32 Bit Float

[boogiemotel]

Does this setting in session setup mean you are recording 32bit audio ? or is just the mixer changing to 32bit float ?
A standard PT converter cannot record at 32 bit so, what does this setting mean to us ?

[jasonthurley]

The engine (audio engine) is 32 bit Floating point... so If your session is set to 32 bit you will have a clean smooth bit transaction, once you change the session to a different bit rate (such as 24bit 48K or CD standard 16bit 44.1K) the software has to convert the # of bits (can use dithering or just truncated (which means you lose bits or information in the conversion) to achieve the setting you want....

Most people record in 24bit 48K and then convert to 16bit 44.1 for CD printing... some people like to use 32bit or even higher... that starts to get into personal preference to me....

So 32bit Floating point is the engine... and you can also set the bit rate of the session (audio files your using and what rate you will record at If tracking in Pro Tools).. make sense?

[VRW]

Me once read in a studio magazine that it could be useful to use a 32bit-session for mixing a formerly 24bit recorded project to gain more headroom in Pro Tools. Is that right? And what´s actually the reason Pro Tools 10 sets 48Khz as a standard over 44Khz? If I´m going to finally put the stuff on a 16bit 44Khz audio CD, does it make sense to change samplerate (which could cause problems sometimes) or is it the better way to stay with 24bit 44KHz?
Please, gimme some info!
Thanks in advance,

cheers, VRW

[jasonthurley]

You just confused the [bleep][bleep][bleep][bleep] out of me.... There are some standards like a CD is 16 bit 44.1K... Most people I know of choose to record at 24 bit 48K and mix in that bit depth and sample rate (as stated above the audio engine is 32bit so some like to use 32 bit sessions so they dont loose bits converting from the 32bit floating point of the engine to the session bit rate. The higher quality (more accurate sample of the analog sound or vibration) is achieved by a higher bit rate and sample rate. Then a lot of people bounce it to a stereo track (so If you recorded at 24bit 48K you bounce at that same rate or you can convert the bounce to CD standard 16bit 44.1 or use the MP3 converter, etc).

How you choose to setup your Pro Tools sessions bit rate and sample rate is a personal preference.

Yes a higher bit rate allows a larger dynamic range (so you have more headroom)... so you also have to watch out as this dynamic range changes when you convert to less bits (like going from 24bit 48K to 16bit 44.1) and If not converted correctly can lead to distortion.

Not sure If that was even close to what you are trying to ask... but I tried to explain...

[CME]

Setting the bit-depth of the sessions sets both the bit-depth of the recorded audio and the bit-depth of the audio pathways through the mixer.

There isn't any point in recording 32-bit float files. There aren't any common interfaces with 32-bit converters. The technology is out there but only select few are using it.

However 32-bit paths can be advantageous. It's not really necessary if you mind you gain-staging, but I like knowing I have virtually unlimited headroom. So what I, and I believe others do, is keep the session at 24-bit float until tracking is done. Once all the tracks are recorded I switch over to 32-bit float.

I really wish avid would give us two separate options. One for audio file format and one for the mix engine. Maybe in pt11. :)

[Top Jimmy]

Originally Posted by VRW

Me once read in a studio magazine that it could be useful to use a 32bit-session for mixing a formerly 24bit recorded project to gain more headroom in Pro Tools. Is that right?

No. The mix engine in LE & etc has always been 32 bit floating point. There is absolutely no gain in potential quality. Besides, what's the point of converting the session when 32 bit floating point files only encode 24 bits per sample anyway.

And what´s actually the reason Pro Tools 10 sets 48Khz as a standard over 44Khz? If I´m going to finally put the stuff on a 16bit 44Khz audio CD, does it make sense to change samplerate (which could cause problems sometimes) or is it the better way to stay with 24bit 44KHz?

Pro Tools 10 doesn't favor one sample rate over another. As for the sample rate you choose, when only dealing with standard rates (44.1, 48) there's no gain in recording at one only to change to the other at the end.



Buttdialed by me & Tapatalk

[CME]

And fwiw pro tools 9 and later has a 64-bit float mix engine.

Bit depth is directly related to dynamic range. I don't want to get too technical but you want your mix-engine (and audio paths) to have enough headroom to do the math needed without clipping.

Float and fixed also makes a difference. Fixed has a defined finite range. Float is sort of like a sliding scale. It is not as finite but still had a limited range.

Sample rate is directly related to high end frequency response. A 44.1khz sample rate goes just above 20khz in frequency response. Which is higher than most people can hear. However we may be able to perceive even higher frequencies. So higher sampling rates allow this. They also allow for steeper transients. How much difference it makes is hard to say, and is greatly dependent on your listening chain/environement.

All that said. I tend to record at the sample rate I'll end up with. 44.1k for music. And as mentioned record at 24-bit and mix at 32-bit.

Oh and don't forget to look up dithering. Why you need it, and how and when to use it.

[jasonthurley]

For clarity...

Pro Tools 9/10 (without Native card): Summing engine=64bit float

Pro Tools 9/10 HD Native (with card): Summing engine=64bit float

Pro Tools HD/TDM 9/10: Summing engine=48bit fixed

However... "Processing" in the engine is at 32 bit floating point.

[sunburst79]

To further muddy the waters a lot of people will record at the sample rate of the finished product.

DVD audio = 48K 24 bit

CD audio = 44K 16 bit. Most record, edit and mix at 24 bit then dither the final product to 16 bit. Recording, editing and mixing at 16bit severely limits your headroom. Going from memory 24bit offers more headroom than 16bit. So you want to take advantage of that.

[Craig F]

16 bit is 96 dB of range
24 bit is 144 dB of range
the best analog can do to date is 138 db of range

but sense most sheep listen to mp3s or the like, does it mater?