentation as you call it please do just don't QRM me and we'll get along fine.

John..

n5rfx

03-12-2007, 02:51 PM

Tim,
Quote[/b] ]I was using Fig 2.2 and Fig 2.4, Intelligibility of normal speech in white noise with a 200hz lower cutoff freq versus signal-to-noise ratio. One without pre-emphasis and one with.
Ok, had to get the magnifying glass out for those, that is why I would like to derive the tabular data.
Quote[/b] ]How are you converting the AI index to a db figure?
10LOG(AI1/AI2) the reason I did this was to put the change in the same form as the change in signal to noise ratio.
Quote[/b] ]'m also interested in how you are coming up with the fact that an increase in AI from wider bandwidth *requires* an increase in signal-to-noise ratio.
I am assuming a radio receiver where I try to closely match the bandwidth of the received signal, with the bandwidth of the receiver. In this case the noise level changes with the bandwidth.

10LOG(BW1/BW2)

This means that as the bandwidth is narrowed, the noise goes down. As an example, when going from 2500Hz to 4000 Hz bandwidth the noise level will increase by 2dB.
Quote[/b] ]Wider bandwidths give you a higher noise figure for sure. But that is really only a penalty if the wider bandwidth does NOT also carry more intelligence which can be discerned.
Correct. To use the same SNR for two bandwidth curves you must understand the change in SNR due to the bandwidth change. What I was comparing is the change to AI vs. SNR when increasing bandwidth. To do this you subtract the AI gain from the SNR loss and you see that in most cases the change in SNR is greater than the change in AI, except when going from 2500 with no pre-emphasis to 4000 with pre-emphasis. In fact the best results come from adding pre-emphasis and not changing the bandwidth. I am comparing the ratios by converting them to dB instead of using factors. dBs are easier for me to look at and compare. This is why I want tabular data, so that we can see what the overall AI gain is. For example using the 2500Hz to 4000 Hz change without pre-emphasis, the AI at 2500 Hz and 10dB SNR is 0.4. Increasing the bandwidth to 4000 causes a 2dB drop in SNR, so the AI at 4000Hz in this case is still 0.4, no change in AI. AI gain is achieved by adding power to the signal. By inspection it looks like the AI gain from 2500 to 6000 with no pre-emphasis and adjustment for SNR is less than 1 dB. Using the 2500Hz to 4000 Hz change with pre-emphasis gives an AI gain. The AI at 2500 Hz 10dB SNR with no pre-emphasis is 0.4; the 4000Hz 8dB SNR with premphasis is 0.58. This is an AI gain of 1.6 dB. I see that I made a typo on my original post; I said the AI gain was 1.1 dB going from 2500 to 4000 with pre-emphasis, when it was actually 1.9 dB.

What I get from this is that speech intelligibility in radio is gained by widening the bandwidth, as long as the signal to noise ratio is also increased or pre-emphasis is added.

73,

Mark N5RFX

W3MIV

03-12-2007, 03:07 PM

Quote[/b] (wa3vjb @ Mar. 12 2007,07:10)]The push comes from one guy at the ARRL who has rounded up his rubber stamp brigade to file on behalf of a group with a vested interest in obtaining federal grant money for "homeland security" communications.
http://www.qrz.com/iB_html/non-cgi/emoticons/rock.gif

http://www.qrz.com/iB_html/non-cgi/emoticons/biggrin.gif

ky5u

03-12-2007, 03:38 PM

Please be careful mixing the current issue regarding automatic stations and wideband data with the enhanced SSB issue. You may see a connection that furthers your goals, but to many it's "...there's those ESSB guys trying to sell their favorite activity."

I do not favor restrictions being placed on ESSB other than what is currently in regulation. Nothing there precludes you from experimentation as long as the band is not crazy. But this has not alot to do with the ARRL scheme. Please comment against the Inmay letter and leave the ESSB fight for another time.

ab0wr

03-12-2007, 03:45 PM

Quote[/b] (KA4DPO @ Mar. 12 2007,07:02)]Quote[/b] (ab0wr @ Mar. 12 2007,07:05)]ka4dpo:Quote[/b] ]Tim, you should read more thoroughly. #If you had read a few posts back you would have read:

"All of my testing and evaluation is focused on very special operational conditions and environments so our data may not translate well to hobby interests but the data that I have speaks for itself. "

So I don't try to compare apples and oranges. #By the way, why are you so defensive about a hobby thing?

WAIT A MINUTE!

You say that the data may not translate well to hobby interests and then turn around and try to say that it speaks for itself?

That's an argumentative fallacy called Amphiboly. You are presenting two valid interpretations with one being carefully substituted for the other while hoping the other person won't notice.

If your data doesn't translate well to hobby interests then it CAN NOT also speak for itself. You *are* comparing apples and oranges if the data doesn't translate well.

I don't consider myself defensive about anything. I do consider myself as being very careful in how I weigh facts and data to come to a judgement about an issue.

To me, this is more than a hobby thing. If you had read the thread carefully, you would have seen what my underlying interest in this is - my hearing loss as I get older. Specifically how the impacts of the bandwidths and gain equations chosen for my hearing aids affect my ability to understand those around me. This translates directly to my ability to understand those I talk to on the radio and I have become very interested in this subject.

Oh, BTW, I notice you didn't specify your testing methodology. Again, that should not be classified. I am very interested in what test method you used.

tim ab0wr
Tim, The data is classified, trust me. #I don't care about extended speech and your blah blah etc and its effects on ham radio. #As I get older I turn up the volume. #Unless someone’s signal is ripped up I never have trouble understanding anyone on the air.

By the way, your transmitting a 6 KC wide signal won't make the other guy more understandable. #Get an MFJ speech enhancer.

You said:

(" I don't consider myself defensive about anything. I do consider myself as being very careful in how I weigh facts and data to come to a judgement about an issue.")

Oh, thats right, you're an expert. #I forgot.

You said:

(" Oh, BTW, I notice you didn't specify your testing methodology. Again, that should not be classified. I am very interested in what test method you used.")

Yeah, must have been an oversight on my part. #I'm sure the Army Staff would just love for me to tell you and can't wait for your expert analysis. #Our team o