jih64
1. Why is it that you use a higher sample rate, it's more taxing on your machine etc, how come you can use lower buffers, get lower latency ? I ould have naturally assumed it went the other way.

 
There's the potential for lower latency, because the machine is running faster. But that doesn't mean you can achieve the lower latency in a practical sense because it is more taxing on your machine. A lot of times it's a wash - if you use X sample buffers at 48 kHz for reliable operation, you'll need to use 2X sample buffers at 96 kHz. 
 

2. I would have thought my machine would have been beefy enough to do 96Khz ? (maybe it would have without the crackles if I upped the buffers a bit)

Higher sample rates do stress out your computer and interface more. For example a USB interface that can stream 8 channels of audio at 48kHz may only be able to stream 4 at 96 kHz and 2 at 192 kHz.

as craig points out, it's all in the maths (whether you can hear it or not ;-)) i.e., generally, the bigger the digital-unit-to-realworld-unit ratio, the better or more accurate it all is

I recently saw a Ted Talks seminar online about how perception affect personal choice. Say you bought a painting and was told it was a Monet. Later, you find out it's a fake and decide to just throw it away. It wouldn't even matter if you've had the painting for years.
 
I record at 44.1 because that's the medium. When the medium changes, I'll change with it. If you are a child, you may hear a difference between the different sample rates but if you think you can as an adult, your kidding yourself.
 
To really know how you hear you have to read up on the anatomy of hearing and get a hearing test and ask for the frequency curve of the results of that test. I asked the technician that did my last hearing test and she thought I was crazy about believing anyone could hear above 10khz. Or was it that no instrument can produce a sound over 10khz. Regardless, we can perceive up to 20khz, maybe 22khz as a child, but for the most part, most can hear up to 14khz.
 
Again, it's all about perception and our minds can trick us into perceiving something that's not reality.

Guitarpima
If you are a child, you may hear a difference between the different sample rates but if you think you can as an adult, your kidding yourself.

Again, what intrigues me is that any difference (assuming there is a real difference) may not be related at all to frequency response, but to some byproduct of using higher sampling rates. The "better sound at 96 kHz for things that don't oversample" is a perfect example of this, as is the contention that some converters offer better performance at 96 kHz so they provide better sound - although it has nothing to do with frequency response.
 
If there is a benefit to higher sample rates that relates to something other than frequency response, then what I would hope is that those benefits could apply to improving sound quality at any sample rate.

44.1 is the final delivery medium, but everything before that should be as hi-res as possible

I'd sum up my view like this: higher sample rates are not a particularly strong influencer of final perceived quality. Certainly, all analogue recordings ever made have lower bandwidth. Lower still once you whack them onto vinyl.
 
I'm not saying that a theoretical low quality of the final medium means it doesn't matter. I'm saying that what might seem, by the numbers, to be lower quality actually garners superb results.
 
44.1 was a well-chosen standard. The point of it was that it was way higher than you'd ever actually need.
 
Back to my opening point. When setting out to record something, I'm concerned about the performances, the instruments, the amps, the sound of the room, the mics, the mic placement, in that order of importance. I'm not sure where sample rate comes in terms of priority, but it's probably not even right behind mic placement. Pres are probably more important, and in truth, pres aren't all that important, once you're at a given level of transparency.

Basically, I decided to stop worrying about it and get on with making records.

There is a fairly cogent argument for using higher than 44.1 for recording, but it's got nothing to do with capturing higher frequencies. It's that it's easier to build the cutoff filter when there's a bit more bandwidth to play with. Not I just say easier, which you can read as "cheaper". There are many filters that work perfectly well in 44.1.
 

I'm at 24 bit / 44.1 KHz.  Honestly, the fidelity from that configuration is very satisfying to my ears and to the ears of other non-techie/non-audiophile listeners who have no preoccupation with things like word size and sample rate.  I'm not convinced that there is a verifiable difference in perception; this has been formally tested on more than a couple of occasions with unconvincing results.  I won't contest that folks who say they can hear a difference actually can, but I think they are probably in the minority of listeners.  I stick with this format because it results in what I consider exceptional audio fidelity while leaving the greatest possible processing headroom from DAW software and PC resources, i.e., great audio fidelity + great DAW stability.  I think if your DAW can keep up with higher sample rates for your projects, then go for it.  But namaste right here...

jih64
3. drewfx1, are you Drew from FXpansion ?

No. 

AT
Like most attempts at high end, if there is a difference between 44.1 and 96 or 768 it can only be heard on a good system and in good rooms by trained ears.  Just like converters and mics and preamps and such.  It is a small diff that most people don't get because they simply can't hear it.  Musicians and engineers etc. w/ ears attuned to the nth degree can tell some differences, but I imagine would be hard pressed if blind tested on ear buds.
 
Of course, many of us get our rocks off on the last inch of quality we can squeeze out of sound whether by hardware or technique, and I warrant many of us get better performances and give greater detail to our work if we believe we can hear such differences.  So it is worthwhile, even if only psychologically.  But that doesn't mean you have to go all DSD or only use the most exotic hardware etc.  The best bet is to do testing yourself on your system and see if you can hear the difference, then decide if filling up hard drives like they are floppies is worth it to you.
 
Me, I record at 44.1 - I haven't heard enough difference to switch to 96 even tho it is becoming standard.  I do make sure the effects upsample and turn the older Cake synths etc. to high render rates.  That I can hear a difference in the smoothness.
 
@

Ear training can indeed allow one to hear small details if they are audible.
 
But the "only with super gear in special room" is nonsense. Different artifacts caused by different things (including listening systems and the listening environment) just do not conveniently line up with each other that way to get masked. 
 
Can't hear an artifact in a given situation (assuming it's real and audible by humans)? Try turning up the volume a little. Or playing a "worst case" signal. Or moving closer to the speaker to reduce the role of the room or environmental noise. Suddenly it's audible with any gear.
 
The stuff that can only ever be heard with special equipment under any conditions always turns out to be imaginary.