Project settings....

R

RAMI

Guest
I just stumbled upon this, and now I'm wondering if I should change it 32bit Floating Point or not. As you can see, my recording setting is at 24 bit.

---------- Update ----------

I don't know why my pics always upload so huge. Hopefully you guys can see the pop-up box in the screenshot.
 

Attachments

  • Untitled.png
    Untitled.png
    344.2 KB · Views: 33
While I'm at it, I might as well ask about this. In the bottom part of the box, there's "Media format for Apply effects, Glue, Freeze", and also "Default media format for project/region". Those are all also at 24bit. Should I be changing all that shit to 32bit FP?
 

Attachments

  • Untitled.png
    Untitled.png
    353.6 KB · Views: 33
Yes, i would be inclined to say if you change the bit rate of anything you should probably follow suit and do it for everything.

I had an interesting instance of a reverb not having a matching bit rate and wreaked havoc on my renders. The playback was fine and i had it mixed with no clipping or anything, but for some reason when i rendered it a wav it was peaking and clipping all over the place. Once i matched all the bit rates to the same thing problem went away.

Also highfive on the skittles looking UI. I wish i spent more time sprucing that stuff up.
 
ShanPeyton said:
Yes, i would be inclined to say if you change the bit rate of anything you should probably follow suit and do it for everything.

I had an interesting instance of a reverb not having a matching bit rate and wreaked havoc on my renders. The playback was fine and i had it mixed with no clipping or anything, but for some reason when i rendered it a wav it was peaking and clipping all over the place. Once i matched all the bit rates to the same thing problem went away.
Click to expand...
Thanx Shan.

But actually, my question is more along the lines of "Should I even be changing the bit rate at all?"
 
RAMI said:
Thanx Shan.

But actually, my question is more along the lines of "Should I even be changing the bit rate at all?"
Click to expand...

Oh. Sorry man. I think i read it wrong as i so often am guilty of. I think that would be a personal thing. Maybe try recording some basic tracks in one project at 24 and the same thing again in a 32F project and see if you hear enough of a difference to justify the increased hard disk space?
 
I record in 24 bit because that is native for my soundcard. Any more is pointless. For freezing and rendering, I do 32 bit floating point because information introduced by plugins can be lost if you go less. REAPER uses an internal 64 bit engine, so I used to bounce to 64, but this is generally thought to be unnecessary, and 32 bit is sufficient. My source on this was Justin. Here is the old thread from the REAPER forum -> WAV bit depth settings - Cockos Confederated Forums
 
ShanPeyton said:
Also highfive on the skittles looking UI. I wish i spent more time sprucing that stuff up.
Click to expand...
You got to try it. I swear drums sound better in red. :D

kindafishy said:
I record in 24 bit because that is native for my soundcard. Any more is pointless. For freezing and rendering, I do 32 bit floating point because information introduced by plugins can be lost if you go less. REAPER uses an internal 64 bit engine, so I used to bounce to 64, but this is generally thought to be unnecessary, and 32 bit is sufficient. My source on this was Justin. Here is the old thread from the REAPER forum -> WAV bit depth settings - Cockos Confederated Forums
Click to expand...
Thanx for that link, buddy, exactly what I was looking for. Looks like I'll keep recording in 24bit, but change the Glue thing to 32bit.
 
My reaper project settings are as yours are: 24 bit PCM. I've never experienced a problem that would cause me to wonder whether they should be changed. So I just leave them as they are.
 
Had to do some research myself. I knew the accuracy would be higher (from a mathematical point), but 24 is pretty damn accurate. From a little research it seems that it would give you a bit more headroom. But I would also figure since it is floating point, the computer would have to do more processing since nothing is fixed (floating point being a variable). Renders would probably take longer as well since it has to take the FP and turn it into a fixed.

Let us know if you find your computer working a little harder (CPU load) under 32FP than 24. I would think it would, and your renders would take longer as well.
 
For the recording setting, you'd be limited by your interface, wouldn't it? My interface goes up to 24 bit, so that would be the limit for recording. Maybe it will take those 24 bits and append 8 zeroes if I select 32 bit, which wouldn't bring any advantage, but would take up more hard drive space. I'm just guessing here.
 
DM60 said:
Had to do some research myself. I knew the accuracy would be higher (from a mathematical point),
Click to expand...

I don't know much about bit depth as it applies to audio, but I do know the CS here. Head room would be a much bigger deal than accuracy. (In fact, at loud volumes, you would LOSE some accuracy)

Floating point means a "decimal" number rather than the whole number of an integer. So a 24-bit integer gives you any value between 0 and (2^24)-1
With a 32 bit FP, 8 bits are spent specifying the exponent (e); the other 24 are the number itself (the significand s). So your range is a number between 0 and (2^24)-1 And that number is multiplied by a power of 2 between -127 and +127. i.e. s=[0..16777215] and e=[-127..127]. With your final value being s*(2^e).

This means that for very large numbers, there will be rounding errors. 16777216 can be represented, but 16777217 can't.
 
You must log in or register to reply here.
Back
Top