Volume Normalization

[a.wright]

Re: Volume Normalization

It's odd isn't it!

My settings for ripping/converting are pretty much default, I am using both XP SP3 & Vista SP1 (same effect on both) and I have not used any DSP effects since I have been using dbpoweramp as I went with the MP3Gain option from day 1! Oh and my version of MP3Gain is the same as yours.

The only other software I use is Tag & Rename which is perhaps doing something with the tags so I'll check to see what happens if I just do a rip/converison and then run straight through MP3Gain. Update: Same, I get MP3GAIN tags and ReplayGain tags when looking at the ID Tag (dbpoweramp extended) properties by right clicking on the file.

If Tag & Rename is somehow interpreting something and writing ReplayGain tags (which I don't think it supports anyway) am I potentially going to have the normalisation applied twice if I play the MP3Gain(ed) files in a player that supports ReplayGain tags?!

I am beginning to think that just using the Volume Normalisation DSP with ReplayGain might be simper & safer, if irreversable!

[xoas]

Re: Volume Normalization

My thoughts are this:

IF Tag & Rename is responsible for showing ReplayGain data for tracks that have had Mp3Gain applied (but have not had ReplayGain applied), then I THINK the issue is trivial and you should NOT end up having normalization applied twice when played on players supporting ReplayGain. My assumption, however, is based on the theory that the misleading information about ReplayGain being applied will not mislead a ReplayGain-compatible player into applying ReplayGain and that this is an assumption on my part.

The same should (I think) be true if the reason that your tags have both Mp3Gain data and ReplayGain data is for any other reason save for the tracks having both ReplayGain and Mp3Gain applied.

However,

The only other software I use is Tag & Rename which is perhaps doing something with the tags so I'll check to see what happens if I just do a rip/converison and then run straight through MP3Gain.

does strike me as a good idea. The other technical possibility might be that you are saving/storing different track information in your tags than I am (but I can't guarantee either that we are saving/storing different track information or that this would account for the difference). Let me know if you find that Tag & Rename is NOT responsible for this and I can see if it makes a difference if I elect to store and display more tag information than I am doing at present.

Perhaps someone more tech savvy than myself might be able to tell us how to access the tag contents directly to see whether the tag contents are as they should be. That would be the truest proof of whether the problem is simply one of misreading the tag as opposed to having had ReplayGain data accidentally introduced for whatever reason.

Best wishes,
Bill

[a.wright]

Re: Volume Normalization

Think I might have a reason....

I think just by loading in files and analysing them in MP3Gain, it writes (fake?)ReplayGain tags so next time you load them in it does not need to re-analyse anything again and it knows what to do to get the desired effect. Tag & Rename does not appear to be doing anything at all.

I ripped an album, loaded all tracks in, analysed them but did not apply any gain changes. I then just loaded in every odd numbered song and applied track gain to them. However, every file has "Undo" tags & ReplayGain tags listed...

Confirmed. The "Undo" & ReplayGain tags are written before gain changes are actually applied.

If I load them all into MP3Gain again, it shows me straight away what the volume levels are without needing an analysis and indeed the ones I track gained are at 92db which I set, the others are still original levels.

That might be an answer but then my question in my last post remains... are they real or fake ReplayGain tags and might they be read by a compatable player to "double normalise"?... for which you have just explained your thoughts as the same time I was posting this!

[xoas]

Re: Volume Normalization

I am running some tests on this now.

I think that if the files that have been ripped and which have been analysed only by Mp3Gain show APEv2 tags, ten it is reasonable to suppose that the ReplayGain information is stored in the APE tag and will not be read as ReplayGain on ReplayGain compatible players.

I'll share more after I have looked at this a bit futrther.

Bill

[xoas]

Re: Volume Normalization

OK. Still a bit confused, but here it is.

Ripped same CD, same Mp3 settings, no dsp with dMC. Applied Mp3Gain analysis only.

This showed ReplayGain in the dMC Audio Properties window and also showed APEv2 tag. ID Tag window showed both Mp3Gain data and ReplayGain data.

Went on to apply gain to these same tracks. Continues to show Replay Gain in the dMC Audio Properties window and both Mp3Gain data and ReplayGain data in the ID Tag window.

Cleared the Mp3Gain tag information (confirming that the Apev2 tag, Mp3Gain and ReplayGain tag information had all been cleared). Re-applied Mp3Gain adjustments with analysis first and then (after clearing Mp3Gain tag information tag information) used Mp3Gain to directly apply gain adjustment. Same results both times (showing ReplayGain in the dMC Audio Properties window and both Mp3Gain and ReplayGain values in the ID Tag window).

The minor mystery is why the same tracks in my original test files with Mp3Gain applied do NOT show ReplayGain in the dMC window and do NOT show ReplayGain data in the ID Tag window.

But I think the major finding is that the fact that using Mp3Gain to remove tag information effectively removes the APEv2 tag and the ReplayGain and Mp3Gain information from the tag information. While I can't absolutely prove this, I feel very confident that the suspect RepalyGain information (whether real or bogus) is confined to the APEv2 tag and thus would not be read by a ReplayGain compatible player.

So my advice, until or unless someone can demonstrate a convincing reason to do otherwise, is to not worry about it at all, and to continue to use Mp3Gain as you have been doing.

Perhaps other Mp3Gain users (and I know you're out there) can confirm that this is normal tag performance with Mp3Gain. I'd be happy to hear from you all.

Best wishes,
Bill

[a.wright]

Re: Volume Normalization

So my advice, until or unless someone can demonstrate a convincing reason to do otherwise, is to not worry about it at all, and to continue to use Mp3Gain as you have been doing.

I shall do! It's certainly the best way to learn about these things and I suppose if more and more software & hardware devices support ReplayGain tags and do read the ones we have found, it's easy enough to reverse things and start again.

Thanks for having a look into it all