Question about project sample rate

[Jose7822]

Originally Posted by Stringtheorist

Awesome. Thanks a lot.

Just so I have this right.... I should record the tracks at 24 bit, then reset my project to 32/64 before arranging and processing tracks, and then reset again to 16/24 before mastering?

Alright, let me explain. The recording bit depth is set in both your audio interface's control panel and in the recording software's (Reaper in this case) preferences menu. This should be set to 24 bits.

By default, Reaper (and most DAWs) process audio internally at 32 bit or 64 bit. Set this to the highest bit depth your DAW supports, again under its preferences menu. This is beneficial because any DSP (Digital Signal Processing) done to audio produces rounding errors which theoretically *could* accumulate to the point of being potentially heard. Hence, the higher the bit depth, the lower the noise floor. Having said that, 32 bit fp has a low enough noise floor to give you very clean and professional recordings. Using 64 bit fp is just an extra layer of protection, if you will (probably not the right choice of words, sorry). But either one is more than enough, really.

The above two parameters should be set once and left alone. The only time you will need to change the bit depth constantly is during export/rendering of the project, and this is usually done after you've finished mixing and mastering your music. In this case, you will want to lower the bit depth to whatever the target format needs to be (usually 16 bit). The exception to this is when sending your individual tracks, or a stereo mix/stems, for mixing and/or mastering at another place. In this case, you will want to keep the files at the highest bit depth possible. Most of the time this is 32 bit, so you would be actually exporting files at 32 bit in this situation. In short, the rendering/export bit depth depends on where the audio file(s) will end up.

As far as the sample rate goes, like I said, stick to the media format's sample rate (44.1 KHz for CD and 48 KHz for film). In your case though, you say you are forced to use 48 KHz, so there's really nothing you can do about that. You definitely don't want to record audio with a sample rate mismatch as that will create artifacts in the recording. It'll make it sound like this:

http://www.youtube.com/watch?v=1CdZQmhiaMg

However, the most common side effects of sample rate mismatch are music that sounds like it went through a chipmunk plugin or music that sounds too low and plays slow (think slow motion clip from your favorite movie). So definitely use 48 KHz for recording, if that's all you have available. Preferably, as other have said, you'd want to recording using a dedicated audio interface. But that's up to you, of course.

Regarding the OS bitness, that is a totally different animal and basically boils down to the amount of RAM the OS can utilize. A 32 bit OS can only access up to 4GB of RAM while a 64 bit OS can access much more than this, up to 192 GB of RAM (depending on the OS version). Using a 64 bit OS is obviously beneficial for those using big sample libraries, like EWQL, and who have more than 4GB of RAM installed in their systems. That basically sums it up, no need to get more complicated than that :-).

Anyway, hope this helps answer your question. Take care!

[Stringtheorist]

Originally Posted by lejmer

Depends on how much ram you have installed. If you have more than 4 GB installed, you will need the 64 bit version to access all of it. If you have 4 GB or less, 32 bit will do.

The reason why they different in the amount of RAM they can access is because of how computers are made. I could get into detail but I don't think anyone would benefit from that here

OK, thanks.

[Stringtheorist]

Originally Posted by Jose7822

Alright, let me explain. The recording bit depth is set in both your audio interface's control panel and in the recording software's (Reaper in this case) preferences menu. This should be set to 24 bits.

By default, Reaper (and most DAWs) process audio internally at 32 bit or 64 bit. Set this to the highest bit depth your DAW supports, again under its preferences menu. This is beneficial because any DSP (Digital Signal Processing) done to audio produces rounding errors which theoretically *could* accumulate to the point of being potentially heard. Hence, the higher the bit depth, the lower the noise floor. Having said that, 32 bit fp has a low enough noise floor to give you very clean and professional recordings. Using 64 bit fp is just an extra layer of protection, if you will (probably not the right choice of words, sorry). But either one is more than enough, really.

The above two parameters should be set once and left alone. The only time you will need to change the bit depth constantly is during export/rendering of the project, and this is usually done after you've finished mixing and mastering your music. In this case, you will want to lower the bit depth to whatever the target format needs to be (usually 16 bit). The exception to this is when sending your individual tracks, or a stereo mix/stems, for mixing and/or mastering at another place. In this case, you will want to keep the files at the highest bit depth possible. Most of the time this is 32 bit, so you would be actually exporting files at 32 bit in this situation. In short, the rendering/export bit depth depends on where the audio file(s) will end up.

As far as the sample rate goes, like I said, stick to the media format's sample rate (44.1 KHz for CD and 48 KHz for film). In your case though, you say you are forced to use 48 KHz, so there's really nothing you can do about that. You definitely don't want to record audio with a sample rate mismatch as that will create artifacts in the recording. It'll make it sound like this:

http://www.youtube.com/watch?v=1CdZQmhiaMg

However, the most common side effects of sample rate mismatch are music that sounds like it went through a chipmunk plugin or music that sounds too low and plays slow (think slow motion clip from your favorite movie). So definitely use 48 KHz for recording, if that's all you have available. Preferably, as other have said, you'd want to recording using a dedicated audio interface. But that's up to you, of course.

Regarding the OS bitness, that is a totally different animal and basically boils down to the amount of RAM the OS can utilize. A 32 bit OS can only access up to 4GB of RAM while a 64 bit OS can access much more than this, up to 192 GB of RAM (depending on the OS version). Using a 64 bit OS is obviously beneficial for those using big sample libraries, like EWQL, and who have more than 4GB of RAM installed in their systems. That basically sums it up, no need to get more complicated than that :-).

Anyway, hope this helps answer your question. Take care!

Great... That's a lot clearer now.

Just one thing... In your first statement you said set the bit depth in the preferences to 24-bit. Why is this (when the project bit depth setting is 32/64)? And do you know where I find this setting in Reaper (as I can't seem to locate it)? The only bit depth parameter I can find is in the project settings.

[peter5992]

Originally Posted by MPDmike

I don't understand this debate. So, if a composer produces a really great score at 44.1, then it will be rejected and they will be asked to submit a new version at 48?

LOL - don't be silly.

If your session runs at a different rate from your samples, your computer / DAW sofware / AD converter has to convert the samples or there is going to be a pitch change. That conversion could result in a (very slight) quality loss, best to keep everything at the same rate.

That said I've used EW samples in many sessions that run at 48khz and never noticed anything strange. But I haven't done any serious A/B comparisons. One thing to make sure is that the deliverable is conform to specs (usually 48khz/24 for film / video).

[admin]

The biggest red flag in this discussion is a soundcard that's stuck at 48k, that suggests it is ill-conceived and other problems may be around the corner.

[Jose7822]

I know, it gets a bit (no pun intended) confusing with all these bit depth settings here and there, but this is what it boils down to. You want to *record* at the highest bit depth provided by your audio interface. Currently, this is 24 bit for ANY audio interface sold today. The truth is that there is no audio interface in existence today that can give you a full 24 bit of resolution due to self noise from internal components in the analog domain. So, in most cases, you're really getting ~17 to 18 bits of resolution. The best DAC goes up to ~21 bits, but it costs + $4,000 USD. Anyway, that's another subject altogether. Basically, 24 bit is the maximum resolution one can record at, so that's why you want to set the record bit depth accordingly. Setting it higher than this won't give you any benefits AT ALL. Only bigger files that will consume your HDD at a faster rate. I don't have Reaper, so I can't really point you to the exact parameter where the recording bit depth is set. But, the project settings you mention sounds about right.

So that's the recording bit depth. Now, you also have the processing bit depth of your DAW's audio engine. You really don't have to worry about this because the minimum resolution used by DAWs these days is 32 bit floating point and, like I said before, that is plenty. Don't sweat this one cause it really isn't one to lose sleep over. In the right hands, more important than any of these bit depth settings, are the hardware used to capture a performance (i.e. instrument quality, mics, preamps, converters, etc) and, even more important, the performance itself and the techniques used during recording (i.e. mic placement, room sound, etc). Those are the things you should lose sleep over because that's what will give you the best sounding recordings possible. These bit depth settings are really secondary in comparison, especially if you think about the quality of equipment we have available today. People were making awesome recordings 40 years ago, with much less. Spend more time learning your equipment (i.e. learning your monitors, mic techniques, how to use EQs, Compressors, etc). That trumps using even the best equipment any day. Believe it or not, you CAN make bad recordings with the best equipment in the world when you don't know what you're doing.

Not sure all of this answers your last question, but I do hope it helps.

Take care!

[gstitt]

The biggest red flag in this discussion is a soundcard that's stuck at 48k, that suggests it is ill-conceived and other problems may be around the corner.

+1

[gstitt]

If your session runs at a different rate from your samples, your computer / DAW sofware / AD converter has to convert the samples or there is going to be a pitch change. That conversion could result in a (very slight) quality loss, best to keep everything at the same rate.

Here is a good SRC comparion:

http://src.infinitewave.ca/

Some of the DAW programs are not real good. If you have Izotope, it looks like it does a really good job. Fortunately, even some of the worst add noise in the -120 db range, which shouldn't be a concern unless you are doing the conversion numerous times. So, it is safe to assume nobody will notice a sample rate conversion on the final exported mix.