The sample rate which alot of people tend to disregard as secondary... is defined by a computer in "dot-to-dot" format. (how many samples it takes per second.) It's not a perfect flowing wave like we all have seen on waveform graphs and such.
Before I blabbler on I will post a link to a good read on this subject. Good for lamen but probably a lil too basic for someone partially educated in the field.

Anyone really should read this before thinking samples rates only matter if you want to hear over 20kHz, there are also well documented attributes to acoustical environments that aren't actually audible in the "human hearing" range. If you are working with mostly samples and adding verbs and such then 44.1 is probably more then adequate. BUT if you are recording an orchestra in the Vienna State Opera you can't possibly think that 44.1 is professional or would get you asked back to record again... do ya?

At least scroll down to the graphs of a square wave and how well defined it becomes as you raise the sample rate... Future Proof Recording Explained

1bit (5.6448MHz) recording... It took me by surprise too when I first read about.

Everything I said in here is meant to be mentaly stimulating and hopefuly make people think. It is NOT aimed at or meant to hurt anyones feelings or say anyone in particular is wrong or right. I am only stating these things as I know them to be true. I could be wrong, I have been before and most likely will be again. I just don't think I am about this.

I don't understand. Why would your mastering project not be at the same sample rate as the imported stereo files?

In short, to avoid upsampling/downsampling. The idea is to keep the original sampling rate of the wave file while at the same time being able to use the plugins that would sound good at higher sampling rates. Just wondering if this would work or not since Sonar is able to have wave files of different sampling rates. What do you guys think?

Well a sample rate played back at the wrong sample rate would cause it to speed up or slow down in comparision to the original. You can, IF the plugin has the ability to process at double the rate which is really a different thing then the recorded sample rate, run them at double... The reason some plugins have this ability is beause of the "truer" wave reproduction/simulation at higher sample rates. They do sound alot better when you click that button eh??

Well a sample rate played back at the wrong sample rate would cause it to speed up or slow down in comparision to the original.

Dang! I forgot about that. So much for my idea . I'm confusing sampling rate with bit depth, duh...I guess I have to rest my head for a little bit and not think too much and doing too much at the same time. Thanks for responding dude. Peace!

ORIGINAL: Jose7822

I don't understand. Why would your mastering project not be at the same sample rate as the imported stereo files?

In short, to avoid upsampling/downsampling. The idea is to keep the original sampling rate of the wave file while at the same time being able to use the plugins that would sound good at higher sampling rates. Just wondering if this would work or not since Sonar is able to have wave files of different sampling rates. What do you guys think?

bitlfipper/Jose7822, I think we have two conversations going on here regarding the "44.1, 48, or higher???" question...

One conversation debates the value of recording frequencies that you can't hear. This is where hardware quality, jitter, and all that other stuff comes in.

The other conversation debates whether there's a "mathematical value" in utilizing sample rates beyond which a typical human can make a distinction -- that is, a value beyond just "can I hear a difference"?

Jose7822's comment above concerns the latter conversation. Dave asserts that the sonitus EQ sounds better when running at 88.2, which is consistent with reasons given elsewhere for upsampling for mastering. If I have a non-linear process that produces harmonics, and those harmonics happen to be around the nyquist frequency, they'll alias back into the audible range. Running the same plug at a higher sample rate spreads the harmonics across the extra bandwidth provided by the higher sample rate. Upsampling in and of itself does not provide any audible enhancement, but provides extra bandwidth for the process and resulting harmonic error, assuming you're using a process that produces harmonic error and doesn't effectively filter it to avoid aliasing. To go a step further: what about upsampling plugins? What's been proposed is that you pay an error penalty when upsampling followed by downsampling, and that error adds up over processing chains.

To address Jose7822's question: do you mean, e.g., import a 44.1 stereo file into a 88.2 or 96 project?

The answer to that is: who do you want to do your SRC? The import will upsample to the project rate -- it has to. I don't think SONAR supports mulitple sample rates in the same project, because there's no such thing. If I have a 88.2 project and I'm outputing audio at 88.2, then if I include a 44.1 stream in that output the 44.1 stream will play back twice as fast as other audio in the stream. At some point, all audio needs to be piped through at exactly the same rate. If SONAR doesn't explicitly do an SRC at import (which I think it does), then it must do it in realtime (unnecessary expense) -- but an SRC is done it some point.

If you're going to let SONAR do SRC at import using whatever it uses (does anybody know?), or you could use something like R8Brain offline, which is supposed to be one of the top dogs currently...

This link posted a few months ago: http://src.infinitewave.ca

Original thread: http://forum.cakewalk.com/tm.asp?m=971550&mpage=1&key=SRC󭑊

ORIGINAL: RnRmaChine
The reason some plugins have this ability is beause of the "truer" wave reproduction/simulation at higher sample rates. They do sound alot better when you click that button eh??

Someone on gearslutz pointed out that "too much SRC can be a bad thing" -- http://www.gearslutz.com/board/mastering-forum/119029-up-sample-not.html

There could be some legitimacy to that... in which case you want to run everything natively at the higher sample rate (regardless of whether or not it's recorded at the higher sample rate).

You know, sometimes I think maybe we're just thinking too hard about this stuff... Maybe it's better to just turn the knobs until it sounds good, and if it doesn't sound good try a different plugin or whatever.

I personally have found no validity in up sampling. It's too late... Record at the higher rate then work at a more considerable rate... not the other way around. In other words record at 96/88.2 and work at the halves respectively.

You know, sometimes I think maybe we're just thinking too hard about this stuff... Maybe it's better to just turn the knobs until it sounds good, and if it doesn't sound good try a different plugin or whatever.

I think you're right. I was just trying to find a way to not have to upsample and at the same time save some resources during recording but I guess this is currently impossible. I just had a brain fart with all the different sampling rates mastering project deal.

I personally have found no validity in up sampling. It's too late... Record at the higher rate then work at a more considerable rate... not the other way around. In other words record at 96/88.2 and work at the halves respectively.

When you say, "record at 96/88.2 and work at the halves respectively", are you talking about in the mastering process or during mixing? i believe this would contradict what Keith just said about the filters wouldn't it?

ORIGINAL: keith

...

The answer to that is: who do you want to do your SRC? The import will upsample to the project rate -- it has to. I don't think SONAR supports mulitple sample rates in the same project, because there's no such thing. If I have a 88.2 project and I'm outputing audio at 88.2, then if I include a 44.1 stream in that output the 44.1 stream will play back twice as fast as other audio in the stream. At some point, all audio needs to be piped through at exactly the same rate. If SONAR doesn't explicitly do an SRC at import (which I think it does), then it must do it in realtime (unnecessary expense) -- but an SRC is done it some point.

...

Actually, this type functionality could be implemented the way that kmixer handles the SRC. You would then need to use drivers that passes kmixer (no ASIO, WDM/KS or hardware accelerated DirectSound then). Would it be 'effective' this way ... and would the SRC quality be good then?

On another forum, I saw writing stating that Vista does SRC for every source it outputs (it was not mentioned if this happens with all driver models Vista supports) even if the system samplerate is set equal with the source samplerate (there were an exampl that even 44.1kHz is resampled to 44.1kHz) so, if this is the truth, then could it be better to leave all additional SRC undone.


Junski

ORIGINAL: Junski
Actually, this type functionality could be implemented the way that kmixer handles the SRC.

The problem is that not all SRC implementations are created equal...

On another forum, I saw writing stating that Vista does SRC for every source it outputs (it was not mentioned if this happens with all driver models Vista supports)

Don't know for sure, but if so could be part of the DRM scheme...