How to tell target level for ReplayGain scanner? Options

[d_headshot]

Currently I have mine set to the default of 89 dB but some other files I applied info to had a target level that was higher or lower. Is there a way of knowing which levels were used for different files so I don't have to rescan the entire folder/library?

[lvqcl]

AFAIK It's not possible to change the target level for foobar2000 RG scanner.

[makotoiwai]

I have to admit--I'm curious how to do this, as well. All of my MP3s are set to 89, which works really well, so when a FLAC comes along, it usually blows me ears into my brain. The only other thing I can think of is to covert all FLACs to MP3s until functionality like MP3Gain happens.

That being said.. is there just no way to target the volume of a FLAC in the same way MP3Gain does this with MP3s?

[db1989]

Um, yes, there is an entire feature: ReplayGain. A little research will reveal that MP3Gain is in fact a subset thereof.

What you are looking for is not a way to determine or alter the target level; it’s merely the standard way to run vanilla ReplayGain, not an MP3-specific variant. This is easy. Try starting here; though outdated, it should give you enough direction.

[dhromed]

so when a FLAC comes along, it usually blows me ears into my brain.

So, um, what's stopping you from scanning the FLAC files?

[makotoiwai]

Yeah, I've seen most of the threads about this, but I really like the way MP3Gain approaches this... it simply allows me to set a target DB for the files. I set mine universally. The ones that clip, I set them a tad lower and remove them... repeat until none of the files clip, anymore.. I use Foobar and Replygain for this, but, again, I like the interface/functionalist of MP3Gain and wish there was a program that worked like this, and exactly like this, for FLACs.

[kode54]

FLAC files do not contain any gain field to be adjusted in this way. The entire file must be decoded, scaled, then re-encoded, and the process is not lossless, since there may be minor quantization errors.

[dumdidum]

FLAC files do not contain any gain field to be adjusted in this way. The entire file must be decoded, scaled, then re-encoded, and the process is not lossless, since there may be minor quantization errors.

i find this interesting, well, actually, surprising. i mean, if i understand correctly, the replaygain is just one number (or two numbers if you compute it for album as well). it shouldn't be hard to write that number in whatever field of the tag. sure, that might not give you cross-application compatibility and that could be an issue (similar to, say, the compatibility issues with WAV tags). but wouldn't it still be better to write in a custom tag than to destructively edit the audio?

This post has been edited by dumdidum: May 1 2012, 00:53

[kode54]

Oh, that. Yeah, there's that. You could also shift-right-click on the file and use the ReplayGain editor post-scan on the relevant files.

Note that the clipping issue you mentioned is not possible with FLAC files, since they only support integer samples, and those can't clip. Or rather, if there was clipping, it happened already.

[mjb2006]

The discussion seems to be getting sidetracked; the OP was asking about MP3s to which he applied RG.

AFAIK It's not possible to change the target level for foobar2000 RG scanner.

Sort of.

Preferences > Advanced > Tools > ReplayGain Scanner > Target MP3 alteration volume level (dB)

When he 'applied' the info to the MP3s, their volume was adjusted to make them actually be approximately that target volume. But the track gain tag in the file will still reflect the gain needed to reach 89 dB. So the answer to his question is yes, assuming he didn't touch the ReplayGain tags, he can simply subtract the track gain values from 89, and he'll be within 0.75 dB of the target. The target can only be an integer (in the fb2k interface), so there will be 2 possible values.

For example, if the track gain tag is +8.32, then the MP3's volume must be 80.68. So the target can be inferred to have been 80 or 81 dB. (We know it wasn't 79 because the MP3's volume would've been 79.18 and the track gain would've been +9.82. We know it wasn't 82 because the MP3's volume would've been 82.18 and the track gain would've been +6.82.)

If he's trying to actually undo the adjustment, though, it's not possible, because the original volume of the MP3 isn't known.

This post has been edited by mjb2006: May 1 2012, 01:19

[lvqcl]

Note that the clipping issue you mentioned is not possible with FLAC files, since they only support integer samples, and those can't clip.

They can clip if their gain is greater than 0 dB.

The discussion seems to be getting sidetracked; the OP was asking about MP3s to which he applied RG.

I'm not sure. The topic title is "How to tell target level for ReplayGain scanner?", and MP3 files weren't even mentioned in the first post.

[mjb2006]

MP3 files weren't even mentioned in the first post.

He said he "applied info to" files using a different target gain. In foobar2000, applying RG, and changing the target for such, is only possible for MP3.