As far as I understand it, some time in the past couple of years Spotify changed the way they manage their users’ playing history, and started only saving tracks playing through their apps and platform.
What that means unfortunately for us is that tracks played from local devices or using the Spotify web player (for example when you play music with Spotify directly on ListenBrainz) don’t appear in the music listening history.
This is extremely frustrating, and I don’t think there is anything we can do about it.
Quoting the Spotify forums:
Spotify only collects statistics on official music from the platform. If you’re listening to your local files - then Spotify acts just like an ordinary music player. These songs and the time you spent listening to the files that you’ve uploaded doesn’t contribute to your music recommendations.
I can suggest two —unfortunately not very satisfying— workarounds :
submit a listen manually from ListenBrainz page; look for this button towards the top of the page:
The main reason we open MB pages in a new tab is to avoid stopping music playback (for those who play music on ListenBrainz).
However, as @UltimateRiff suggested, this is going to change soon when we introduce artist and album pages directly in ListenBrainz.
On the technical side, we also want to make the LB website into a single-page application, to avoid loading an entirely new page when a user navigates around the website. This will help tremendously with the issue I describe above and we will then change our links to open in the same tab as you would expect.
I think my other major suggestion - and I don’t know if this is already on the cards or not possible - is that I’d like to see the mapper be more intelligent about artist/track aliases. In quite a few cases there’ll be artists on MB that use one name, but the name on streaming services like Spotify in other countries may be different. Both are official names for the artist, but currently it seems like no aliases are accounted for. Similarly for track titles where sometimes a translated title might also be used in official contexts but the primary track title on MB is the untranslated one
Actual listens browsing is core functionality to me. And editing of listens would complement it perfectly (maybe MB track linking is already sorta that, not quite sure)
I recently started using the android app, and it seems to be working well, capturing listens from the various apps I’m using to listen to local music. I noticed it also picks up YouTube videos and Google Podcasts (based on notifications, I assume). Apps like the official Last.FM scrobbler allow you to disable scrobbling specific apps - is there any way to do this in the ListenBrainz app, or are there any plans to implement this feature?
Ooh good catch, podcast listens are unlikely to be matched right now (because they are a bit of a nightmare to add to MusicBrainz), but I still track all mine for the day that they will be!
An optional feature I’d like would be a display of the how similar an auto match is. Title and artist identical or something it would show 100%(Likewise if alternate versions in different scripts are a perfect match), or if not whatever lower percent.
That or maybe a page that shows a list of tracks that aren’t a perfect match that I can verify myself similar to the missing data page.
I was listening to a song on loop then imported my listens from last.fm and noticed this. If I hadn’t had it looping I wouldn’t have noticed as I listen to the incorrectly matched song as well. Just made me wonder if there are others I don’t notice since I don’t closely monitor my listens while listening.
Something I would like is to be able to link MusicBrainz private collections and use that as sources for either new suggestions (additionally to listens) or to (more importantly for me) exclude tracks from new suggestions.
I have a collection with most of my CDs, and ListenBrainz keeps suggesting me to listen tracks I have in this collection but that I don’t have in my listens. It would likely improve recommendations made to me if they were considered as previously listened.
I agree, but I think there should be options with how collections interact with collections. for example, I’d like to be able to include stuff in my collection in recommendations, as I don’t always listen to albums when I get them, so I might miss a hidden gem or two here and there
Ok I’ve sort of mentioned this elsewhere but putting it in here too for maximum visibility
I would very much like for there to be a way to (at the very least) link to a release or release group. Even better if it can just use the release MBID submitted with tracks if present. In general I feel like Listenbrainz is very bad when it comes to handling releases. It just doesn’t work. It seems that it selects a random release for a recording and calls that a day, but this has a lot of issues. For example: a recording can exist across multiple release groups. Say it’s released as a single first and then included in an album later. Right now Listenbrainz has no way of distinguishing the two. It will just pick one and say that’s the one. This is, quite frankly, incredibly frustrating. Doubly so when recently most of my new listens are not actually mapping to a release automatically leading to those releases not showing up in my stats
I submit release MBIDs with my tracks and it actually displays the correct release on the listens page (this is the only way I know of currently to specify a release), however it is seemingly not possible for this to then be reflected in the stats page which continues to use the auto-matching result
I hope it’s clear why this would frustrate me or any other potential user - these statistics are wrong and there’s no way to correct them seemingly
My main playlist that I play on shuffle has grown past 1,000 tracks. A few times now I had someone mention, or stumble on information showing how a lot of tracks I enjoy were written by or composed by the same person despite the tracks being performed by different artists. With the amount of tracks on in my “library” it would be hard if not impractical to try and find writers and composers who’s music I tend to enjoy.
I think a nice feature would be for ListenBrainz to not just understand which tracks I like based on my listening, but find commonalities like writers and composers and such and present them to me. I am unsure how or where this should be displayed. Maybe just extra recommendations like “We noticed you enjoy these songs written by ___ here is a playlist of others they wrote”?
I think it’d be cool if we could implement from either importers like stats.fm and the like or import CSV files from Spotify history directly in ListenBrainz.
A functionality to transform/migrate a recording series from MusicBrainz into a ListenBrainz playlist. Or something similiar, like the “Play on ListenBrainz” button for releases.
Thanks for the suggestion, I quite like the idea
I opened a ticket to track this: [LB-1355] Allow importing/playing a MusicBrainz Series - MetaBrainz JIRA
We might not get to it for a while, but we have already discussed various ways to make playlists/radios from curated MusicBrainz data so this fits right in !
I tried to search and couldn’t find a previous mention of this:
I would love a feature added to the stats page showing a persons top discoveries for the period, similar to the feature on https://lastfmstats.com/
Mine for example
Average scrobble date for tracks with at least 10 scrobbles, sorted descending
These may not show up in a persons normal top listens for the period, so it would be different from existing top lists on the stats page. It would be great to browse similar users and see their top discoveries to find new songs to check out that they like.
