APE tags not supported

Back in Dec '20 someone expressed that APE tags were not supported on export and that the team didn’t have the manpower to undertake such a programming effort.
Is there a chance that you may reverse that decision as APE is the only format that supports Replay Gain metadata which is crucial to us Web radio operators???

I realize that one could always reprocess the exported files with something like Mp3Gain, but why should we have to do the same job twice?.. Haven’t tested but we probably also lose the ability to reverse Replay Gain changes in the process…

All of my MP3s are ReplayGain tagged (not MP3Gain) and I did that in Winamp. It also works with FLAC and most other formats.

Winamp doesn’t seem to support tags (of any kind) in WAV files, and in general WAV tags are not well standardized or well supported.

ReplayGain is usually a player feature and Audacity isn’t intended to be an everyday audio player.

Agreed, Replay Gain is a player feature and AutoDj, Sam Broadcaster and most other streaming engines are ‘players’ of sorts. I happen to normalize my media files in Mp3Gain which doesn’t re-encode the file, is completely reversible and saves it’s metadata in APEv2 format. The fact that Audacity doesn’t conserve APE metadata on export causes the loss of the replay gain data that streaming engines need to output a stream at a steady volume and loses us the reversibility of the volume changes. Adding APE support to Audacity would save us radio operators much headaches and reduce the need for tedious reprocessing all the time…

DVDDoug; a question if I may; why do you deem it ok that Audacity would ‘lose’ important data produced and used by other audio programs? If it’s there, leave it, no skin off of Audacity’s nose…

I assume that you are referring to “losing” the ReplayGain tags in APE files.

There are two parts to this. The first is a rather pedantic distinction about what Audacity does:

Audacity does not process APE files (or any other audio format). It processes data that is read from the file. That data comprises the audio data, and a handful of common metadata tags (if present). Assuming that the format is valid (does not raise an error), all other data (such as video data, images, hyperlinks, unsupported tags, …) is ignored. Audacity does not “lose” ReplayGain tags, it never extracted them from the file in the first place.

The second reason is more pragmatic:

Let’s say that Audacity did retain all APE tags.

  1. Import an APE file (including the ReplayGain tag).
  2. Amplify by -12 dB.
  3. Export as an APE file, including all metadata that was imported.

You exported APE file now contains the ReplayGain tag, but the tag has the wrong ReplayGain data for the modified file.

Another example:

  1. Import 2 APE files
  2. Export as a new APE file (a mix of the two files)

Which of the imported file’s tags should be in the exported file?

Who’s talking about processing APE files, if there ever was such a thing?
I see the point of your examples where the Audacity reprocessing includes changing the volume of the media. In my case, I have a media collection well over 200,000 strong, all normalized with replay gain. Why would I want to change the volume, it’s already set exactly where I need it to be and with the metadata to process them easily in my streaming engines.
I’m using a nice little Audacity plugin to remove leading and ending silence on each file because some streamers do a pith-poor job of truncating to allow for a clean crossover.
The fact that I lose APE metadata in this process means that I have to re-inflate the files to max volume in Audacity (using normalize at 0), trim the ends and export them back out. Then I have to re-normalize them with an app that encodes them with the replay gain metadata that I lost.
Since Audacity doesn’t yet take advantage of multi-core technology, you can guest that this job is gonna take forever and a day…
But thanks anyways for your insights…

APE tags are part of the (Monkey Audio) APE file format and most commonly found in APE files.
What file format are you referring to?

I’m talking about mp3’s normalized with Mp3Gain or other normalizer that saves Replay Gain settings to APEv2 metadata

OK, right… Audacity is NOT the best for metadata… I doesn’t seem to be a priority of the Audacity developers. Audacity still doesn’t support embedded artwork.

It’s “difficult” because different the formats use different tagging standards (MP3 uses ID3 tags, FLAC uses Vorbis Comments, etc.). And ReplyGain tags are not part of the ID3 standard. It’s an additional tag. (Song ratings are also non-standardized.)

With Audacity, you can’t edit the metadata without creating a whole new file or completely overwriting the old one. With lossy formats (like MP3) that means an additional generation of lossy compression and some “damage” accumulates. You may not hear any quality loss and you may not have a choice, but it’s something to be aware of and you should try to avoid it and compress once if possible. Otherwise, try to minimize the number of times it’s re-compressed.

Mp3Tag can copy the tags from your old file to the new file. And you can set-up a batch process. (I’ve never done that.) ReplayGain shows-up as “Extended Tags”. mp3Tag is only updating the tags. It doesn’t “touch” the audio data. (It works on all of the common audio formats, not just MP3.)

Which also means if you have MP3s you are going through another lossy compression cycle.

mp3DirectCut can do limited editing without de-compressing/re-compressing. It’s a little “odd” to use and I haven’t used it that much but it should retain the tags.

FYI - MP3Gain doesn’t use ReplayGain tags. It changes something in the frame headers that sets the volume so it “works everywhere” and doesn’t depend on the player supporting ReplayGain. It also doesn’t go through a decompression/re-compression cycle. There is a limitation that the volume can only changed this way in 1.5dB steps, which means your tracks are usually within 0.75dB of each other, but not as precise as the real ReplayGain.

You can use them both together but you have to run MP3Gain first, and then ReplayGain is based on that new loudness. If you do it the other way around, MP3Gain will ignore the ReplayGain tag and the tag will be wrong for the new-modified file.

If you edit an MP3Gain’d file in Audacity, the new exported file will have the same loudness but without the modified headers, so from that point on it’s not reversable.

…I don’t think ReplayGain is an APE tag, but maybe so.

ReplayGain may be stored as an APE tag. It can also be in an mp3infotag, or directly in the header of a Musepack file, or as VorbisGain metadata, or as a Vorbis Comment, or other formats. It is part of the specification for APE and WavPack formats, but for (most?) other formats it’s non-standard.

Thanks for pointing me to that great little app. A batch trimming job in this is 100x faster than in Audacity, which will save weeks on my current project. It doesn’t preserve APEv2 metadata either, but I don’t mind reprocessing with Mp3Gain, thanks to the time saving.

I’m still keeping Audacity for what it’s really good at… Cheers!