IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
OGG & Replay Gain
krsna77
post Dec 8 2001, 23:44
Post #1





Group: Members
Posts: 74
Joined: 29-September 01
Member No.: 35



With all this talk abt Musepack gaining new found support for Replay Gain (which is really cool IMO), I was wondering if such a thing is planned for Ogg Vorbis (or even possible), and if so, is anyone currently working on it, or just thinking abt it?... smile.gif

I would really LOVE to see Replay Gain for Ogg Vorbis (I only wish my coding skills were enough to help this come to fruition...).

Just curious. smile.gif
Go to the top of the page
+Quote Post
2Bdecided
post Dec 10 2001, 13:17
Post #2


ReplayGain developer


Group: Developer
Posts: 5266
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



It's possible - and since Frank has implemented it for mpc and released the source code I would expect the task to be comparatively easy.

Is anyone going to do it? No idea, but anyone who wants to is welcome to try.

Cheers,
David.
http://www.replaygain.org/
Go to the top of the page
+Quote Post
JohnV
post Dec 10 2001, 14:59
Post #3





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



Question is, does Vorbis use similar kind of global gain factor like MP3.

I'm not sure if MPC kind of approach is possible (any more) with Vorbis, since the decoding part is said already to be finalized. MPC doesnt have MP3 like global gain factor, instead Frank implemented gain adjustment by modifying file header and decoder. I believe this will be part of the MPC SV8 bitstream specs, so that SV8 compliant decoders can take advantage of gain value modification values.

But how is the situation with Ogg Vorbis?


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
ErikS
post Dec 10 2001, 15:35
Post #4





Group: Members
Posts: 757
Joined: 8-October 01
Member No.: 247



QUOTE
Originally posted by JohnV
But how is the situation with Ogg Vorbis?


it should be possible/easy to make a tag with name "REPLAYGAIN" and make special decoder that reads this value and modifies output. Anyone here who feels like doing this?
Go to the top of the page
+Quote Post
JohnV
post Dec 11 2001, 10:02
Post #5





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



Sure, of course with special decoder anything is possible. But the point is whether it can be made official or not...

Every mp3-decoder is compatible with mp3 gain modification. I believe every SV8 MPC-decoder will be compatible with MPC gain modification.

But will Vorbis gain modifications only be compatible with few specific decoders, and so for example possible hardware implementations may lack it?


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
2Bdecided
post Dec 11 2001, 10:18
Post #6


ReplayGain developer


Group: Developer
Posts: 5266
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



I realise the vorbis decoder is "finalised", but how many people are using vorbis now compared to how many are hoped to use it in the future? Wouldn't now be a good time to add this to the decoder? Surely it can be added in a way that doesn't break compatibility?

We haven't even got a v1.0 final encoder yet - surely the "new" decoder (with ReplayGain) could ship with that version of the encoder (which everyone will download) when it arrives.

I think the ReplayGain support for vorbis should work exactly as with mpc - this is how I always intended it to work - mp3gain is simply a nice hack to allow people to use ReplayGain with players that don't support it. The original concept is described fully on
http://www.replaygain.org/


I took ReplayGain to the vorbis developer list when I first published the idea, but nothing came of it. Maybe now it's been accepted for mp3 and mpc, this is a good idea to mention it again. I'll leave that to someone who regularly contributes.

Cheers,
David.
http://www.David.Robinson.org/
Go to the top of the page
+Quote Post
JohnV
post Dec 11 2001, 19:56
Post #7





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



Well, Monty commented this:
<xiphmont> I like ReplayGain (the idea) I want to use it myself. Someone else is going to have to implement support. It belongs int he apps, not the codec libs.

Basically this means there's not gonna be codec level support for Vorbis' ReplayGain, and every (player) application have to implement it separately (not likely at least for future hardware players).


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
2Bdecided
post Dec 12 2001, 09:57
Post #8


ReplayGain developer


Group: Developer
Posts: 5266
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



If a place isn't designed for it in the bitstream, and if there's no global gain value with which to "bodge" it, then it won't happen.

So, if that's a definate "no" to defining a place for the data in the bitstream, then Vorbis support of ReplayGain is ruled out. Relying on individual (incompatible) player support has obvious disadvantages.


The latest Media Jukebox (still in BETA test) includes ReplayGain, which can be used with ALL the file types it handles. The ReplayGain data is stored within the Media Jukebox database, so it is not included within the audio files themselves. This approach means that standard accepted gains cannot be added to the files "at source", which is a pity, since this is one of the reasons I created ReplayGain: to ultimately encourage content providers to include this valuable meta data with their recordings. It shouldn't be the responsibility of users to fix levels - the recording studios should embed the ReplayGain data on CD (or DVD-A). Just a pipe dream, but I can hope...

Cheers,
David.
Go to the top of the page
+Quote Post
JohnV
post Dec 12 2001, 15:36
Post #9





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



Yeah, in my opinion even though the bitstream is "final", there could be bitstream extension. Old bitstream compliant decoders wouldn't understand the gain adjustement (like it's the case with mpc), but newer would. It wouldn't have to brake anything but would make it able the use the gain adjustment.

It's really unfortunate if Vorbis will lack support for ReplayGain. Maybe some people like Garf could try to influence Monty's opinions...


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
Garf
post Dec 12 2001, 17:15
Post #10


Server Admin


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



>Maybe some people like Garf could try to influence Monty's
>opinions...

He listens to reasonable arguments, so I would just recommend you guys to write a well-argumented post to the mailinglist.

--
GCP
Go to the top of the page
+Quote Post
JohnV
post Dec 12 2001, 18:33
Post #11





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



Well, he has read this thread already, and commented it (xiphmont quote)...


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
xiphmont
post Dec 18 2001, 08:57
Post #12


Xiph.org


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



Hi,

I don't object either to ReplayGain, or even an effort to encourage it. I do have reasons to avoid mandating its use, or making it an absolutely necessary official piece of Vorbis. I'll go into them if needed.

I have a few questions; is the exact technique ReplayGain uses to do limiting/compression documented in an easy/reliable place I can link? I know the general technique is well understood, I'm more interested in knowing if Joe Random hacker can go out, find all the needed, gritty details and implement entirely correct ReplayGain support in his own player apps himself. I know I'm showing some research laziness by needing to ask the question.

If that's possible, I'm happy to encourage ReplayGain support through an 'official' tag (as official as a tag can be in our informal system; it will get a spot in the spec) and mention that app developers should consider supporting it.

Monty
Go to the top of the page
+Quote Post
2Bdecided
post Jan 11 2002, 16:14
Post #13


ReplayGain developer


Group: Developer
Posts: 5266
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



I missed this post - sorry.

There isn't an "official" limiter described on the Replay Gain site. The limiter only kicks in if a user insists on pushing the levels up too high and introduces clipping.

However, everything else is fully documented - and at least four people have used the information from the site to make independent Replay Gain implementations.

Garf seems to have introduced an RG tag into Ogg - will this become official? Can you send me the specifics please Garf?

Cheers,
David.
EDIT: please reply here:
http://www.hydrogenaudio.org/forums/showth...?s=&postid=5913
(sorry for bumping up an old thread)
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: 24th November 2014 - 15:18