IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
add "MP3+V" to the title of this forum
cbuchner1
post Oct 17 2001, 20:03
Post #1





Group: Members
Posts: 46
Joined: 30-September 01
Member No.: 82



so the board name will be "MP3/MP3Pro/MP3+V"

There's a new technology out there. It's called Plus-V (or MP3+V).

It follows a similar approach than MP3 Pro. It creates an MP3-compatible baselayer bitstream and adds another layer of proprietary high frequency coding. Of course this is NOT high quality.

What they do is a parametric approach to encode noise power in high frequency subbands and separately encoding additional harmonics.

I'd love to hear how it will sound. Check out their website.

http://www.plusv.org/intro.php

Unfortunately, they got patents on this.

[ This is a double-posting and also appeared on r3mix.net ;-) ]
Go to the top of the page
+Quote Post
TrNSZ
post Oct 17 2001, 21:31
Post #2





Group: Developer
Posts: 717
Joined: 25-September 01
Member No.: 20



What seems interesting is that the tools are implemented as a LAME wrapper set, and work externally of the encoder.

I'm going to encode some stuff here and in MP3Pro and see how they compare at 64kbps to 96kbps.
Go to the top of the page
+Quote Post
Garf
post Oct 18 2001, 11:30
Post #3


Server Admin


Group: Admin
Posts: 4886
Joined: 24-September 01
Member No.: 13



I had a little listen. It doesn't have the metallic sound of MP3Pro as much, but it makes much larger mistakes.

As far as I am concerned, it doesn't really provide an actual quality enhancement.

--
GCP
Go to the top of the page
+Quote Post
TrNSZ
post Oct 18 2001, 11:34
Post #4





Group: Developer
Posts: 717
Joined: 25-September 01
Member No.: 20



Also, at this point, I'm really unimpressed with the software in general, and wasn't able to get any serious testing done.

Maybe it will get better later.
Go to the top of the page
+Quote Post
Garf
post Oct 18 2001, 12:39
Post #5


Server Admin


Group: Admin
Posts: 4886
Joined: 24-September 01
Member No.: 13



QUOTE
Originally posted by TrNSZ
Also, at this point, I'm really unimpressed with the software in general,


They don't even have a decoder smile.gif

You need to generate the plus part from the original wav, encode that wav with lame at the bitrate you want, decode again, and then readd the plus part, producing the output wav.

Extremely primitive.

--
GCP
Go to the top of the page
+Quote Post
TrNSZ
post Oct 18 2001, 12:51
Post #6





Group: Developer
Posts: 717
Joined: 25-September 01
Member No.: 20



Yeah, so to playback the file, I wonder if the +V data can be demultiplexed on the fly, or if they are gonna have to scan the whole MP3 and save the +V data and then remultiplex with the decoded WAV during the output..

That would suck for a long (2 hour) MP3.. err, M3V.
Go to the top of the page
+Quote Post
leopold
post Oct 18 2001, 13:23
Post #7





Group: Members
Posts: 3
Joined: 18-October 01
Member No.: 315



PlusV doesn't require decoding the full bitstream before playback can start. The example programs do so only because they don't include MP3 encoders or decoders. This is done to avoid any hassle with MP3 patent holders and their license fees.

The PlusV data has been interleaved to the MP3 file in such a way that a MP3+V frame can be decoded immediately after it has been read, just as with plain MP3.

Yesterday (2001-10-17) a WinAmp PlusV input plugin has been submitted to http://www.winamp.com/ . When that plugin gets through WinAmp's screening process, you'll see that no intermittent files are needed. This can also be seen http://www.plusv.org/download/plusv.pdf
Figures 13 & 14 (MP3 Bitstream Organization and
MP3+V Bitstream Organization).

If you have any questions about PlusV, please ask them at http://www.plusv.org/ instead of here - at least if you want to have a proper answer. I noticed this question purely by luck, and it could well have been left unanswered.

Kind regards,
- Leopold
Go to the top of the page
+Quote Post
YouriP
post Oct 18 2001, 16:48
Post #8





Group: Members
Posts: 66
Joined: 22-September 01
Member No.: 7



I think this is an amiable effort in reproducing SBR. Any chances of seeing an implementation for AAC at a later time?
Go to the top of the page
+Quote Post
cbuchner1
post Oct 18 2001, 17:58
Post #9





Group: Members
Posts: 46
Joined: 30-September 01
Member No.: 82



QUOTE
Originally posted by leopold
If you have any questions about PlusV, please ask them at http://www.plusv.org/ instead of here - at least if you want to have a proper answer. I noticed this question purely by luck, and it could well have been left unanswered.


It's definitely nice to have a Plus-V developer here! I hope you don't mind if we discuss the format and its quality implications of plus-V here. ;-) This board was meant to be a place for discussing a variety of audio formats.

But it's good to know whom to ask when detailed technical questions arise.

Someone on the OGG-Vorbis mailinglist criticized that your library links to FFTW - which is GPL'ed software. Strictily speaking, you would have to make your software GPL as well. Maybe you better get an independent and free ("Public Domain") FFT implementation.
Go to the top of the page
+Quote Post
leopold
post Oct 19 2001, 16:32
Post #10





Group: Members
Posts: 3
Joined: 18-October 01
Member No.: 315



>It's definitely nice to have a Plus-V developer
>here! I hope you don't mind if we discuss the
>here. ;-) This board was meant to be a place for
>discussing a variety of audio formats.

No, I don't mind at all. Just notice that I have _tons_ of things to do and it may very well be that I don't see all the questions I might be able to help with.


>Someone on the OGG-Vorbis mailinglist criticized
>that your library links to FFTW - which is GPL'ed
>software.

That someone was right, and I am to blame. The download packages were removed immediately when I found out of the problem this morning. Now the packages are there again, but FFTW is replaced with a general, vanilla 512-point FFT. I had a little e-mail conversation with FFTW guys Steven G. Johnson and Matteo Frigo, and I must say that they were real gentlemen when it came to solving the issue. In addition to that, they gave me some invaluable advice regarding patents and licenses.

Thanks to them, I believe that we are going to be able to define our official "legalese" license policy by the end of the next week. Meanwhile: all free software developers will for certain be allowed to use the source codes and the format freely, as long as they don't intentionally break it ( = make it incompatible with itself).

Kind regards,
- Leopold
Go to the top of the page
+Quote Post
leopold
post Oct 19 2001, 16:39
Post #11





Group: Members
Posts: 3
Joined: 18-October 01
Member No.: 315



>I think this is an amiable effort in reproducing
>SBR. Any chances of seeing an implementation for
>AAC at a later time?

If someone makes an AAC implementation, then why not?

The only thing that must be made certain of is how to embed PlusV data to an AAC bitstrem. MP3 has a very convenient pointer that allows for the encoder to force the MP3 decoder to skip some data (See for the technical spec at http://www.plusv.org/ for details). This force-skipped data can then contain PlusV data. Everything is 100% compatible with vanilla MP3, old MP3-only decoder's don't see anything wrong with the bitstream, and generally everyone is happy.

Unfortunately I don't know AAC at all on the bit-level, so I cannot say if it has as convenient hooks where additional data could be put in a compatible way.

Kind regards,
- Leopold
Go to the top of the page
+Quote Post
Ivan Dimkovic
post Oct 19 2001, 21:13
Post #12


Nero MPEG4 developer


Group: Developer
Posts: 1466
Joined: 22-September 01
Member No.: 8



Menno (FAAD) and I thought about AAC implementation + V - Leopold please e-mail me on co.

Regarding AAC bitstream - in AAC you can put up to 256 bits in one frame via DSE (data stream element).
Go to the top of the page
+Quote Post
cbuchner1
post Oct 20 2001, 06:49
Post #13





Group: Members
Posts: 46
Joined: 30-September 01
Member No.: 82



The WinAmp plugin V1.01 for Plus-V is here.
http://mbnet.fi/~gridle/plusv.exe

It was posted by Santeri_Saarimaa on remix.net:
Re: New technology for improving fidelity at low rates
Reply #3 on: 10/19/01 at 09:23:41

what a shame this discussion is torn apart between two message boards wink.gif One developer here, one betatester there.
Go to the top of the page
+Quote Post
ff123
post Oct 20 2001, 07:16
Post #14


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



The plusv winamp plugin and other downloads can be found from the plusv website: http://www.plusv.org/

I have so far not gotten the encoder (either plusv.exe or lamepv_low.exe) to work. Oh, lamepv_low.exe will produce an .mpv file all right, but it shows no evidence of having any higher frequencies than 11 kHz at 64 kbit/s when played via Winamp using the plugin.

Plusv.exe -e doesn't encode a playable file.

Plusv.exe -d doesn't decode the sample file (dream.m3v) on the plusv.org website, nor does it decode a file made with lamepv_low.exe. Also, dream.m3v doesn't have frequencies higher than 8 kHz. Same as if you play it in winamp after renaming it to dream.mp3.

Highly frustrating.

ff123
Go to the top of the page
+Quote Post
cbuchner1
post Oct 20 2001, 09:06
Post #15





Group: Members
Posts: 46
Joined: 30-September 01
Member No.: 82



QUOTE
Originally posted by ff123
The plusv winamp plugin and other downloads can be found from the plusv website:  http://www.plusv.org/

Also, dream.m3v doesn't have frequencies higher than 8 kHz. 

I don't see the WinAmp plugin link working on the official site, yet.

Right. Winamp reports a sampling rate of 16kHz.

QUOTE
Originally posted by ff123
The plusv winamp plugin and other downloads can be found from the plusv website:  http://www.plusv.org/

Same as if you play it in winamp after renaming it to dream.mp3.

Highly frustrating.

ff123

Nope. After using the DiskWriter plugin with the MP3 and the M3V version, you will notice a difference (subtract both samples). Note the difference will be in the second half of that particular sample only.

But I can't hear the difference. And the MP3's at 48 kbit sound so bad...

More details in my posting on r3mix...
Go to the top of the page
+Quote Post
ff123
post Oct 20 2001, 12:42
Post #16


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



QUOTE
Nope. After using the DiskWriter plugin with the MP3 and the M3V version, you will notice a difference (subtract both samples). Note the difference will be in the second half of that particular sample only.


Yes, I saw the differences in the mp3 vs. the m3v using dream.m3v. However, I did not see any frequencies higher than 8 kHz. Diskwriter is not DirectSound, so that takes care of that argument. So far, you, Wish, and I have not been able to make any of this stuff work.

ff123
Go to the top of the page
+Quote Post
Wish
post Oct 20 2001, 23:57
Post #17





Group: Members
Posts: 206
Joined: 25-September 01
Member No.: 19



http://mbnet.fi/~gridle/in_plusv.zip

The other plugin was a broken version. With this one, it does report the sample m3v from PlusV.org as 32khz instead of 16khz. Seeking is slow though. And the time display is messed up also, it changes from 1:19 to 1:22. Seriously, please debug and fix all these problems first before releasing any plugin/encoder/decoder. Ugh.

I didn't hear any difference with the new plugin. Maybe someone else can.
Go to the top of the page
+Quote Post
cbuchner1
post Oct 21 2001, 03:05
Post #18





Group: Members
Posts: 46
Joined: 30-September 01
Member No.: 82



QUOTE
Originally posted by Wish


The other plugin was a broken version. With this one, it does report the sample m3v from PlusV.org as 32khz instead of 16khz


Yes, the new version indeed reports 32 kbit. However, it does not come with the nice Pimp Installer or a README file. Kick me. I still don't hear a difference in fidelity.

I think the plusv-guys should address the issue of getting a decent quality 56 kbit MP3 "base layer" first. Try using Fraunhofer's FastEnc, for example.
Go to the top of the page
+Quote Post
YouriP
post Oct 21 2001, 19:28
Post #19





Group: Members
Posts: 66
Joined: 22-September 01
Member No.: 7



Well, these are probably just some growing pains that plague all new software and will probably be fixed sooner than we think. After all, this idea isn't more than a few weeks old (publicly at least), so we should give them some time.
Go to the top of the page
+Quote Post
xiphmont
post Oct 22 2001, 01:20
Post #20


Xiph.org


Group: Developer
Posts: 176
Joined: 24-September 01
Member No.: 16



I visited the www.plusv.org forums because I had several questions to ask about the license (ie, saying "You don't need a license... it's free!" doesn't wash, especially with 'patent pending' displayed prominently), but the 'submit post' button takes me to a blank page. I sent mail asking for help, but no response yet...

(Don't worry, I know my own email management skills are rather poor when I'm busy, no offense taken ;-) I'm mentioning it mainly to improve the chances one of the plusv people will notice.
Go to the top of the page
+Quote Post
jianxin yan
post Oct 22 2001, 14:09
Post #21





Group: Developer
Posts: 73
Joined: 8-October 01
Member No.: 245



hi.
you'd better register at first. than you will get an email which include a password for logging in. and than you can post a new thread, otherelse you can only reply after old threads.


regards.
Go to the top of the page
+Quote Post
xiphmont
post Oct 26 2001, 06:30
Post #22


Xiph.org


Group: Developer
Posts: 176
Joined: 24-September 01
Member No.: 16



QUOTE
Originally posted by jianxin yan
hi.
you'd better register at first. than you will get an email which include a password for logging in. and than you can post a new thread, otherelse you can only reply after old threads.


regards.


Yes, I know, I registered. The forum didn't work. No error, no message, just a blank page (and my message wasn't there when I went to look for it).
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: 21st October 2014 - 06:31