ALERT: Possible problem with audio fidelity on Windows 7

[sullivang]

There appears to be a problem with audio fidelity on Windows 7 (may affect Vista too), in some applications. So far, the most important one (to me, at least) is YouTube, and I suspect, any web site that uses FLASH content.

The problem occurs when the sample rate of the audio content is different to the sample rate that the audio interface is set to. For example, on my netbook, the default sample rate is 48kHz, and YouTube audio produces quite noticable artifacts. When I set the audio interface to 44.1kHz, the problem goes away.

The problem does not seem to affect Windows XP. (and I've actually tried quite hard to induce the problem).

I am documenting the problem on the MSDN Pro Audio Developers Forum:
http://social.msdn.microsoft.com/Forums/en/windowspro-audiodevelopment/thread/725546ce-57bf-40d0-b7aa-47e51de9c3ae

The first discussion to occur about this issue that I am aware of is on this blog site:
http://blog.szynalski.com/2009/11/17/an-audiophiles-look-at-the-audio-stack-in-windows-vista-and-7/

Greg.

 

[COLGeek]

I would tend to look at the netbook's anemic performance, compared to a more capable notebook or desktop, as the culprit. I have been using Win 7 since Beta 1 and on multiple systems and have not seen what you are describing.

The netbook's processing power probably has more to do with your observation than Win 7. Just something to add to the discussion.

Have fun!

 

[sullivang]

Two other people, at least, have noticed the prolbem as well, however I have not yet asked them what hardware they are using.

Have you actually listened carefully to the test tone that I mentioned on MSDN, with your sound card set at 48kHz? If not, give it a try. If you still don't hear anything, be sure to compare the result with your soundcard set to 44.1kHz. Even then, if you STILL don't notice a difference, you should record the output and then inspect the frequency spectrum.

It might also be worth trying a test tone generator, such as this one:
http://www.esseraudio.com/en/test-tone-generator-windows-software-generate-test-signal-sine-pink-noise-crest-factor.html

This has the problem (on my system), if my audio interface is set to 48kHz. I can hear the aliasing very clearly. Use a low frequency sine wave, such as 100Hz.
Switch back and forth between 44.1kHz and 48kHz if you don't notice the difference.

You might be right. I asked on the blog whether Windows may detect the CPU capability, and use a lower grade of sample rate interpolation for low performance systems. I also asked whether higher grades of Windows might be enabled for higher quality. That's when someone running Windows 7 Professional answered that they had the problem, although as I say, he didn't specify his hardware.
(but we know that it's probably not related to the grade of Windows)

Greg.

 

[sullivang]

Just by the way, on my old clunker Windows XP Pentium M 1.4GHz laptop, I do not have the problem. That's the system I used for the XP testing.

Also, my netbook is CAPABLE of performing high quality sample rate conversion, because the Windows Media Player works fine. (as does iTunes, although I am relying on another user for that information)

There are some test tones for download here: http://www.mediacollege.com/audio/tone/download/
Download a 100Hz tone (the longer 30s one is probably best), and then play it with Winamp. Compare the result using DirectSound (should be ok) to that when you use WaveOut, for the output renderer.

Also, Inmatrix (the Zoom Player developers) have confirmed that their player produces the problem, when Waveout is selected. Here is their post: http://forum.inmatrix.com/index.php?showtopic=11609&view=findpost&p=43349 I am guessing that they used something a bit more powerful than a netbook, but I haven't actually asked yet.

Greg.

 

[COLGeek]

Keep in mind how CPU intensive Flash can be. Regardless, I'll dig into it further when I get some spare time.

Good stuff to think about. It will be interesting to see what others say on the subject.

Last, not surprised that your clunker handles well.

 

[sullivang]

Thanks. (note that I've added some stuff to my previous reply - please re-read)

Greg.