Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[JimH]

For the last few builds, MC could do a manual lookup of cover art, director, actors, etc from TVDb, Rotten Tomatoes, and Wikipedia by right clicking on the movie.

Now this lookup is automatic when movies are imported.  If you have existing data, it will not be replaced.

You can turn this off in the options for auto import.  Screenshot attached.

[Matt]

The new auto-import option 'Get movie & TV info' runs against the entire library.

It looks up metadata for files matching the following criteria:

1) The file is a movie or TV show (based on media sub type)
2) A lookup for the file hasn't been automatically performed in the last year
3) The file does not already have metadata (defined as a filled description, actors, or director for movies / filled description or actors for TV shows)
4) For TV shows, the file must have a valid season and episode number

(p.s. I've posted this a few times, but this seem like a good resting place for it)

[MrHaugen]

You know I love this new auto scraping system, but there if a couple of minor things that is not super. Would it be possible to consider this changes (hopefully minor ones)?

- Add scraping of Rotten Tomatoes scores, or preferably IMDB rating. This is something I believe MANY users want.
- Change the source of keywords. As it is today it only downloads rather useless keywords.
- Add a way to add TVDB ID to TV Shows, so we have an easier time to change wrongly detected TV Shows.

[glynor]

- Add a way to add TVDB ID to TV Shows, so we have an easier time to change wrongly detected TV Shows.

They did this for the next build.  Awesome-sauce.

- Add scraping of Rotten Tomatoes scores, or preferably IMDB rating. This is something I believe MANY users want.

Rotten Tomatoes scores would be good.  IMDB doesn't allow you to use their data without paying the piper (or hacking it), so I suspect this is unlikely.

[MrHaugen]

They did this for the next build.  Awesome-sauce.

That's nice!

Rotten Tomatoes scores would be good.  IMDB doesn't allow you to use their data without paying the piper (or hacking it), so I suspect this is unlikely.

Yea. I think we'd need some custom plugins if that should be strictly legal. It's not THAT important anyway. I think that the TheMovieDB scores is almost as good. I would actually prefer those over Rotten Tomatoes. I find the Rotten Tomatoes rating to be a bit off sometimes. But that might just be me? The positive thing is that they have two types of ratings though. Critic rating and user rating...

[glynor]

I find the Rotten Tomatoes rating to be a bit off sometimes. But that might just be me?

No, I agree.

I feel like Rotten Tomatoes' score is often pretty good, especially with more "indie" or less mainstream films.  The problem is (I think) that there is a bit of a "big hit penalty" to their user rating system.  Any big-budget film brings out reviews from a percentage of film buffs who are going to give any $180M film a bad review just because they'd rather be watching French art-house flicks all the time, which tends to kick those scores down.

Those same users probably "grief" the ratings on other films too, but when it is a big blockbuster title, it is popular enough that it brings them all out at once.

[SamuelMaki]

Good news I like to see automatic coverart also for movies... Now it detects right movie and get information, but the image is just a thumb instead of coverart (so, I have to run get TV/Movie information again to get the picture...). This should be an easy fix

[pauly139]

The auto-import of TV episode descriptions etc was a total failure for me - even though the process took around an hour. By failure I mean no episode descriptions were brought in, and my existing images were lost.

I think it would be great if we could manually select say a season of episodes and tell MC to go and get data for those, including options such as chosing to overwrite existing data, or images, or both.

The way it is now we have no control (that I am aware of) of what videos it will update, and what data relating to those videos will get updated.

So I believe you have done a good first step, but would love to see more.

Paul.

[alspoll]

The auto-import of TV episode descriptions etc was a total failure for me - even though the process took around an hour. By failure I mean no episode descriptions were brought in, and my existing images were lost.

I think it would be great if we could manually select say a season of episodes and tell MC to go and get data for those, including options such as chosing to overwrite existing data, or images, or both.

The way it is now we have no control (that I am aware of) of what videos it will update, and what data relating to those videos will get updated.

So I believe you have done a good first step, but would love to see more.

Paul.

+1

[JimH]

What happens if you do a single lookup by right clicking on the show?

[pauly139]

That works fine.

I've been playing, and found that if I clear out metadata such as Description, Artists, Publisher and then run an auto-import it works. I guess this makes sense, and means I can get it to 'repair' my TV episodes that need it.

Can someone tell me exactly how it chooses what metadata to import? Does it 'fill in the blanks' so to speak, or is there a subset of fields it looks at to decide whether to import or not?

If it checks a subset, can we know what they are so I can simply clear those out on the episodes I want it to import?

Paul.

[pauly139]

Can anyone tell me how I can force MC to run the metadata import again for certain files? I've tried clearing the contents of all the episode related metadata, changing filenames, but now it refuses to do the metadata capture again.

It does work fine when I do the same files individually.

[pauly139]

To follow-up I found I had to delete all those files from the library in order to get their metadata imported from the internet. In my opinion, that is too drastic a requirement just to get metadata updated for multiple files.

[pauly139]

Saga continues

Although I delete all the files from the library, and import them again, the TV Series M*A*S*H does not get metadata imported - in fact based on the import progress window MC does not even try and import the data.

A typical filename for the episodes is "M*A*S*H S01E02.avi"

Extremely frustrating.

[bil1010105]

The auto-import of TV episode descriptions etc was a total failure for me - even though the process took around an hour. By failure I mean no episode descriptions were brought in, and my existing images were lost.

I think it would be great if we could manually select say a season of episodes and tell MC to go and get data for those, including options such as chosing to overwrite existing data, or images, or both.

The way it is now we have no control (that I am aware of) of what videos it will update, and what data relating to those videos will get updated.

So I believe you have done a good first step, but would love to see more.

Paul.

+1

I had the same issue last night - auto-import "broke" a large number of my video files by overwriting previously created thumbnails, keywords, descriptions, etc... luckily I noticed what was happening so was able to limit some of the damage to my library by stopping the lookup. Despite that, I restored a backup, but the thumbnails of ~400 videos are gone for good.  

I've disabled the option until it can be:

• Enabled on a folder-by-folder basis. Not all my import folders have the same kind of video files.
• For each folder, allow me to specify exactly what it should replace or update from the Internet

I would also like to see proposed changes from the import put into some kind of "holding" area for review, before actually updating the library.

[JimH]

I've disabled the option until it can be:

Enabled on a folder-by-folder basis. Not all my import folders have the same kind of video files.[/li][/list]

You can do that now.  Set up auto import to watch different folders according to media type.  Then edit each one to include the types you want.

For each folder, allow me to specify exactly what it should replace or update from the Internet[/li][/list]

It should not fill any tags that have information. 

[MrHaugen]

You can do that now.  Set up auto import to watch different folders according to media type.  Then edit each one to include the types you want.It should not fill any tags that have information. 

I think he's talking about setting different auto import process tools on a per solder basis? Today you can only do Audio Analysis, Movie/Series Scraping, Cover Art downloading etc on everything or nothing. This is something I supports as well. I think this can introduce some problems. There is Video files I DO want to get automatic data for, and there is import folders that I do NOT want this for. I think this options should be saved for individual folders to limit frustration.

[bil1010105]

You can do that now.  Set up auto import to watch different folders according to media type.  Then edit each one to include the types you want.  

What about different folders with the same media type? All I see is a "Get Movie/TV Info" checkbox, which is either enabled globally across the entire auto-import list, or not at all. I have several video folders, some I would like to auto-update with information, others I most definitely would not.

It should not fill any tags that have information.

It would be better if I could specify which (if any) tags I want updating at all. For instance, I have everything organised using Keywords. It's a mainstay view for videos, and I use the fact that I can have a list seperated by ';' and '\' to allow me to drill-down into various categories.

When I ran auto-import last night, it generated a vast number of these keywords for things I haven't gotten around to categorising yet (which would normally then appear in 'Unassigned', which is fine), and it made a real mess of the view.

I would like it to leave Keywords always alone, and never ever ever attempt to update that tag from the Internet. Not ever. If there's some data it wants to give me, that's fine, but please let me specify where it should go.

[glynor]

I agree that we need to be able to "protect" fields from metadata writing.

My suggestion in another thread was a new option under the Options -> Library & Folders -> Manage Library Fields dialog, where you can enable/disable cloud metadata writing on a per-field basis.

That way, you could just "protect" your Keywords field and be done with it.

[bil1010105]

I agree that we need to be able to "protect" fields from metadata writing.

My suggestion in another thread was a new option under the Options -> Library & Folders -> Manage Library Fields dialog, where you can enable/disable cloud metadata writing on a per-field basis.

That way, you could just "protect" your Keywords field and be done with it.

It would certainly be better to protect fields. However, what would be better still IMO is if we could specify which fields are updated on a per-folder basis.

I also think the ideal would be to give you a "review" screen, where you can verify in a list what changes it's going to make. The fuzzy logic it uses to match a video's details to the data it finds on the Internet is, well, fuzzy. Quite often, it just gets things plain wrong. I'm not suggesting that's JRiver's fault, because the APIs work in the same way. You are, to a large extent, at the mercy of what they bring back.

Having an automated import is great (brilliant in fact, and long overdue) but we need more control & visibility over what it's doing. Well, I do anyway.  The computer does not always know best.

[rick.ca]

I agree that we need to be able to "protect" fields from metadata writing.

This pertains to what we were discussing elsewhere—about how meta data sources should be configurable. If all meta data elements were mapped to fields designated for them, conflicts would be impossible. You would then do whatever you wanted with the meta data, including combining it with that of another field. By not mixing it with data from other sources, it's also clear where the data came from and whether the retrieval was successful. For any such field, the user would still have the option of mapping directly to any other field—a sensible option if there is no potential conflict.

One significant advantage that's forgone by rejecting this approach is the ability to easy "fix" and update data from a particular source. When the data is arbitrarily mapped to different fields also update by other sources (including, potentially, the user), it's not possible to do either without losing other data.