Help - Search - Members - Calendar
Full Version: ReplayGain For Ogg Vorbis?
Hydrogenaudio Forums > Lossy Audio Compression > Ogg Vorbis > Ogg Vorbis - General
Lawless
Does anybody know of a utility that can be used to apply ReaplyGain to ogg files? Is such a utilty even possible or being developed? I'm personally concerned about using ogg whilst such an important functionality is not available...
Garf
QUOTE
Originally posted by Lawless
Does anybody know of a utility that can be used to apply ReaplyGain to ogg files? Is such a utilty even possible or being developed? I'm personally concerned about using ogg whilst such an important functionality is not available...


Such an utility does not exist nor do I know of anyone making one.

However, it's perfectly possible to make one, if you want to.

There has been discussion before in other threads about ReplayGain support for Ogg, so you might want to browse around a little.

BTW. Surprised this is 'important functionality'. MPC even only got this very recently.

--
GCP
Lawless
QUOTE
Originally posted by Garf


BTW. Surprised this is 'important functionality'. MPC even only got this very recently.



Hi Garf,
I believe you answered my same question on the r3mix forum as well, thanx for the reply! I guess a better statement would have been 'important functionality for me'. If I had not stumbled across the ReplayGain site, I would not have had gave it a second thought, but the concept of ReplayGain struck me as one of those great ideas, that is quite simple, yet no one had thought to implement it before. That said, adjusting the volume on the stereo between discs is not exactly the end of the earth :-) But it is a function that many users would greatly appreciate.

It's particulary useful for me, as I'm considering putting on an 80's hardrock / heavy metal night & I was thinking it would be a good idea to encode all of my CDs & connect the PC up to the PA & using playlists for a good half of the night, instead of lugging 100's of CDs to the venue & flipping CDs all night. As you can imagine, this could involve playing tracks from 100's of CDs & hence a possible volume prob. Remasted sources would be too loud, turn it down & old CDs would be too low...

I'm sure ReplayGain will become part of Vorbis eventually, but unless there is an offical field part of Vorbis such as Global Gain for MP3, it might not be 'offically' possible to reliably implement it. If no such field is available, implementing it would require a modified plugin (such as MP+?) & would not work in portable devices. I could be wrong, as I've only started looking into the whole audio compression thing, as I've got 400+ CDs to encode, so I'm really trying to puzzle out which is the best format & settings for encoding, as I really dread the thought of having to re-encode that number of CDs a 2nd time!

It's a real shame that people don't flock to the best quality / value products (mp+ & ogg?), I mean there's even money change WMA will become a defacto standard reguardless of it's quality, just because Microsoft has a large degree of control over the end user market...
Garf
This is what Monty had to say about it:

http://www.hydrogenaudio.org/forums/showth...ight=ReplayGain

As far as I understand, this should not be much harder than writing something that decodes ogg's, applies a set of filters and some more math magic to get the gain values, and writes them to a tag.

In the decoding apps, check for the tag, and multiply everything by the correct value.

Sounds simple enough that I could probably manage it myself, with a little help. (perhaps after the exams?)

If anyone is looking for a spare project to waste time on, this is a nice pick.

Fixing XMMS to support this should be no problem, and if you talk to Peter P. who does the WinAmp plugins I think he could get WinAmp to support it too. That takes care of all important apps. (read: the ones I use)

--
GCP
Garf
giancarlo@garf:~/replaygain > ./replaygain < smack.ogg

Bitstream is 2 channel, 44100Hz

Decoded length: 1487744 samples
Encoded by: Xiphophorus libVorbis I 20010813 (Experimental GT1)

Initializing Gain Analysis...
Recommended Title Gain: +5.35 dB
Done.

(this was a quiet track)

Well, looking good so far smile.gif

--
GCP
Garf
CODE


giancarlo@garf:~/replaygain > ./replaygain < ref_pink.ogg



Bitstream is 1 channel, 44100Hz



Decoded length: 220500 samples

Encoded by: Xiphophorus libVorbis I 20011231



Initializing Gain Analysis

Analyzing...



Recommended Title Gain:  +6.02 dB       Peak Level:       9523 (0.291)

              Scaling:   2.00 x        New Peak Level:  19046 (0.581)

Done.


--
GCP
Scooby-Doo
Wow, you already whipped all of that up? Good going Garf, you keeping making me love Vorbis even more.
NickSD
QUOTE
Originally posted by Garf

BTW. Surprised this is 'important functionality'. MPC even only got this very recently.
-- 
GCP


I was under the impression that ReplayGain was the preferred way to eliminate clipping. Wouldn't that be considered important?
Garf
QUOTE
Originally posted by NickSD


I was under the impression that ReplayGain was the preferred way to eliminate clipping.  Wouldn't that be considered important?


Hmm, I remember Monty explaining once that the ogg encoder had built in clipping prevention. I'm not aware of the details though.

ReplayGain support is as good as done now. (only need to make sure the tags aren't written twice)

--
GCP
NickSD
QUOTE
Originally posted by Garf


Hmm, I remember Monty explaining once that the ogg encoder had built in clipping prevention. I'm not aware of the details though.



True - I'm not so sure if clipping is a problem with Vorbis. I guess I was referring more to MPC - sorry smile.gif

QUOTE
Originally posted by Garf

ReplayGain support is as good as done now. (only need to make sure the tags aren't written twice)



Thanks! That's great!
Garf
[img]c190338e21[/img]
[img]c190338e21[/img]

I have a new plugin for XMMS that has full ReplayGain support.

--
GCP
2Bdecided
I don't know - nothing happens for six months, and then TWO ogg implementations of ReplayGain come along at once!

Yes, that's right, someone else emailed me today to say they'd implemented ReplayGain in Ogg!

Anyway, Garf, can I have the source code for what you've done please? It'll be meaningless to me (since I can't program and only run windows 95), but I'm certain that people would be interested, if you'd let me put it up at replaygain.org? Or provide a link? Or are you not ready to release it yet?

Cheers,
David.
EDIT: I forgot to say - great work Garf - that looks really cool!
EDIT2: please reply here:
http://www.hydrogenaudio.org/forums/showth...?s=&postid=5913
krsna77
@Garf: I suppose that I understood that implementing replaygain in OGG should be relatively uncomplicated in principle...

But to completely implement it successfully in ONE DAY? Nifty.

As I sit here struggling to finish "C++ Primer Plus"... wink.gif
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.