"Evil rock music impresario Lars Mangros puts hypnotic signals in the music done by his best-selling band Andromeda that incites young fans to riot and rebel against authority. Andromeda are going to perform their first big concert which will be broadcast on a global basis. It's up to Buck Rogers and Twiki to thwart Mangros' dastardly plot to warp the minds of the youth all over the universe during said concert."
 
Amazing.
 

Buck always deliberately spoiled everything, didn't he?

mike_mccue
It seems like an actual comparison would require a protocol featuring side by side capture, mixing, 
mastering, (all at the respective specifications) a final export to the distribution specification, and perhaps a digital to analog conversion.
 
If you arrive at the same conclusion after that, your conclusion will be based on an actual comparison rather than an implication that a comparison was made.
 

  
 
I didn't have the luxury of doing a side by side recording so couldn't do that test. However I'm curious why you consider my test to have been not valid. I compared a downsampled version with the original 96K recording by phase inverting snd mixing the two wave files. If there was something  special  in the original 96K recording the phase invert and mix test should have pulled out just the differences in the two recordings, right? I'm no mastering expert so I'm happy to be corrected here.
 
BTW I agree that the some of statements in that article were a bit over the top like the one about it being bad for the environment :)

"BTW I agree that the some of statements in that article were a bit over the top like the one about it being bad for the environment :)"
 
Its been proven countless times that high sample rates depletes the Ozone.  LOL But then so does breathing!

The guy writes in a semi-humorous style. I can think of worse crimes.

Noel Borthwick [Cakewalk]

mike_mccue
It seems like an actual comparison would require a protocol featuring side by side capture, mixing, 
mastering, (all at the respective specifications) a final export to the distribution specification, and perhaps a digital to analog conversion.
 
If you arrive at the same conclusion after that, your conclusion will be based on an actual comparison rather than an implication that a comparison was made.
 

  
 
I didn't have the luxury of doing a side by side recording so couldn't do that test. However I'm curious why you consider my test to have been not valid. I compared a downsampled version with the original 96K recording by phase inverting snd mixing the two wave files. If there was something  special  in the original 96K recording the phase invert and mix test should have pulled out just the differences in the two recordings, right? I'm no mastering expert so I'm happy to be corrected here.
 
BTW I agree that the some of statements in that article were a bit over the top like the one about it being bad for the environment :)

I'm no mastering expert either The Scientist Blog seemed to be focused on comparing the relative merits of choosing different sample rates for analog to digital conversion and work flow. It even emphasized how one should try to use the best converter for each respective sample rate when working with said rate because it's unlikely that one converter will be ideal for two different rates.
 
When you sample at 96 and then down convert to 44 and decide it sounds great then all you have actually done is figured out that you can down sample to 44 and it sounds great.
 
It isn't that your test didn't have some useful application. The point was that you didn't compare the results of a 44 workflow to a 96 workflow so all you got was a chance to observe that a down convert from 96 sounded really good. There's a small irony in this, because the scientist spent a lot of time explaining that the very same work flow you used has all kinds of gotchas to beware of. Luckily, your music made it through just fine. :-)
 
 
all the best,
mike
 
 

Noel Borthwick [Cakewalk]

mike_mccue
It seems like an actual comparison would require a protocol featuring side by side capture, mixing, 
mastering, (all at the respective specifications) a final export to the distribution specification, and perhaps a digital to analog conversion.
 
If you arrive at the same conclusion after that, your conclusion will be based on an actual comparison rather than an implication that a comparison was made.
 

  
 
I didn't have the luxury of doing a side by side recording so couldn't do that test. However I'm curious why you consider my test to have been not valid. I compared a downsampled version with the original 96K recording by phase inverting snd mixing the two wave files. If there was something  special  in the original 96K recording the phase invert and mix test should have pulled out just the differences in the two recordings, right? I'm no mastering expert so I'm happy to be corrected here.

I don't see anything to correct, unless you're testing whether something recorded, mixed, and mastered at 44.1kHz is going to sound better than something recorded, mixed, and mastered at 96kHz or recorded, mixed, and mastered at 96kHz and then downsampled to 44.kHz. Your test gives 44.1kHz the "benefit of the doubt" by feeding it the [supposedly] higher-quality 96kHz source material.
 
One area where people might not take matters into account is the output filtering. I don't know if it changes for different sample rates, but for 44.1kHz, you have to brickwall pretty heavily (e.g., 96dB/octave) to keep the clock out of the output. With 96kHz, you can use a 48dB/octave filter and obtain even a bit more rejection.
 
In a comparison test at AES several years back between a 30 ips analog master tape and PCM reproductions at various sample rates as well as DSD, to me (and others in the audience) DSD sounded more like the reference music tape than 44.1kHz PCM. I've wondered if it's the technology, the fact that DSD can use gentler output filters, or something else altogether.

I don't understand that comment about output filtering. I don't even understand what you mean by it in this context. Can you explain or link to an article that would explain it?

The article Noel linked to in post number 1 of this thread describes what Craig has referred to as "output filtering" under the headings:
 
Improvements at 44.1: Fixing the Filters
 
and
 
When More is Better: Making The Filters Even Better