Hydrogenaudio Forums > Musepack (MPC) SV8 beta is out

kwanbis

Sep 25 2007, 19:12

QUOTE(michtar @ Sep 25 2007, 23:57)

Is there any way to batch convert from SV7 to SV8? BTW, using * in output name results in a crash.

just in case, devs recommended not to use it just yet, for normal use.

mr budzone

Sep 27 2007, 14:58

Phreakin awesome!
Musepack <3
I wonder what Andre Buschmann is doing nowadays

Keep up the good work

12345

Sep 28 2007, 08:07

Congrats! This is really great to see. It must have been already way back in 2002 that people were talking about this SV8 thing . It's nice to know that Musepack is still being developed. I still have about 300 records in --insane.

Antonski

Sep 29 2007, 14:06

QUOTE(WaldoMonster @ Sep 24 2007, 22:44)

I will repeat my findings with mpcdec.
mpcdec doesn't decode to stdout but writes to the file: -

CODE

mpcdec test.mpc - | …

Is decoding to stdout not implemented yet?

According to the timeline, pipe decoding was implemented on 27th September, however the link to the windows package still points to the one from 23th September.
Maybe somebody can compile the latest source?

indybrett

Oct 1 2007, 09:30

Been searching the forums for the old high bitrate listening tests. Granted, Vorbis and LAME have improved since then. Does anybody have a link to a more recent test than this one?

gaekwad2

Oct 1 2007, 11:33

QUOTE(indybrett @ Oct 1 2007, 17:30)

Been searching the forums for the old high bitrate listening tests. Granted, Vorbis and LAME have improved since then. Does anybody have a link to a more recent test than this one?

This one, a year later.

indybrett

Oct 1 2007, 11:40

QUOTE(gaekwad2 @ Oct 1 2007, 13:33)

QUOTE(indybrett @ Oct 1 2007, 17:30)

Been searching the forums for the old high bitrate listening tests. Granted, Vorbis and LAME have improved since then. Does anybody have a link to a more recent test than this one?

This one, a year later.

Thanks. That's the one I last remembered reading, but couldn't find it.

razer

Oct 3 2007, 11:18

Fantastic news.

novocane

Oct 8 2007, 17:08

I knew this day would come...the "eternal awaited" SV8

vinnie97

Oct 9 2007, 21:20

QUOTE(indybrett @ Oct 1 2007, 09:40)

QUOTE(gaekwad2 @ Oct 1 2007, 13:33)

QUOTE(indybrett @ Oct 1 2007, 17:30)

Been searching the forums for the old high bitrate listening tests. Granted, Vorbis and LAME have improved since then. Does anybody have a link to a more recent test than this one?

This one, a year later.

Thanks. That's the one I last remembered reading, but couldn't find it.

More proof that has failed to quench the seemingly widely held belief that "MPC is the most transparent lossy format at 160 kbps and above" unfortunately. Calling them anything other than equal would be disingenuous, IMO.

indybrett

Oct 10 2007, 09:41

QUOTE(vinnie97 @ Oct 9 2007, 23:20)

QUOTE(indybrett @ Oct 1 2007, 09:40)

QUOTE(gaekwad2 @ Oct 1 2007, 13:33)

QUOTE(indybrett @ Oct 1 2007, 17:30)

Been searching the forums for the old high bitrate listening tests. Granted, Vorbis and LAME have improved since then. Does anybody have a link to a more recent test than this one?

This one, a year later.

Thanks. That's the one I last remembered reading, but couldn't find it.

More proof that has failed to quench the seemingly widely held belief that "MPC is the most transparent lossy format at 160 kbps and above" unfortunately. Calling them anything other than equal would be disingenuous, IMO.

I don't think anybody is claiming MPC superiority anymore. At least, not in this forum. If they are, I haven't seen it.

greynol

Oct 10 2007, 11:16

QUOTE(indybrett @ Oct 10 2007, 08:41)

I don't think anybody is claiming MPC superiority anymore. At least, not in this forum. If they are, I haven't seen it.

http://www.hydrogenaudio.org/forums/index....st&p=518844
http://www.hydrogenaudio.org/forums/index....mp;#entry516550

indybrett

Oct 10 2007, 11:31

QUOTE(greynol @ Oct 10 2007, 13:16)

QUOTE(indybrett @ Oct 10 2007, 08:41)

I don't think anybody is claiming MPC superiority anymore. At least, not in this forum. If they are, I haven't seen it.

http://www.hydrogenaudio.org/forums/index....st&p=518844
http://www.hydrogenaudio.org/forums/index....mp;#entry516550

In the first thread you linked I didn't see anyone claim MPC superiority, only one person joking about "bowing down to MPC". In the 2nd, I saw one poster who claimed it was better than MP3.

greynol

Oct 10 2007, 11:36

QUOTE(indybrett @ Oct 10 2007, 10:31)

In the first thread you linked I didn't see anyone claim MPC superiority. In the 2nd, I saw one poster who claimed it was better than MP3.

Now bowdown to MPC! (Well, I thought it funny at least. It was posted shortly after he finished with the other thread.)

You need to read through the other one to see that he also suggests that MPC is superior to AAC and seems to prefer it over OGG as well.

indybrett

Oct 10 2007, 11:44

QUOTE(greynol @ Oct 10 2007, 13:36)

QUOTE(indybrett @ Oct 10 2007, 10:31)

In the first thread you linked I didn't see anyone claim MPC superiority. In the 2nd, I saw one poster who claimed it was better than MP3.

Now bowdown to MPC! (Well, I thought it funny at least. It was posted shortly after he finished with the other thread.)

You need to read through the other one to see that he also suggests that MPC is superior to AAC and seems to prefer it over OGG as well.

Yeah, I didn't see the the Ogg part. Oh well... Most of us know

greynol

Oct 10 2007, 11:52

He's ok with ogg's quality, just that mpc was better for his purposes.

Anyhow, I thought it was funny. Sorry for going off-topic.

indybrett

Oct 17 2007, 09:30

Just curious how the beta is doing. Would like to see this out of beta, and support in Rockbox for the new stream version.

xmixahlx

Oct 17 2007, 11:03

there has been some upstream dev since the beta to address some of the issues brought up here like:
* piped i/o for mpcdec
* fix for id3v2 seek bug
* mpc2sv8 output options (overwrite with -o and to directory)
* quantizer change (also helps in rockbox integration)

and preglow will be working on the rockbox changes when he is able.

later

indybrett

Oct 17 2007, 11:32

Great news, thanks for the update.

WaldoMonster

Oct 17 2007, 13:41

QUOTE(xmixahlx @ Oct 17 2007, 19:03)

there has been some upstream dev since the beta to address some of the issues brought up here like:
* piped i/o for mpcdec
* fix for id3v2 seek bug
* mpc2sv8 output options (overwrite with -o and to directory)
* quantizer change (also helps in rockbox integration)

and preglow will be working on the rockbox changes when he is able.

later

Nice work.
Are there win32 compiles of these updates available?
If so I will do some testing with the pipe output of mpcdec.

xmixahlx

Oct 17 2007, 15:37

no - but the windows project files are updated so you can checkout and compile at a whim!

later

ProtectYaNeck36

Oct 20 2007, 10:48

wow, I never thought I would see the day sv8 came out! time for another listening test guru?

yandexx

Oct 20 2007, 10:53

There are no changes to psychoacoustic model, as stated. So, no tests needed, at least for now.

ProtectYaNeck36

Oct 20 2007, 11:00

heh, yeah I just read that, but you replied before I could edit my post

WaldoMonster

Nov 26 2007, 14:50

QUOTE(xmixahlx @ Oct 17 2007, 18:03)

there has been some upstream dev since the beta to address some of the issues brought up here like:
* piped i/o for mpcdec
* fix for id3v2 seek bug
* mpc2sv8 output options (overwrite with -o and to directory)
* quantizer change (also helps in rockbox integration)

and preglow will be working on the rockbox changes when he is able.

later

Decode to stdout still not working with the latest binaries.
Is there a special command for stdout or can I use - ?

http://r2d.musepack.net/files/musepack_bin_20071123.zip

Antonski

Nov 26 2007, 17:09

QUOTE(WaldoMonster @ Nov 26 2007, 22:50)

Decode to stdout still not working with the latest binaries.
Is there a special command for stdout or can I use - ?

http://r2d.musepack.net/files/musepack_bin_20071123.zip

It works fine for me. Yes, you should use - for stdout.

CODE

D:\Util\MusePack\sv8>mpcdec.exe -h
mpcdec - Musepack (MPC) decoder v0.9.2 (C) 2006-2007 MDT
Built Nov 23 2007 16:27:47
Usage: mpcdec.exe [-i] [-h] <infile.mpc> [<outfile.wav>]
-i : print file information on stdout
-c : check the file for stream errors
(doesn't fully decode, outfile will be ignored)
-h : print this help
you can use stdin and stdout as resp. <infile.mpc> and
<outfile.wav> replacing the file name by "-"

WaldoMonster

Nov 26 2007, 18:30

CODE

E:\Console\codec>mpcdec input.mpc - | lame -V5 --vbr-new --quiet - output.mp3
mpcdec - Musepack (MPC) decoder v0.9.2 (C) 2006-2007 MDT
Built Nov 23 2007 16:27:47
2446080 samples decoded in 342 ms (162.18x)

This wil create a mp3 file with very much noise.
Decoding directly to a wav file is working without any problem.

Also decoding without transcoding creates a wav file with very much noise.

QUOTE

E:\Console\codec>mpcdec input.mpc - > output.wav
mpcdec - Musepack (MPC) decoder v0.9.2 © 2006-2007 MDT
Built Nov 23 2007 16:27:47
2446080 samples decoded in 405 ms (136.95x)

xmixahlx

Nov 26 2007, 19:36

stdout works for me on debian - both output to file and to pipe.

and since others on windows have verified this there must be something wrong with your situation specifically.

later

WaldoMonster

Nov 27 2007, 03:24

When trying decoding to stdout on another system it gave me the error message that msvcr71.dll was missing.
After copying msvcr71.dll to the same directory as mpcdec.exe it gives the same disturbed output as on my other computer.

PS
Decoding to file works perfectly on both systems.

Antonski

Nov 27 2007, 07:11

QUOTE(WaldoMonster @ Nov 27 2007, 02:30)

CODE

E:\Console\codec>mpcdec input.mpc - | lame -V5 --vbr-new --quiet - output.mp3
mpcdec - Musepack (MPC) decoder v0.9.2 (C) 2006-2007 MDT
Built Nov 23 2007 16:27:47
2446080 samples decoded in 342 ms (162.18x)

This wil create a mp3 file with very much noise.

Now I see.
Verified here with 3 versions of lame as well as with Ogg Vorbis. Same results (noise).
~

robert

Nov 27 2007, 07:46

I haven't looked at it yet and might do it later this evening, but what's the output format of mpcdec when writing to stdout? wave or raw pcm? What LAME version are you using?

robert

Nov 27 2007, 15:18

Ok, seems a simple mpcdec x.mpc - >x.wav results in noise too. Maybe some line ending translation is going on here? (LF -> CR-LF)

Another small issue: the mpcenc seems to disturb my console color setting.

Niwatori

Dec 1 2007, 17:51

Years of wait... thanks, now it nearly dream will be ture...

( missing Hydrogenaudio.. )

xmixahlx

Dec 1 2007, 23:17

i just noticed - r2d fixed the stdin/out for y'all windows folks in SVN.

later

Antonski

Dec 2 2007, 17:33

QUOTE(xmixahlx @ Dec 2 2007, 07:17)

i just noticed - r2d fixed the stdin/out for y'all windows folks in SVN.
later

Confirmed

mpcdec (build 20071201)
~

WaldoMonster

Dec 2 2007, 17:55

QUOTE(Antonski @ Dec 3 2007, 00:33)

QUOTE(xmixahlx @ Dec 2 2007, 07:17)

i just noticed - r2d fixed the stdin/out for y'all windows folks in SVN.
later

Confirmed

mpcdec (build 20071201)
~

Thanks for the updated binaries.
The first tests are looking promising.
Also the terminate bug (link below) from the previous 1.95e decode is fixed.
http://www.hydrogenaudio.org/forums/index....showtopic=54889