Forums | Search | Archives

All Forums  Cool Edit  Crackle and sample rate.

Author	Topic
terran Posts: 12	Posted - Tue May 20, 2003 5:06 am  A customer wanted some recordings transferred from audio cassettes to cd-r. No problem the kind of thing I've done a thousand times before, set it up monitored a bit of the recording sounded quite good (metal cassettes). Then when I played back the files I noticed a slight crackle rather like some clipping effects you can get when the input is set too high.(This was not present when monitoring the recording). Double checked everything just could not see what the problem was. Then I noticed the sample rate had been set at 44100, generally I use 48000 unless there is some reason another rate is needed. Tried recording again, perfect! Now I am just curious as to why crackle is introduced as a function of sample rate? Anyone know?
ozpeter Location: Australia Posts: 3200	Posted - Wed May 21, 2003 3:49 am  Something must be interfering with the clocking of your soundcard somewhere. System sounds can (SteveG informs us if I remember rightly) upset clocking. What's your system configuration esp soundcard? - Ozpeter BTW some cassettes have static problems, so playing them generates static clicks heard on playback - though you say you couldn't hear the problem during transfer, I note.
Tuskin Location: Finland Posts: 56	Posted - Wed May 21, 2003 4:45 am  I've had the exact same problem. I'm still looking for a cure. It seems to come and go, sometimes I get crackle and sometimes not. I think lowering the samplerate to around 16K got rid of it. But what on earth would I do with recordings on that samplerate. _________________ Tuskin Hiljaa hyvä tulee.
terran Posts: 12	Posted - Wed May 21, 2003 5:08 am  Thanks for the replies, No absolutely fine on transfer, the sound card is on this machine just a SB 5.1 but I have always found it perfectly adequate for what is needed. Running XP Pro (dual booted with W98SE ) on a venerable but again adequate AMD 750 driven homemade box. The strange thing is when booted to 98 the problem disappears. Since is does not occur at 48000 on either o/s it really is not a problem I was just curious to know if anyone knew why.
toodoom Posts: 7	Posted - Wed May 21, 2003 6:37 am  I agree with ozpeter,it must be something with clocking.Did jou try to remove jitter?Waves idr is a good plugin that deals with this kind of problems. hope it helps
SteveG Location: United Kingdom Posts: 6695	Posted - Wed May 21, 2003 7:14 am  Well firstly, you can't write a CD at 48k, so if your file was sampled at this rate, it would have been converted on the fly to 44.1k anyway. If it's all in the PC you can forget about jitter - this won't have anything to do with correct writing at all. You are transferring a file into an app that converts it to CDA code, which it sends as variable-length signals to a laser, at a rate that is locally determined by the spin rate of the disc. How on earth does jitter enter into this? Answer - it doesn't, because the system is designed to cope with the mechanics of this. Jitter is a purely electronic phenomenon that occurs in A-D and D-A conversion - not really anywhere else, and certainly not in CD-writing! Really, we need to know a few things. Like what app you are using to write the CD-Rs, was anything else happening on the machine, what blanks were they, what was the peak level of the file (almost all professional mastering is done to -0.1dB (ref highest available code) unless the client specifically requests stupidity - all those clipped CDAs usually have clipped masters!) On the face of it, the most likely explanation is that the disc was written rather more slowly when the on-the-fly conversion was happening, and quicker when it wasn't. So what speed did you write the disc at? _________________
terran Posts: 12	Posted - Wed May 21, 2003 7:36 am  It never got to the writing stage. The problem was detected before that when the wav file was reviewed and checked for any restoration that might have been required. There was no problem with the subsequent burning of disc, the file first being converted to 44.1k for the writing software (something that is done as a matter of course due to the differing formats I handle) SteveG wrote: Well firstly, you can't write a CD at 48k, so if your file was sampled at this rate, it would have been converted on the fly to 44.1k anyway. If it's all in the PC you can forget about jitter - this won't have anything to do with correct writing at all. You are transferring a file into an app that converts it to CDA code, which it sends as variable-length signals to a laser, at a rate that is locally determined by the spin rate of the disc. How on earth does jitter enter into this? Answer - it doesn't, because the system is designed to cope with the mechanics of this. Jitter is a purely electronic phenomenon that occurs in A-D and D-A conversion - not really anywhere else, and certainly not in CD-writing! Really, we need to know a few things. Like what app you are using to write the CD-Rs, was anything else happening on the machine, what blanks were they, what was the peak level of the file (almost all professional mastering is done to -0.1dB (ref highest available code) unless the client specifically requests stupidity - all those clipped CDAs usually have clipped masters!) On the face of it, the most likely explanation is that the disc was written rather more slowly when the on-the-fly conversion was happening, and quicker when it wasn't. So what speed did you write the disc at?
SteveG Location: United Kingdom Posts: 6695	Posted - Wed May 21, 2003 7:55 am  Okay, so it's not a CD problem.... :) Did you do the file conversion in CEP? Jitter still doesn't enter into this... I don't think that ozpeter's reference to my mention of system sounds is likely to be the cause of this either - this just causes huge errors on playback - like making files play at twice the rate, etc. It is true that some soundcards 'prefer' to work at some rates rather than others, but this is usually quite consistent - but I'm not aware of SB5.1 cards 'crackling' at 44.1k. If you are recording in CEP, then it might be worth altering the record buffer sizes around to see if this alters the situation - having incompatible settings here can cause all sorts of little recording problems. _________________
terran Posts: 12	Posted - Wed May 21, 2003 8:10 am  Yep, CEP. But just to clarify, 44.1k works ok in 98, 44.1k in Xp(dual boot on the same machine) with the same settings gives the little crackles. I normally use 48k for line in recordings, there are no crackles at this setting in either system, it was only when it was set to 41k in Xp that I noticed this, so it is not a problem. I am not too bothered about this, just curious. SteveG wrote: Okay, so it's not a CD problem.... :) Did you do the file conversion in CEP? Jitter still doesn't enter into this... I don't think that ozpeter's reference to my mention of system sounds is likely to be the cause of this either - this just causes huge errors on playback - like making files play at twice the rate, etc. It is true that some soundcards 'prefer' to work at some rates rather than others, but this is usually quite consistent - but I'm not aware of SB5.1 cards 'crackling' at 44.1k. If you are recording in CEP, then it might be worth altering the record buffer sizes around to see if this alters the situation - having incompatible settings here can cause all sorts of little recording problems.
SteveG Location: United Kingdom Posts: 6695	Posted - Wed May 21, 2003 11:41 am  terran wrote: ...44.1k works ok in 98, 44.1k in Xp(dual boot on the same machine) with the same settings gives the little crackles. I normally use 48k for line in recordings, there are no crackles at this setting in either system, it was only when it was set to 41k in Xp that I noticed this, so it is not a problem. I am not too bothered about this, just curious. Could still be buffering problems. It has been said that a totally stripped down to the bare bones version of XP runs at about the same speed as Win98 - that's what they call 'progress'. There have been all sorts of little niggles with XP since it was launched, and your problems could be as much to do with drivers as anything. VxD works fine, but WDM has definitely had its moments... Patient:"Doctor, Doctor, if I put one hand over my shoulder and the other one up my back and clasp them together and pull, it hurts..." Doctor:"Don't do it then!"[/list:59b2151977] _________________
DeluXMan Location: Canada Posts: 330	Posted - Sat Aug 09, 2003 2:44 pm  I have the dubious task of mastering a mix with obvious clocking crackle(sounds like kindling but brighter), created by an idiot i guess, but this is all i have to work with, so it poses the question - Has anyone had any success removing the clicks using CEP? I found this thread on my quest, and realize this falls under polishing a turd, but hey, you can buy polished turds on the internet and in jewelry stores can't you? (not to mention in record stores). I have had great results removing these clicks one at a time, by filtering a few ms. only, where the click resides, but it is tedious. This file has way too many clicks to remove by hand. :D The reason i think there is a way is that these are all single sample clicks. On the other hand that means the impulse covers the spectrum... TIA
DeluXMan Location: Canada Posts: 330	Posted - Sat Aug 09, 2003 8:51 pm  DeluXMan wrote: I have the dubious task of mastering a mix with obvious clocking crackle(sounds like kindling but brighter), created by an idiot i guess, but this is all i have to work with, so it poses the question - Has anyone had any success removing the clicks using CEP? I found this thread on my quest, and realize this falls under polishing a turd, but hey, you can buy polished turds on the internet and in jewelry stores can't you? (not to mention in record stores). I have had great results removing these clicks one at a time, by filtering a few ms. only, where the click resides, but it is tedious. This file has way too many clicks to remove by hand. :D The reason i think there is a way is that these are all single sample clicks. On the other hand that means the impulse covers the spectrum... TIA :P Well so far i managed to hammer the high frequency spikes down using the basic de-esser technique as a de-spiker instead by raising the band to a high-pass and zero attack with 1 ms. look-ahead and 15 ms. release time. Now the cracks sound like much quieter 'lumps' but they are somewhat masked by the compressed highs.
AndyH Posts: 1425	Posted - Sat Aug 09, 2003 11:42 pm  Basic Click and Crackle removal doesn't work?
DeluXMan Location: Canada Posts: 330	Posted - Sun Aug 10, 2003 3:49 pm  Basic click removal leaves about 20-30% of these clicks untouched, but i will play around with the setting some more.
	Topic

Page: