44.1/16 or higher?

N

Networker

Well-known member
My recorder will let me record (wav) in 44.1, 48, or 96KHz, all at 16 or 24 bit. I seem to remember that 48KHz is the preferred rate for video use, but I'm not sure about the bit depth. I would think 24 bit should be the preferred setting. Or am I missing something?
 
44/16 is for CD quality audio. 48/24 is for video in general. Higher, such as 96/24, are best for highest quality sound recording.
 
Just remember: any time you raise the sample rate or bit depth on a recorder, the recording times on any recorder media decrease, sometimes dramatically. 48/16 is a very good compromise for film.
 
24 bits gives you a wider dynamic range with as Doc said less recording time as a compromise. Using 16 bits for dialogue will work fine, when recording sound effects usually 24/96+ is preferred... depending on the sounds characteristics.
 
I should have known there wouldn't be one single, easy, answer :)
Thanks guys, I will now change the rate as needed. Just got my Octava, Mixpre-D, and some cables, and I'm starting to experiment. Having a blast too. Just finished "The book of Audacity" by Carla Schroder, and have just started "Audio Postproduction for Film and Video" by Jay Rose. Should keep me busy for a little while.

I will probably be back with more questions as I get further into this...
 
I do 16-bit for dialog, 24-bit for sound effects recording. I record in 48Khz. I've recorded in 96kHz before to do tests, but I don't hear any appreciable difference in the recording (nor do I see it after analyzing deeper). 96kHz is overkill. It sounds different mostly because the sound noise floor sounds different running with different oscillator settings in the ADC, but not to say one had less noise. In fact, the SNR ratio looked to be the same in my measurements.
 
44.1/16 is the CD standard, and so is a common music rate. 44.1 isn't used in video.
48/16 is the video standard.
48/24 gives you more resolution that can have the effect of giving you more headroom (at the same quality as 16 bit) or just higher resolution. You can get in to long arguments on which your actually getting and it's really a how you deal with it issue. IF you record at the same levels you would with 16 then you get better resolution. If you record lower you can record at the same resolution as 16 and have more headroom. Either way there generally is no good argument for not recording at 24 bit. You are not using significantly more card space than 16 bit and you are getting some good benefits.

96/24 is not really that useful other than for some FX recording and, on much higher end systems than your probably using, for music.

It may be the "new thing" in FX. It is used by some I know, but mostly for short sounds like gunshots etc. The main reason in FX is for slowing things down. It's argued, quite validly, that unless your system is is well into the five digits your just recording high quality noise. For the FX work that really doesn't matter. What we are looking for is a very smooth sound when slowing down an FX so if it's mostly filler noise that really doesn't matter much. 96K will cut very significantly into your record time.
However most FX are still recorded at 48/24 for the simple reason that that is what they will be used at. If you record at 96 every file will need to be converted when you bring it into your ProTools timeline. So FX that are not going to be heavily manipulated are almost never recorded at 96k. An exception might be if your recording to sell FX. Then the marketing hype of 96k is probably worth the file space.
 
Some also recommended this to me:

"ProTools for Video, Film and Multimedia "

While it is ProTools based (and I actually use 3 different DAWs depending on who I am working with), the information in the book is GREAT. Especially on setting up your mixing space and getting to calibrated properly for mixing dialogue.
 
Noiz2 said:
44.1/16 is the CD standard, and so is a common music rate. 44.1 isn't used in video.
48/16 is the video standard.
48/24 gives you more resolution that can have the effect of giving you more headroom (at the same quality as 16 bit) or just higher resolution. You can get in to long arguments on which your actually getting and it's really a how you deal with it issue. IF you record at the same levels you would with 16 then you get better resolution. If you record lower you can record at the same resolution as 16 and have more headroom. Either way there generally is no good argument for not recording at 24 bit. You are not using significantly more card space than 16 bit and you are getting some good benefits.

96/24 is not really that useful other than for some FX recording and, on much higher end systems than your probably using, for music.

It may be the "new thing" in FX. It is used by some I know, but mostly for short sounds like gunshots etc. The main reason in FX is for slowing things down. It's argued, quite validly, that unless your system is is well into the five digits your just recording high quality noise. For the FX work that really doesn't matter. What we are looking for is a very smooth sound when slowing down an FX so if it's mostly filler noise that really doesn't matter much. 96K will cut very significantly into your record time.
However most FX are still recorded at 48/24 for the simple reason that that is what they will be used at. If you record at 96 every file will need to be converted when you bring it into your ProTools timeline. So FX that are not going to be heavily manipulated are almost never recorded at 96k. An exception might be if your recording to sell FX. Then the marketing hype of 96k is probably worth the file space.
Click to expand...

I couldn't have said it any better. :) I tend to use my 24-bit for headroom. Of course, I also record dual channel with one set to -12dB as well. Just in case we get an awesome take and don't want to ADR later if I don't have too.
 
Rick Burnett said:
I do 16-bit for dialog, 24-bit for sound effects recording. I record in 48Khz. I've recorded in 96kHz before to do tests, but I don't hear any appreciable difference in the recording (nor do I see it after analyzing deeper). 96kHz is overkill. It sounds different mostly because the sound noise floor sounds different running with different oscillator settings in the ADC, but not to say one had less noise. In fact, the SNR ratio looked to be the same in my measurements.
Click to expand...

We did a bunch of tests with a NagraV and some Neumann mics and came to pretty much the same conclusions. WE could kind of sense 96k but it was always a guess. 24 bit VS 16 bit very clear difference. This is all FX so YMMD. However slowing a 96 recording by 50% VS a 48/24 by 50% showed 96 to produce a smoother sound, as expected. So the things we know we are going to slow down, creaks and groans etc. we tend to record at 96, everything else at 48.
 
Noiz2 said:
We did a bunch of tests with a NagraV and some Neumann mics and came to pretty much the same conclusions. WE could kind of sense 96k but it was always a guess. 24 bit VS 16 bit very clear difference. This is all FX so YMMD. However slowing a 96 recording by 50% VS a 48/24 by 50% showed 96 to produce a smoother sound, as expected. So the things we know we are going to slow down, creaks and groans etc. we tend to record at 96, everything else at 48.
Click to expand...

I bought an Apogee Rosetta 200 and a Lucid Genx6 (generates very stable word clock) and compared internal files and even files to a Motu 896HD recording both 48Khz and 96kHz. I used a Blue Kiwi mic for the recordings. Did both vocal and acoustic guitar. While I could sense a difference, it wasn't a quality difference to me either. In fact, I had a set of people do a blind comparison to tell me which was better. None could between the Apogee, Motu, 48Khz and 96kHz. I used acoustic guitar because the transients I have been told make a huge difference.

I also tested different mic pre-amps in the mix as well. $500 to $1500 range. (JUST using the amplification, not the compression, filtering) and there too, no one could tell the difference. Anything mic based was dominated by room tone. And even in a vocal booth, there was still enough movement of air by the performer that it dominated any noise from the ADC or preamp.

As for slowing sounds down, good point. I am going to have to play with that some more. I generally hadn't slowed FX down because of the degradation and 'pixelation' of the sound, if you will. Not sure why it didn't click to just record those at a higher sampling rate! Haha. Thanks!
 
The other tools that work well for slowing things down, other than tape which is fantastic, are Soundminer (not cheap but a very nice db and FX tool), an E4 hardware sampler (very smooth), and if you have the $$$ an old Synclavier (they sample at 100k). The later is complicated to get to work since you also need an old Mac to get a decent interface. The E4 is pretty cheap on eBay and is "platform independent"! If you do a lot of FX you want soundminer anyway so...
 
You must log in or register to reply here.
Back
Top