Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[yonkiman]

Is anything being done on the MC front to create ReplayGain tags that my Rockbox can recognize?  I'm sure MC's proprietary replay gain tags are wonderful, but half the time I'm listening to my music it's on my Rockbox iPod, and now that my library's being converted to FLAC and OGG, my old solution (mp3gain) doesn't cut it.

MediaCenter does so much, and it does it very, very well.  I don't need any third party programs for anything audio or video playback/tagging related except for MC.  That's why I keep paying for it and the upgrades.  I can't believe you guys at JRiver can't engineer an elegant solution that will allow replay gain to work on most/all the portable devices out there.  Isn't it just a tagging option?

Can someone update everyone on when/if this might happen?

Thanks,
Fred

[Matt]

Media Center was one of the first pieces of software to support ReplayGain.  There weren't obvious tag standards at that time for it so we chose what seemed reasonable to us.

Now the LAME header can include the replay gain (I think), but this isn't too slick since it's essentially a third place to store information for an MP3.

Do you know what tag field the RockBox is looking for?  Perhaps there's more of a consensus on what ID3v2 tag field to use at this point.

Then there's the business about what the "correct" reference noise level is.  We adopted the original standard.  Some others have changed.  In hindsight, it would have been smarter for replay gain to be an absolute value instead of a value relative to some magic noise level.

[Alex B]

AFAIK, JRiver's MJ8 was the first player that had a replay gain analyzer and playback correction about 5 years ago.

At about the same time the first replay gain aware playback plug-ins for Winamp were introduced (third party plugins). A year or two later some developers changed to use the 89 dB reference level value instead of the safer 83 dB that was originally proposed. Probably this was done because the users wanted louder files (many had underpowered PC speakers and portables).

Rockbock uses the same standard that is used with Foobar and Winamp. The reference level is 89 dB and the tag fields are as follows (for example):

MP3 (iD3v2):

TXXX replaygain_album_gain -2.66 dB
TXXX replaygain_album_peak 0.923768
TXXX replaygain_track_gain -2.74 dB
TXXX replaygain_track_peak 0.832581

Ogg Vorbis & FLAC

replaygain_album_gain=-2.66 dB
replaygain_album_peak=0.923768
replaygain_track_gain=-2.74 dB
replaygain_track_peak=0.832581

If someone would like to convert MC's replay gain tags to this format the following things would be needed besides just copying the field values:

- add +6 dB to the replay gain values
- add " dB" after the replay gain values
- fill the album peak value from the album file that has the highest track peak value (MC does not store the album peak value separately)
- use six decimal places in the peak values (EDIT: This may be unnecessary. The number is treated as a numeric value and the exact amount of decimal places should not matter.)
- if a comma is used as a decimal separator change it to a dot in the converted tags. (MC writes the decimal separator that is defined in the Windows Locale settings and it can read both separators correctly.)

Personally, I have just used Foobar for calculating a second set of replay gain tags. Both tag styles can coexist without problems.

Though, with MP3 files Foobar causes two problems: It destroys the multi-line comments that MC uses for many tags (including the replay gain tags) and the current foobar version writes ID3 v.2.4 UTF-8 tags by default. ID3 v.2.4 UTF-8 tags are not compatible with all devices and programs. So after analyzing MP3 files with foobar I always convert the ID3 v2.4 tags back to ID3 v.2.3 ISO-8859-1 with Mp3tag and then reapply all MC tags from MC's library. These steps are not needed with OGG or FLAC files.

Edit

The replay gain value that LAME encoder adds is obsolete. I have not seen any programs that use it.

[yonkiman]

Do you know what tag field the RockBox is looking for?  Perhaps there's more of a consensus on what ID3v2 tag field to use at this point.

Hi Matt,

Alex has summarized the answer to that question in much more detail than I could. 

Thanks for taking the time to consider this again.  It would be so wonderful if you could make it work with Rockbox and anything else using the Foobar / Winamp tagging format.

-Fred

[Araj]

It would be so wonderful if you could make it work with Rockbox

Seconded !

[yonkiman]

Any plans to address this?  It seems like Alex has laid out all that needs to be done to be compatible with Rockbox, Winamp, Foobar, etc...

[Jim L]

It would be so wonderful if you could make it work with Rockbox and anything else using the Foobar / Winamp tagging format.

Thirded 

[MadJewDisaster]

Yes , this is one of the unpleasant aspects of MC .
To much things working only in MC .

If you have a player , or use another jukebox at work or girlfreind's the Mc replaygain is of no use.

Means you need to replaygain TWICE your files.

For this raison i do not use MC to play - I send albums to another player who is reading the ''standart' ReplayGain tags.--As far as the word Standart can be use...

I do not feel like to MC replaygain all my files .

I do not know if possible , but i would like MC to :

-Or be able to read no MCReplaygain- Do not need to be upfront , can be a setting in DSP
-Or to write ReplayGain tags as foobar and Winamp and all small progs like MPC Gain or OGG gain write them.

There is no need to change the MC replay value , people who want  the 89 value just need to put permanent adjustement of +6

Concerning foobar , there is a side effect:
Updating any tag in foobar REMOVE ALL SLEEVES IN FILES in MP3
And if you replaygain for use in more than foobar , it means not puting values in data , but write them in tags ....

Now , Winamp can replaygain all kind of formats[MP3-OGG-Mp4-MPC-APE-FLAC and so on ] without any of the foobar probs .

[gummbah]

Sorry to bring up this old thread.
But is there any chance that MC will become compatible with most other players like Foobar, Winamp and Rockbox?
That would really be helpful!

[Robo983]

I am going to bump this up agian. Any plans to make this compatible with Rockbox? I use it on all my players that it supports, iRiver 120, 320, and IPOD 5G.

I was so happy to discover Rockbox so I could/would use the IPOD I received as a gift. I have been even happier to discover MC and all the tagging, managing, playing,...etc. it does as well as manage media on my IPOD. I only use the iPod firmware for videos since Rockbox can't use the hardware encoders in the IPOD. I have taken the time to add the replay gain for my music that is played with Rockbox now to be disappointed now that MC doesn't write a tag supported by Rockbox. I haven't tried the iPod firmware to see if it reads it but that wouldn't help much on my iRiver portables.

I probably use replay gain more in Theater view where I do not walk around with my remote to adjust volumn. On my portables volumn is always in finger reach but it would be nice to have it working.

[skeeterfood]

I'll second the bump.  Just got my $29.99 refurb Sansa E250R from buy.com and will soon be installing Rockbox on it.

[NoBS]

Bumped in agreement 

[llafriel]

I'll bump this one too. It seems the rest of the world has settled on a standard wich differs from that of MC. Maybe now is the time to change?

[gummbah]

Yep, changing MC's replay tagging into the de-facto standard would be my first priority at the moment as well..

[ADDiCT]

I started yesterday collecting info about the replaygain stuff, and there are some things to consider IMHO. Here's a quick rundown of what i've found (forgive me for possibly incorrect usage of technical terms):

- If the MC devs decide to utilize "standard" replay gain tags, we could have a conflict with mp3gain. mp3gain writes own tags into the ID3V1 tags (at the "end" of a mp3 file), such as "MP3GAIN_MINMAX", "MP3GAIN_ALBUM_MINMAX", "REPLAYGAIN_ALBUM_GAIN" and a few others. I see a potential problem there with MC overwriting this information for albums that already have been processed with mp3gain.
- foobar 2000 uses ID3V2 tags for replay gain information storage. When using both mp3gain and foobar replay gain analysis on a mp3 file, you end up with different replay gain values in the ID3V1 and ID3V2 tags. This is bad, because the acutal volume values are different, probably because of different ways of determining the volume levels.

Some general thoughts:
- I think there is no such thing as a "general accepted standard" for replay gain. There are some "de-facto standards", but i wouldn't call the handling of replay gain a "standard" per se. The only "real" standard i see is the one implemented by FLAC.
- I can't understand all the fuss about RockBox. I mean, it's a very interesting project, and i appreciate the effort that's put into it. But you have to see what RockBox can do, and what it can't. On the iPod, using RockBox has some major drawbacks. For example. RockBox can't use the "standard" iPod database, which means you either have to build a special database within RockBox based on the file tags, or use a folder/filename based browsing approach. This is a major step backwards from the original iPod design IMHO, and makes RockBox pretty much incompatible with iTunes and MC in terms of database compatibility (think smartlists, for example). Another major RockBox flaw is that is doesn't support power saving functions on many iPod models. This means shorter runtime, and possibly higher strain on the batteries - a large problem IMHO, because it's not easy, and/or expensive to replace the iPod battery. Just have a look at the RockBox website, there are some detailed reports about RockBox' status. For the iPod, i think RockBox is just not "mature" enough yet, and i think it will take a long time before RockBox can be considered a feasible alternative to the original iPod firmware (I've been following RockBox' development for two years now, and it's advancing quite slowly). For other players, the situation might be different. I believe RockBox is more of a toy for experienced users, than a "mainstream" piece of software. It's a bit like Windows vs. Linux, IMHO - the latter holds much promises, but the former is the de-facto standard that can be used easily by "Joe User". I think it would make sense for JRiver to consider if it's worth investing precious development time into a platform that may be used only by a fraction of the potential customers.

My conclusion for the whole replay gain shebang is to use "hard" replay gain. foobar 2000 has a wonderful function where you can apply "hard" replay gain when converting files. The basic workflow is that you first analyze the audio, and apply replay gain tags. After that, you convert the files, and foobar will apply the replay gain settings to the resulting files. This means that, no matter on which device, and with which software you play the files, you will always get a "corrected" volume level. For mp3 files, mp3gain is my tool of choice. Using MC's replay gain functionality is just icing on the cake for me, for playing files in MC. With that approach, i lose an album's "original" volume setting, but i gain complete compatibility for each hard- or software, audio-cd's, and so on.

[Alex B]

- If the MC devs decide to utilize "standard" replay gain tags, we could have a conflict with mp3gain. mp3gain writes own tags into the ID3V1 tags (at the "end" of a mp3 file), such as "MP3GAIN_MINMAX", "MP3GAIN_ALBUM_MINMAX", "REPLAYGAIN_ALBUM_GAIN" and a few others. I see a potential problem there with MC overwriting this information for albums that already have been processed with mp3gain.
- foobar 2000 uses ID3V2 tags for replay gain information storage. When using both mp3gain and foobar replay gain analysis on a mp3 file, you end up with different replay gain values in the ID3V1 and ID3V2 tags. This is bad, because the acutal volume values are different, probably because of different ways of determining the volume levels.

What you said in the above quote is not exactly correct. I could explain this thoroughly, but it would be slightly off-topic on this forum.

The important thing is that MC12 does not destroy, change or even care about the replay gain tags that are written by MP3Gain (APE tags, not ID3v1) or Foobar2000/Winamp (ID3v2.3 or 2.4, APEv2 or Vorbis/FLAC comments).

If you alter the file volume levels with foobar2000 or MP3Gain you just should do it before analyzing the files with MC12. Then everything should work fine.

[ADDiCT]

The important thing is that MC12 does not destroy, change or even care about the replay gain tags that are written by MP3Gain (APE tags, not ID3v1) or Foobar2000/Winamp (ID3v2.3 or 2.4, APEv2 or Vorbis/FLAC comments).

Yes, that's the current situation, and i have to admit it's what i prefer for now (; . MC stores its own replay gain tags in the "comment" field, together with other MC specific tags. What i meant is the possible future handling of replay gain tags by MC, as suggested by other user in this thread.

If you alter the file volume levels with foobar2000 or MP3Gain you just should do it before analyzing the files with MC12. Then everything should work fine.

That's exactly what i'm doing now, and it works great (for me).

Btw, Alex, what i was describing in the paragraph you quoted is based on my own "research", which i accomplished simply by testing all apps, and viewing the results with a hex editor. I would be most interested in comments about my observations. If you have the time and motivation, maybe you can explain what's incorrect about my statements. I'm still not done with "researching" and thinking about a perfect solution, so your input is appreciated. If you know only half as much as you do know about MC's expression language, for example, you should be my replay gain guru (; . As there is no "correct" forum on Interact for the type of info we're discussing, we could as well discuss it here.

[Robo983]

If you alter the file volume levels with foobar2000 or MP3Gain you just should do it before analyzing the files with MC12. Then everything should work fine.

So, even though this isn't a Rockbox forum will this work in Rockbox? If running it through MC after doesn't alter the info Rockbox is using then this is a good comprimise since I do not view this a high priority for MC.

I too thought this was a good thing but the more I use Replay Gain the less I like it's behavior. I find that it lowers the volumn too much and I have to really crank up my amplifier. If I for some reason forgot I did this to play MP3s then play something else I blow out my window in the house.

I have been using MCs built in tool to increase the volumn's of the offending songs/album that are extreamly lowerer than others. What I found is that even with Replay Gain, albums that were 20+ years old may have only few narrow spikes in the song and therefore use that as the peak causing the RG to be set high reducing the overall volume even lower. About the only time I have a problem now is if I am using the Play Radio feature.

[Alex B]

So, even though this isn't a Rockbox forum will this work in Rockbox? If running it through MC after doesn't alter the info Rockbox is using then this is a good comprimise since I do not view this a high priority for MC.

Yes. You can have two separate sets of replay gain tags in the files (as I said in my earlier post a year ago.)

I too thought this was a good thing but the more I use Replay Gain the less I like it's behavior. I find that it lowers the volumn too much and I have to really crank up my amplifier. If I for some reason forgot I did this to play MP3s then play something else I blow out my window in the house.

With modern overly compressed recordings this is true. You may want to try the "Automatic" correction setting. It increases the overall volume level of the Playing Now list as much as possible without causing clipping.

I have been using MCs built in tool to increase the volumn's of the offending songs/album that are extreamly lowerer than others. What I found is that even with Replay Gain, albums that were 20+ years old may have only few narrow spikes in the song and therefore use that as the peak causing the RG to be set high reducing the overall volume even lower. About the only time I have a problem now is if I am using the Play Radio feature.

Actually, it is the other way around. Replay Gain measures the average volume level (*). The maximum peak level is registered, but that info is used only for clip preventation. Old dynamic tracks are detected as quieter and without a manual correction MC plays them louder.

Of course, if you use a big positive correction setting and the clip prevention feature MC may not be able to apply the full correction to dynamic tracks because that would increase the volume level above the original level and make the tracks clip. On the other hand, any track can be corrected at least as much as its negative replay gain value is. Usually a new loud and compressed track has a big negative replay gain value. For example:

If a manual correction of +12 dB is set:
-- an old uncompressed track from 1985. Peak Level 100%, Replay Gain -3 dB => correction can be +3 dB at max.
-- a newly mastered, compressed track. Peak Level 100%, Replay Gain -14 dB => correction can be +12 dB (it will still be 2 dB quieter than without using RG).

The new track will be 9 dB louder than the old track when a +12 dB correction setting is used. A better way would be to not use a manual correction or set it to automatic. The automatic setting would correct both RG values +3 dB and fix the original, 11 dB, difference. I.e. the old track would not be adjusted at all and the new track would be adjusted -11 dB.


EDIT

(*) -- It isn't a direct mathematical average. The calculated value is based on a complex formula that tries to simulate the human auditory system and estimate the perceived subjective loudness.