Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Jerry Tubb]

Bruno Putzeys wrote on Tue, 18 December 2007 08:53

The only reason why we're juggling steepness vs aliasing is because 44.1kHz leaves insufficient room to put a proper filter in.

Excellent explanation Bruno,

So perhaps there isn't a perfect SRC algorithm after all.

With specs acceptably close between SRC applications, choosing by  sound & musicality is the ticket.

So yet again we've come full circle, judging quality by ear.

JT

[George_]

Quote:

The only reason why we're juggling steepness vs aliasing is because 44.1kHz leaves insufficient room to put a proper filter in. A nonaliasing filter - certainly one designed to cut off the alias range left by halfband filters elsewhere in the chain- rolls off inside the audio band. A slower filter will leave a significant deal of alias products lurking just at the end of the audio band. In neither case do we need a fancy hypothesis to explain why it's audible.

yes, very good explanation!  appreciated.

[Patrik T]

Jerry Tubb wrote on Tue, 18 December 2007 19:58

So yet again we've come full circle, judging quality by ear

What else to do when it comes to music?



P

[Kees de Visser]

Patrik T wrote on Wed, 19 December 2007 04:04

What else to do when it comes to music?

Of course the ear is the final judge, but . . . knowledge about the theory can help.
Wikipedia: http://en.wikipedia.org/wiki/Music_theory
quote: "Composers who analyze music regularly often find an increase in the quality of their musical compositions."
IMO theory can help interpreting what you think your ears are telling you.

[Patrik T]

Kees de Visser wrote on Wed, 19 December 2007 13:29

IMO theory can help interpreting what you think your ears are telling you.

And the other way around as well! That one can think it's audible because theory, or the eye, says so.

It's all a bit messy.

Patrik

[Kees de Visser]

Patrik T wrote on Wed, 19 December 2007 23:50

And the other way around as well! That one can think it's audible because theory, or the eye, says so.

Very true! Our hearing has "changed" dramatically since we've started looking at audio (waveforms, FFT, Spectrograms etc.)
Perhaps we need a (golden ear?) Psy model to put things into perspective.

[Kees de Visser]

Roland Storch wrote on Sun, 09 December 2007 23:38

I would like to see the graphs for the old Sonic Solutions HD system. The SRC is regarded as one of the best SRCs by a lot of mastering engineers including Bob Katz (if I may name him here). This is impressive because this SRC is a lot of years old and allowed to choose different filter settings.

The three Sonic HD SRC results are online now: steep, gentle and gentlest filter. The published results are those without dither, just 48 to 24 bit rounding. I don't know how old the design is, but it seems well done.

[Roland Storch]

Looks like a used Sonic HD is worth the money only for the SRC.

[Alexey Lukin]

Good news: the impulse response graphs have been added to SRC comparisons at Infinitewave.ca. Some of the graphs are truncated in time, because certain impulses can obviously be quite long or even infinite.

[Andrew Hamilton]

Hey Alexey, what happened to the problem that AutoSonic had with the whacker test?  Was it just my rig that failed, or did you alter the test since my results were originally posted?


Thanks,
    Andrew