I agree with the 24 bit and 44.1 or 48 kHz rates for recording because the higher processing cost of the double and quadruple depths don't have benefits I can detect for the music I'm mixing.
 
I do think there is value in getting a good interface. There is an audible difference between 24 bit 44.1 on a $100 USD interface and on a $1000 USD interface. Interfaces that work at higher rates fairly well tend to improve their lower rates as they push the higher ones for quality.
 
I'm guessing here...
When you have an interface that does 192, but set it to 48, I think you just get every 4th sample. I believe the interface itself is still sampling at 192. For my RME, that means that the RME internal compession, EQ and Reverb will work on the 24 bit 192 signal, even when the device is set to work at 48. So, when I run sound with the device and send to the mains directly from the interface, I'm working at 192 even when I'm recording at 48.
 
It's also possible for me to record all the tracks plus the mains I'm sending to the PA, giving me a 48 kHz stereo track of the live mix done at 192.
 
Please, jump in and correct me if I'm wrong.

While I also agree that better interfaces are simply better, I believe that has more to do with the pre-amps and the quality of the A/D & D/A converters than the mere fact that a given interface happens to have 192k available and you are sampling at a lower rate.  Some research out there I recall seeing concluded that some vendors have 192k sample rates on one or more interface models, but as more of a marketing ploy suggesting a better interface than it is - where even with the 192k sample rate available they have the best quality produced at either 48k or 96k, for example.
 
If I recall correctly, the OP had gotten his answers earlier in the thread, and at this point folks are engaging in the endless sample rate debate with no hope of convincing anybody of anything (as it has usually gone every other time I have ever seen this discussion in any prior thread).
 
:)
 
Bob Bone
 
 
Bob Bone
 

gswitz
 
I do think there is value in getting a good interface. There is an audible difference between 24 bit 44.1 on a $100 USD interface and on a $1000 USD interface. 

Agreed.  
 
A typical audio interface has a noise-floor somewhere between -103dB and -107dB.
The higher cost units have a noise-floor somewhere between -114dB and -117dB.
Say (very roughly) that's a 6dB difference in noise-floor.
Not overly obvious on a single track of audio.
But... multiply that 6dB difference across 24 or 48 tracks of audio... and it's a substantial difference.
Akin to a veil of noise being lifted from the material...

Here's a link if you want to brush up on over-sampling:
http://en.wikipedia.org/wiki/Oversampling
 

gswitz
I'm guessing here...
When you have an interface that does 192, but set it to 48, I think you just get every 4th sample. I believe the interface itself is still sampling at 192.

 
Slightly more complicated than that:
 
Almost all mainstream modern converters oversample to varying degrees, but when downsampling to the output sample rate it needs to be filtered digitally before dropping the extra samples (dropping the extra samples is known as decimation). So you have two filters, an analog antialiasing filter and a digital decimation filter. This allows for a much gentler analog filter, a gentler filter being much easier to implement in the analog domain than a steep one, and a steep digital filter.
 
And in some cases, particularly with DAC's, if you read the specs you will see that a 24bit converter often will state that it's using "Xx oversampled 20bit" conversion or something like that. But this doesn't mean it's a really only a 20bit converter, as it's actually quite possible, though not intuitively easy to understand, to trade bit depth for sample rate (at a cost of some added noise).

The theoretical dynamic range of 24Bit audio is 144dB.
The highest quality A/D D/A units yield about 120dB dynamic range.
The theoretical dynamic rage of 20Bit audio is 120dB.
 
Essentially, we're working with 20Bit resolution

Jim Roseberry
The theoretical dynamic range of 24Bit audio is 144dB.
The highest quality A/D D/A units yield about 120dB dynamic range.

There are actually a few converters that can go higher than that (from a noise perspective), such as this:
 
http://www.stagetec.com/en/audio-technology-products/standalone-converter.html
 
Typically things like this run a number of individual converters in parallel with varying degrees of attenuation and then output the one that is near the top of its range without clipping and digitally compensate for the attenuation.

Jim Roseberry

bobguitkillerleft
 
I'm very sorry if this offends[or worse doesn't],it's just,as an "experiment" I changed everything in "Preffs.-Audio Data" to the highest-64[even on my lowly Dell xps L502x i7 Sandy Bridge mobile cpu 2630qm,and [of course!]the Highest Sample Rate on my 09 SONAR V-Studio 100,loaded an old project,remixed with the NEW settings and yes,I hardly noticed a difference.
 

That's not overly surprising...
The source material was captured at lower sample-rate.
(ie: Sample-rate converting from 44.1k to 96k won't add what wasn't originally captured.)
The only advantage you'd have is the higher resolution when mixing/processing.
 
Working at higher sample-rates is a judgement call balancing quality vs. performance.
IME, The sound-stage is more 3D and detailed when working at higher sample-rates (not simply more high end).
That said, no record has been (or ever will be) purchased solely because the sample-rate at which it was recorded.
Far too many other variables affect the final quality of your work.  Most importantly, the song... (performance/arrangement/acoustics/mic position/etc)
 

Awesome reply Jim Roseberry,Thank You!
 
Those last 3 sentences,but especially["That said...,and extra specifically for me IMNSHO[at times,it seems :(] "Most Importantly,the song... (performance/arrangement/acoustics/mic position/etc)-the BIG BIG part many of us sometimes[all the times myself?!] really really LACK[to be brutally opinionated,and perhaps stupidly honest about "That Stuff"]
 
Was/IS super-to be intelligently reinforced that a "Great Song" especially made WAY better by smart,concise arrangement[to me possibly a lot more important/lacking with my own material that Iv'e actually "whipped together/recorded" almost insanely quick,[and IS ALL my soundcloud-Stuff!]once I became even barely able to record[all done with the $13.33 each[3 pack Behringer XM1800S mics]-$39 something USD in the U.S,[?]AFAIK-$79 AUD in Oz,that I paid,about 4 years before I'd ever typed/used any computer,let alone knowing any more about digital recording,other than-"Pro Tools is what people use" on "those computer things"let alone EVEN knowing the term DAW,hence why I first bought in Feb 2011 the now discontinued/superceeded Tascam 2488[quickly sold Thank Goodness-if you're a killer songsmith/producer,I'm sure they're GREAT,but],and all ordered from the P4 IBM an incredibly generous friend gave me at first,late 2010.
 
I found out about SONAR from The CDR of 8.5LE inside the AT2020/2021 pack[still these mics are embarrassingly unused yet-info no one really needs to know AT all,though I feel ATM to "expose"....why is that Dr.???]
 
All my available time has been spent learning X1 studio/Producer,X2a and "futzing"-Hence Jim your"The source material was captured at lower sample-rate"makes a BIG deal AND "why I can't seem to better it ANYMORE"!!!!221 re-mixed versions of the absolute first take "nutshell" soooLOL now-1 XM on the 4x12,another to sing with,at the same time" with first takes[of absolutely everything],as computers,with virtual mixing desks,of what I was so thrilled "allowed to see and use"[not touch-except when my guitar sound was overly AWFUL-and engineer,didn't seem to notice!]back in 86-92 LOL,so getting something-anything newer on soundcloud-[first it was a new intergration of X1d?] was paramount.
 
Re-mixing,re-masterring,learning a few more SONAR possibilities,rather than really writing,re-recording,[with better mics than the dynamic Behringers!]is NEXT I hope![though they're alright next to a battered[like fish? sorry,it's late here] SM-58 in reality[?].
THANK YOU
Sincerely
Bob

Well, I'll be humble enough to stay out of the discussion, as there are areas where I know my stuff, but admittedly this isn't one of them.
 
To simply answer the OP, I usually just go with 44.1K/24 bit.
 
If I think I'll be doing anything for film, I'll change that to 48K.
 
Otherwise, most things seem to default to 44.1k which, for me is fine almost all of the time.
 
As state, there are a number of other variables that will affect your recordings & mixes far more outside of that anyway, so that's where I'm learning as much as I can for now.
 
I will admit that I'm surprised that it was mentioned 3 times above to use 48k for DVDs.
 
I had always heard 96k/24 bit, and hence why some were using those instead of CDs for their final burns - at least as physical stored & archived master copies.
 
Perhaps things have changed lately.

 
 

Jim Roseberry
 
That's not overly surprising...
The source material was captured at lower sample-rate.
(ie: Sample-rate converting from 44.1k to 96k won't add what wasn't originally captured.)
The only advantage you'd have is the higher resolution when mixing/processing.
 
Working at higher sample-rates is a judgement call balancing quality vs. performance.
IME, The sound-stage is more 3D and detailed when working at higher sample-rates (not simply more high end).
That said, no record has been (or ever will be) purchased solely because the sample-rate at which it was recorded.
Far too many other variables affect the final quality of your work.  Most importantly, the song... (performance/arrangement/acoustics/mic position/etc)

Thank You again,your answer [After Thinking It Over And Over,the last few days...especially] cleared up a [Huge-seemingly now!... a whple other"mess of Questions" I had,but were "believe it or not!- too woosed out"badly"[of course] to try and get another answer,from the actually Very [IMNSHO at that time] "ESTEEMED Bob O." on GS!
 
Instead of...............this particulary NSHIS especially much now....  hmmmH  Bob here!on Cakewalk's  forum........perhaps........for?
 
 Cheers  [also "Very ESTEEMED!-IMNS..]  "Jim Roseberry!" 
 Sincerely
Bob Sattler