Digital Noise in Raw Files: Pre-Master

[mercurydime]

good day, all

i have taken in a some files to master of bagpipes, brass, and drums that were recorded live at a local university. these files have an interesting occurrence of what seems to be a digital anomaly:

in the lowest dynamic passages, notably at tail-out, but also at points where a phrase or passage begins, there are low-level garbled, dropout-like glitches that occur. they are evident in the wave forms. this is obviously a result of either the recording or bouncing/processing aspects in mixdown; but i've never encountered anything quite like it...even in the loads of restoration work i've done.

any ideas from this sage group as to what might cause this?

i just may have to get the engineer to revisit these files since i'd like not to have to de-noise these if at all possible.

any insight is appreciated, and i thank you in advance!

as ever,

[jdg]

was it recorded to a minidisc?

sounds like a low bit rate issue.

[mercurydime]

no, it's quite surprising knowing the engineer, the equipment and facilities. i've certainly seen my share of things done on the cheap, but this is not one of them!

the work was done at a state university...not sure what the recording platform is (just looked again at their facilities/equipment list), but they are using macs, maybe PT or logic?

at any rate(the files have come to me as 24/88.2), and they do sound good outside of a few niggling but fixable concerns...

best,

[Elk]

I've seen this result with some reverb plug-ins that seem to "bounce" with whatever sounds are present in the tail - but this clearly is not what is going on here.

[mercurydime]

indeed. recorded in a very large hall...23 pieces between the brass, pipes, and drums...no 'verb needed, really.

but the incidences do occur right at those very most dynamic of places...surprised the engineer did not pick up on it. i will acknowledge he has been 'goaded' somewhat by the principals involved to get finished. he missed these anomalies, though...

may just have to suggest a re-visit to these before trying to master. by the time i give these files more gain, these boo-boos will be quite noticeable.

best,

[24-96 Mastering]

Do you have a sound file of a tail? Otherwise it's pretty much guesswork.

[Adam Dempsey]

Originally Posted by mercurydime

i have taken in a some files to master of bagpipes, brass, and drums that were recorded live at a local university. these files have an interesting occurrence of what seems to be a digital anomaly:

Does it sound like accumulated truncation distortion? Could possibly be from some earlier stage prior to 24/88 conversion..

Just for the fun of it, reverse or play a file backwards.. the bagpipes will sound almost the same .

[PatrikT]

Cheap gating?

Best Regards
Patrik

[mercurydime]

it is my understanding that the material was recorded at 24/88.

i will post a couple clips at mid-day today that has the noise in it...

thanks for all the input so far...

this is my third time working with bagpipes. just my lot in life, i guess

best,

[mercurydime]

problem solved. it would appear that i was experiencing some type of buffer under-run in my system. not yet sure of the cause, but i suspect it is software related. a slight buffer adjustment and a computer re-start have corrected the glitching. first time i had experienced anything quite like this in my audio work. i was not stressing the system, but it could be one of the plugs i was using to spec out some possible (analog) eq changes i would make later as i listened to the material for the first time...

i perused the material in toto last night with nary a problem.

the hall in which this material was recorded has a sweet ambience with a nice tail-out, but it's a little 'thick' 130hz-200hz range. nicely, though, that bump there does help the tuba and euphonium 'speak' more in the space. the bass drum will need a little taming. i have a track with a trap kit, and that kick drum is eating most of the rest of the material alive on that particular tune; it's presence so pronounced that it's also skewed the stereo width and field depth...

and, i am pleased to report the receipt and deployment of a manley vari-mu and an api 5500 yesterday. sweet, they are!

thanks to all again for the insight.

as ever,

[mercurydime]

as a quick update the problem was discovered to be the Elephant Limiter running at 4x oversampling...

best,

[trmchenry]

Originally Posted by mercurydime

as a quick update the problem was discovered to be the Elephant Limiter running at 4x oversampling...

best,

I don't know your history, so don't let this post come off as condescending.

I personally find it sounds best to only use the oversampling in Elephant if the file is 44.1/48k

If your file is 88.2k or above you should turn the oversampling off.

Even then, you should probably resample your lower sample rate rate files to a higher rate before you begin your processing. If you don't have the budget for something super sweet like Weiss Saracon, you can try the free version of r8brain from Voxengo Sample rate converter tool software - r8brain - Voxengo

[mercurydime]

Originally Posted by trmchenry

I don't know your history, so don't let this post come off as condescending.

I personally find it sounds best to only use the oversampling in Elephant if the file is 44.1/48k

If your file is 88.2k or above you should turn the oversampling off.

Even then, you should probably resample your lower sample rate rate files to a higher rate before you begin your processing. If you don't have the budget for something super sweet like Weiss Saracon, you can try the free version of r8brain from Voxengo Sample rate converter tool software - r8brain - Voxengo

thanks, and you have not been condescending at all. i appreciate it.

best,

[dave-G]

Originally Posted by mercurydime

this is my third time working with bagpipes. just my lot in life, i guess

Someone should make sure you're kept from belts, shoelaces and sharp objects for the next few weeks.. Maybe seek out a support group.

Originally Posted by mercurydime

as a quick update the problem was discovered to be the Elephant Limiter running at 4x oversampling..

Wow. Yet more reinforcement of the "don't turn your back on digital" adage.

-dave

[mercurydime]

Originally Posted by dave-G

Someone should make sure you're kept from belts, shoelaces and sharp objects for the next few weeks.. Maybe seek out a support group.


Wow. Yet more reinforcement of the "don't turn your back on digital" adage.

-dave