Odd distortion

[kozikowski]

Here. I intentionally damaged a sound file...
That peak flattening is most unusual in a sound file and creates very bad distortion.
Koz

[JulianHJ]

Apologies, didn't mean to sound snappy

I'd burnt the CD a little while ago to play the tracks in my car, which is when I first noticed the problem (as I'd not listened back to the tracks on my PC before playing the disc in my car).

I know what you mean in terms of the tops of the wave forms flattening out - as though they are clipping.

Here's a snip of a recording I've just made:



This is a close-up of one of the drum beats where the distortion is obvious:



I'm not sure how I could upload an audio clip so you could get an idea of the issue first-hand.

[steve]

Could you post a bigger picture of that second image. There appear to be some odd little spikes but I can't quite make it out.

Short (less than 1MB I think) MP3 or WAV files can be uploaded when you are making your post - look for the "upload" tab below the message compose box.
A short audio sample would probably be better than another picture.

[JulianHJ]

Apologies for the delay, here is a single drum noise, recorded normally. It sounds fine 'live' but distorted on playback.

The original image can be viewed here: http://www.flickr.com/photos/julianhj/4 ... 5/sizes/o/

[steve]

It's those little spikes that are the problem - that is not normal at all.
It isn't caused by "clipping", but looks more like data corruption.

It's difficult to tell if the spikes are evenly place - they look to be around every 400 samples or multiples of that - possibly bad RAM buffering the audio input? Other ideas anyone?

I'm beginning to think the issue is with either the PC itself or the soundcard. I'm going to try moving the card to a different PCI slot as well.

That's looking like a possibility. Have you tried that yet? If it's easy for you to do then I think it's worth testing.

Has anyone mentioned sound card drivers yet? Driver bugs can cause all sorts of weird problems - if you've not done so already, check that the sound card drivers are up to date.

[JulianHJ]

I moved the sound card to a different PCI slot today, and the computer refused to see it - I assumed it would have to be completely reinstalled, so moved it straight back.

I checked my M-Audio driver, and it's up to date. I even unplugged and dragged my AV receiver up to the study to see if an alternative method of pre-amping would solve the issue. It didn't.

Plan B is about to be implemented - the purchase of a NAD PP3 phono pre-amp. This has a USB output, so not only can I by-pass the M-Audio card, I can also exclude my aging desktop from the equation and use my much newer laptop. It does initially mean I lose the gain control if I go deck > NAD > laptop, however the NAD also has a line in. If the 'soft-clipping' of the NAD's DAC proves a problem, I can run he deck through the DJ mixer to maintain gain control.

Thanks for all your responses so far, they've stopped me tearing my hair out!

[waxcylinder]

... It does initially mean I lose the gain control if I go deck > NAD > laptop, however the NAD also has a line in. If the 'soft-clipping' of the NAD's DAC proves a problem ...

Maybe, but maybe not. I always used to think that with a USB device like this connected than you lost the ability to use the software gain slider in Audacity. Cetailny this was how my earlier ION iTTUSB behaved and how my external USB soundcard (Edirol UA-1EX) still behaves.

But I am assured by Gale Andrews that with some USB devices and the correct drivers then the Audacity software input gain control will operate. He assures me that his later ION TTs work this way.

He believes that my inability to use the input slider in Audacity is possibly down to the fact that I use my Edirol with the Windows supplied drivers - and that if I installed the Edirol drivers then the slider should work. I am reticent about trying this as my current setup works extremely well and I am loathe to change this - I don't need the software gain control as I have hardware gain controls both on the Edirol soundcard and on my phono pre-amp (I use the ARTcessories DJ-PreII). I did consider the NAD PP2 (which is the OO£ but without the soundcard&USB) on my shortlist when I was buying, but the lack of gain control put me off. But the NAD certainly looks to be a good piece of kit.

Once you have implemented plan B and got the NAD working with Audacity perhaps you could contribute a brief review of the device in this sticky thread we have re. soundcards that work well with Audacity (which Stevethefiddle started a while back): http://forum.audacityteam.org/viewtopic.php?f=27&t=9477

WC

[kozikowski]

It's not clipping or flat-topping. That's serious data distortion. It's not just the spikes, either. Look at how some of the graceful up and down progression of the waves is interrupted like someone took a razor to it. Look about 5.150. That's apart from the spikes which are very serious.

I'm just looking closely. There's several of those discontinuous offsets. There's another one at 5.190 and then again at 5.215.

That damage happens at a different period from the spikes. So there's two data errors.

Koz

[steve]

If I had to guess I'd say faulty sound card, though I'd like to test it, either in a different PC (preferably with a different OS) or in a different recording program using ASIO drivers (preferable on a different machine).