192kHz considered harmful

[Mnyb]

Some one have been kind enough

http://src.infinitewave.ca/

Yea I don't hear past 18k anymoore and absolute phase is debated if it is really audioble on music signals and if it is it the effect is also gear dependent .

(The 18k to 20k rollof and phase inversion of r8brain free )

Over to use SoX (now it is installed on my main computer )

I urge any other audiophile to try kind of nerd fun and put things into perpective...

[Mnyb]

Update I used with SoX with 2 suggested comandlines

The -v option was in my case not really necessary only 1 out 5 files complained about clipping and it was at most 16 samples so it should be used on case by case basis.

Unless you *cough* produce loudness war music but then you won't come here and harass poor old audiophiles about sound-quality *cough*
You would roll in $$$ blowing the ears of the young generation....

One thing caught my interest here:

http://src.infinitewave.ca/

Many brand name name resamplers don't do a spotless job ??

No wonder there is some fuzz about it and many engineers get third party plugins with better performance , A little surprised that the big names in DAW don't care more . But maybe it's examples from older versions .
But anyway it's just math why don't do it rigth .

This also again point out SoX as one of the better choices actually better than what is used professionally in some cases, and that should cure some hypochondria about the resampling inside LMS .

Some people claims that re-sampling 192 > 96 with LMS/flac/SoX sounds "bad"
This is about as BS as it can be.
And constantly whines about that the bad bad naugthy squeezebox only supports 96kHz

I'm going to listen some more on these files with my crappy hifi that actually only supports 96k, (192k would be re-sampled at the input) , the designer thought it was a total waste of power to run all the DRC and surround modes and digital xovers at 192k .

[Wombat]

One thing caught my interest here:

http://src.infinitewave.ca/

Many brand name name resamplers don't do a spotless job ??

The pictures are nice to watch but as mentioned before you´ll have a hard time to find even the worst resamplers tested there to sound wrong
Something like the noise down 180dB isn´t exactly needed at all but most use the stopband so low becaus with todays processing power it comes easy. Saracon even uses more then these 180dB even but don´t expect any change especialy when we cut down to 16bit with this resampling.

[Mnyb]

SoX was not that hard to use , the hard part is to choose settings afaik as SoX is very flexible and probably can be used with anything but sane settings if you wan't .

Thank you for suggested settings

Preliminary the results are a null ( as expected ) I will convert some more files too try further
Haven't picked the exact track to try to ab rigth now I 'm fumbling with my i pad casually just tapping somewhere in the playlist.
To ab I would mass copy my chosen track and place each one in a folder of their own then I can have the same file name then I can build a playlist that I can shuffle and thus hide my own activities to myself.
This will not go down perfectly as I know what I'm testing and already have an opinion so I'm biased even if I blind it .

Ideally I should invite a friend and let him/her listen unknowingly of what's being tested .
But the experience of actually trying on a lot of different tracks put things in perspective.

I Recommend anyone else to try too but use SoX as it is very good, the paranoia risks are very low as it yields excellent results , I did a whole zoo of different transcodings but as the diff is very very small go for 24/96 ( 192 ) to 16/44.1 direct so you can try worst case .

This is actually very much on topic , I you're unsure on what to think about the topic just try it for kicks.

This way you can be sure to compare apples with apples commercial disc with the "same" music in different formats can be compromised.

[TheOctavist]

I wouldn´t choose r8brain free, cutting as low as 18kHz is not necessary. It even inverts phase.

this makes no audible difference whatsoever. r8 brain free offers the best phase performance with *zero* aliasing..even better than the pro version!!

besides the new version does not flip the phase(it was inaudible anyway!!)

id bet that you couldn't id them blind.(r8 brain gratis vs pro vs saracon vs uv22 vs mbit maxx vs POW-R! Hell, I doubt that anyone could..maybe some of the really badly implemented ones on a certain type of music.. but doubtful.

[TheOctavist]

Mynb... if you want to have some more fun, download THIS

Open your DAW, generate a 1Khz sine wave, then open up Christians tool and see what is going on.


It is quite interesting to see the behavior of some of the analog modeling suites, EQs, and others.!

[Wombat]

this makes no audible difference whatsoever. r8 brain free offers the best phase performance with *zero* aliasing..even better than the pro version!!

All resampler that have a flat line there have perfect phase response. How is r8brian best then?
Avoiding aliasing is common. If you don´t say sox to alias, it doesn´t. You should better ask yourself if we should allow a controlled amount of aliasing like Sarcacon for example.

Edit: Last time i checked r8brain free had only flat dither. Has this changed?

[TheOctavist]

Brother, I can't tell any difference with any of them.

I just use the Samplitude 11 inbuilt one..


BTW.

Dither Test.

http://www.filedropper.com/dithertest

[Wombat]

Brother, I can't tell any difference with any of them.
http://www.filedropper.com/dithertest

Yes, i also think it is in practice very unlikely you´ll notice a difference.
Thats why i wonder when some audiophiles dig out the big vocabularies to describe all the audible things tiny changes in the settings of a resampler do

Thanks for the dither samples, now others can have a listen how it sounds.
Did you amplify the noise by +48dB or around that? If so i still prefer the sox Shibata dither curves over the ones in these samples.

[TheOctavist]

http://www.ethanwiner.com/dither.html

http://www.24-96.net/xchange/test_dithered.wav

http://www.24-96.net/xchange/test_undithered.wav

http://www.24-96.net/xchange/test_di...compressed.wav

http://www.24-96.net/xchange/test_un...compressed.wav

http://www.gearslutz.com/board/attac...-110-powr3.wav

http://www.gearslutz.com/board/attac...0no-dither.wav


A dynamic range of 118 dB is determined necessary for subjective noise-free reproduction of music in a dithered digital audio recorder. Maximum peak sound levels in music are compared to the minimum discernible level of white noise in a quiet listening situation. Microphone noise limitations, monitoring loudspeaker capabilities, and performance environment noise levels are also considered.
Author: Fielder, Louis D.
Affiliation: Ampex Corporation, Redwood City, CA
AES Convention:69 (May 1981) Paper Number:1772
AES E-Library: Dynamic Range Requirement for Subjective Noise Free Reproduction of Music
http://www.aes.org/e-lib/browse.cfm?elib=11981





"CONCLUSIONS
This article has reviewed the issues surrounding the transmission of high-resolution digital audio. It is suggested that a channel that attains audible transparency will be equivalent to a PCM channel that uses:
· 58kHz sampling rate, and
· 14-bit representation with appropriate noise shaping, or
· 20-bit representation in a flat noise floor, i.e. a ‘rectangular’ channel
This conclusion has the following obvious implications:
· The CD channel with 44.1kHz 16-bit coding (even with noise shaping to extend the resolution) is inadequate
· Even 48kHz sampling is not quite high enough
· Sampling at 88.2kHz or 96kHz is too high, and therefore wasteful of data
· The use of sampling rates above 96kHz to convey a wider audio bandwidth cannot currently be justified"
http://www.meridian-audio.com/w_paper/Coding2.PDF - check figure 20


Martin Kantolas excellent 24 bit vs 16 bit TPDF test that showed 16 bit dither to be audible using a _tape source_ recorded to digital, with normal loud music.

http://www.nordicaudiolabs.com/samples/24.wav
http://www.nordicaudiolabs.com/samples/16-hpt.wav

http://www.youtube.com/watch?v=Sl0U_L3tb_M

http://www.cranesong.com/products/dither/index.html