I added a new Compilation CD with Various Artists and I had several problems to understand the MusicBrainz workflow and I think some things should be optimized
1.) Title list: Entering a song title does not cause open a suggest dropdown. As an example: I entered “Dance Monkey”. This song already exists in the database, so why do I need to enter the artist manually? I mean having a specific title length and a title name should give us a 99% hit rate for the dropdown. The same could be done the other way around. This means if I enter “Tones and I” and the length fits only to some songs from “Tones and I” they should be listed in the dropdown. Example:
2.) Entering multiple artists is not explained. The button shouldn’t be called “edit” I think. Instead it should be called “add multiple artists” or similar
3.) The popup to add multiple artists should show multiple lines by default. It does not make sense that the user needs to press the “+” button. Every click saved, saves time.
4.) Why is it needed to add the “link” phrase manually? Why no dropdown with:
" , "
" and "
" feat. "
" feat "
" & "
Custom
And if the user selects “Custom” he can enter a phrase manually.
5.) There is no option to add a second/third medium to an existing release. The only option I found was to add a new release and finally merge it. Is this the usual way? If yes, please add an “add medium” button to the release page to save this step.
6.) If a new release is added some options should be pre-selected like “Status = Official”
7.) If I select “German” as the language it should pre-select “Latin” as font (same goes for other languages). The user is still able to select a different font if needed.
8.) If the package dropdown is opened a help text with a link to this page is shown. Instead it should show Mini-Thumbnails of the different packagings directly in the help text.
9.) “Add Medium” is missing if you use Picard to transmit a CD-TOC:
10.) If its not allowed to add Spotify to CDs (phyisical releases), it should be simply not possible or at least a warning/hint should be displayed if a user tries to add a Spotify link to a CD.
The answer to a lot of this is so that people entering data don’t get lazy and enter something inaccurate because it’s easier. The general philosophy here is that no data is better than bad data.
In the case of 2 and 3, there are many cases where it is used to enter a specific artist credit rather than multiple artists.
I like 4, although also possibly vulnerable to the first point. 5 does exist already. I also like 7, but what about languages which could have more than one script?
The interface is complex as it covers so many different use cases, you’ll get used to it.
Attempts have been made to redesign it to something more modern\usable but I think those developers ended up locked away gibbering at the enormity of the task…
The artist does give you a drop down choice. Track names are too likely to have been repeated on multiple releases. Would expect too many clashes. (And the current search at MB really isn’t up to the task) Also these are more often be imported from other sources.
2 and 3) That is a multi-purpose edit. Majority of the time you only have a single artist who is copied down a release. Some people need to hit EDIT as they have a different alias for the artist they need to type.
The + already puts " & " in automatically.
Add Medium is there on the Track List page. And lets you add in multiple ways. (Bottom right of the page)
Yep, agree. But this also can lead to too many people leaving it wrongly selected. You’ll find the MB way is “no data is better than wrong data”
Agree here as it would certainly make sense to assume a standard for most European languages. I have never understood why those two boxes are not linked. How often is English \ German \ French not Latin?
This means I entered “Stupid Love” and it returns only “Lady Gaga” as a artist suggestion, because its the only “Stupid Love” recording which has a length of 3:14. Of course the user can still enter the artist manually if there is no matching entry in the database.
You are right. And it is intelligent as well as it adds commas for more than 2 artists. Nice. But again something, which is not explained. Maybe the “link”-column should be hidden as long the user did not click on the + symbol. And if he clicks it, the “link”-column is shown with the ampersand in it. By that the process is self-explaining.
If my suggestion #3 is realized, it could become visible after the first char has been entered in the artist field of the second row
The complete process, registration, editing existing entries to understand the workflow etc is so huge, that I don’t think you will ever motivate such lazy people to use MusicBrainz.
Ok, then it (2) could be called “edit artist(s)”. Maybe I’m the only one, but for me it was not clear that this button is to edit the artist name and/or adding multiple artists.
And having multiple fields (3) should not be a problem at all as empty fields shouldn’t be added to the database.
The user should be still able to edit the field. Its only pre-selected and not greyed out.
The search has problems even finding “Money” as it matches multi-word. I can see what you are asking for, but I would also expect it will be adding a chunky load to the server if you want it to start matching on the first letter.
Also how will it have a track time? That is usually typed after the name.
I think it is a lot of Dev time for a smaller use case.
There are also quicker ways of adding track names using the < Track Parser > button or < Add Medium >
Yeah - it is a bit hidden. Go back and edit the track list and look at the bottom of the page
Just had a look at the edits and can see how the confusion has occurred. Merging is the best fix this time, but if you poke around with that Add Medium button you’ll see you can also link your CD2 across that way and add it to this Release.
Good. They are no loss. Data like this needs properly checking.
I add only entries for CDs I own. This means I rip it with EAC and if it does not return an entry I add it to MB. And this means opening Picard and being forwarded to the website with the Disc ID and title lengths. Thats why they already exist when I start to add the title names.This is the easiest way as it saves me to enter those values, too
Server load is the same as it is with artists suggestions. Its even lower if you compare the title length.
Do you really think its small? If it works with artist<>title and title<>artist this could save a massive amount of typing work I think. But of course it depends on the priorities of the dev team
Thank you. Maybe the same link should be added to the release overview page.
Much of the trouble this time has been because you jumped in at the deep end with a compilation and therefore many different artists.
Did you type the track names into EAC? If you did, you can drop the ripped files into Picard and “Cluster” them on the left. Then Right click \ Run Scripts \ Copy Cluster to Clipboard. This can then be pasted into the Track Parser window.
So many ways of adding this data… think that’s why search is not automated in the way you are requesting.
I can’t talk about the Dev team - let them answer for themselves. But I know they are overloaded with changes anyway. They have a Ticket system for that kind of request.
The “Add Medium” button does not make sense on the Release page as usually you will be looking at all the CDs already completed on that page. The changes happen behind the < Edit > tab.
I think it is because you came into MB via Picard “Submit a DiscID” that made it different for you. I think the more normal flow would be to add the release first, with its multiple CDs, and then
Picard can also be used to upload track names from the ripped files.
My method for adding a new multi-disk release is approx as follows:
Rip CD1 in EAC. Manually type in the track names in EAC and add the basic details (title \ artist)
Drag these ripped and tagged FLAC files into Picard. CLUSTER on the left.
Start Ripping CD2 in EAC. Manually typing in track names again.
While that rips, back in Picard I look at that Cluster. Right Click, Run Scripts, Add Cluster as Release
This starts my new Release in MusicBrainz. All the track names are now in place, along with approx track times. Transferred automatically from Picard
I fix the Artist choices as a list - one by one, just going down the list of artists using the artist lookups.
CD2 finishes ripping. I drop this into Picard and again make a CLUSTER, but this time I right click, Run Scripts, Copy Cluster to Clipboard. (careful to only have the CD2 tracks)
Swap to the website and Add Medium on the Tracklist page.
Press Track Parser button, and PASTE in my tracks from the clipboard. Tweak the tickboxes as needed. Bing - I have track names, artists, and times
Now I have two mediums in place.
Next to associate the Recordings one by one if these already exist, or make a new batch.
Finish the new release. Save and submit it to MusicBrainz
Now I go back to Picard. CD2 is still in the drive. So I Lookup CD and Submit the DiscID.
I have my new release to attach it to.
Swap CDs in drive and repeat - Lookup CD, Submit DiscID.
After submitting that all the CDs I’ll check the DiscID tab and make sure I have Set Track Duration for both CDs. This will set the track times to 1000ths of a second, overriding the times from Picard.
Now I look for that Green TAGGER button to push that new release back to Picard.
Picard now has my new release on the right. So I can drag my files across.
Check all is well, and do a SAVE
Bonus points - I now generate Fingerprints and Upload AcoustID
And finally artwork…
But everyone does things different.
Some will use the Add Medium button to copy track details and recording links from another Release with a similar layout.
Many people will be importing data from elsewhere - this is why your method is slower. That Tracklist page is designed to handle so many different methods. (We haven’t even touched on the scripts for scraping websites like iTunes, Discogs, etc)
When done like above all the artists are fixed at the same time. After fixing one, you click next to go to the next artist. Just a list of artists. I don’t think I have ever entered details like you did the first time where you are swapping track \ artist \ track \ artist \ etc.
Yes, all of this is overly complex when you first use it, but after a while you find the power in some of the odd options here.
Thank you for posting your workflow. Finally this adds more work as you need to type in the full artist names and you have to correct them afterwards although you already fully typed them. And I don’t Drag & Drop or sync any data between Picard and MusicBrainz. Instead I have much less steps:
open CD in EAC
open Picard if no MetaData has been found by MusicBrainz
send CD TOC to MusicBrainz
if release exists, attach it, if not, add release. As the track lengths are already present, I only need to add the artist names (which are auto-correct/auto-suggested) and track names. In the same step I add EAN, Language, Cover Art, etc
retry CD in EAC a few days later. Done.
this is why your method is slower
As you can see in to steps above, my method is much faster for single-disk releases. The only problem is, that my method does not allow adding an additional medium to an existing release as the website has no option for this:
After that I wanted to add the CD TOC to this placeholder, but sadly it’s not shown through the attach CD TOC process (Medium 3 is missing):
And this must be an error as this list should only contain entries without CD TOCs. But as Medium 1 and 2 already have TOCs, only Medium 3 should be displayed.
So I tried to “hack” the process. I checked the HTML source-code of Medium 3, to obtain the Medium ID and overwrote the “Medium 1” entriy it in the Attach CD TOC form:
The solution is simple: Do not enter any (temporary) track lengths (just the correct amount of tracks - else a TOC can not be attached as you already have noticed) and attach the TOC, the track lengths will be set automatically. If there already are placeholder lengths you have to explicitly tell MB to overwrite them with the lengths from the TOC.
I have also cut out details in the above list as it was long anyway. It is what works for me. I have many variations on that routine depending on what type of release it is. I also tend to be focusing on single artist releases.
If I am being impatient then I will pre-fill a release via a DiscID from Picard. If it is a really long boxset I will probably have copied the track list in from elsewhere. Maybe a Discogs import, maybe a copy\paste to the rather useful Parse Tracks button. I do a lot of prep in notepad++ on the really messy releases.
I also don’t understand your need to “hack” the process. As already explained, adding a DiscID is there and easy to do. The interface is complex partly because it covers so many different ways to do things.
-=-=-
Actually now I re-read my details above that is the simplified version. I was trying to be clear.
I also often submit discIDs from the EAC logs. Putting these into place AFTER everything else is in order. This page needs an EAC log copy\pasted to it (at least the TOC from the file) https://eac-log-lookup.blogspot.com and I use that to throw in the multiple CDs to attach as by now the physical CDs are back in the box on the shelf. I’ll have the Release pre-seeded with approx times open in another window as I’ll need to copy\paste the Release ID from that page into the DiscID box when it appears to speed up my searches.
Also most release are I enter are single artist, single CDs and fit the above workflow better.
Haha - if I fully documented how I exactly doing things I doubt your need for speed would like it. Notice there is no mention of artwork scanning in the above details? And the AcoustIDs. Then the discog links, annotations about different editions, matrix details… I was trying to keep simple.
Here is my workflow to add releases with multiple discs (which is as simple as it is possible with the current website AFAIK):
Find a text-based tracklist somewhere on the internet and feed it into the track parser of the release editor (for each disc). Now you have all the textual information already available and do not have to type anything yourself (in the best case).
Do not care about the track lengths now and leave them empty, you can still add them later.
If this is a release by a single artist and you have already entered the general release information in the first tab you can now submit your edits. For releases with various artists you have to select the correct artist for each track by pressing the magnifying glass icon.
a) If there are multiple artists per track you have to press the edit button on the right instead and add them all on a separate credit line (by duplicating the pre-filled text-based credit for each new artist before pressing the magnifying glass icon).
b) This popup dialogue also provides a checkbox to change all matching artists in one go which is very useful if a credit occurs on more than one track of the release.
After you have submitted the release you can lookup all the CDs and attach their TOCs to the release. The track lengths will now be set automatically.
I think this is where @mgutt you are getting confused with your Add Medium post. Just start with basic data in the release and then expand. Add three mediums with correct track counts in your initial edit and only at the very end try and upload the full set of DiscIds.
“Add DiscID” is more about “align this discID to a current release” and not really about “add this new CD”
That Track Parser is excellent. Especially when combined with Notepad++. If I have something especially long I’ll work in Notepad and make the track list up in there. It lets me copy\paste from all kinds of sources and trim the text down. I then feed the track parser, following @kellnerd’s steps, and follow up with hitting the Guess Case button.
You can massage a new release in the first 24 hours and it will update your edits immediately. The database expects you to come back and tweak track titles.
