Replaygain code

[MrSinatra]

splitting this off from elsewhere...

but I am curious about the RG plugin.

in an older ver of winamp, it used to list this link:

http://www.nullsoft.com/free/ReplayGain

but nothing ever showed up there. even now credit for the plugin is [seemingly] given to:

David Robinson and Glen Sawyer, 2001

but it is unclear if the credit is for the plugin, or the code the plugin is based on.

(David Robinson developed the RG standard)

DrO says:

the percentage bug has been mentioned in another thread or two from memory (i think one of the bug reports - not 100% sure without checking my bug list archive), but it's more from non-forum methods that we've been made aware of it.

as for "the RG plugin is supposedly open source" - it's the library used which is just a straight compile of the source code for that library which is open source and as it's not been modified, there is no need to provide what is already available. unless you're mixing up and expecting ml_rg.dll to be open source as well, if so then that's not possible (and would somewhat defeat the point of having paid to get it).

which confuses the hell out of me.

so the code to do RG is open source, but the plugin itself is not, and the plugin was paid for by winamp to someone apparently?

anyway, my point is that now might be a good time to update the RG plugin, and to make the whole thing totally opensource. it would allow others to find elusive bugs as well as allow the progress in RG implementation, and it shouldn't necessitate any expenditures.

r128 seems to be the newest, and best way, to calculate RG values and is becoming the new de facto standard.

opensourcing could allow the plugin to give users options they don't have now, like picking which implementation to use, adding soundcheck tags, and so on.

so I guess I don't understand why the RG can't or shouldn't be totally opensource?

http://en.wikipedia.org/wiki/Replaygain

http://www.nullsoft.com/free/mpc (this was something else I never saw the docs on)

http://www.nullsoft.com/free/ (I took "free" to mean open sourced)

http://www.nullsoft.com/

[DrO]

you're really mixing up things.

ml_rg.dll is the Winamp coded plug-in that uses ReplayGainAnalysis.dll and the open source code was what makes ReplayGainAnalysis.dll. i can't remember what was meant to be on the nullsoft.com site (archive.org has nothing that i can find for it).

so the credit is for the code that makes up ReplayGainAnalysis.dll (which is freely available as gain_analysis.c and gain_analysis.h). and yes there was the old link but as it was never properly setup and seeing as nullsoft.com went done ages back, it was removed as it was broken, and so only the text reference remained on the preference page.


as for open sourcing native Winamp plug-ins (as ml_rg is), in this specific case it's not known what causes the odd percentage issue and for we know it's an issue in winamp.exe or a slew of things causing it - which if only ml_rg is open sourced would not help.

sadly the whole 'open source fixes all issues' view is flawed as you need people to know and understand the code even if it was freely available. Winamp is closed source and is pretty much going to stay that way afaik.


and with respect to supporting R128, there is absolutely nothing stopping someone from having implemented a plug-in to do it already. and i'm more than aware it would be nice to have that in Winamp natively, but considering everything that has gone on and is still going on, there's a lot of things that can and probably should be added / changed in Winamp, but there are far bigger issues to be resolved such as even being in a position to have a feature compatible Winamp client post-AOL.

so for the time being, other than tracking down and fixing the bug, there's little else to update in the existing plug-in as there's no changes to the replaygain spec (that's why there's now R128 as it's simpler to provide a new spec instead of altering an old one and leading to a mis-match in implementation of the same feature) and there's no need to open-source it.

[MrSinatra]

I don't disagree there are bigger fish to fry, so just consider this convo academic:

but one thing I think you are missing, is that if the RG plugin was opensource, it could at least allow others to potentially eliminate it as the cause of the bug, which frankly is a pretty worthy cause, imo, b/c then you would look elsewhere for the bug.

it could also allow others to provide patches and fixes and updates, so the devs could concentrate on those bigger fish. a guy like TT could say "here you go" and you guys could have a r128 option. an intrepid 3rd party dev could even figure out how to convert those values to soundcheck values, so those kinds of tags could be added as well. remember the mp4/m4a bug a ver of RG introduced? that could have been avoided.

I am interested in what you said about that too... are you saying the tags for r128 would be separate and distinct from original RG tags? i assumed r128 tags would just overwrite existing tags as is. I don't know enough about r128 spec to know how its supposed to be setup/implemented. i can see pros/cons to each way, meaning separate tags vs overwriting tags.

I think its on the user however, to know if RG is using 83db, 89db, or r128, meaning, which exact implementation of RG. its important to have consistency on playback / across all your files, but that is the responsibility of the user.

anyway, i guess I'm surprised that an app like winamp doesn't want to try to get the possibility of help, by opensourcing more of its native functions / plugins like RG, esp since the dlls are open source. doesn't it benefit devs to offload the work potentially? i guess i just don't see the argument for keeping the plugin part closed?

[DrO]

i'm just not going to go down this open-source route you're pushing. it was was done to death when the shutdown notice went up and i'm plain tired of going over the same points ad-nausea.

and i'm still waiting to see where all of those wonderfully enthusiastic coders who wanted to save Winamp are... as i've seen absolutely sod all in an uptick in user-generated plug-ins (which if there are so many people around who wanted to save Winamp and all that, that's what they should be working on providing).

and the issue which this has all been triggered by is not in ReplayGainAnalysis.dll so having the code (which i've already provided what needs to be googled for) would not help. and with that, i'm leaving this thread (as this is one of those moment where i wish i was more like the old dev team and just didn't pay attention to the forums).

[MrSinatra]

well, I'm now only talking about the plugin, not winamp, so that's different. and the plugin is based on open source DLLs which you stated, so its hardly a big reach either.

not sure why that irritates you so much, since i think i made solid points here, but i have np that you disagree, i just wish that you would not get all huffy when disagreeing.

something else 3rd party folks might be able to contribute, would be multi-threading so when you do whole MLs of files its faster.

but anyway, I'm still left wondering: i don't see the argument for keeping the plugin part closed?

[MrSinatra]

and i'm still waiting to see where all of those wonderfully enthusiastic coders who wanted to save Winamp are... as i've seen absolutely sod all in an uptick in user-generated plug-ins (which if there are so many people around who wanted to save Winamp and all that, that's what they should be working on providing).

that was added during my post, so i'll just point out that TT, a 3rd party dev, has expanded winamp by adding opus format support. without his effort, its unlikely winamp would currently support opus.

that's just one example.

all i am trying to say, without it being argumentative, is that i think TT (and others like him) would appreciate it if more of the code that can be open sourced, was open sourced, so that when a bug is reported to him, he could check it out. he could then contact the devs and say 'i looked at the RG plugin, but did or did not see xyz,' which potentially could save the devs a lot of time. he may then even get motivated to otherwise improve / expand the RG plugin, or others that are open, and suggest patches for you all to review/include.

i don't know why that's so aggravating to you, but please believe me i do not wish to aggravate you. its just what i think.

[Aminifu]

all i am trying to say, without it being argumentative, is that i think TT (and others like him) would appreciate it if more of the code that can be open sourced, was open sourced, so that when a bug is reported to him, he could check it out. he could then contact the devs and say 'i looked at the RG plugin, but did or did not see xyz,' which potentially could save the devs a lot of time. he may then even get motivated to otherwise improve / expand the RG plugin, or others that are open, and suggest patches for you all to review/include.

That is true. It is also true that many devs do not want to give away the fruits of their labors. Things hard won thru years of learning and experimentation. Intellectual property is personal property. We all are protective of our personal property. Also, in the US, we respect personal property rights and are not allowed to give away or seize the property of others without compensating them.

At some point certain things pass into the public domain. But that is a whole other discussion as to what things and when.

[MrSinatra]

and many do want to give it away, b/c its not a big deal. look at TT and opus. look at all the 3rd party free code given to squeezebox by its fans. the point isn't who doesn't want to give it away, the point is appealing to those who do.

I also suggested btw, that a winamp store would be a good way to encourage 3rd party dev for those who DON'T want to give it away. I'm for both approaches. I don't know why you feel the need to say things like "in the us blah blah blah" when that's obvious and not really speaking to this limited issue, and totally ignoring the whole thrust of the wildly successful open source movement.

[Aminifu]

I don't know why you feel the need to say things like "in the us blah blah blah" when that's obvious and not really speaking to this limited issue, and totally ignoring the whole thrust of the wildly successful open source movement.

It's not a big deal to you. Code I developed on my own was a big deal to me.

I added that comment because Winamp is used worldwide and supported by people worldwide and people in some parts of the world do not respect intellectual property rights (different customs).

I've yet to see open source be as successful as commercial private/public enterprises. But we could have different definitions for what is success.

[MrSinatra]

ur right, Linux, android, apache, etc, they should just shut down now.



and again, each person could decide on their own. I'm not for forcing code out of anyone. just b/c u don't want to share doesn't mean others wouldn't, that should be obvious.

[pbelkner]

with respect to supporting R128, there is absolutely nothing stopping someone from having implemented a plug-in to do it already.

Instead of using a build-in EBU R128 / ATSC A/85 / ITU-R BS1770 scanner or plug-in you may use the external scanner R128GAIN writing tags compatible with Winamp:

Home: http://r128gain.sourceforge.net/
Download: http://sourceforge.net/projects/r128...iles/r128gain/

[Aminifu]

i don't know why that's so aggravating to you, ...

His reasons should be obvious too (by now).

I have nothing against open source projects. I just don't think they are wildly successful, except for Android (with mobiles) and Apache (with servers), open source projects still don't have major market shares in their respective areas (in terms of users or money generated).

It's great that some devs and coders are willing and able to donate to what they enjoy. Everyone should donate something if they can (even if it's only money) to support what they enjoy.

[garetjax]

Open Source - Generally, is only as successful as those willing to contribute to it. I think there are solid points as to why this likely would not prove as successful as we'd like it to be if the plugin portion of code were to go open source.

Yes, I imagine people would get all the code...then start up derived versions of Winamp and add the plugin code to sources other than the intended version.


Android, Unix, Linux, Mandrake etc etc all suffer from Fragmentation and everyone working on their own version of a solution that may or may not be the intent for a given source.

The success of Android - for phones is amazing...however, it is run up against a wall just as any other mobile device is....The costs for using the devices, is in its use and problems.

I think I disagree with you on what is interpreted as 'success'. There are as many failures on the road of open-source as their are successes otherwise. And a great deal of those 'failures' cost users; and they get nothing as a result of the inherit open-source of their system. Private Companies offer a product that is provided with some warranty of service, software and solution. As a trend they are a solid platform for which they are built, as only the best of these software products make it to mass consumption.

My thoughts do not reflect Radionomy, Nullsoft and/or Winamp. They are my own, and many have stated both ideas so many times that there is little else to comment about.


FYI: I maintain a website, make it public and allow anyone to post anything for free. While I allow freely connected users without advertisements and etc. I would never open source the code I use to generate the content for my site. Nor, would I use open-source for the site, applications or etc. Unless, it was something for 'no value' add.

[MrSinatra]

just FYI, I never said open source guarantees success, but it does guarantee that it will always be available to be updated.

and secondly, other apps ALREADY use winamp plugins. keeping them closed does little to prevent that if the other apps copy the API.

[garetjax]

That doesn't mean that they will work, or there is a guarantee that the author of the other programs are not doing something incorrect. As a rule, when I use repositories, applications and etc. that are open source, I do not download/install software from 'community' submissions without a thorough understanding as to what they impact.

Though, to each their own. I also must disagree to your statement guaranteeing that a given open-source solution will always be available for updating; unfortunately there are so many reasons why this is an opinion. Most open-source community driven solutions fall by the way-side as a result of lack of contribution and eventually, are lost for the same reason.

Additionally, hobbyist coders and etc. under the best of circumstances do things with the software that may/may not be for the best. Additionally, in the contributions tend to lag behind in terms of technological trends, technology as a whole and best patterns/practices that closed sources thrive with.

On this note, I'd rather pay for Winamp while getting the best out of the software from any platform that I'd like to use it on. Also, while there are open source alternatives to Winamp...there is a reason why people come back to Winamp and try to find a way to run it. It isn't by chance that everyone emulates it, or tries to get features of it into their products.