Test your soundcard for clipping, with this sample Options

[lvqcl]

I have tested UDIAL.WAV encoded into 320 kbps MP3 [...]
distorted = I can hear cracking (clipping?) AT ANY VOLUME LEVEL
What is going on here?

This clipping is the result of encoding into MP3, nothing with your hardware.
Download LAME 3.97 and encode UDIAL.WAV as

"lame.exe -b 320 -k udial.wav udial.mp3"

[nonreality]

Just thinking about how many people might have accidentally left this in their library over the years and had it come on when listening to random music with the sound cranked up.

[WonderSlug]

Hello all. New member and decided to respond to this thread with my first post.

Anyway, I downloaded the original APE file in the first post and tested it on my system.

Here is my setup:

Windows XP SP2
Creative Soundblaster Live! PCI Value
DirectX 9.0c

All Windows XP Updates as of February 2008 except for DirectX (I don't have DX10 installed yet). I also have the latest drivers I could get for all my hardware.


Software Media Player: XMPlay 3.4.2.1 (by Un4Seen Developments) with Monkey Audio Plugin (xm-ape.dll)

using Output Device: DirectSound - SB Live! Audio [D000] with 16-bit stereo.

Also, using a set of "average" on-the-ear headphones plugged into the SB Live! They are "average" because I spent just $25 on them. So I guess they are "mainstream consumer" headphones. They are rated at 16 Ohm impedance and frequency range 20 - 22000 Hz with a SNR of 98 dB.

I turned off all reverb/chorus/equalizer and any other DSP effects in XMPlay. I also set the main WIndows Mixer volume to 50% with Wave volume to 100%. Within XMPlay I set the volume slider to 10% and then moved it up slowly until I reached a light comfortable listening level which was achieved at 30%.

With the regular non-altered 44100 Hz output, I get the starting 3 "beeps" of touch tone dialing and then a really high pitched razzing noise, like a ray gun firing or alien spaceship, that overpowers the rest of the beeps.

However, if I tell XMPlay to resample to 48000 Hz, that razzing noise goes away and I get only the touch tones, which I presume is the 'correct' sound.


If this is correct, then XMPlay's resampling to 48000Hz for Creative cards works very well. Before this test, I have always had XMPlay resample all sounds to 48000 Hz before outputting to the sound device because I had already read about the subpar upsampling that many Creative cards do.


My motherboard also has a built-in C-Media Wave AC'97 sound device, but I rarely use that thing so I didn't test it.

This post has been edited by WonderSlug: Feb 7 2008, 08:25

[krabapple]

The old links to udial.flac don't work, so I'm rehosting the file (I converted this to flac from the .ape file, whose link still does work)


udial.flac

This post has been edited by krabapple: May 8 2008, 17:10

[v.43]

audigy 2 zs, windows vista, foobar direct sound output, PPHS resampler. it sounds ok only if resampled at 32KHz OR if replaygain clipping prevention is applied and the volume is then lowered of about 20db

This post has been edited by v.43: Jul 14 2009, 23:29

[Arnold B. Kruege...]

For everybody fighting with occasional clipping (or perfectionists wanting to test everything) you should try the attached sample, udial.

Testing:

• play the sample in your favourite player (decoded or not, having disabled all DSPs and EQs that can interfere)
  
• if the output sounds weird at any time you should:
  • lower the output vol. in your soundcard config (should be the speaker icon in your system tray).
  • For some soundcards 48kHz is better (audigy 2 etc.).
  • lower the vol. till the sample doesn't sound weird at any time.

The output will sound really weird if your settings clip.
Your output can occasional clip without you knowing/hearing it. But for us perfectionists this will theoretically give a better quality output.

ATTENTION: Play this sample at a low volume anytime, even if you hear nothing special! It can be very harmful to equipment and/or your ears.
It's strongly recommended to use some very cheap (PC) speakers if you want to test this,
otherwise you might really ruin your tweeters (it has happened several times already).

*pedantic hat on*

Well yes it works, but it begs the question what sort of a signal makes the best possible audible test for clipping.

IMO the best possible test is the one that gives a clear audible indication with the least possible clipping.

I've long thought that the best way to do this would be to mix two tones that are < 22 KHz but high enough to be *every* hard to hear when they are pure tones. When clipped, there would be numerous spurious products smack dab in the middle of the the range where the ear is most sensitive.

I was just playing with such a pair of tones that I quickly generated in CEP/Audition. I could easily hear a difference when just 2,000 of 400,00 samples were clipped. Intrestingly enough, no single spurious response was above - 90 dB FS, but there were a lot of them.

With Udial, I had to clip about six times the percentage of samples before I could detect clipping as easily. The problem with Udial was that the stimulus tones were in the same frequency range as the spurious responses, and clearly audible. So they tended to mask their own distortion The spurious responses were few in number, but only about 40 dB down.