IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
hybrid coding mode
Pixar
post Dec 16 2010, 11:50
Post #1





Group: Members
Posts: 11
Joined: 16-December 10
Member No.: 86567



Hello all =)
I tested a hybrid coding mode (the best-miqhhb5) ...
I was delighted with the highest quality.
Programs designed for the detection of differences in sound do not detect a difference lossy and lossless.

What removes from the sound?
I did not find any similarity with other lossy codecs))
How have you achieved such excellent results? blink.gif

This post has been edited by Pixar: Dec 16 2010, 11:57
Go to the top of the page
+Quote Post
pdq
post Dec 16 2010, 14:19
Post #2





Group: Members
Posts: 3305
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



The lossy version is slightly less accurate than the oiginal, so in effect it is a little bit noisier.

I do hope you will change your avatar. sick.gif
Go to the top of the page
+Quote Post
bryant
post Dec 16 2010, 16:09
Post #3


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



The hybrid coding mode just adds broad spectrum quantization noise during encoding, and this is very similar to the the natural noise floor of most recordings. For the command line you mention, this added noise would be very low level, and might even be below the noise floor of the recording. This would be almost impossible to detect with any program (mush less hear), and especially for programs that are looking specifically for the kind of things that regular lossy codecs do (like breaking up the audio into separate frequency bins, which WavPack hybrid never does).

If you tried a lower bitrate with the hybrid mode (like -b3, which gives about 270 kbps) then you might be able to hear noise in some music (especially very tonal or electronic music) because at lower bitrates WavPack hybrid will generally not do as well as regular lossy codecs. However, because the only artifact is this smooth added noise, those detection programs might still fail even though the degradation is audible.
Go to the top of the page
+Quote Post
Pixar
post Dec 16 2010, 18:19
Post #4





Group: Members
Posts: 11
Joined: 16-December 10
Member No.: 86567



Degradation does not occur))
With your codec I do not need lossless (only to spend too much space) ...
Optimum price / quality ratio.

I have a question.
How do I glue together multiple files into one (without transcoding)?
Analogue to Ogg Vorbis
copy / b 1.ogg + 2.ogg + 3.ogg = image.ogg
In WavPack is not working.

Thank you for your work bryant.
WavPack brilliant creation of genius smile.gif
Go to the top of the page
+Quote Post
bryant
post Dec 17 2010, 04:50
Post #5


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



QUOTE (Pixar @ Dec 16 2010, 09:19) *
How do I glue together multiple files into one (without transcoding)?
Analogue to Ogg Vorbis
copy / b 1.ogg + 2.ogg + 3.ogg = image.ogg
In WavPack is not working.

You can concatenate WavPack files like you describe, but the only way to decode the result is to use the -b flag with wvunpack (“blind” decoding). I have done this to combine tracks, but you really need to transcode to make the result playable, seekable, etc. Of course, lossless transcoding is not a problem (and it works great with pipes), but I would not do it with hybrid.

One of the things I would like to create is a WavPack file “fixer” that will modify the headers of files like these in place (without transcoding) because there are other situations where these files can occur, but I haven’t gotten around to it.
Go to the top of the page
+Quote Post
pdq
post Dec 17 2010, 14:16
Post #6





Group: Members
Posts: 3305
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



QUOTE (pdq @ Dec 16 2010, 09:19) *
I do hope you will change your avatar. sick.gif

Thank you, that's much better. smile.gif
Go to the top of the page
+Quote Post
Pixar
post Dec 17 2010, 16:47
Post #7





Group: Members
Posts: 11
Joined: 16-December 10
Member No.: 86567



QUOTE (pdq @ Dec 17 2010, 15:16) *
QUOTE (pdq @ Dec 16 2010, 09:19) *
I do hope you will change your avatar. sick.gif

Thank you, that's much better. smile.gif

You have amused me laugh.gif

QUOTE (bryant @ Dec 17 2010, 05:50) *
One of the things I would like to create is a WavPack file “fixer” that will modify the headers of files like these in place (without transcoding) because there are other situations where these files can occur, but I haven’t gotten around to it.

It would be excellent. Thank smile.gif
Go to the top of the page
+Quote Post
DARcode
post Dec 31 2010, 16:03
Post #8





Group: Members (Donating)
Posts: 679
Joined: 10-January 05
From: Italy
Member No.: 18968



QUOTE (bryant @ Dec 17 2010, 04:50) *
[...]
One of the things I would like to create is a WavPack file “fixer” that will modify the headers of files like these in place (without transcoding) because there are other situations where these files can occur, but I haven’t gotten around to it.
That'd be awesome, count me among the very very interested in that, clearly after you're done merging all the executables into a single one tongue.gif ...


--------------------
WavPack 4.70.0 -b384hx6cmv/qaac 2.32 -V 100
Go to the top of the page
+Quote Post
Cavaille
post Feb 6 2011, 03:02
Post #9





Group: Members
Posts: 370
Joined: 20-May 06
Member No.: 30963



QUOTE (Pixar @ Dec 16 2010, 18:19) *
Thank you for your work bryant.
WavPack brilliant creation of genius smile.gif


This is being said quite too seldom around here in my opinion. Indeed, WavPack is the best lossy compressor available. It produces precisely predictable results when used with several bitrates and it always works. Stable as a rock. One can rely upon it. I use it very often with 96 kHz files together with the -q switch. It´s just perfect for that: you´ll get wonderful "master" files to create other things from and you save space at the same time. Brilliant!


--------------------
marlene-d.blogspot.com
Go to the top of the page
+Quote Post
shadowking
post Feb 6 2011, 03:30
Post #10





Group: Members
Posts: 1523
Joined: 31-January 04
Member No.: 11664



I find 4 bps very reliable , saves a lot of space over lossless (2 ~ 3 times), average is 370 kbps for cd audio. I use -b4hhx and could never abx anything in my collection of 400+ cd's. I often transcode these lossy files to mp3 with great results too.


--------------------
Wavpack -b450x1
Go to the top of the page
+Quote Post
bryant
post Feb 8 2011, 19:29
Post #11


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



Thanks guys; I’m glad WavPack is working out for you.

I also find the hybrid lossy mode useful for certain things. Lately I have been doing some recording from local HD radio broadcasts. I first record the analog HD radio signal onto a Rockboxed iRiver H120 using WavPack lossless, then I copy those to the PC for conversion. If I don’t care too much about quality I’ll just transcode them to MP3, but if I am concerned about possible lossy to lossy artifacts (because the HD radio uses a lossy codec) then I transcode them to WavPack lossy at about 300 – 400 kbps.

I have thought about putting 4 bps hybrid lossy mode encoding directly into the Rockbox code, but I’m afraid it would just confuse people and I would be the only one using it... smile.gif
Go to the top of the page
+Quote Post
Didjeridoo
post Nov 10 2011, 19:56
Post #12





Group: Members
Posts: 16
Joined: 8-June 10
Member No.: 81308



Hello to all!
I am interested in the further development of the hybrid mode.
Is it possible to integrate the psy acoustic model from musepack to wavpack.
Or possible to encode sound to make with no noise?
Thanks & sorry for broken English))


--------------------
MPC --quality 10 --tmn 20 --nmt 20 - %d || WV -miqhnb5x3 - %d
Go to the top of the page
+Quote Post
darkbyte
post Nov 10 2011, 21:46
Post #13





Group: Members
Posts: 107
Joined: 14-June 11
Member No.: 91517



QUOTE (Didjeridoo @ Nov 10 2011, 21:56) *
Is it possible to integrate the psy acoustic model from musepack to wavpack.

I'm hoping in a stronger integration with the LossyWAV project. It would be awesome to encode the lossy part with the ultraportable setting plus have a correction file for lossless use with the already well supported hybrid file format of WavPack.
Go to the top of the page
+Quote Post
Didjeridoo
post Nov 11 2011, 06:52
Post #14





Group: Members
Posts: 16
Joined: 8-June 10
Member No.: 81308



QUOTE (darkbyte @ Nov 10 2011, 22:46) *
I'm hoping in a stronger integration with the LossyWAV project. It would be awesome to encode the lossy part with the ultraportable setting plus have a correction file for lossless use with the already well supported hybrid file format of WavPack.

It would be nice.
Unless of course lossywav better musepack.


--------------------
MPC --quality 10 --tmn 20 --nmt 20 - %d || WV -miqhnb5x3 - %d
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 18th April 2014 - 12:53