Lend me your ears!

cornbb

I'm an audio software developer by trade and the product I'm currently working on is producing a very dodgy crackling noise. Its not a clipping issue. Its also not a crappy connection anywhere or any hardware issue as the problem is occurring on numerous machines and in rendered sound files. I was wondering if anyone here could have a listen to the following clip and see if they could identify the noise, eg wordclock issue, buffer problems, a crappy reverb, or what have you. My head is wrecked!

http://www.box.net/shared/htlu1gdk4f

Thanks!

squibs

Can't access the file due to firewall, but just curious. Is it volume dependent? Have you zoomed in on the waveform to the point where you are getting down to individual samples? If the problem can be reproduced so that you can get the crackles on top of digital silence, or failing that maybe a perfect square wave, looking at the signature of the waveform for the crackling might point you in the right direction. Without knowing more about the nature of the software, it's hard to be more specific. Any chance of a job?:D

cornbb

Cheers squibs, I've checked the waveform and there's plenty of headroom left. It doesn't sound like clipping anyway, its more of a crackling noise then a distortion.

The software is a sequencer (of sorts), the crackling problem actually seems to be occurring within an AU plugin (a fairly well-known sampler) which is hosted by our software.

Don't think they're looking for people here atm but I'll drop you a PM if that changes

Oriel

It certainly sounds like a volume issue to me. I don't think it would be hardware related.

booooonzo

Sounds similar to a wordlclock/sync issue alright or but it could be any number of things. Is it just on play back of the sampler in your software. does your software play other files ok or other Virtual instruments.

440Hz

Hullo Can I join this one lol!!

Its def not a volume issue, I mean its certainly not an issue with clipping.

At first the problem seemed to be machine specific, but that has since been ruled out as have 3rd party cards, OS versions, and for the most part software conflicts.

Also it is worth noting that the 'crackling' or 'distortion' (reluctantly describe it as this) is not consistent. A file could be played back a few times with no crackling and then the crackle will reappear on a subsequent playback without any change in the software.

Actually I have probably just confused things a bit more.

@booooonzo.. cheers for that comment, its a good point. cornbb - that is something to chat about tomorrow, it is just on the playback engine as the audio from the input file type is free from crackles. Which also rules out any hardware issue.

squibs

Has anybody actually taken a capture of the audio complete with crackles? You can usually see the crackles if you zoom in enough. Could it be a problem with zero crossings or cycles being retriggered before previous cycle completes?

Instant jumps in signal amplitude (even relatively small ones) will cause a click. For example if you splice 2 loops but the loops have been topped and tailed badly and don't start/finish at exactly 0 amplitude you will hear the click at the transition point where there's an instantaneous jump in the amplitude. Many clicks together would make a crackle I imagine.

cornbb

Have taken a close look at the waveform and it looks pristine. No nasty jagged edges, jumps in amplitude etc. It sounds like an analogue crackling which is really weird, as its all happening in the digital domain. This leads me to believe it might be a problem with a dodgy reverb or something.

jimi_t

Probably too obvious but... have you considered it might be to do with wireless networking/the OS in general?

440Hz

Cheers jimi_t I have checked all of that out and ruled it out. Def not the OS and would be very doubtful that its a network issue. Thanks for the suggestion though.

440Hz

Machines vary from Mac Mini's to Mac Pro's and they all have demonstrated the problem. Machines with up to 4GB RAM have this crackle. Have monitored CPU activity in relation to this and it doesnt seem to correlate

cornbb

Tried it on a number of machines with different specs, from an older G4 Powerbook with 512mb RAM to a new Mac Pro with 4gb. Have found no correlation between the problem and the amount of welly the machine produces!

Regarding the bouncing thing, they are being rendered as 16 bit 48kHz PCM files (WAV or AIFF). The unusual samplerate is to cater for a specific type of customer. Regardless of this, the problem is occurring during real-time playback too.

Thanks for the ideas, keep em coming!

cornbb

440Hz wrote:

Machines vary from Mac Mini's to Mac Pro's and they all have demonstrated the problem. Machines with up to 4GB RAM have this crackle. Have monitored CPU activity in relation to this and it doesnt seem to correlate

Damn, need to get out of bed early to keep up with you

frobisher

This is interesting. I don't know much about plug in design but audio is audio right? My first thought on listening to it was that is was a buffer or cpu issue. Especially considering that 440hz has said that the problem isn't consistent. But if you have tried it in several machines that have different settings then that should rule that out. UNLESS the file you are transferring is carrying data that will change the preferences and options of your sequencer so therefore while they may be different machines they're still at the same settings. I guess that depends on what program you're using.

I'm listening on cheap headphones on a PC with a crap sound card but there's a few things I notice about the sound. The crackling (I'll just stick with calling the offending noise that) seems to follow the pizzicato strings more than the piano so I'd look in that direction first. Although I'm prepared to be wrong on that as the 16th note feel of the piano is fairly consistent in dynamics so maybe it's just not as noticeable. If it is the strings then what is on the strings only or more than the piano? Is there any limiting/compression/EQ on it? Even if it's set so that it's not doing anything it could be that. Is it me or does the crackling get louder along with the strings but very, very slightly after the peak? That could point again to the reverb maybe having a pre-delay and if it is the source the crackling it is coming in after whatever millisecond pre-delay is programmed into it. It's extremely hard to reverse engineer these things just by listening, especially if you don't know or aren't familiar with the set up. If it isn't a system settings issue and there isn't a stray plugin or routing option etc at play I'd look to the reverb. It even sounds like it could be as simple as overload on input.

Be sure to let us know!!

frobisher

Niall - Dahlia wrote:

Maybe it's a combination, a convolution reverb by any chance, that's completely hogging system resources?

By the by, are corn and 440Hz one and the same? I can't quite figure it out in my slightly inebriated state...

I think the work or go to college together. Something along those lines.

artvandulet

Sounds to me like a reverb issue. The noise is only on the reverb of the pizzicato. Is this an onboard or outboard reverb unit? If you bypass the reverb im assuming it goes away yes? I had very similar sound issue when I used specific reverbs on a nanoverb I used to have.

440Hz

cornbb and I work in the same company! so working on the same problem! Im a regular reader but not a regular poster so when I saw the thread I thought I should join in

brettzy

I think artvandulet is on the right track. It sound to me like a reverb fault!! Is it possible the transient of the pluck is causing an internal clip within the reverb unit??

If you increase the pre-delay does the noise move in relation to the pre-delay time??

cornbb

Just so ye know - the problem has been fixed. It was caused by a bug in the plugin we were using and we got over it by disabling multiprocessor support.

Crisis averted - phew

frobisher

What plug in?