| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0013756 | MMW 5 | General | public | 2016-12-15 04:11 | 2022-09-14 14:39 |
| Reporter | peke | Assigned To | |||
| Priority | urgent | Severity | feature | Reproducibility | N/A |
| Status | feedback | Resolution | fixed | ||
| Target Version | 5.2 | Fixed in Version | 5.0 | ||
| Summary | 0013756: Add/Rescan: Add additional Column of preffered/expected content type | ||||
| Description | To speed Scanning files to Library e should consider adding additional parameter to Add/Rescan Dialog named "Preffered/Expected Type" It would serve the purpose to easily detect file type, Filename format, Metadata import (external files included) and would also speed up handling and importing files into MM by at least half especially on video files. It would be served as dropdown with these selections: Automatic detection -> Default would work as existing Music Music Video Video TV ... Later we can limit specific scans to Audio/Video formats and count based on extension to show in Scan report something like "not imported Videos/Audio: xxx" | ||||
| Additional Information | related to TICKET ID RSI-193-85593 | ||||
| Tags | No tags attached. | ||||
| Fixed in build | 2079 | ||||
|
|
I don't understand how would this help, to me it rather looks like it would just be another UI option for the user to read/consider, while offering nothing significant. |
|
|
I disagree, but on the first hand it looks like you pointed. Taking in consideration that many Video apps like PLEX, KODI, Sonarr, SickRage, EMBY, and others already automatically downloaded 99% of metadata (saved in external files) that MM just need to pick up. When there is Selection of preferred/assumed type for that folder MM would skip many iteration and try directly to handle Available Metadata and increase scanning process considerable. I uploaded Metadata for several series that are based on definitions from 0012481 and as performance test I used "Star Trek Voyager" that have large number of files to compare Scanning speed and results ("Star Trek Voyager" Folder contain both Metadata and MM.DB archives). Where MMW showed as way inferior (unusable) to other apps. For testing I scaned "Star Trek Voyager" on NAS which took almost 15 minutes and also showing incorrect Metadata, where on other hand each other app (Sonarr/SickRage/PLEX) took <1 minute for same folder and showed correct metadata Other examples on FTP observe same behavior. Scanned files are all classified as Video instead TV which additionally take time to correct. We have users with hundreds of TV series and thousands of episodes which would took days to correct. Also Metadata contained in definitions of 0012481 (tvshow.nfo and series.xml) contain INFO/ID of IMDB that additionally improve TV Series handling. |
|
|
If we talk about _automatic_ metadata extraction from filenames and other files in the folder then yes, I'm all for it. However adding a new UI option just doesn't look like a good idea - I suppose that any slowness that you observe can be removed by optimizations. In order to make progress here, I suppose that you could prepare a list of all common naming conversions, so that we can create/update algorithms for inferring metadata from filenames, including categorization by Media Type (TV vs. Movies). OK? As for externally stored metadata, I suppose that it's more or less covered by 0012481, right? Feel free to add anything missing there. |
|
|
User defined scan settings per Type would be useful and is something Plex employs by having scan settings per Collection. 1) User would need to be able to define location(s) per Type. Most users would have the different Types of content in different Locations. This would ensure that Type is correctly defined (Movie Type would need to be added). If not set MM would default to auto-assigning Type as in MM4. 2) User can define what Masks to apply per Type to facilitate tagging from Filename (many video files are poorly tagged/have no good tagging standard). It would be more flexible if several Masks per Type can be defined with priority. 3) User can define which external metadata files (NFO/XML) can be read per Type. Ideally user could set which format such file is (whom created it) so it's accurately read. 4) User can define which metadata lookup services can be automatically used per Type post scanning and their priority (as some will overlap in data they provide). 4a) Addon developers need a method to add their Addon to lookup services option. |
|
|
@Lowlander 1) I Agree completely as it will make mm5 better manage files per collection and also allow user to force some files into specific type if needed (Currently there is limited Music Video support in MMA and they are better presented if synced as Music) 2) I agree completely, priority is debatable as I would prefer using in order of used, but by the current state looks like 0012481:0048059 covers most of filename masks for TV series. 3/4) I would add ability to assign specific add-on to type that can be executed on add/rescan like Subtitle search or Poster search. |
|
|
As Seen in added images many apps uses predefined types to easily classify scanned files type. Which can in combination of metadata scrappers highly improve accuracy of type detection and metadata search/download. Example Bad filename for TV series like <Season#>\<Episode Name>.ext can correctly detect what series and Episode if it searched Episode Name in TVDB and compare result with Season # to Get Series Title and ID to fetch Full Metadata afterwards. In Addition to above I would add MUSIC VIDEO also or we can tweak Music Detection to Automatically assign Music Video type to Video Files as essentially Auto tag from Filename should be same except file extension. |
|
|
reassign for triage |
|
|
Ok, I agree that the addition of Content type per scanned folder might be useful. We'd probably want to keep the current model of 'Auto' type though, i.e. that types are auto-detected. Users could select a particular type in order to get better matches. Assigning to Rusty for feedback and possible preferences re. UI changes. |
|
|
Current MMW collections are: Music Classical Music Podcasts Audiobooks Music Video Video TV Current MMW Types are: Music Classical Music Podcasts Audiobooks Music Video Video TV Video Podcast Missing Type (and probably Collection as well) is: Movie To review, the purpose of Type is primarily to: - differentiate metadata lookup - differentiate properties view Collections are based on type, but have the flexibility of including additional content. They primarily affect views. One could argue that if we are scanning to a Type, that there's no real point of using collections any longer. That would generally be true except in cases where it's not possible to directly scan to a Type. e.g. for Classical Music (since it'll usually be merged with other music); or for podcasts and audiobooks that may be saved along with Music to a /download directory; or for content that is not clearly one type or another (e.g. symphonic music or instrumental music). So we'll continue to maintain both. The scanning process should give users the ability to associate a directory with any one of the Types described above. The simplest approach would probably be to modify the Add/Rescan process as follows: 1) Add rescan dialog -- more or less the same as it is today, except that it lists Type 2) Choose folders button, currently opens 'Add Media' and then 'Add folders'. It would make sense to change this along the following lines a) 'Add Media' b) 'Add folders' (New folder chooser would only allow a single folder to be added). Ideally this should be in the same 'Add Media' dialog. At the bottom of the 'Add folder' dialog, the user would choose settings particular to that folder e.g. Type, Frequency, etc. c) Metadata lookup rules can probably be configured separately in the settings since in most cases this won't be changed e.g. Add/Rescan files to the Library... -------------------------------------- Scan the following folders for media: [Choose folders...] .[x]..list... [ ] Look up track information [ ] Import files... [Options...] [OK] [Cancel] --------------------------------------- Click [Choose folders...] ---> Add media -------------- Local storage Network Google drive .... .... .... [OK] [Cancel] -------------- Click [Local storage] ---> Local storage opens within the dialog. ie. Add media --------------- Select the folder you want to scan for media files: | folder chooser (allows selection of 1 folder) | | Media Type: [Various, Music, Classical Music, etc..] | Scan frequency: __[One time], at startup, continuously, at startup and continuously__ | | . . . . . . [OK] [Cancel] ---------------- Note: we would also need to modify the Sources section of the metadata lookup options along the following lines: Type: . . . . Source: . . . Music . . . . . ._MusicBrainz_ Classical Music ._MusicBrainz_ Movies . . . . . _The Movie DB_, NFO files, XML files, etc. TV . . . . . . ._The TV DB_, NFO files, XML fiels, etc. ... |
|
|
Fixed |
|
|
Tested 2088 and it seems to work correctly in the sense that content is scanned into different collections. However, the application of metadata lookup rules based on different content types can't be tested since isn't yet implemented since #14571 isn't yet implemented (including both Type for Movies vs home Video and lookup rules by Type). |
|
|
Confirmed Verified 2089 for now implemented part |
|
|
Verified 2286 Confirmed for 5.0.1 revision |
|
|
Assigned for revision triage due the 0013756:0048733 attached current UI for comparison with proposed tweaks/redesign |
|
|
image.png (40,219 bytes) |
