Re: [ARSCLIST] Compressed files & differences (was something else)

["Richard L. Hess" <arclists@xxxxxxxxxxxxxxx>]

At 01:33 PM 1/22/2006, Don Cox wrote:
> On 21/01/06, Mike Richter wrote:
> > Frank Strauss wrote:
> >> Hi Tom-I am a long time lurker on this list, trying to learn what I
> >> can, and have been very interested in this thread. I have put about
> >> 500 cd's into iTunes, using mp3 format at 196 kbs. I love music, and
> >> have always thought I had reasonably good ears. I am disturbed to
> >> think that with iTunes(and by this,
> >>
> > It is possible to detect the differences depending on encoder,
> > bitrate, and most significantly listening closely, particularly when
> > one knows what to listen for. Among the differences, imaging is
> > distorted, though that matters little on a studio recording.
>
> I also found audible differences between different decoders, on my
> computer.


Funny story department...or why you should never do anything ad-hoc 
at a seminar.

I was giving a presentation on audio archiving and had a demo CD 
prepared and was playing selections. Someone asked about MP3s. I 
happened to have had one of the selections on the demo CD also as an 
MP3 on my Palm Tungsten T3 so I swapped the 1/8" miniplug from the 
Sony CD Walkman (with LINE out) and the Palm's headphone out. The MP3 
sounded terrible.

When I got home, I took the MP3 and loaded it into Samplitude, 
aligned it with the WAV file and intercut the two and burned it out 
to the new version of demo CD. You can barely (and I mean barely) 
tell the difference. There is a subtle difference on the bass. I 
think this was a 192 kb/s MP3, but I could be wrong.

I emailed the conference organizer and explained this and asked her 
to email all the attendees explaining the error in the demonstration: 
we were demonstrating the difference between the Sony CD Walkman's 
D-A and output section vs. The Palm's.

So, in my testing, MP3s can sound very close to the original WAV file 
in the 192-256 kb/s region and there are FAR more differences in the 
D-A and downstream of that.

I would like to suggest that the same differences occur when someone 
upgrades from 44.1/16 to 88.2/24 and hears a world of difference. 
Perhaps what is being heard is the new cleaner converters on the 
higher resolution device (or the analog stages). IMHO the only fair 
test (and sometimes that's not even fair) is to run 44.1/16 AND 
88.2/24 or 96/24 through the same converter and then compare, 
preferably by intercutting two sound files and playing back a stream.

I remastered and released one of Marie-Lynn Hammond's albums whih HAD 
been out on CD previously. I took a digital clone of the previous CD 
and my version from the 30 in/s 2-track masters and intercut. 
Marie-Lynn heard the difference immediately on a moderate system. Her 
voice sounded much cleaner. I wonder what A-D had been used in 1990 
to make the original CD.

Cheers,

Richard

Cheers,

Richard

Cheers,

Richard