Replay Gain utility codec - odd results

[Spoon]

Re: Replay Gain utility codec - odd results

Will mark as a bug for the next . release

If you have one file which cannot be done when using R13.3 (on its own with the replaygain utility codec) then please email that file to:

http://www.dbpoweramp.com/email.htm

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

Thanks for looking into this.

I'm not sure a single file is going to demonstrate the issue. The problem is that all the files in a large group get the RG track values, but only some get the RG album values. I believe I've gotten good results when selecting a single album, or maybe two or three albums, so I thnk the number of files in the batch may determine whether this issue surfaces or not.

I think each album either gets album values for all tracks, or none of the tracks, rather than a mix of both for a particular album. I'll do some more tests to find out for sure, and also to find out if it misses album values on the exact same albums every time and whether R13.3 will miss album values for a single file in some cases but not others.

If it looks like the problem only comes about when doing large groups of files, are you interested in getting multiple files via some other route than e-mail?

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

I did some more tests and found the following.

Sometimes part of an album has album RG values, while the rest of the album does not. I previously thought that all songs on an album either had album values or not. In one test of 4 albums, the first 2 albums in alphabetical order did not have albums values. The 3rd album did not have album values on the first 4 songs, but the 5th song onward, as well as the 4th album, all had album values. It's as if it picked up in the middle of the list when applying album values.

Results are not repeatable. I ran another test of the same 4 albums mentioned above. I copied the original files (before RG was applied) to another folder and reran the test. This time, all songs on all 4 albums had album values.

I'll send song 4 from the 3rd album mentioned above - the point just before where album values began to be appear in the first test. Note that I ran RG on this file alone (using the original pre-RG file), and album values were applied this time as well.

[Spoon]

Re: Replay Gain utility codec - odd results

Could it be security software is stopping dBpoweramp from accessing the files?

[Christian_S]

Re: Replay Gain utility codec - odd results

Could it be that you accidentally did not wait for Music Converter to finish its job? In particular when converting large numbers of files, the converter seems to sit there doing nothing after it has analysed all the files.

But you have to wait until it is fully finished. When you start with the Batch Converter windows, you will have to wait until it reappears.

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

I looked into the security software issue. At first, it looked like that could be the problem. I tried a small scale test of about 20 albums and it always worked when the security SW was turned off, and failed most times, but not always, when the security SW was on.

Then I did a test with all 700+ albums, security SW off, and only about 24 albums got album values.

[moley6knipe]

Re: Replay Gain utility codec - odd results

I had a similar problem ages ago > http://forum.dbpoweramp.com/showthread.php?t=17732

I've kinda resigned myself to running RG album in smaller batches, such as per Artist folder

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

I'm responding to the last couple of posts here.

1) I did another test to make sure I wasn't stopping dMC prematurely. I let it run all night even though the track value stage was only projected to run for under 2 hrs. The interface clearly showed that it was finished in the morning, and only a few albums had album values.

2) For a workaround, I have been able to get album values for all songs by running dMC R13.1 on an older computer, running the same RG codec. It just takes longer because the old computer is much slower. If I get a chance, I'll try uninstalling R13.3 on my newer computer, then installing R13.1 and see if I get the same results on a different computer.

The only other piece of information I have on this right now is that the 24 or so albums that do get album values on R13.3 are at the end of the list alphabetically by album artist. All the songs that worked had an album artist that started with Y or Z.

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

After more testing, this problem is different than what I previously thought.

I tried adding RG to all my files on the old computer, and it was successful using dMC R13.1 and R13.3. I'm not able to get all album values on the new computer with either version.

Previously, I based my conclusion on testing with the old computer using R13.1 vs testing on the new computer using R13.3. Since it worked on the old, I assumed there was a problem with R13.3. But now I know that the new computer is the problem, somehow.

Both computers are running XP Pro with the latest updates. The new computer is a quad core, while the old has a single core.

Any ideas on how to find out what's happening with the new computer?

[Spoon]

Re: Replay Gain utility codec - odd results

Something might be blocking dBpoweramp from opening the files for writing.

[mbone@kiva.net]

Re: Replay Gain utility codec - odd results

I finally got this resolved. I tracked the problem down to the security software, as you originally suggested.

I was using BitDefender Total Security 2009. Turning off real time virus protection did not solve the problem. Couldn't even get things to work if I tried to disable the services and startup programs in msconfig. I had to uninstall the security software, or restore to a backup taken before it was installed to get ReplayGain to work.

I downloaded the free upgrade to the 2010 version, and now everything works fine if I temporarily turn off the realtime virus protection.