Topic: Crackle on Pipe Organ Concert recording..

[RebelRebel]

The building is ancient, so who knows. I believe in spirits, for sure.

OK, I'll give you that, but broken metal straps and ancient rat bones can count for something, too.

Are you sure that there were no gear resonances ?
I always wonder whether my stand, stereo bar, mic-pre connection, cables, etc. might cause some strange stuff at higher SPLs.

Im not sure Bob, no..but Ive been reading up on pipe organs since I will be archiving all the concerts there from now on to try to sidestep any future issues. The instrument produces a lot of frequencies/harmonics out of the range of human hearing...so I think that fact may have something to do with it the weirdness as well. I am not ruling out the gear issue by any means, but from what ive read on pipe organs lately..it seems that  the inaudible frequencies may be a factor...This is uncharted territory for me, hopefully Ill get it figured out. Thanks for all the tips everyone.

[prof_peabody]

Are you running any sort of low pass filter? (What filter are you running at the high end?)  I agree with others that there may be some high pitched harmonics.  If you have a relatively sharp filter at the high end you could produce some aliasing which would sound like noise.  This is just a stab in the dark.

EDIT: looking at the pics that skalinder posted, it is obvious there is a very sharp frequency filter at the high end.  With the energy right at your cut frequency you will get some aliasing, hence the crackling you hear (but barely, it is hardly noticable to me).  You may want to try some sort of filter in post processing with a nice wide taper to cut out the high end.

EDIT AGAIN: Having cleaned 19th century organ pipes and checked out a few old organs in my time, I wonder if there is the possibility that a small leak on a lead to a pipe could cause a high frequency whistle.  Do you notice this high frequency energy elsewhere in your recording?

[RebelRebel]

The crackle I heard wasnt a charactaristic of the recording at all, but rather a latency issue with my PC. The recording is fine sonically. There is some energy present outside of the range of hearing, but I am almost certain that it is due to harmonics ...Pipe Organs produce a lot of frequencies that we cannot hear...The Cathedral's acoustic properties probably added to the harmonic effects..I dont run a low pass filter, it normally isnt necessary. The recording is fine, I just wanted to see what the source of that unexplained energy was...I dont really notice anything with my ears, but using the FFT Analyzer you can see it.

Are you running any sort of low pass filter? (What filter are you running at the high end?)  I agree with others that there may be some high pitched harmonics.  If you have a relatively sharp filter at the high end you could produce some aliasing which would sound like noise.  This is just a stab in the dark.

EDIT: looking at the pics that skalinder posted, it is obvious there is a very sharp frequency filter at the high end.  With the energy right at your cut frequency you will get some aliasing, hence the crackling you hear (but barely, it is hardly noticable to me).  You may want to try some sort of filter in post processing with a nice wide taper to cut out the high end.

EDIT AGAIN: Having cleaned 19th century organ pipes and checked out a few old organs in my time, I wonder if there is the possibility that a small leak on a lead to a pipe could cause a high frequency whistle.  Do you notice this high frequency energy elsewhere in your recording?

[prof_peabody]

The crackle I heard wasnt a charactaristic of the recording at all, but rather a latency issue with my PC. The recording is fine sonically. There is some energy present outside of the range of hearing, but I am almost certain that it is due to harmonics ...Pipe Organs produce a lot of frequencies that we cannot hear...The Cathedral's acoustic properties probably added to the harmonic effects..I dont run a low pass filter, it normally isnt necessary. The recording is fine, I just wanted to see what the source of that unexplained energy was...I dont really notice anything with my ears, but using the FFT Analyzer you can see it.

You might want to look into the aliasing issue anyway.  You probably don't notice it on playback, but it will introduce mostly random noise into your recording.

Everywhere I said aliasing... I meant ringing (Gibbs phenomenon).  I am reading into it again now... it ha been a few years.

[prof_peabody]

Apparently the low pass filtering of digital recordings is a big issue in the pro audio world.  I think I am going to have to withdraw from making insights.

My final remark is that it seams really odd that there is suck a rapid cut at the Nyquist limit.  In my line of work that is not a good thing (I work with a different type of audio...)

[CHURCH-AUDIO]

Frequencies vary from song to song we are talking about amplitude, harmonics, duration and sympathetic resonation of the mic diaphram or somthing near the mic diaphram can cause distortion on a recording when the song before it was fine. One of both of the DPA mics could be bad.
Also here is a big problem with gear most distortion figures are based on 1k not multiple frequencies. So you might have recorded something that was at the right frequencey and amplitude to set off this distortion. If you look at a real distortion graph even based on 1k with broad band fft you will see spikes at other frequencies these are harmonics of 1k. These are also high in distortion so you can have one frequencey that has a harmonic on one of these "spikes" that will set off the distortion. The only way to tell for sure is to get a source transducer that has very little distortion and do a frequencey sweep analysis most likely the distortion area will appear as a bump in frequencey response in the sweep if the source transducer is flat.

IN OTHER WORDS TURN DOWN THE GAIN OR see if you have a bad mic capsule or a bad cable or a bad input on a preamp or recorder.

I am trying to get this 2nd piece from the other day tracked, etc...and I notice some crackle during the louder portions of the piece...What is interesting is that in the Bach piece that was recorded right before it,,crackle isnt an issue(at the very end it shows up , but not at all like this one..this piece(By Charles Widor) is much louder and a lot more spirited than the Bach. The recording chain was:
DPA 4061s>>>mma6000>>toddR mini to 2XLRm cable>>>mytek>>>722

This was in a cathedral...no PA or anything...just the instrument and my gear....So it doesnt really seem to be a cable issue I dont think, but maybe some weirdness created by the reverb in the hall or perhaps the intensity of the output of the pipes...Low Frequency problems..(the Widor definitely hangs around the underbelly of the Bass Clef moreso than the Bach)..Or as Moke has said before "Cathedral Hall Reverb Anomaly" heck, I dont know really...I just know I have another Organ Recital Saturday, a Chamber Ensemble+Vocalist recording sunday and Rutters Requiem Monday...So I want to avoid this at all costs. I will be using the DPA/Jdisc/mytek and 722 again for the organ concert and then add in the MGs or u87s the following days..

http://s12.yousendit.com/d.aspx?id=1TJ3KTAWGHHCW01F41CDGGRJMD  <<File

Thanks all for any input. I want to do right by these folks at the Cathedral for allowing me this wonderful opportunity.
Many thanks.

Teddy

[prof_peabody]

Frequencies vary from song to song we are talking about amplitude, harmonics, duration and sympathetic resonation of the mic diaphram or somthing near the mic diaphram can cause distortion on a recording when the song before it was fine. One of both of the DPA mics could be bad.
Also here is a big problem with gear most distortion figures are based on 1k not multiple frequencies. So you might have recorded something that was at the right frequencey and amplitude to set off this distortion. If you look at a real distortion graph even based on 1k with broad band fft you will see spikes at other frequencies these are harmonics of 1k. These are also high in distortion so you can have one frequencey that has a harmonic on one of these "spikes" that will set off the distortion. The only way to tell for sure is to get a source transducer that has very little distortion and do a frequencey sweep analysis most likely the distortion area will appear as a bump in frequencey response in the sweep if the source transducer is flat.

IN OTHER WORDS TURN DOWN THE GAIN OR see if you have a bad mic capsule or a bad cable or a bad input on a preamp or recorder.

Interesting +T.  I am curious if you have any comments about the "brick wall" like drop in frequencies at 48,000 Hz (like a brick wall filter... not like brick-walling).  I know it's the Nyquist limit, but the sudden drop has raised some questions in my mind.

[CHURCH-AUDIO]

I have seen this before with files I think its something to do with the PCM clock not being sync'ed properly beween the device that recorded it and the sound card that is analyzing it. On good analyzing programs like WINAUDIOMLS allow you to change this sync so that it is correct. One way to see if this is the problem is to do a propogational delay test of the wave form and see where the actual clock rate is. I also think its somthing to do with the fact that the file being analyized is 48k the program doing the analyzing is set to 96khz or 192khz thats my theory and I am sticking too it

Frequencies vary from song to song we are talking about amplitude, harmonics, duration and sympathetic resonation of the mic diaphram or somthing near the mic diaphram can cause distortion on a recording when the song before it was fine. One of both of the DPA mics could be bad.
Also here is a big problem with gear most distortion figures are based on 1k not multiple frequencies. So you might have recorded something that was at the right frequencey and amplitude to set off this distortion. If you look at a real distortion graph even based on 1k with broad band fft you will see spikes at other frequencies these are harmonics of 1k. These are also high in distortion so you can have one frequencey that has a harmonic on one of these "spikes" that will set off the distortion. The only way to tell for sure is to get a source transducer that has very little distortion and do a frequencey sweep analysis most likely the distortion area will appear as a bump in frequencey response in the sweep if the source transducer is flat.

IN OTHER WORDS TURN DOWN THE GAIN OR see if you have a bad mic capsule or a bad cable or a bad input on a preamp or recorder.

Interesting +T.  I am curious if you have any comments about the "brick wall" like drop in frequencies at 48,000 Hz (like a brick wall filter... not like brick-walling).  I know it's the Nyquist limit, but the sudden drop has raised some questions in my mind.

[prof_peabody]

I have seen this before with files I think its something to do with the PCM clock not being sync'ed properly beween the device that recorded it and the sound card that is analyzing it. On good analyzing programs like WINAUDIOMLS allow you to change this sync so that it is correct. One way to see if this is the problem is to do a propogational delay test of the wave form and see where the actual clock rate is. I also think its somthing to do with the fact that the file being analyized is 48k the program doing the analyzing is set to 96khz or 192khz thats my theory and I am sticking too it

Frequencies vary from song to song we are talking about amplitude, harmonics, duration and sympathetic resonation of the mic diaphram or somthing near the mic diaphram can cause distortion on a recording when the song before it was fine. One of both of the DPA mics could be bad.
Also here is a big problem with gear most distortion figures are based on 1k not multiple frequencies. So you might have recorded something that was at the right frequencey and amplitude to set off this distortion. If you look at a real distortion graph even based on 1k with broad band fft you will see spikes at other frequencies these are harmonics of 1k. These are also high in distortion so you can have one frequencey that has a harmonic on one of these "spikes" that will set off the distortion. The only way to tell for sure is to get a source transducer that has very little distortion and do a frequencey sweep analysis most likely the distortion area will appear as a bump in frequencey response in the sweep if the source transducer is flat.

IN OTHER WORDS TURN DOWN THE GAIN OR see if you have a bad mic capsule or a bad cable or a bad input on a preamp or recorder.

Interesting +T.  I am curious if you have any comments about the "brick wall" like drop in frequencies at 48,000 Hz (like a brick wall filter... not like brick-walling).  I know it's the Nyquist limit, but the sudden drop has raised some questions in my mind.

The plot thickens... interesting.