Windows bug causes decoding performance with certain Soundcard Interfaces.

NT1K, here is the information from the hardware ID:

USB\VID_08BB&PID_2904&REV_0100&MI_00
USB\VID_08BB&PID_2904&MI_00

I have had this signalink for a couple of years now.

Tom
K5TEE

 

That is one of the chips in question that has the bug. I am not saying you are wrong or anything. But if you messed around with the sliders and the output gain on your radio in the past, its possible that the levels look correct but it's very possible that you put a band aid on it. It's possible that you are just attenuating an already distorted signal. Reason why I am saying this is because so many people that have the exact same chip have the bug. Unless there is something else involved (like different drivers, hardware, conflicts) that is causing certain users to not see the issue or remedy it.

It's not like someone is crying wolf. There is something up.

 

I just upgraded to Windows 10 from Windows 8.1. When using Windows 8.1, I have never had a problem with decoding JT-65 signals using JT-65HF and WSTX. However, upon upgrading to Windows 10 , the software rarely decodes. If I am lucky, it will decode one signal when there should be 6 or 7. I have tried the fixes that other users have listed in this thread. None of them work. Anyone have any suggestions? I am running an ICOM-7410 via USB direct from rig to computer (no Signalink). I've spent a couple hours messing with the sound settings, no luck.

 

I do not know what sliders you are talking about nor the output gain of the radio.
This is the RAW audio data from my Signalink USB soundcard.
The Signalink is connected to the radio via the ACC port on the radio. As far as I know, that audio output is not adjustable.
If it was distorted before the audio scope, then there would have been flat peaks on both channels, there are none in the data I provided.
I understand that some have the same chip and have the distortion, but not at my station.
A Signalink does not have any drivers, it is listed as a sound device. There are, however, windows drivers for the sound devices.
My sound drivers are the ones that came with the windows 7 64 bit package.
I guess not everyone has the bug mentioned.

Tom
K5TEE

 

Update: After spending another couple hours messing with the settings, I finally fixed the problem. I had to move the recording level down to -11dB in the sounds settings. I'm not sure why I had to make this change upon upgrading to Windows 10, but I did and it seems to work now.

 

Not really a bug as such; the chip in question isn't WLK certified for versions of Windows after XP, which means that the behavior seen is actually correct, since the input on the PCM290xB is in fact a microphone input not a Line-in. This was changed in the PCM290xC revision. So if you want it to function as Line-in, you need to set the number of inputs to two and leave it there.

 

Is there any replacement for this chip ??

 

Toggling the channel count only gets rid of the gain (and makes windows see it as a line in device) until the device goes idle and it comes right back. So if you go from 1 channel and set it to two channel the gain will be gone, until the device goes idle, then its right back.

I've spoken with texas instruments, they have confirmed that they already knew of the issue, and there is a much simpler and permenant fix. I made an update video with this info that's much shorter and I highly recommend watching, and please send it to anyone you know using the old fix, as this is MUCH easier:

 

Forgot to mention - if you don't like the video form, the reddit post has been updated, the entire fix is now outlined in a single small "Fix" paragraph for those short on time/patience. Thanks! https://www.reddit.com/r/amateurrad...windows_bug_ruining_your_receive_performance/

 

In other words, follow the example of Joe, W4TV and others that use software that has a visual indication of safe audio levels? Rick KN6KB and the hams that use the WinMOR TNC and other software modems that he has written have never seen this bug, mostly because we don't allow the input to reach the saturation level to begin with.

 

Why has no one posted a Reg Hack ?

Those settings are in there.

Have Fun.

 

I can't say as I don't remember what Joe's example was. All I remember is him saying this bug doesn't exist and was "fixed" on win8/10, which as you can see from TI themselves is incorrect. The winmor users might not have seen the bug, that doesn't mean windows isn't sitting there adding the gain silently and they've just cranked their signalink down to compensate. I've heard from more than a few winlink users that this helped their success out quite a bit, but not sure what software they were using. Just trying to help! If your setup is working well enough for you as is then more power to you, I don't mean to stir up any controversy.

K4REF (who makes winlink help videos) said it improved his performance and the performance of his viewers quite a bit on winlink. Just one report though, who knows. After all the harassing emails saying I secretly work for kenwood, I'm certainly done investigating it though. So feel free to use it, or not

 

Announcement from Tigertronics, makers of the SignaLink interface:


We will put out an official response on our website soon, but in the
mean time I wanted to contact each of you that emailed us about the
Windows/PCM29xx "bug" report as quickly as possible so that this issue
can be put to rest.

The person who originally posted the report has just posted a new
you-tube video explaining how you can easily and permanently fix this
"bug" by simply adjusting the Window's software input level slider for
the SignaLink's sound card. Here is the link:



The fix posted in the above video is exactly what we were planning to
release once we received official confirmation from TI that our own
observations were correct. In short, the "bug" is permanently fixed
simply by lowering the "Level" slider for the SignaLink's "Microphone -
USB Audio Codec" sound card input from "+30db" to "0.0db". We will
update our Windows Vista/7/8 instructions with this information shortly,
but for now you'll find step-by-step text instructions for doing this below.

With all of the unwarranted negative attention and panic that this "bug"
report has caused for us and other manufacturers, we would sincerely
appreciate it if you would get the word out to your fellow Hams that
this issue has been resolved with a simple, common-sense fix. Thank you.

Regards,

Jason Yilek
Manager
Tigertronics


1 - Right-click the white colored speaker icon located in the
lower-right corner of your desktop and select "Recording Devices" from
the pop-up menu.

2 - In the new window that opens, click one time on the SignaLink's
"Microphone - USB Audio Codec" sound card to select it and then click
the "Properties" button.

3 - In the Properties window that opens, click the "Levels" tab.

4 - Right-click the percentage display to the right of the Level slider
and then select "decibels".

5 - Lower the Level slider to "0db" or as close as you can. This is
"-0.4db" in Windows 7. It might be slightly different in Vista, Windows
8 and 10, but in any case, the closet value to 0db will work just fine.
Note that you can use the left/right arrow keys to move the slider
once you've clicked on it. This might be easier than using your mouse.

6 - Click OK, then click OK on the Recording Devices window.

7 - That's all there is to it!

 

Or tigertronics could start using new revision chips.

 

"Or tigertronics could start using new revision chips."

They may have already. When I did the "fix", my SignaLink USB made the JT-65 program pretty- much unusable for me. No matter how far left or right I turned the RX knob on the device, I couldn't get it to center on zero or anywhere near. As a matter of fact, I couldn't get it to even come out of the red negative signal reading area that's to the left of zero.

I put it back to the original +30 setting and all is fine and back to normal now. All of the digital programs that I use ran fine before the "fix".

I guess if it aint broke...