Rendered File Bit Depth in Commit Preferences

[Thenewexhibit]

Under the "Commit" section in preferences, the "Rendered File Bit Depth" looks like the default is "Always Use 32-bit". The other option is "Follow Session Settings".

Is it necessary to always use 32-bit for commit? I generally track in 24-bit, so I'm wondering, is that option for committing tracks even really that necessary and is it worth the extra space it'll take up vs 24-bit?

[tope d]

Best to set your render bit depth to the setting of your session. Best to select the session's bit depth.

[Darryl Ramm]

Originally Posted by Thenewexhibit

Under the "Commit" section in preferences, the "Rendered File Bit Depth" looks like the default is "Always Use 32-bit". The other option is "Follow Session Settings".

Is it necessary to always use 32-bit for commit? I generally track in 24-bit, so I'm wondering, is that option for committing tracks even really that necessary and is it worth the extra space it'll take up vs 24-bit?

Nothing is *necessary* here. You only need to use 32 bit float files if you want the additional dynamic range/clipping avoidance. By defaulting to 32-bit float Pro Tools is simply ensuring doing a commit will not cause signal clipping. And given how commit works there might be corner cases where a signal might clip in a commit where that might not be obvious to a user say watching track meters. I think this is a kinda neat thing that Avid did. The size of the commit files is not likely to be an issue for most users. I would leave it at the default.

Edit: Maybe worth pointing out that the internal data path in Pro Tools between plugins and plugins and the rest of the system is 32-bit float. When you say commit a track or commit up to a plugin on a track, 32-bit float preserves all the data that was possibly there as before the commit, any other commit file format has potential to lose data, esp. clip at 0dBFS. The default choice Pro Tools is making makes sense, that choice is matching the internal data format used between plugins, it is not anything directly to do with the session file formats.


But I would turn the question back around... why not just use 32-bit float for all files?

[XJENSEN]

Originally Posted by tope d

Best to set your render bit depth to the setting of your session. Best to select the session's bit depth.

I don't agree with this statement. Much safer to have it commit to 32 bit, for the exact reason stated above by Darryl (avoid clipped commited files) so the commit files also remain usable for export to other applications.

Also, mind you there are still some plugins in Pro Tools that are incompatible with 32 bit float and that will clip internally, irrespective of 32 bit setting (e.g. the signal generator plugin can't deal with levels over 0DBFS and will clip at 0, even in a 32 bit float session).
This is not strictly related to the OP's question, though, but I thought it worth mentioning.

Also see:
https://duc.avid.com/showthread.php?t=411839
https://duc.avid.com/showthread.php?t=407525

[Franklyn]

I am using always 32bit. Because when I commit or do Audiosuite render with very quiet Clips the sound stays 100% the same.

The Mixer of the DAW also works with 32bit/64bit. And that makes sense, because when your signal goes through 20 plug ins and different gain stages 32 is your best friend. :)


I always render may master files/Mixes in 32bit. so the headroom doesn't t matter any more.

[Darryl Ramm]

Ha I also wish signal gen would do > 0dbFS, if only to help with quick demos/explanations of clipping/gain staging in Pro Tools.

[Thenewexhibit]

Originally Posted by Darryl Ramm

Nothing is *necessary* here. You only need to use 32 bit float files if you want the additional dynamic range/clipping avoidance. By defaulting to 32-bit float Pro Tools is simply ensuring doing a commit will not cause signal clipping. And given how commit works there might be corner cases where a signal might clip in a commit where that might not be obvious to a user say watching track meters. I think this is a kinda neat thing that Avid did. The size of the commit files is not likely to be an issue for most users. I would leave it at the default.

Edit: Maybe worth pointing out that the internal data path in Pro Tools between plugins and plugins and the rest of the system is 32-bit float. When you say commit a track or commit up to a plugin on a track, 32-bit float preserves all the data that was possibly there as before the commit, any other commit file format has potential to lose data, esp. clip at 0dBFS. The default choice Pro Tools is making makes sense, that choice is matching the internal data format used between plugins, it is not anything directly to do with the session file formats.


But I would turn the question back around... why not just use 32-bit float for all files?

Hey, you bring up some great points here that I didn’t think of! If things could clip, it’s like a safety net. The processing for plugins is something I haven’t thought of either! I never knew this existed, so I thought I’d ask about it, but based on the stuff you brought up, that makes total sense, so I’ll keep it at the default. Thanks for chiming in!