Tidying up random and messy tags on live, radio or studio DJ mixes and live rave PAs

Greetings MusicBrainz…
Firstly, maximum respect for all the hard work you guys and girls have been putting in providing extensive metadata to help me on a personal level organise my own music collection of retail released albums and singles better than it was.

I’m an avid collector of rave tapes which often contain no official track lists as they are unofficial releases or bootlegs and in addition to digitising my own collection, I download other people’s rips. Often these are inadequately tagged for my liking so I find myself re-tagging their rips with what I believe to be the most useful standard for archiving live sets or radio shows…

TITLE:
YYYY-MM-DD - ‘Performing Artists’ - ‘Show’ - Part ‘Track No’

ARTIST/COMPOSER/PUBLISHER:
‘Promoter/Station/Studio’

I also embed the front of the event flyer/poster, or the radio station logo as the Album Cover.
I’ve often got track listings for these mixes elsewhere but haven’t embedded these as I haven’t decided how best to go about it.

Can I upload my work so far to the database, or is there a way to format this data into better suited ID3 tag fields so they follow your standards (if you have any)?

Thanks.

3 Likes

There are likely to be some tweaks to fit the guidelines.

“Performing artists” are put into the artist field, with commas or feat. separating them.

A Date is usually followed by a colon, with comma after the show name and before Part Number.

Artists: ‘Performing Artists’
Track Name: YYYY-MM-DD: ‘Show’, Part ‘Track No’

Maybe give us some full examples?

On a second note, the guidelines of the database can be a little fussy to fit in with everything else already here, but it is also possible to use Picard to create your preferred tags for your own files. A bit of scripting is needed, but then you have best of both worlds.

2 Likes

Here is a thread to look at as a guide.

Someone putting Radio 1 sessions into the database. They get results like this:

In those cases they are aimed at a DJ and a mix.

Notice how they also make use of Series to keep them all together:

3 Likes

Here is some of the work I’ve done so far…
I’ve halted this project until I can reach a consensus on this so I don’t have to reupload everything after a format change again! LOL

https://archive.org/download/RavePreservationLibrary

2 Likes

Interesting project there. I believe the above example is an ideal pattern to work with.

It is labelled as “bootleg” due to it being a “taped from radio” release and not official. The layout of the Title is slightly different, but not a major divergence from your tracks.

I think the main difference is the DJ is the Artist. And the radio show goes into the title instead.

As an example I downloaded:
1989-12-12 - 808 State - Part 1 - Sunset 102 FM Manchester - 128kbps - 43706242.mp3

Album - Manchester
AlbumArtist - Sunset Radio
Artist - Sunset 102 FM Manchester
Composer - Sunset Radio
Copyright - Sunset Radio
Genre - Promo
Publisher - Sunset Radio
Title - 1989-12-12 - 808 State - Part 1
Year - 1989

Following the above guide, I think that becomes:

Album - 1989-12-12: Sunset 102 FM Manchester
AlbumArtist - 808 State
Artist - 808 State
Copyright - Sunset Radio
Type - bootleg
Genre - Promo
Publisher - Sunset Radio
Title - 1989-12-12: Sunset 102 FM Manchester, Part 1
Year - 1989-12-12

Part 2 would be track two from the same Release.

I’ve stuck a link in that other thread asking for thoughts from there. It looks to me like you have some very similar data. The main difference being you don’t have actual full tracklists.

As to how to get this to the MB database - it will need a bit of trickery on the scripting. I know how I’d do it to shuffle the tags around using MP3TAG, but not enough of a Picard expert to do it with that tool. Still needs work to add a “Release per show”. I’d also tie them all together with a Series.

Edit: Maybe the titles would be 1989-12-12: 808 State, Sunset 102 FM Manchester, Part 1 if it is 808 State’s show on that station? The tricky bit is the example above is the name of the show and station combined. Your files seem to be DJ + Radio Station

1 Like

I see 2 types of recordings here:

For the latter ones I think the format should probably the standard for Bootlegs of concerts listed here: Style / Specific types of releases / Live bootlegs - MusicBrainz

So the above album title would be:

1996-10-05: The Colosseum, Stockton on Tees, UK

And the track title should also be the same.

3 Likes

Thanks for this. I use a combo of iTunes and MP3Tag to do tagging and ideally would pull the metadata from my library if that was possible…

Yes, I’m with you. This looks doable…
With releases which end up for sale on Discogs they don’t like corrections, and instead insist everything is how it was on the box, typos, false line-ups and all so I’m reluctant to add it all to notes because the point is to make the digital versions better historic records going forward.

My metadata for this release, for example is fixed to include people uncredited as guests etc

1 Like

MB also likes to keep typos in the Track lists, but get things correct for the actual Recording. The Track list being what you read, but the Recording being what you hear.

It is the Tracklist that tagging apps then pickup. Whilst internally the MB database will link back to the correct items with the Recording (where known)

With the example cassette you show - is your recording a copy of that? Or do you have the original radio show? If you have a slightly different source you’d end up with two items in the Release Group - one for the Cassette edition, one for the Digital Media edition.

Best thing to do is pick a few examples and add them to MB. Then post links here for people to poke at to get a pattern sorted for uploading these. We can also show you how to make best use of the Edit Relationships pages to add all those other uncredited people you know about who worked on the recordings.

The problem with the format suggested is when it comes to retail releases (albeit technically ‘unofficial’ despite being released by event organisers themselves to ticket outlets and underground music stores) of entire stages at events (all night raves) is you’re gonna end up with 12 different concerts. With these 8-pack tapes, you’d essentially get a concert and a half per cassette (8 x 90 minute cassettes with one act, usually a DJ with multiple guest MCs per hour).
So if I was a whizz kid with all this backend stuff it would seem more time saving to submit the acoustic fingerprints to the MP3s in circulation and then identify the contents and which release they were sourced from.
Sets often find their way into my collection loose, and it’s only after establishing the artists and venue that I can usually find the poster for the event and therefore the tape or CD pack which was later released, so I kind of work backwards when assimilating other people’s rips into my collection.
I basically want to know the most effective way of submitting my ID3 tags somewhere online which will help the public sort out their versions automatically as the vast majority of collectors of the genres I’m interested in aren’t too hot with embedding metadata and very frequently I have to “rescue” manually named files which actuially have Unknown Artist - Track 1 embedded therefore often becoming overwritten with useless tags when media players re-organise and name stuff based on tags rather than filenames, if that makes sense.
I’m trying to share my fixes so if and when other collectors discover Picard, it will recognise their rips too and offer to give them my/our tags instead of ‘Unknown Title’ stuff.

1 Like

If I understand the mission, it is more important to get your Digital Media copies uploaded and documented as these are the bootlegs that are more common to be found by people than the physical cassettes. And that is what you now have the data documented for.

I don’t think you need to worry about 12 different concerts as the focus will be on the Release package of 8 tapes.

This is why I am suggesting to dive in at the deep end and throw a couple of the different types of releases up. We can then help get the data into shape, and then find a way to speed up the data upload.

By ‘concerts’ do you mean releases? If the tapes were sold together then they should all be in MB as a single multi-tape release, as in your Discogs Hysteria example.

You can apply the bootleg concert style to these packages if you think it’s appropriate, but a title like ‘Hysteria 11: Jungle Drum & Bass 8 Pack’ is also ok imo. Particularly if they sometimes contain more than one concert.

:raised_hands:

Yes that’s correct… The vast majority of mixes are ripped as individual files containing the whole set, or in two parts for each side of tape. Some stuff are in APE or WV formats with track-marks embedded but to this day I have never got round to learning how that format is read and/or edited.
I was following the format suggested following the BBC Essential Mix ‘view’ pretty much, up until the last couple of years when for some reason a huge batch of my library lost ID3 tags and because my library was using iTunes all the most important stuff, ie. the dates and specific events, got lost so I was left with folders full of ‘01 - Live Mix.mp3’ and an absolute nightmare to fix because by that point I’d moved them from directories created by iTunes so lost the information in what would be the Album field altogether. PAIN DETECTED…
So now I’ve been doing everything with iTunes in mind as I imagine it is or will be the most common media player around the world soon so if other people experience this glitch (which I still don’t know how to prevent happening again), the MP3s/M4As etc will at least have the date in the filename and can probably survive without the utter head-desk-inducing disaster I had.
And yes, I do have my drives mirroring each other but I didn’t spot this sudden disappearance of all the tags until after I’d restored iTunes to factory settings and moved all the files into one big folder! Fuming… Talking thousands and thousands of shows.
Made me realise that however obscure these recordings are, they are not as valuable as my bloody time these days! I’d like to do all this once and for all and save everybody’s time.
The Warez Scene rules for live shows likewise don’t account for the randomness of improv type unplanned DJ sets with unadvertised vocalists turning up so I haven’t been following that standard either.

1 Like

Hi,
Okay, you might think I’m taking this too far but I’m very OCD and ultimately would quite like my media player to group audio files together by the actual event rather than say "New Year’s Eve, Vol. 1’ and so on…
Also there was a time in the 2000s when it was now affordable for manuafacturers of tape packs to also release a CD pack version… As these have surfaced online, there are sometimes portions which were only on one format and not the other so I tend to put the jigsaw together if I can so that if I play the whole event, it plays back like recreating the entire night from start to finish recorded across multiple arenas and in the order the artists played rather than the order the cassettes end up arranged for release.
Here is an example -

https://archive.org/download/HelterSkelter-DenbighLeisureMiltonKeynesBuckinghamshire-1998-12-31

But the sources for these are potentially incomplete because sets were often truncated to fit on cassettes but if the DAT masters still exist, there’s a chance of new unreleased or missed/forgotten sets might surface too (in the case of many other events at least). So what I’m getting at is can collections be open-ended in case of more to be added later?

Another thing is in the 90s rave scene there were bootlegs of bootlegs which sometimes had totally different details on the inlay from what they actually were!

Bit of a mission!

1 Like

Lots to cover, only a little time this morning. I’ll be back later in the day.

The main point is - most of your worries and questions are covered by the design of the database that MB uses. You are not the first person with a mission like yours, so the ideas have already been sorted through way deeper than you realise. They just need some explaining. It will make most sense when we start adding some data using the examples you have dropped.

The database has a structured way to look after your musical releases. And when we follow that, then any decent media player will play along.

As to OCD - why else do you think this place is here? No sane person spends this long on sorting out music. :crazy_face:

Starting at the end:

This is a MusicBrainz “Release Group”. Under one Release Group you will have:

  • The Release on Cassette media with track list broken down.
  • CD reissues.
  • Radio broadcasts of the event.
  • You’ll have a “single file per side” bootleg on Digital Media as is found on the archive.org link at the top of the thread.
  • Newer DAT releases that appear of the Whole Gig.
  • This would also include your “re-edits” of the CD and Tapes if these are in that same publicly accessible archive.

The idea of a Release Group is editions of the show on different Media are kept together. The theme being the event is in the same group, and each Release will have their own variation of a title.

A Release Group is more normally used to keep all versions of a “normal” release together: CD, Vinyl, Cassette, 10th Anniversary reissue, Deluxe editions.

You are talking “Bootlegs” here really, but the same concept is used. You are kinda lucky that the archive.org folder is publicly accessible as that makes these legit bootlegs that MB will document. If these were just private mix tapes they would not be admissible to the database.

Backups - a quick side step. Go buy an external drive or two NOW. Music is precious. Make snapshots of your music folder and keep them offline. Mirroring hard disks is for hardware failure, not protection against virus or human error.

You also want to take a snapshot now of your archive before you start. As you already know, there will be times errors are made and it helps to be able to go back in time.

iTunes - I’ll keep this short, but don’t build things around that player. It is one of the most awkward out there as it is really just an old shop front. And will soon be abandoned by Apple as they go more online. (Seriously - when was the last time it had an update?) Other threads cover how bad it is handling tags, images and updates. I don’t want to be distracted with that old rant here. :zipper_mouth_face:

Music storage - a flat file system is never good. iTunes is a shop selling artists and doesn’t really know what to do with your music. This is why we will find a better way of keeping it on the hard disk. But first we will be looking at getting the data into MusicBrainz. Later we talk Picard and what it can do for you with fancy file renaming scripts. MusicBrainz stores way more info about your music than any player can currently handle. I am a KODI fan and the music side of that media centre is built around MB’s tags. Winamp, Foobar2000, many other better players out there. But lets not go down that tangent yet. We need some examples in the database first. :slight_smile:

1 Like

MB has at least one of these Helter Skelter releases.

As many more are going to join it, we need to work out what should and should not be in the titles. Clean up the template.

Discogs is also clearly confused as it seem to have three different ways of naming things in your example. Not surprising for those Left Pondians as they often get confused about the UK.

Are they all the same gig “The Final Countdown 1998-1999”? That may be a good example of three that would be in the same Release Group together.

In that set of Discogs example you have a compilation set of 8 tapes, but also two separate tapes. How do they relate? Are they from the same 8-pack, or separate stand-alone releases?

1 Like

You’re in a community of people who spend thousands of hours adding music data minutae… don’t worry about it :stuck_out_tongue:

Re. how you’d like to title releases to sit in your own collection, or merge them into new releases, that has to take a back seat in the community database, where it’s more important that everyone adds data consistently/predictably. For instance putting the event name in the title might not suit these releases.

However as IvanDobsky has said, this DB is very capable of handling most of the data you can throw at it, and then a lot of that can be used in your tagging setup. The mixture of releases, release groups, collections, and events (that you can link releases + recordings to) means that you can probably go as deep as you would like. Your knowledge sounds invaluable so I’d say the quickest way is to just jump in with a Discogs import script and add a couple of releases! Stuff like bootlegs of bootlegs with different details sounds like one of those things that’s super satisfying to disentangle and get all tidy on MB!!

2 Likes

The Discogs examples are both the official release (which is considered unofficial, won’t have had barcodes) plus some bootlegs of those not released by Helter Skelter. I think the different 8-packs and 12-packs, also 4-packs and double packs released by Helter over their years in business were sometimes straight recordings of one arena, ie. the Technodrome, but other times collections of all genre sets between two main arenas. This event used to have drum & bass in arena 1, the Sanctuary, while happy hardcore was being played in Arena 2, Rollers, but then switch genres every hour, but I’ve been reassembling all the different sets and using the track tags to indicate which hour of which arena rather than which side of tape or disc. This is where rules get complicated for MB’s database and mine as the tapes or discs were often compiled in the order of most notable artists first rather than the warm-up acts.

1 Like

I have been meaning to learn scripting but so far have just been probably doing everything far more manually than is healthy so I will have to take directions, if anybody can point me to the right instructions. Thanks again.

As a start, you might want to check out the Scripting section of the Picard User Guide.

2 Likes