Recording: 24bit vs. 32bit float

[WombatStudio.Org]

Originally Posted by mikeyman

why do you record at 24 instead of 32floating

This question came up in another thread, but as some of us know, this should have a dedicated thread to keep the other on topic.

If you're doing tracking (recording live sources), you're going through a 24bit converter, 32bit float will record and reproduce the 24bit clipped signal.

BUT, if you're recording internally, check out this video:
http://www.youtube.com/watch?v=Qt-EJhDDHUI

Also, if you're clipping that badly - the problem lies elsewhere really.

I used to think there was no question, to record at 32bf - but then I did some research and concluded that recording people through my converter that I will gain nothing since the converter is only able to use 24bits of headroom.

What say you?

[Emcha_audio]

Even 5hough the converters are at 24 bit, there's still the advantages of not having to worry about gain staging within the plugins if I'm not mistaken.

[Bob Olhsson]

Everything gets converted to floating point in native DAWs anyway so recording 32 float gains nothing and I understand even costs you one bit. You just need to remember to switch Pro Tools to 32 float before applying audiosuite.

[JFreak]

There is one benefit using 32bit float.

Usually, VI's output way too hot signals and may be close to clipping 24bit fixed. Doesn't matter whether it's VI or real audio, but if it clips it clips. Now 32bit float is different in one aspect. You can fix that clipping later.

Try this:

#1: open up a session that does not clip. want to be sure? set up a limiter that doesn't allow over -6dBFS out. bounce it out for a test.
#2: create a 32bit float session and import/convert that bounced test file into that session
#3: drive the faders up and/or use whatever plugins you need to get it 100% surely clipping and sounding bad. then bounce it out as 32bit float.
#4: create another 32bit float session and add that clipping bounce into the new test session
#5: use clip gain to bring levels down and prepare to be amazed.

[YYR123]

Originally Posted by JFreak

#5: use clip gain to bring levels down and prepare to be amazed.

Next level sh|t !!!!






Sorry I couldn't resist

[Bob Olhsson]

A VI is not coming from a 24 bit a to d converter so it indeed is best written at 32 float.

[Top Jimmy]

Russ stated in his video that you get unlimited headroom. I wouldn't call 1500+ dB of idiot-proofness "unlimited headroom" since no matter where in the floating point word the signal resides, it never has more that 144 dB dynamic range.

Bob, you are correct in that a 32-bit floating point word only encodes 23 bits of signal. The most significant bit is always implied since it can never be anything other than 1. There's no need to encode it if it never changes.

[mesaone]

Originally Posted by Top Jimmy

Russ stated in his video that you get unlimited headroom. I wouldn't call 1500+ dB of idiot-proofness "unlimited headroom" since no matter where in the floating point word the signal resides, it never has more that 144 dB dynamic range.

Wouldn't 32 bit floating point offer 192 dB of dynamic range? Or did I misunderstand what you are saying?

[Raoul23]

Just reading this thread and I'm on TDM rig I record at 24bit but isn't TDM 48bit ??? Maybe I'm confused so some one can put me right :) what I wanna know is can it be clipped or not ???

[Craig F]

Originally Posted by WombatStudio.Org

I used to think there was no question, to record at 32bf - but then I did some research and concluded that recording people through my converter that I will gain nothing since the converter is only able to use 24bits of headroom.

You are correct recording 32 bit from a 24 bit AtoD converter does not gain any headroom and cost you disk space