If you read the thread you will see what we found was at the higher sampling rates the VST's that especially showed the more obvious changes put out a smoother less top end type of sound. Even after rendering at 96K and down sampling to 44.1 K.
 
Sounds coming in through the soundcard did not behave the same way as VST's did operating/rendering at either 44.1K or 96K
 
Here is a fresh link to the original test I did with the two files.
 
https://www.hightail.com/...ZUcwN3RZWlR3NUlLSk5Vag
 
And the second test with the K5000 synth creating similar additive sounds but coming in via the soundcard.
 
https://www.hightail.com/download/ZUcwN3RZWlQyWGR3SGNUQw
 
 
 
 
 
 
 

In short, NO! no one can distinguish between the 2 and if they do, they got lucky. 
 
If you want to bog down your DAW and waste Disk space, then YES! It's worth it!!! 
 
Stick with 88.2KhZ or 44.1kHz
 
Now, if a plugin has an oversample possibilities, then this is different, as oversampling in the limiting stage, (if your limiter has it) is a good idea at times. But this is oranges and kangaroos or grapes and apples or somehting.
 
Cj

some say higher rates sound better, others don't hear a difference.....it's not a massive sound difference I guess.
but higher rates make lower latency.
so when bouncing and especially exporting a song, maybe better to set high rate as the song will get more latency anyway later on (mediaplayers/cd players converters can add latency) and might come to a latency level where it really starts to sound flat.
so to minimise that, bounce /export at high rate....just for latency sake.
this is just what I read on the web.....so don't shoot please you forummers looking for a target (you remind me of my mother in law).
 
GL
 
 

There are different modes remember you can be thinking about this.  One mode bringing signals in through your sound card and create sessions at either 44.1K and 96K. And perhaps rendering the 96K version down to 44.1K.  I am not so sure there would be so little difference even in doing that.   Many plugins and things will probably sound better at the higher sampling rate too.
 
The other mode is using your virtual instruments and playing and sequencing them.  But when it comes to playing and performing on a virtual instrument there most certainly is a definite improvement working at 96K with some of them.  Still after doing even more experiments with this I still prefer the 96K sound.  I have become a bit hooked on it now.  My next setup will most definitely be 96K and 24 bit all the way. Apple's Mac Pro will handle it all with ease though.  The only way to do it.  But you need a DAW that runs on a Mac.
 
I can see the advantages of working all at 96K and 24 bit right from the get go and just convert the final master down to 44.1K and 16 bit at the last minute.  It always has been and still is the best and most professional way to produce quality audio.  The lower latency is also going to reveal itself in the way it feels under the hands of a great player.
 
We can work at playback resolutions now more so all the way through the audio production process and get great results. But once you get used to working at higher resolutions I think you will agree it sounds better and feels different.  We used to do it all with high end tape machines and vinyl was the playback medium at the end of that. (vinyl is inferior as well compared to a big multitrack playing back at 30 ips)
 
We can still do it in the all digital world too.  Work at 96K 24 bit all the way and create our playback medium last in the chain. The good news is the playback medium is much better now.  Why would working at a higher res not sound a little nicer like it always did.  It does.  But we can work at payback res now all the way from start to finish because many projects we do sound great that way and would hardly benefit from working at a higher res.  But then other things certainly do, like playing and creating music with 'Prism'.  (Native Instruments Virtual synth and a very unusual one at that!)  It is very obvious.  Just listen to the two files in my first link posted above and I think you will agree.
 

Wouter Schijns
some say higher rates sound better, others don't hear a difference.....it's not a massive sound difference I guess.
but higher rates make lower latency.
so when bouncing and especially exporting a song, maybe better to set high rate as the song will get more latency anyway later on (mediaplayers/cd players converters can add latency) and might come to a latency level where it really starts to sound flat.
so to minimise that, bounce /export at high rate....just for latency sake.
this is just what I read on the web.....so don't shoot please you forummers looking for a target (you remind me of my mother in law).
 
GL
 
 

I don't mean to "shoot" you but I don't think you have a good idea what latency is or what it does. It is true that higher sample rates (with similar buffer settings) will result in lower latency. This has however zero effect on bouncing your song, and certainly is there no way this kind of latency can in any way lead to your song sounding flat.

Sander, great to read your experience/thoughts, hope Sharke will get something from this thread
 
under Preferences, when you set higher rate/lower buffers.. Sonar specifies a roundtip with lower latency
I understand that as latency you will have in music out of Sonar (plugin latency not calculated inn).
I think I hear a latency when exporting a song, then A/B listening to Sonar and the exported file.
the wav file sounds not as real/live as the Sonar project, it has slower attack to my ears....the difference is just small...can only hear it on headphones.
those things don't come through latency ? or maybe the Media player reduces the sound quality ?
a saw a post somewhere from a mastering engineer, he likes the overall DAW latency under 10 m/s
 
 
 
 
 

This is indeed an interesting thread. One thing that bothers me slightly though is: why is it not possible to change the sample rate of a project once it's started? Is there some technical reason for this, aside from the obvious that any audio clips would have to be converted? I have some really synth-heavy projects in 48kHz which I'd love to try converting to 96kHz, but I'm not that keen on the amount of work it would take to recreate them from scratch, especially given the amount of automation on some of the tracks. 

This is indeed an interesting thread. One thing that bothers me slightly though is: why is it not possible to change the sample rate of a project once it's started? Is there some technical reason for this, aside from the obvious that any audio clips would have to be converted? I have some really synth-heavy projects in 48kHz which I'd love to try converting to 96kHz, but I'm not that keen on the amount of work it would take to recreate them from scratch, especially given the amount of automation on some of the tracks. 

It is possible to change the sample rate once the audio is converted see http://forum.cakewalk.com/FindPost/2934357

scook
It is possible to change the sample rate once the audio is converted see http://forum.cakewalk.com/FindPost/2934357

That too involves starting a new project. I realize you can change the rate of audio and import it to a new project - I just wondered about any technical reasons for Sonar not being able to change the sample rate of the same project. Is it just a feature limitation? I imagine it would be a non-trivial feature to implement, but without any information about why it's not possible, I don't know, it seems possible.....