Sample rate question

[natenajar]

I am on Mac osx 10.7.4 and evaluating reaper. I am using a sound devices usb pre2 for my sound card. It is a class compliant device that works off of core audio. It will respond to sample rate requests from a host program independent of the setting in "audio midi setup."

So I setup the device in reaper under "preferences/device" and I set sample rate to 48000 and the buffer to 128. I made some test recordings and they sound nice.

Then I found the "project info" option under the file menu. It said 44100 for project sample rate. But there is a checkbox next to it unchecked.

So I am confused. At what sample rate did I record my test recordings? And what is the proper way to set the sample rate? And how do I get to see or know for sure what sample rate the program is operating at at that time?

Thank you.

[mrelwood]

Originally Posted by natenajar

So I am confused. At what sample rate did I record my test recordings?

The recorded files were made in 48K that you set up for the sound card. But the Reaper project is running at 44.1K, so it is resampling all audio streams going in and out, on the fly. That is a bit tasking for the CPU when the project gets bigger. The sample rate of the project is independent of the sample rate of the recorded files or the sound card. It is of course usually best to use the same rate in all of them, but you do have to set them up separately. A 44.1K project can record and play 48K files as you are now doing, it is just using more CPU for converting in the fly.

The checkbox next to sample rate settings in both project settings and Reaper preferences is confusing, I give you that. I can't even tell you the exact reason for them. What I do know is that my sample rate setups have always gone as expected when the boxes are all checked.

Also do note that there is rarely reason to use 48KHz unless you're making music for a movie. The less conversions the better, so 44.1KHz should be the default option.

[compasspnt]

Originally Posted by mrelwood

Also do note that there is rarely reason to use 48KHz unless you're making music for a movie. The less conversions the better, so 44.1KHz should be the default option.

Except that many recordists believe that 48 is higher quality, and they prefer to record at that higher quality.

And many at 96k of course.

These final mixed files are often sent to mastering, where SRC happens at a highly professional level.

[Europanaut]

Originally Posted by compasspnt

Except that many recordists believe that 48 is higher quality, and they prefer to record at that higher quality.

And many at 96k of course.

These final mixed files are often sent to mastering, where SRC happens at a highly professional level.

Very true.

Welcome compasspnt! It's great to have you here. The experience that you bring to this forum is valuable and quite appreciated.

[compasspnt]

Thanks, looking forward.

[mrelwood]

Originally Posted by compasspnt

Except that many recordists believe that 48 is higher quality, and they prefer to record at that higher quality.

I don't see that one being a valid reason if it is based on belief, not facts. The difference is too little, and is negated by the extra conversion stage as well as other surprises it might cause due to us being humans and making mistakes.

And many at 96k of course.

88.2K makes sense if one really want's to extend the spectrum.

These final mixed files are often sent to mastering, where SRC happens at a highly professional level.

I don't think that will be in the original poster's agenda any time soon. And no matter how professional the SRC is, it's still a conversion, an unnecessary step in my mind.

[ned]

Originally Posted by natenajar

At what sample rate did I record my test recordings? And what is the proper way to set the sample rate? And how do I get to see or know for sure what sample rate the program is operating at at that time?

The project's sample rate is continuously displayed in the menu bar in os x (beside the 'window' menu). You'll be able to tell when a media item is using a different sample rate from the project by the blue info icon at the top left of the item in the time line (when using the same sample rate it's just grey) mousing over that icon will give you a popup telling you from what sample rate the audio is being resampled.



As mrelwood pointed out, best practice is to have both the program's preference and project setting's sample rate the same. If you already recorded @ 48kHz just change the project setting to that sample rate and everything will be cool.

Originally Posted by mrelwood

The checkbox next to sample rate settings in both project settings and Reaper preferences is confusing, I give you that. I can't even tell you the exact reason for them. What I do know is that my sample rate setups have always gone as expected when the boxes are all checked.

The 'Request sample rate' check box in the program preferences makes sure the sample rate you enter gets used by the system audio and not overridden by the audio/midi setup program (core audio).

The check box beside the sample rate in the project settings locks the project's grid to the chosen sample rate (i.e. selections, cursor/playhead snap to the nearest sample within your chosen grid increments). Depending on how you want your grid to behave this is optional.

[natenajar]

this is great, thanks Ned

[compasspnt]

Originally Posted by mrelwood

I don't see that one being a valid reason if it is based on belief, not facts. The difference is too little, and is negated by the extra conversion stage as well as other surprises it might cause due to us being humans and making mistakes.

Most professional recordings are done at 48 or higher, many these days at 96.

Almost all professional mastering is done with playback @ file rate into an analogue chain, then captured at the final 44.1.

This conversion happens, almost every time. So why not record at a higher, better quality rate? Most pro's do.

BTW, I often mix from 48 or 96 into DSD @ 5.6MHz / 1 bit, that plays into the analogue chain, just like a tape machine.

88.2K makes sense if one really want's to extend the spectrum.

Yes, and if that makes sense, then 96 makes even more sense. I guess it depends upon how much sense you want to make.

Of course, the higher you go, the more strain on the system, fewer plug-ins, etc.

Here's some good reading about the SR subject...

http://www.sendspace.com/pro/dl/gkksh5

Best regards!

[mrelwood]

I can't comment on how the professionals work in your country, but in mine most work is still done at 44.1K. But that is not the point. The point is that there is no reason to advice a beginner to work at 48KHz when the audio quality will not make a difference, and can cause various problems at several stages of the workflow.

Besides, adults do not usually hear anywhere near 20KHz.

[serr]

Originally Posted by mrelwood

I can't comment on how the professionals work in your country, but in mine most work is still done at 44.1K. But that is not the point. The point is that there is no reason to advice a beginner to work at 48KHz when the audio quality will not make a difference, and can cause various problems at several stages of the workflow.

Besides, adults do not usually hear anywhere near 20KHz.

Going to the higher sample rates (88.2k, 96k) has absolutely nothing to do with increasing the usable frequency range beyond human hearing when working with audio. This is a strawman argument! The higher sample rates simply move the sampling noise FAR away from the audio band. It's RIGHT next to it at the lower rates. (Filtered yes, but not perfectly.) You will readily hear this in a studio setting and high end (audiophile) consumer equipment.

Of course, using 24 bit over 16 bit still makes the biggest difference and you'll hear this on just about any system excluding the cheap portable devices. And at any sample rate (even 44.1k).

44.1k is fine for 'beginners' as mentioned but use 24 bit word length.

Better:
Record at 88.2k or 96k and keep it there from tracking to mixing to mastering. Then deliver the final 24/88.2 or 24/96 master to your consumers.
You can do that here in the 21st century. There's no reason to reduce the quality to deliver to consumers anymore. Then make an additional reduced quality master at lower res for the portable device listeners and anyone else that still insists on listening to their old CD player. And for further destruction by mp3 compression. Make this version louder to accommodate the reduced dynamic range of these formats and devices.

You want to make your music accessible to everything and everyone but you don't want to throw the baby out with the bathwater and reduce your master to the lowest common denominator.