• scrobble
• getArtistInfo2
• getTopSongs
• getSimilarSongs
• getAlbumInfo

There should be a way to disable the integration, for those users concerned with privacy.

This will be independent from the current iTunes scanner

I know this is already on the roadmap, but I wanted a way to vote for it, so I created this issue :)

As far as I'm concerned, either a binary or a Docker image that works on Raspberry Pi (Raspbian Buster) would be great!

Browsing, playlists, covers, everything seems to be working, but it does not play songs.

Hello! First of all, thanks for your work! Looks very promising.

Actually using Airsonic.

I wanted to ask if it will be possible to have a playlist management (i saw smart / dynamic playlists on the road map) but i'd like to be able to manage standarf m3u / m3u8 playlists with the ability to import - export them as mine are very precious to me :-)

Added by @deluan to track progress:

Tasks:

• Add playlist REST endpoint
• Create/Delete playlists
• Browse/Play/Add to Queue playlists
• Remove tracks
• Edit playlist details (Name, Comment, Public)
• Reorder tracks
• Import M3U/PLS files
• Export to M3U files
• One-way sync of imported playlists
• View as albums, artists, songs. Similar to iTunes Will be implemented as part of a larger overhaul of the UI

Hi,
I think navidrome not support ogg files. All my album on ogg are name [Unknown Album] and i can't listen it

And navidrome can't get metadata from ogg

Add an option in the login window to keep the session valid for 1 week of inactivity

In 0.7.3 doesn't work transcoding.
Docker, ultrasonic.

Music Stash is using an API that was introduced in API version 1.11.0 (I think) to get artist info (getArtistInfo2) when browsing by artist and choosing an artist to view. Navidrome returns a 404 for getArtistInfo2 and getTopSongs.

I know Music Stash is not necessarily the only client being targeted, and that Navidrome is advertised as using API version 1.8.0, so if this is just a known "won't fix", then please close this ticket.

Also revisit image resizing, benchmark different options

This project sounds amazing!

If we could get smart playlists based on the skip to play ratio, we had a perfect dynamic playlist, that adjusts to our momentary listening Habits.

Lets say i have 6 root playlists :

Pacing: calm, active, neutral
Style: acoustic, electronic, mixed

All my music is sorted into these lists.

Now i do a smart list from calm and acoustic, which is dynamically updating by the skip to play ratio. Then songs with a ratio of < 1 will be played more often and songs with a ratio of > 1 will be played less.

If some one can recommend any other client based server, where this is possible please let me know.

This album is a V/A compilation. When I filter on the artist though:
http://localhost:4533/app#/album?filter={%22artist_id%22:%22cd4187b19abbcad3526c0e87460c19c8%22}&order=ASC&sort=year
I see only the regular albums of the artist:

The compilation tracks are not included, indicating that the filter only works on Album Artists, not Artists. Should probably be fixable with a WHERE EXISTS in the underlying sql query, so both Album Artist and Artist fields are included (and possibly in the future, fields like Composer, Arranger, Producer, etc too)

Many compilation albums (both multi-artist and single-artist) have songs from different years on them. So you tag them with their individual years, and that year will be used in searches, filters, smart playlists etc. That's the theory.

Navidrome does store individual years per track in its DB, but in the album view only shows the highest number. This can be fixed by not having a single Year attribute for an album, but two fields (query of the max and min of the track years) and for the UI a simple embedded if-query to display either the range ("1976-1989") or a single year.

Just built from master and starting ./navidrone yields

[ORM]2020/01/26 16:09:24 register db Ping `default`, unable to open database file: no such file or directory
panic: register db Ping `default`, unable to open database file: no such file or directory

just creating the data dir fixes it, but I wonder if it shouldn't be created automatically.

A similar error a second or two later if ./music doesn't exist:

ERRO[0002] Error importing MediaFolder                   error="open ./music: no such file or directory" folder=./music

adding it to the README could also be a possible resolution, but at least the first one should probably be a better error message, maybe after creating the dir?

Otherwise I can report that building with go 1.13 / nodejs 13.3 on FreeBSD 11.3 works flawlessly.

So I'm using a custom web UI that supports subsonic API servers. My client mimics the classic iTunes UI so to populate the library I have to make a lot of API requests to retrieve all information. It looks like every time a Subsonic API endpoint is hit, you're updating the DB with 'last access time'. I'm not sure if this is the only issue, but a good portion of the API calls I'm making are failing due to this locking issue. (Some are also 'Invalid login' messages, but not sure if that's related or something else).

If you want to test it out, here's the docker command to run my client. My library is pretty massive (over 80k songs) so this issue may only be a problem on larger libraries.

version: '3'
services:
  qooxtunes:
    container_name: qooxtunes
    image: alexphillips/qooxtunes
    ports:
      - 80:80
      - 443:443

With LastFM integration ( #18 ) right around the corner, it would be nice to have a preference to disable external lookups: I'm not at ease with that my music player leaking information to third parties :/

Had a folder called ./music/Band Name/Album Name (2008) with some MP3s (also with spaces) in them.

First I got

ERRO[0002] Error importing MediaFolder                   error="open music/music/: no such file or directory" folder=./music
ERRO[0002] Errors while scanning media. Please check the logs              
ERRO[0002] Error scanning media folder                   error="errors while scanning media" folder=./music

then I renamed Band Name to Band_Name and I got

ERRO[0002] Error importing MediaFolder                   error="open music/music/Band_Name: no such file or directory" folder=./music
ERRO[0002] Errors while scanning media. Please check the logs              
ERRO[0002] Error scanning media folder                   error="errors while scanning media" folder=./music

if I just copy one of the files to ./music/test.mp3 this is found.

This is FreeBSD, but I can't imagine the folder reading to be really different than on Linux.

My initial idea is that the first time you run the app, it will always redirect you to the /user/create page and won't let you navigate to anything else while you don't create an admin user.

A majority of the entries in the Albums and Songs view show up as [Unknown Artist] and [Unknown Album] for me. I think I've narrowed it down to the bracket characters [] being present in the filepath. I organize folders like "Artist/[Date] Album/#. Title.ext." The only items that show up correctly in these views seem to be the stragglers that I haven't yet cleaned up naming scheme on.

Interestingly, in most cases the top level of the Artists view is fine, even if every song by a particular artist has a bracket in the filepath (i.e. the album artists seem to be picked up from the tags). Same story as before if I click on an artist though.

Also allow Annotations (Starred, Rating, LastPlayed, PlayCounts, etc..) per user

As @deluan said:

Because Navidrome uses Material-UI, it is just a matter of generating a new theme (a json file) and using it.

It would be nice to document this somewhere, to enable people to create/share themes, maybe done via generators like this one.

It looks like parsing the year for some of my music is failing. I think I've tracked it down to the use of TDRC vs TYER tags. I use beets to manage my music collection, and pull metadata from MusicBrainz. beets (and other MusicBrainz based taggers, I think) will use the TDRC tag for release date rather than the TYER tag (which is just the year). ffmpeg then reports the date metadata output using the TDRC field, which can be a date string rather than a simple integer year.

It looks like the format in TDRC can be: yyyy, yyyy-mm, or yyyy-mm-dd. Lots of taggers that use MusicBrainz will use TDRC rather than the separate TDAT, TIME, TRDA, and TYER. The fact that TDRC also incorporates the TIME tag means it might be possible to find one with a time stamp added as well, but I've never run across one yet.

This looks great! Would love to see playlist support.

The metabrainz foundation is providing a couple of interesting services via api:

• musicbrainz to get information about artists/tracks/albums/…
• covertarchive to get covers of albums
• listenbrainz and to build smart playlists and to scrobble things

I might try to write an optional integration with navidrome as an alternative to lastfm.

• Implement repositories
  • Artist
  • Album
  • MediaFile
  • Property
  • Index
  • NowPlaying
  • Playlist
  • Checksums
• Add transactions
• Remover LedisDB Repositories (or rewrite it?)
• Refactor interfaces
  • GetStarred
  • Optional QueryParams

Is m4b support planned? I added my audiobook library but Navidrome doesn't seem to recognize m4b files.

Now that we don't interact with iTunes anymore, playlists must be maintained in the DB

Context: https://www.reddit.com/r/selfhosted/comments/et585q/navidrome_music_server_and_streamer_compatible/fffks9c/?context=3

In my test library for this I have 152k songs just a hair over 4tb. After about 40-43k songs processed it starts to noticeably slow down processing. I think the spawning of ffmpeg is taking its toll maybe as the gap between spawns is growing. its been running for about 5hrs and at 78k processed so far db size is 75mb. after I do this i will turn on trace and rescan again. i wanted to let this run its course for science ! heh

I havent looked at the db etc for structure or indexes (I am a db guy in real life) so ill have a peak at queries after its done too. staying about 130-138mb mem usage during processing according to ps in the docker. If you like ill just use this to keep a running list regarding initial load

Need to embed ffmpeg and ui bundle. Also document build steps for various platforms

Packaged versions should be generated in the pipeline, for the following platforms:

• MacOS
• Linux
• Windows
• Docker

Hi there, i tested navidrome and really love it, simple light weight and awesome addition to my home server setup.

I have one suggestion / issue

The ability to add a base url to navidrome so it can be accessed at a sub directory

I.e. myawesomedomain.com/music

I did attempt this with nginx but this failed to work (as a work around i grabbed a free subdomain from my provider to go to music.myawesomedomain.com) however i would love to have it as a /music

I can post the nginx config i used if needed but im sure its down it a call back to root instead of a base url.

This was running on a Pi4 b 4Gb ram with nginx 1.17.6

Thanks and great work

Implement a simple UI, using React-Admin with the following features:

• Manage users
• Simple Library Browsing
• Music Player
• Manage Playlists
• Configuration
• Rescan (with scan status indicator) - covered by #130
• Simple Statistics (Dashboard) - maybe later

I've tried a couple of Android subsonic clients (D-Sub and Music Stash so far), and D-Sub will show cover art where Music Stash doesn't. Looking at Navidrome logs, it looks like D-Sub is making separate getCoverArt calls where Music Stash does not... both are calling getAlbumList2. When connecting to AirSonic, both clients get cover art.

I've tried digging into the Subsonic API schema to see where the disconnect is but I'm unable to find it. I can try to provide logs if that will help as long as I know what I'm looking for.

While the web player in Navidrome is very simple it fits my needs perfectly. The only issue I have with it is if I have to pause for a while when I come back and press play I'm automatically logged out which is just a bit annoying.

Understandably this is a security feature, but would love it if there was a way to disable or increase this timeout to prevent this from happening.

It would be really cool to be able to browse by genres in a Subsonic client (I use play:sub) and in the web interface.

Is this planned soon?

I spun up the docker and upon going to http://localhost:4533 or the URI I set up, I am not prompted to create an admin account. My yml is below and a screenshot of the page.

services:
  navidrome:
    image: deluan/navidrome
    container_name: navidrome
    restart: always
    environment:
      ND_MUSICFOLDER: /music
      ND_DBPATH: /data/navidrome.db
      ND_SCANINTERVAL: 1m
      ND_LOGLEVEL: info
      ND_PORT: 4533
    ports:
      - 4533:4533
    volumes:
      - ${USERDIR}/assets/navidrome:/data
      - /mnt/media/music:/music
    networks:
      - traefik
    labels:
      - "traefik.enable=true"
      ## HTTP Routers
      - "traefik.http.routers.navidrome-rtr.entrypoints=https"
      - "traefik.http.routers.navidrome-rtr.rule=Host(`navidrome.${DOMAINNAME}`)"
      - "traefik.http.routers.navidrome-rtr.tls=true"
      - "traefik.http.routers.navidrome-rtr.tls.certresolver=dns-cloudflare"
      ## Middlewares
      - "traefik.http.routers.navidrome-rtr.middlewares=noauth-chain@file"
      ## HTTP Services
      - "traefik.http.routers.navidrome-rtr.service=navidrome-svc"
      - "traefik.http.services.navidrome-svc.loadbalancer.server.port=4533"

networks:
  traefik:
    name: traefik
    external: true```

That's how some clients talk to the Subsonic API. Ex: play:Sub

Example request:

curl -v -X POST -H 'Content-Type: application/x-www-form-urlencoded' 'http://localhost:4533/rest/ping.view' --data 'c=client&v=1.8.0&f=json&u=user&p=password'

• Implement as io.ReadSeeker
• Redo transcoding/downsampling
• Transcoding per user
• Transcoding per player
• Opus encoding
• Caching

Version affected: v0.4.1 (22d57a7)

Description:
Running navidrome with docker-compose with log level DEBUG and scan music folder mounted with goofys

$ docker-compose logs -f navidrome
navidrome_1              | panic: runtime error: invalid memory address or nil pointer dereference
navidrome_1              | [signal SIGSEGV: segmentation violation code=0x1 addr=0x38 pc=0x922c97]
navidrome_1              | 
navidrome_1              | goroutine 11 [running]:
navidrome_1              | github.com/deluan/navidrome/scanner.(*Metadata).Size(...)
navidrome_1              | 	/src/scanner/metadata_ffmpeg.go:43
navidrome_1              | github.com/deluan/navidrome/scanner.(*TagScanner).toMediaFile(0xc00017e6f0, 0xc000285bc0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
navidrome_1              | 	/src/scanner/tag_scanner.go:245 +0x487
navidrome_1              | github.com/deluan/navidrome/scanner.(*TagScanner).loadTracks(0xc00017e6f0, 0xc00040ac80, 0x40, 0x1937c38, 0x0, 0x0, 0x0, 0x0)
navidrome_1              | 	/src/scanner/tag_scanner.go:216 +0x125
navidrome_1              | github.com/deluan/navidrome/scanner.(*TagScanner).processChangedDir(0xc00017e6f0, 0x0, 0x0, 0xc00040ac80, 0x40, 0xc00031f8c0, 0xc00031f800, 0x0, 0x0)
navidrome_1              | 	/src/scanner/tag_scanner.go:154 +0x351
navidrome_1              | github.com/deluan/navidrome/scanner.(*TagScanner).Scan(0xc00017e6f0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xc00003e5e8)
navidrome_1              | 	/src/scanner/tag_scanner.go:61 +0x3ab
navidrome_1              | github.com/deluan/navidrome/scanner.(*Scanner).Rescan(0xc000186e00, 0xc00002404f, 0x6, 0xc000001b00, 0x0, 0x1)
navidrome_1              | 	/src/scanner/scanner.go:37 +0x18e
navidrome_1              | github.com/deluan/navidrome/scanner.(*Scanner).RescanAll(0xc000186e00, 0x0, 0x5, 0xc00003e7a8)
navidrome_1              | 	/src/scanner/scanner.go:50 +0xcf
navidrome_1              | github.com/deluan/navidrome/server.(*Server).initScanner.func1(0xc000186ea0, 0x34630b8a000)
navidrome_1              | 	/src/server/server.go:79 +0x5a
navidrome_1              | created by github.com/deluan/navidrome/server.(*Server).initScanner
navidrome_1              | 	/src/server/server.go:76 +0x172

So I would really want to submit a proper issue text here, but this issue is fairly small. There is currently no way to configure the windows binary at all, as the software does not automatically create a config file (navidrome.toml) in where the changes should be made, so I hope you can fix this.

Best Regards, -Loke

Musicbrainz uses the sort fields (specifically the 'Sort Artist' and the 'Sort Album Artist' fields) to store romanised versions of the content in the 'Artist' and the 'Album Artist' fields respectively for records that use non-Latin scripts. Making these fields searchable would immensely help in navigating media that do not have romanised data in any of the main tags.

Putting the UI search aside, piping them to work with the Subsonic API might require merging the sort fields with their non-sorted counterparts as the Subsonic API does not have separate fields for the sort tags. A possible implementation can be found in a modified fork of Airsonic, named Jpsonic which specifically caters to this issue.

The Musicbrainz styling guideline, however uses the field in a last_name, first_name format for individual artists. The field is used as-is compared to its non-sort counterpart if it's an alias or a band name. Articles are handled differently as well (The Strokes would become Strokes, The). The complete styling guideline can be found here. Regardless of these factors, the comma is always the delimiter; and so it might be a possible idea to tokenise and flip the tokens based on the position of the comma, check if the resultant is the same as the non-sort counterpart, and append to the sanitised string only if it's different, which would almost certainly be the case for non-Latin tags.

Listing the fields that should be searchable:

• Sort Album Artist
• Sort Artist
• Sort Album
• Sort Title

Just spun this up in docker and it automatically started scanning. It finished but only made into the 'B's. Hitting refresh on the UI doesn't seem to kick off a new scan. If it does, then it finishes almost immediately with no new files.

Compilation albums seem to get inserted into the Album list (sort by Artist) according to the first track artist on that compilation.

Compilation albums should either get their own category in the Library tree on the left (preferred way to handle them I think), or get sorted in the Album list either first before A or last after Z (preference #2), or least ideal, under the "V" for Various Artists.

It would be super useful to either have a little "Playing" icon or highlight the row or something of the song that's playing in the main tracklist (not the "Now Playing" popup list). Just makes it easier to see where I'm at in an album without needing to open up the Now Playing list which can cover the track lengths.

Source: https://www.reddit.com/r/navidrome/comments/fs2wow/playing_icon_in_track_list/

The navidrome db stores Genre for individual tracks. Many albums have songs from different genres on them. In the Album view, the navidrome UI only displays one genre, even if the album has tracks with different genres tagged (note: not multiple genres per track, just a different genre for each track). The UI should display it as a list, for example "Pop, Disco, House, Hip-Hop".

I was just trying to build this on a new vm and I noticed that

check_go_env:
        @(test -n "${GOPATH}" && hash go) || (echo "\nERROR: GO environment not setup properly!\n"; exit 1)

could be a bit problematic. I've not done a lot of go for 2 years, but the docs currently say

Create your workspace directory, $HOME/go. (If you'd like to use a different directory, you will need to set the GOPATH environment variable.)

But even if I put the checkout in this correct folder, this will fail if you don't set GOPATH, but this shouldn't be needed when it's in exactly this folder?

It would be nice to be able to force navidrome to rescan the music folder, while ignoring the timestamps.

Bonus point if this is made accessible to admin users via a button in the web interface.