[`danny]

Is there a place where I can download a comparison of say 96khz/24bit vs 44.1khz/16bit

 

[cconnaker]

Linn records and iTrax are good places to start. I know they have sample files, not sure if comparisons are available or how productive down-converting a sample 24/96 file would be.

 

[Joeywhat]

The problem with comparing the two is that one of them will have had been converted. Typically if you have a 24/96 recording, the 16/44 will have been converted from the higher rate. I do not know of anywhere that has true comparisons available. In order to do that one would have to simultaneously record into two separate sessions, one at 44.1 the other at 96...I do not know of any other way to do it.

You can't just rerecord everything a second, separate, time since the live instruments may not be miked exactly the same, effects processors may have been changed, and many other factors.

 

[ru9]

here's one:

Soundkeeper Recordings Format Comparison

 

[`danny]

Thanks for that sample, that is the first time I have actually heard something in 96khz/24bit. I am wowed by it, you can literally tell which instrument is playing which it is incredible. I wish CDs were this way

 

[HFat]

You shouldn't hear a big difference. Most people wouldn't be able to hear any difference.

Can you tell that 24/96 sample from the same downsampled to 16/44 in a blind test?

 

[Tarkovsky]

Everything on CD at 44.1/16 has been recorded originally at 96 or even 192 in 24 bit and downsampled.

 

[Tarkovsky]

Well, almost everything.

 

[Joeywhat]

Quote:

Originally Posted by Tarkovsky /img/forum/go_quote.gif Everything on CD at 44.1/16 has been recorded originally at 96 or even 192 in 24 bit and downsampled.

Not really...maybe a few of the real large bands do it, but otherwise there is no need. Very few bands even want to release an album at those rates, let alone be able to afford it. This is especially true since, if the end medium will be largely CD's, any good engineer would keep the session @ 44.1.

 

[sejarzo]

Quote:

Originally Posted by Joeywhat /img/forum/go_quote.gif Not really...maybe a few of the real large bands do it, but otherwise there is no need. Very few bands even want to release an album at those rates, let alone be able to afford it. This is especially true since, if the end medium will be largely CD's, any good engineer would keep the session @ 44.1.

I've been told by a number of engineers that they all track at 24 bit so any post-processing can be done at the higher resolution before downsampling to 44.1. Even an amateur friend of mine with a ProTools rig and a notebook does everything at 24/96.......high school bands, choirs, local theater group musicals, whatever.

 

[hciman77]

Quote:

Originally Posted by HFat /img/forum/go_quote.gif You shouldn't hear a big difference. Most people wouldn't be able to hear any difference. Can you tell that 24/96 sample from the same downsampled to 16/44 in a blind test?

I downloaded the two tracks and compared them in Audacity - the waveforms were identical at the highest possible levels of zoom, casually they sounded identical.

I loaded them into Foobar 2000 and used the ABX utility, I scored exactly 5/10 i.e pure chance, admittedly through a laptop sound chip, but if 24/96 is really that much better I should be able to hear some difference ? surely ?

EDIT: I just did the ABX thing again this time I scored 13/15 - I cranked the volume right up and one semed to be slightly louder than the other , I coiuld only tell by instantly swapping between the two at only at the start of the track, to be honest I thought I was still guessing a lot and I would not have noticed the difference any other way, listened to as music they seem identical , I wont be rushing out to buy a high res player....

EDIT2:

This is perhaps the difference I was hearing

 

[mr. nice]

man im not sure what your are hearing cuz you need some serious gear in a nice acoustically treated room to hear differences in recordings done at 44.1k 96k or even 192k and that's highly debatable (in all reality i think the mind is playing tricks on us). most engineers keep it at 44.1 24bits if going to CD, some even 44.1k 16bit. the extra 8 bits just gives an engineer that much more headroom to work with. 44k 96k and especially 192k is all pretty much debatable. Dan Lavry believes 192k is BS and simply marketing.

read this article 16 vs. 24 bit Audio Recording Demystified

you wanna drive yourself nuts and read that it is impossible for humans to listen to what we are even debating check this out

Nyquist frequency - Wikipedia, the free encyclopedia

 

[Joeywhat]

Quote:

Originally Posted by mr. nice /img/forum/go_quote.gif Nyquist frequency - Wikipedia, the free encyclopedia

What people tend to forget about Nyquist's Theorem just says the highest frequency that can be reproduced...it doesn't mention anything about the lower frequencies. The more times a frequency is sampled, the "better" or more accurate it will be reproduced. So at 44.1 KHz at 20 KHz frequency will only be sampled around two times...but the lower frequencies will be sampled more, and the lowest frequencies will be sampled 100's, if not 1000's of times (I didn't do that math, so I might be off a bit).

So how's that relate to higher sample rates? A sample rate of 192 KHz will sample 20 KHz more times then a 44.1 rate, thus producing a more accurate signal.

Is there a difference in the two sample rates? I think so...but I will say that I believe it's mostly psychoacoustics. I'm sure the signals will be different when looked at closely, but not enough difference for it to be night and day. Most people I ask do say they like the 192 KHz songs better, although they typically can't say why. This is also usually done with an original recording at 192, then downsampled to 44.1, FWIW.

 

[HFat]

Quote:

Originally Posted by hciman77 /img/forum/go_quote.gif I downloaded the two tracks and compared them

I was thinking about downsampling it (or any other sample) yourself... I wouldn't trust one to be a mere downsampling of the other.

Quote:

Originally Posted by hciman77 /img/forum/go_quote.gif I cranked the volume right up and one semed to be slightly louder than the other

In foobar, you need to use replaygain for volume matching in blind tests, right? I'm not sure it's accurate enough though.

 

[12Bass]

Something to bear in mind is that D/A converters are imperfect. Thus, even though Nyquist's theorem may be correct in principle, the benefits from using higher sampling rates can be audible in practice. One effect seems to be from the change in the frequency of the antialiasing filter. From my own listening, I've found less coloration in the treble frequencies at higher sampling rates - which seems to be due to less brickwall filtration artifacts making their way into the audible spectrum at the higher rates. So, I'm not sure if it is really the higher rates that are making the difference, or if it is just that the higher rates have less audible artifacts when the signal is reconstructed in the D/A conversion....