Print

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

[ADDiCT]

There are several posts about build 518, and i'd like to add some observations.

I'm running XP/SP3, and did some heavy tag editing with 518. I've also synched my iPod Classic 160GB (latest firmare) several times. Auto-Import is disabled, but i've imported a lot of albums with 518. So far, i didn't see any library corruption (knock on wood!), and the iPod database seems to be in order, too.

What i did notice, however, is that MC adds a "replaygain_track_gain 6.000000" tag to _every_ mp3 tag it "touches" (writes, updates). This value is not visible in the "Replay Gain" field inside MC's tagging window, but it can be seen, for example, by playing the file with Winamp, or by viewing it with a hex editor. The tag is written no matter if the file has been analyzed, or not. I hope the next build will remove these "bad" tags automatically, otherwise i'll have to wade through a _lot_ of mp3's to get rid of the incorrect tag.

Another thing i've noticed is that cover flow on the iPod is behaving strange. I'm not 100% sure, but i think "multiple artists"-albums were grouped together in cover flow (i don't use cover flow too often, mainly to "show off" (; ). Now, i have multiple entries for "multiple artists"-albums, one for each artist.

[Alex B]

I reported this problem on the beta board and it should be fixed in the next public build:

From: Alex B   Re: Media Center 12.0.518  « Reply #21 on: June 27, 2008»   

MC saves a +6 dB track gain value when it adds tags to a fully untagged and unanalyzed MP3 file.

This is obviously the correction value that is used for compensating the difference in the 83 dB vs 89 dB replay gain reference values, but if the library does not contain a replay gain value MC should not add a replay gain tag.

I noticed this because I converted some old unanalyzed APE files to MP3 and forgot that I had disabled tag writing last time I did some testing. I saved tags from the library and MC added an incorrrect +6 dB replay gain value to all files.

EDIT

Actually the file doesn't need to be untagged. It is enough that it isn't analyzed. When MC saves a tag change a +6 dB replay gain value is added. This can cause serious clipping distortion if some other program or device uses the value. MC reads the tag as 0dB because of the used correction, so it does not do anything inside MC even when the library is updated from tags.

However, I have not noticed that MC would write an incorrect +6 dB replay gain value when the library has a correct value. In addition, I think this problem has been around since the replay gain tag writing system was changed a few months ago (though, I have not tested the behavior with an old MC build.)

EDIT

You should already be able to fix the tags by analyzing the files. After the fixed build is available you can either analyze the files or remove all tags and resave them from the library.

I don't think the fixed MC build will automatically remove existing, possibly incorrect, replay gain tags from the files. It would somehow need to know if the existing tags are correct or not. MC would need to analyze the files and compare the replay gain values. That would be a very complex solution. It is better to just analyze the unanalyzed files and let MC save the correct values.

[Matt]

We believe the new builds (520 and later) do not write a replay gain tag if MC doesn't have a replay gain.  They also remove existing replay gain values in the tag during a tag save if MC has no replay gain for the file.

Let us know if you see something to the contrary.

[Alex B]

They also remove existing replay gain values in the tag during a tag save if MC has no replay gain for the file.

Oh, I see. That is a reasonable approach. Though, it is possible that the user has analyzed MP3 files with some other tool after originally importing them to MC. In that case the user should make sure that the tag changes are imported to MC's library before writing any other tag changes with MC12.

[ADDiCT]

Well, the "replaygain_track_gain" seems to be removed when updating tags. But instead, MC added a "replaygain_track_peak" tag with the value of "1", which is displayed as "Peak Level 100%" inside MC, to some of the files. I'm not sure whether this is better or not.

[Alex B]

... But instead, MC added a "replaygain_track_peak" tag with the value of "1", which is displayed as "Peak Level 100%" inside MC, to some of the files. I'm not sure whether this is better or not.

I don't think that is intended and I couldn't reproduce the behavior in a quick test with a few unanalyzed MP3 files.

If possible, could you explain how the problem can be reproduced?

[ADDiCT]

Alex B, i had the effect on a few files only, the rest went fine. I'll try to reproduce the effect.