Music Library Import - Emby changing Metadata & Wrong Artist/Album Artist (I've tried everything!)

[DarrenClimie-Newman 6]

11 minutes ago, Happy2Play said:

No issues adding those test files either in 4.5.4.0.

As for the command line you need quotes around path.

ffprobe -i "C:\Users\Media\Desktop\Videos\Music - emby metadata\Completed\UK Garage\DJ EZ - Ez Intro 5.mp3" -show_streams -show_format -print_format json

So you need to either cd to the folder containing ffprobe or type out path to ffprobe.

C:\Users\Media\AppData\Roaming\Emby-Server\system>ffprobe -i "C:\Users\Media\Desktop\Videos\Music - emby metadata\Completed\UK Garage\DJ EZ - Ez Intro 5.mp3" -show_streams -show_format -print_format json

 

Sorry Happy2play, i'm struggling to get this to run. I've moved the files to here:

ffprobe is in : C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system

What do i need to type in at the '\Desktop>' location? 

Sorry, I appreciate your help on this! 

[PenkethBoy 2063]

put the full filename in quotes as you have spaces

and if you dont have ffmpeg/ffprobe in the same folder that needs a full path to the exe as well in quotes

Edited March 11, 2021 by PenkethBoy

[Happy2Play 8282]

8 minutes ago, DarrenClimie-Newman said:

Sorry Happy2play, i'm struggling to get this to run. I've moved the files to here:

ffprobe is in : C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system

What do i need to type in at the '\Desktop>' location? 

Sorry, I appreciate your help on this! 

first type  "cd C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system" to change to that directory

ffprobe -i "C:\Users\Darren & Gemma\desktop\DJ EZ - Ez Intro 5.mp3" -show_streams -show_format -print_format json

Edited March 11, 2021 by Happy2Play

[DarrenClimie-Newman 6]

2 minutes ago, PenkethBoy said:

put the full filename in quotes as you have spaces

and if you dont have ffmpeg in the same folder that needs a full path to the exe as well in quotes

2 minutes ago, Happy2Play said:

first type  "cd C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system" to change to that directory

I cant seem to get to  that path location:

I think it has something to do with the Username having an '&' symbol......:

[Happy2Play 8282]

2 minutes ago, DarrenClimie-Newman said:

I cant seem to get to  that path location:

I think it has something to do with the Username having an '&' symbol......:

Yes I don't see this issue on Windows 10 but you need quotes whenever there are spaces.

cd "C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system"

but the long method should work also

"C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system\ffprobe" -i "C:\Users\Darren & Gemma\desktop\DJ EZ - Ez Intro 5.mp3" -show_streams -show_format -print_format json

 

[DarrenClimie-Newman 6]

Finally! Sorry about that! 

ffprobe version 4.3.0-emby_2021_02_27-g8f222573e9+634 Copyright (c) 2007-2021 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 9.2.0 (Rev2, Built by MSYS2 project)
{
Input #0, mp3, from 'C:\Users\Darren & Gemma\desktop\DJ EZ - Ez Intro 5.mp3':
  Metadata:
    album           : Ez Intro 5
    artist          : DJ EZ
    album_artist    : DJ EZ
    genre           : UK Garage
    title           : Ez Intro 5
    date            : 2000
  Duration: 00:04:42.33, start: 0.025056, bitrate: 322 kb/s
    Stream #0:0: Audio: mp3, 44100 Hz, stereo, fltp, 320 kb/s, Start-Time 0.025s
    Metadata:
      encoder         : LAME3.99r
    Stream #0:1: Video: mjpeg (Progressive), yuvj444p(pc, bt470bg/unknown/unknown), 600x537 [SAR 1:1 DAR 200:179], 90k tbr, 90k tbn, 90k tbc, Start-Time 0.025s (attached pic)
    Metadata:
      title           : Cover
      comment         : Cover (front)
    "streams": [
        {
            "index": 0,
            "codec_name": "mp3",
            "codec_long_name": "MP3 (MPEG audio layer 3)",
            "codec_type": "audio",
            "codec_time_base": "1/44100",
            "codec_tag_string": "[0][0][0][0]",
            "codec_tag": "0x0000",
            "sample_fmt": "fltp",
            "sample_rate": "44100",
            "channels": 2,
            "channel_layout": "stereo",
            "bits_per_sample": 0,
            "r_frame_rate": "0/0",
            "avg_frame_rate": "0/0",
            "time_base": "1/14112000",
            "start_pts": 353600,
            "start_time": "0.025057",
            "duration_ts": 3984261120,
            "duration": "282.331429",
            "bit_rate": "320000",
            "disposition": {
                "default": 0,
                "dub": 0,
                "original": 0,
                "comment": 0,
                "lyrics": 0,
                "karaoke": 0,
                "forced": 0,
                "hearing_impaired": 0,
                "visual_impaired": 0,
                "clean_effects": 0,
                "attached_pic": 0,
                "timed_thumbnails": 0
            },
            "tags": {
                "encoder": "LAME3.99r"
            }
        },
        {
            "index": 1,
            "codec_name": "mjpeg",
            "codec_long_name": "Motion JPEG",
            "profile": "Progressive",
            "codec_type": "video",
            "codec_time_base": "0/1",
            "codec_tag_string": "[0][0][0][0]",
            "codec_tag": "0x0000",
            "width": 600,
            "height": 537,
            "coded_width": 600,
            "coded_height": 537,
            "has_b_frames": 0,
            "sample_aspect_ratio": "1:1",
            "display_aspect_ratio": "200:179",
            "pix_fmt": "yuvj444p",
            "level": -99,
            "color_range": "pc",
            "color_space": "bt470bg",
            "chroma_location": "center",
            "refs": 1,
            "r_frame_rate": "90000/1",
            "avg_frame_rate": "0/0",
            "time_base": "1/90000",
            "start_pts": 2255,
            "start_time": "0.025056",
            "duration_ts": 25409829,
            "duration": "282.331433",
            "bits_per_raw_sample": "8",
            "disposition": {
                "default": 0,
                "dub": 0,
                "original": 0,
                "comment": 0,
                "lyrics": 0,
                "karaoke": 0,
                "forced": 0,
                "hearing_impaired": 0,
                "visual_impaired": 0,
                "clean_effects": 0,
                "attached_pic": 1,
                "timed_thumbnails": 0
            },
            "tags": {
                "title": "Cover",
                "comment": "Cover (front)"
            }
        }
    ],
    "format": {
        "filename": "C:\\Users\\Darren & Gemma\\desktop\\DJ EZ - Ez Intro 5.mp3",
        "nb_streams": 2,
        "nb_programs": 0,
        "format_name": "mp3",
        "format_long_name": "MP2/3 (MPEG audio layer 2/3)",
        "start_time": "0.025056",
        "duration": "282.331429",
        "size": "11374575",
        "bit_rate": "322304",
        "probe_score": 51,
        "tags": {
            "album": "Ez Intro 5",
            "artist": "DJ EZ",
            "album_artist": "DJ EZ",
            "genre": "UK Garage",
            "title": "Ez Intro 5",
            "date": "2000"
        }
    }
}
 

I cant see 'Agent X' anywhere in the output of that???? 

[DarrenClimie-Newman 6]

This is the output for the Todd Edwards Track:
 

C:\Users\Darren & Gemma\AppData\Roaming\Emby-Server\system>ffprobe -i "C:\Users\Darren & Gemma\desktop\Todd Edwards - Echo of The Past.mp3
ffprobe version 4.3.0-emby_2021_02_27-g8f222573e9+634 Copyright (c) 2007-2021 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 9.2.0 (Rev2, Built by MSYS2 project)
[mp3 @ 0000029bf214cb00] Estimating duration from bitrate, this may be inaccurate
Input #0, mp3, from 'C:\Users\Darren & Gemma\desktop\Todd Edwards - Echo of The Past.mp3':
  Metadata:
    album           : Echo Of The Past
    artist          : Todd Edwards
    album_artist    : Todd Edwards
    genre           : UK Garage
    title           : Echo Of The Past
    date            : 2012
  Duration: 00:06:04.75, start: 0.000000, bitrate: 321 kb/s
    Stream #0:0: Audio: mp3, 44100 Hz, stereo, fltp, 320 kb/s
    Stream #0:1: Video: mjpeg (Progressive), yuvj444p(pc, bt470bg/unknown/unknown), 410x410 [SAR 1:1 DAR 1:1], 90k tbr, 90k tbn, 90k tbc (attached pic)
    Metadata:
      title           : Cover
      comment         : Cover (front)

[Guest]

It would sound like Emby is using the previous metadata, but the metadata being wrong and being inconsistent bothers me. I have the same kind of files in my archive ( without the issues )

One mistake that I did see earlier in your thread, generating an albumartist from artist is going to cause problems..

for DJ EZ's file..  my suggestion which should work for more than Emby would be... 

Structure = \\Music\@VA - Warner Music ( or Pure Garage )\1999 - EZ - Pure Garage III\<---files...

                    \\Music\Todd Edwards\2012 - Shall Go EP\<---files...

File information= see attached..

If I identified it correctly.. It was a compilation which is copyrighted and created ( final approval of tracks by.. ) Warner Music.. I put the Series in as album artist as they get final say on what gets mixed. I started to put in the Record Label that released it until I realized it was a series like fabriclive, DJ-KiCKS, .... credits to the DJ exist on the album name, and you could put them into 'remixedby' fields. Regardless you may need to point the file to something it can find or something it can't which exceeds the curse of the 'Various Artist' entry which plagues da earth.. It should rely then on only information within the file.

I follow this sort of thinking about rights and knowing a little about the industry to label my stuff, BEYOND the information available online as data returned differs and causes issues.. like Various Artist.. Also discographies are more pure to the artist actual creations, and credit given to original artist names. I use record labels which are creating compilations when a series is not present, and rarely will there be anymore DJ's releasing mix CD's because professionally you get into lawyers, usage rights per artist.. and actual credit goes to remixing/mixing the tracks/disc. Song credits always go to the original artist even if its remixed. Usage rights have to be addressed and usually a collab done with the artist to get it released... writing a book.. trying to help you.. sometimes our systems in place don't cover it really though and we are just using machines..

I have files that I literally have to go online and search for information to have it correct. Generating album info only helps when its correct and I don't recommend it.. is useful when You have no metadata to narrow down results ( usually done from file name info ) But you will probably narrow down the results and get a better result with not just Emby.. but other programs as well with the metadata I suggested.. and if you are not working from a Professional DJ Folder inside your mobile for your software and equipment, I would suggest the structure for the archive.. Your working DJ Folder should be one level deep with \\Music\Genre\<---files.. Each file should have its full metadata, stripe, BPM, Key so on.. Speed of accessibility and finding what your want from your collection to quickly load the next file and cue...when you roll those over and add new stuff and are ready to pull them out they should go into the archive with the structure format I suggest above. Easy filing.. the same throughout and easier to find.. continuity for speed.. Those files can still be copied out for playlist creation for a set.. or sampling.. what not. ( even put into a resource or sampling folder after cut and scratched- chopped and screwed.. whatever you want to say ).. Most programs will allow you to choose by Genre in the end,, DJ scenario, professionally is different.

A suggestion

Edited March 11, 2021 by Guest
Forgot attachments..

[ginjaninja 536]

On 10/03/2021 at 19:55, Luke said:

One compromising idea might be to prefer the Artists tag, but only when there are actually multiple artists. When there's just a single artist, use the Artist tag. This would get us closer to what most other apps display, while still supporting the Artists tag for the situation in which it is needed.

@ginjaninja

@Vicpa

can you clarify what problem you are looking to improve/overcome, i havent followed yet so cant comment well.

my understanding is emby might potentially  look different to other apps (which dont support  music brainz tagging), when

•  musicbrainz tagging (artistS & mbzids) is employed AND multiple artists are present typically,

but in this case the difference is intended,welcome (presumably).

maybe a checkbox on the music fetcher config "use musicbrainz tagging" defaulting to off would suffice.

like i say i havent appreciated what the issue is yet.

 

 

 

[Luke 37068]

2 hours ago, ginjaninja said:

can you clarify what problem you are looking to improve/overcome, i havent followed yet so cant comment well.

my understanding is emby might potentially  look different to other apps (which dont support  music brainz tagging), when

•  musicbrainz tagging (artistS & mbzids) is employed AND multiple artists are present typically,

but in this case the difference is intended,welcome (presumably).

maybe a checkbox on the music fetcher config "use musicbrainz tagging" defaulting to off would suffice.

like i say i havent appreciated what the issue is yet.

 

 

 

When there's only a single artist, the Artist tag is much more commonly used than Artists, so the change is to have the server prefer that, unless there are multiple artists, then prefer Artists tag first.

[DarrenClimie-Newman 6]

Thanks for the suggested recommendations on using Album tags, however, I've already tagged 3400 tracks, and I'm not going back through all of that to fix an issue, that will clearly be resolved in 4.6.

Similarly, I want the original release year of the track, not many years later in a compilation album, which is why I've used Discogs for the original year and artwork, most of my tracks are original vinyl rips from my collection.

@Luke - How long until 4.6 will hit GA? I'm happy to move to the 4.6 beta with a fresh install on my QNAP, as I can pull my watch history in TV and Movies from Trakt, so no big problem there. Are there many bugs in the 4.6 beta today?

Edited March 12, 2021 by DarrenClimie-Newman

[Luke 37068]

 

[ginjaninja 536]

4 hours ago, Luke said:

When there's only a single artist, the Artist tag is much more commonly used than Artists, so the change is to have the server prefer that, unless there are multiple artists, then prefer Artists tag first.

artist is more commonly used than artistS...nothing uses artistS apart from musicbrainz advanced tagging (afaik)...sound to me.

But when artistS is present it is almost certainly there on purpose by people who know what it means and manage their tags.

so the suggested change seems unnecessary albeit built on a sound reason BUT ONLY

as long as emby can 100% identify multiple artists, which as we know is very difficult to get right 100% of time, indeed the only cast iron method is looking at musicbrainz's artist_mbzid tag. (no room for discrepancy here on how many artists there are, as although taggers might use an arbitrary separator, mbzds are fixed length)

Spoiler

(or possibly use id3v2.4 with something other than ffprobe for mp3 - not realistic in medium term).

 

so if the suggestion is to

use artist generally, and only use artistS when multiple artist_mbzids are present

then that sounds very robust to me. (albeit i am not seeing the necessity for the change)

 

Edited March 12, 2021 by ginjaninja

[Luke 37068]

44 minutes ago, ginjaninja said:

artist is more commonly used than artistS...nothing uses artistS apart from musicbrainz advanced tagging (afaik)...sound to me.

But when artistS is present it is almost certainly there on purpose by people who know what it means and manage their tags.

so the suggested change seems unnecessary albeit built on a sound reason BUT ONLY

as long as emby can 100% identify multiple artists, which as we know is very difficult to get right 100% of time, indeed the only cast iron method is looking at musicbrainz's artist_mbzid tag. (no room for discrepancy here on how many artists there are, as although taggers might use an arbitrary separator, mbzds are fixed length)

  Reveal hidden contents

(or possibly use id3v2.4 with something other than ffprobe for mp3 - not realistic in medium term).

 

so if the suggestion is to

use artist generally, and only use artistS when multiple artist_mbzids are present

then that sounds very robust to me. (albeit i am not seeing the necessity for the change)

 

yea the issue is that a less commonly used tag easily becomes a forgotten tag, and when users use other editing software it may not get updated, and therefore left with incorrect values. And then of course it becomes our problem due to all of the troubleshooting. So that's why I think when there is only one artist, sticking to the Artist tag makes sense.

[Guest]

I would agree that all standard tags for the data/music/filetype as in standard ID3 v2.4@UTF8 should be supported ONLY..

Making caveats or considerations for every program, database, update, or evolution in databases ( such as musicbrains id's which are good for implicitly identifying it within musicbrains only,  so on ) Is impossible to effectively cover and would mean constant change instead of a baseline/standard to how that data is used and aggregated. In the next ten years you could see so many changes in how this is done that it isn't even funny. 20 years ago there wasn't much of a standard in some cases not even a way.. Those methods and standards have changed immensely and most of the databases that started aren't even used today or around... The way programs use it. OR rather if they even supported it at all.. is even in question looking at it..

How we move forward and actually look at that area should probably start with a standard when creating a program to work with that data for compatibility across the board...

 

[PenkethBoy 2063]

You cant apply a standard to something where as this thread has highlighted - people have very varying quality of Metadata Tags any app needs some level of flexibility

As all you will do is create another set of problems

Emby has been very flexible in its ability to read data from files and also using the file structure make generally a good guess as to what the music is etc with 4.5 and before

the newer beta implementation can be more strict - depending which structure option you choose - but i your metadata works with 4.5 then there is a high chnace it will work with 4.6

Either version works fine (excepting a few bugs that still need to be ironed out in 4.6) - IF you have consistent accurate metadata 

As for MBz ID's they are used by other apps/web sites - e.g. AudioDB so not limited to just MBz

 

[Guest]

I think you miss my point in what I am saying.. but its okay..   

[PenkethBoy 2063]

nope - you could have written it more clearly though

[Guest]

Maybe this will help a little in what I mean... I look at my tags differently as far as standard..

Keep in mind that because of extensibility, ID3v2.4 and ID3 format are flexible ... Like I was mentioning above, any 'Tag' may be created and filled in with any data. Like musicbrains, which is a custom ( versus standard tag entry ) and aggregates or is used to aggregate from an ID or information that is not usable with any other online database or database resource. IF ( forbid ) musicbrains went by way of the dodo.. lost funding, was destroyed in a nuclear blast = the tags become useless. Saying therefore, they are specific for musicbrains. Going through 139,000+ songs ( now as of today ) and modifying those files your are curating the data for an information database. A lot of work if you never plan to pass on the record shelf, or if you plan to go and do it again in the future.

SO.. my thoughts ( for my archive ) were to make sure that standard tags are a better solution to doing it right the first time, giving ANY program/mediaserver the ability to gather information from the files - to identify with ANY online resources, and be used within programs in the 'OFFLINE' sense, to display information.

This would state universal compatibility. My approach on it. IF an app or program uses standard tags to query databases or display information.. and those tags are filled in within the files it is likely that the data is usable for a longer amount of time and more compatible, as a default. 

I think I actually may be repeating myself in my idea.. The two following links show you some technical information about the structure of the internals to the tags from ID3.org and the link for exiftool.org shows the 'standard' list of tags. I still do not need ALL of the standard tags  and I do keep my metadata fairly clean and simple. I usually have no issues with my files identifying/displaying correctly and I always strip all other tags completely out of my files after I am done with plugging in the online source ( however it may be, even multiple ); I then edit the data with the same type of notation/capitlization/genre indication ( several things I check ) - so that only those tags remain.

id3v2.4.0-structure - ID3.org

ID3 Tags (exiftool.org)

[ginjaninja 536]

39 minutes ago, Hxemby001 said:

Maybe this will help a little in what I mean... I look at my tags differently as far as standard..

Keep in mind that because of extensibility, ID3v2.4 and ID3 format are flexible ... Like I was mentioning above, any 'Tag' may be created and filled in with any data. Like musicbrains, which is a custom ( versus standard tag entry ) and aggregates or is used to aggregate from an ID or information that is not usable with any other online database or database resource. IF ( forbid ) musicbrains went by way of the dodo.. lost funding, was destroyed in a nuclear blast = the tags become useless. Saying therefore, they are specific for musicbrains. Going through 139,000+ songs ( now as of today ) and modifying those files your are curating the data for an information database. A lot of work if you never plan to pass on the record shelf, or if you plan to go and do it again in the future.

SO.. my thoughts ( for my archive ) were to make sure that standard tags are a better solution to doing it right the first time, giving ANY program/mediaserver the ability to gather information from the files - to identify with ANY online resources, and be used within programs in the 'OFFLINE' sense, to display information.

This would state universal compatibility. My approach on it. IF an app or program uses standard tags to query databases or display information.. and those tags are filled in within the files it is likely that the data is usable for a longer amount of time and more compatible, as a default. 

I think I actually may be repeating myself in my idea.. The two following links show you some technical information about the structure of the internals to the tags from ID3.org and the link for exiftool.org shows the 'standard' list of tags. I still do not need ALL of the standard tags  and I do keep my metadata fairly clean and simple. I usually have no issues with my files identifying/displaying correctly and I always strip all other tags completely out of my files after I am done with plugging in the online source ( however it may be, even multiple ); I then edit the data with the same type of notation/capitlization/genre indication ( several things I check ) - so that only those tags remain.

id3v2.4.0-structure - ID3.org

ID3 Tags (exiftool.org)

i dont think any one would argue against using standards where ever possible, for compatibility and longevity as you say.

bear in mind

even the linked examples have some application specific tags...but agree the basics are broadly agreed upon now across most applications.

 

with no standard separator prior to 2.4, identifying multi-value tags is not 100%, and there is still alot of 2.3 or earlier 'out there'.

so it was great when emby introduced support for musicbrainz tagging as a way of ensuring 100% multi value detection along side 100% matching for art and biographies and disambiguation of multiple artists on the same name (or even disambiguation of multiple albums by same album artist) .

So yeah there is a risk that mbz will go the way of the dodo but i think its a case of doing the best with whats available at the time. And ensuring support for sophistication doesnt impare the experience for people who dont care for it and/or need music to look right in other applications...I always thought by using and preferring (when present) separate tags Emby struck this balance well.

and through its tag reading preference engine, emby can do its own thing (if it feels there is a strong benefit/case), whilst supporting 'advanced' norms of other software whilst falling back to the core basic agreed tags (with potentially less functionality/accuracy same as other softwares). so almost no reason the best of all worlds cant be achieved.

[Guest]

Seems to be an onslaught here no matter what I post people seem to think I am taking away there musicbrainz..  or denying people who choose not to identify their stuff with metadata out in the cold.. I'm not..

SO.. lets just leave me here.. at this point.. ya'll carry on..

I really have been ALL over this forum for a few weeks trying to help with a few things and seem to be completely incapable of a proper suggestion. Nor anyone to understand why I suggest it. 44 posts and probably 35 of them have been about inconsistency in metadata and variables in online results.. with this sort of outcome. I could really be doing other things with my time .. Like getting somewhere with my work on said subject with incoming media.

Thanks..

[DarrenClimie-Newman 6]

Arguably there are 'Standard' tags which TagScanner highlights, and distinguishes them from more 'Advanced/Extended' Tags.

And Emby itself only provides some of those 'Standard' Tags to search by, meaning additional Advanced Tags kind of irrelevant anyway.