I realized that the first track I recorded in Reaper was captured at 32bit floating point but my audio interface supports 24bit. Now, the sampling rate matched up (I had both set to 44.1), but will having Reaper set at a higher bitrate negatively impact the sound quality or how the audio was captured?
32bit floating pt.
- Thread starter AndrewC
- Start date
Gidge
Lapdance Test Dummy
no..........
That's good to know. Should I even bother changing it to 24bit then? Also, when I render the song what bitrate should I select-- 24 or 32?
It seems that even when I change the settings to 24bit, Reaper continues to display "32bit floating" in the upper right hand corner.
Finally, what do you recommend in terms of the "allow projects to override device sample rate" option under audio device settings? On or off?
It seems that even when I change the settings to 24bit, Reaper continues to display "32bit floating" in the upper right hand corner.
Finally, what do you recommend in terms of the "allow projects to override device sample rate" option under audio device settings? On or off?
Farview
Well-known member
The program does all of it's calculations at 32 bit float. There is no such thing as 32 bit converters, so the audio that you are recording is 24 bit.
If you have the session set to 32 bit, any time you do some destructive processing, it will be written as a 32 bit file. If you don't do any destructive processing, it will remain 24 bit audio with 8 extra zeros at the end.
If you have the session set to 32 bit, any time you do some destructive processing, it will be written as a 32 bit file. If you don't do any destructive processing, it will remain 24 bit audio with 8 extra zeros at the end.
Farview
Well-known member
The only reason to use 32bit fp is if you are doing a lot of destructive editing on the original files. Otherwise, you are just wasting hard drive space.
I don't do destructive edits, so I don't bother with 32 fp. No, more bits is not necessarily better when you are talking floating point.
I let the session over-ride the hardware, otherwise you can get yourself into trouble with the session playing back at the wrong speed and accidently recording things with the session not playing back correctly. It can turn into a mess pretty quickly.
I don't do destructive edits, so I don't bother with 32 fp. No, more bits is not necessarily better when you are talking floating point.
I let the session over-ride the hardware, otherwise you can get yourself into trouble with the session playing back at the wrong speed and accidently recording things with the session not playing back correctly. It can turn into a mess pretty quickly.
Farview
Well-known member
The processing is always 32 bit.
When you choose 24 bit, you are writing 24 bit files to your hard drive. The software is still processing at 32 bit, and you can't change that.
Farview
Well-known member
Unless you are working directly on the files, you are only processing them as they are playing in the mix. There is no point in recording them to hard disc at 32 bit because there is only 24 bits of information coming from the AD converter.
The only real reasons why you wouldn't render the mix at 32 bit is because different programs have a slightly different way of encoding 32 bit files, so they aren't universal. If you send it to a mastering engineer that uses analog equipment (and you should), there will only be 24 bits coming out of the converters on its way to the analog goodness because there is no such thing as a 32 bit converter.
Besides, the bit depth is a very small detail. The difference between 24 bit and 32 bit is not something anyone will notice. It is more of a theoretical advantage than an actual one.
But internally in the DAW wouldn't files you edit with plugins be more accurate with 32bit floating pt. (or even 64) than 24 and the subsequent rendered song represent that editing better at 32 or 64? Maybe, as you said, even if this holds true the advantage is negligible.
Due to financial restraints and the fact I change/update my songs sometimes, I'll simply be using T-racks 3 to "master" (I know it's not ideal) my tracks. More often than not they will be heard as mp3s online anyhow. Since T-racks allows either 32 or 64bit floating point files to be processed, that won't be an issue for me. And I can always go back and export songs at 24bit if I do someday send them out for professional treatment.
Due to financial restraints and the fact I change/update my songs sometimes, I'll simply be using T-racks 3 to "master" (I know it's not ideal) my tracks. More often than not they will be heard as mp3s online anyhow. Since T-racks allows either 32 or 64bit floating point files to be processed, that won't be an issue for me. And I can always go back and export songs at 24bit if I do someday send them out for professional treatment.
Farview
Well-known member
I don't really know. It would probably depend on the vst instrument that is creating the sound.
If these instruments are sample based, they would have the bit depth that they were sampled with.
I know that with my sample collection for Drumagog, I sampled them at 24 bit. Drumagog turns them into a 32 bit file, but it is still only 24 bits of information.
If these instruments are sample based, they would have the bit depth that they were sampled with.
I know that with my sample collection for Drumagog, I sampled them at 24 bit. Drumagog turns them into a 32 bit file, but it is still only 24 bits of information.
Farview
Well-known member
It won't make any difference unless you render the midi as audio.
Even then, it won't make a difference.