Sample Rate Question

[secutanudu]

Hello,

Fairly new here, I have a question about sample rates.

Currently, I have:

• Focusrite Saffire Pro 26 Interface (goes up to 96k)
• PreSonus DigiMax LT Digital Mic Preamp (goes up to 48k)

I'm really not too familiar with sample rates, word clocks, how they work, and how they all affect the final product.



A few questions:

1. Will I notice any difference if I just do everything in 48k?
2. If I get an external world clock, will I be able to use the above hardware in 96k?
3. Will I gain anything else by getting an external clock?

Thanks.

[albee1952]

You may get differing opinions on some of this, but here's my 2 cents; Unless you are recording orchestral stuff for major release, 48K will be fine(I've been doing everything at 48K for several years). Many lightpipe devices only operate up to 48K(remember how many hit records in the eighties and nineties were done on 16 bit.44.1K ADAT machines). Unless both units in your 2-box setup have SMUX capability, 48K is as high as your gear will allow(my rig CAN go to 192K, but I still work at 48K) Also, higher sample rates will chew up a lot more computer power, and double the drive space.

External wordclock won't change your capability. Will it change your sound? Maybe. If you can borrow an Apogee Big Ben to try, try it and see. I clocked my old rig with a Lucid GenX 192 and really didn't notice a sonic improvement.

[Jim Morris]

The only reason, in my opinion, to record at a sample rate lower than 96k is if your hardware doesn't allow it. That could be for several reasons: 1) It just doesn't support higher rates ... or 2) You don't have the processing power. 3) You are concerned about file size.
In my experience, the difference in fidelity at the higher rates is quite noticeable. Even though we most likely will deliver our product at 44.1k, the mixing algorythms are far more mathmatically precise at the higher rates.
I advise you to try your own test. Try an entire recording at 96k. Mix it. Then make a session copy at 44.1k. Print the mix without changing anything from the 96 version (leave the plugin settings unchanged). Export both mixes at 44.1k. Compare them. There are definite differences. What you are hearing is the difference in accuracy of every algorythm used in the mixing process. If you mix using any analog outboard gear or summing, the differences are much larger.
Jim Morris

[TimothyJohn]

If you're main audience is consuming CD's, and you have the disk space, use 88K. If you are mixing for video, or if everything is going to be converted to mp3's of some sort, use 96k, again if you have the disk space. If not, 44 for CD's and 48 for video. My 2 cents.

[wwittman]

There is no relationship between the session sample rate and the desired final product.
Or to put it another way, there is no 'advantage' to working at 88.2 versus 96k other than disk space.
The idea that it's mathematically 'related' to 44.1 is simply not how sample rate conversion works.

To the original question: no one can say what YOU will notice.
I can say only what I notice. So I work at 96k unless it's impossible.

[Bob Olhsson]

Twenty years ago there was a hardware sample rate converter that only did 88.2 to 44.1 and sounded better than anything else. Ten years later it had been surpassed. Considering that almost nobody buys CDs anymore, video is probably the most important market quality-wise. That means working at least at 48k if not 96. The reason for working at 96 is that a lot of converters and plug-ins simply sound better up there. Obviously there are exceptions that sound fine at 48k but the cost is trivial at this point. I've also heard plug-ins sound better at 48 than 88.2 but not 96.

[Park Seward]

[QUOTE=Jim Morris;2343680] the mixing algorythms are far more mathmatically precise at the higher rates.


Hmm. "More mathematically precise"? How can you be "more precise"? Could you show how?

[Park Seward]

Here is a great YouTube video about sampling rates. Great for beginners or anyone else who wants to understand digital audio better:

https://www.youtube.com/watch?v=cIQ9IXSUzuM

[Park Seward]

Originally Posted by Jim Morris

I advise you to try your own test. Try an entire recording at 96k. Mix it. Then make a session copy at 44.1k. Print the mix without changing anything from the 96 version (leave the plugin settings unchanged). Export both mixes at 44.1k. Compare them. There are definite differences.

I'd be interested in exactly what those differences are. Level, frequency response, noise, distortion, anti-aliasing artifacts or what?

Is it the SRC between 96 and 44.1 that is causing the difference? Why would a sound, let's say a 10k tone at 96k, would sound different after being converted to 44.1 than recording directly that 10k tone at 44.1?

[wwittman]

Originally Posted by Bob Olhsson

...The reason for working at 96 is that a lot of converters and plug-ins simply sound better up there...

Yes.
that's still the bottom line... despite all the theory in the world (or on the internets)