Identify to default to video filename

[justinrh 175]

I record a lot of movies from TV.  I have to Identify each recording to get all the extra metadata.  I'd like the filename (sans the extension) to be defaulted into the Title search field, then I can just hit Search and get the desired result.  As it is, I have to copy/paste for each Identification (I'm glad the filename is on the search page!).

The same issue occurs for any movies I rip from my disc collection.

I know Luke has said before that some users don't like this idea, but I'm thinking 'when would you not want this'?  The only time Identify has value is when it is a known/public title and not private videos like "My 25th birthday at Aunt Martha's house.mp4".

I'd think this would be the 90% case anyway.  So, now you can tell me why this be a bad idea...

[Luke 37153]

Quote

I know Luke has said before that some users don't like this idea

Hi, it has nothing to do with users liking or not liking it, and has everything to do with our observations of how users use and understand the Identify feature.

The reason it is blank is to force you to force you to know what you're doing by having to figure out what to enter. When it defaults to something, what we've noticed is that users don't change anything, and then have no idea why the Identify lookup is not finding any results.

Further, the regular library scan already does filename searches on initial import, so what you're doing to repeat the process manually with the Identify feature is not intended nor does it accomplish anything. If you really think you need to do this, then let's look at why and get to the bottom of that.

[justinrh 175]

Hey @Luke.  Emby does a match (I get the movie poster), but I don't get the ratings or actors.  You have to Identify to get everything.

Obviously I don't know what you know about other users' use of this, but I don't see how you can search w/o a title and get anything useful.  Just year does no good.  The only other option is a  cloud db ID - so even if you used this + title you should get what you want or at least reasonable results to choose from, right?

I'm just not seeing the downside.  Only Emby has the inside scoop and can make the best call here, but I've voiced my frustration and enhancement request.

[Luke 37153]

Quote

but I don't see how you can search w/o a title and get anything useful.

You can leave the title blank and use an external id.

The intended use of Identify is to fix incorrect matches. Defaulting to the file name isn't going to help you fix an incorrect match because it will just end up producing the same result.

[justinrh 175]

I know I can use only an ID.  That's probably a low-frequency use case.  But even if you had the title along with an ID you can get good results (that is, it doesn't hurt).

Ah, but it will not always end in a bad match.  Think of how videos are misidentified.  They are usually very close (wrong year or similar titles), and I'm still going to use the filename to verify what I'm searching for.  Thus, the search results will likely contain the exact match.  Again, no loss but there is a gain.

3 hours ago, Luke said:

The intended use of Identify is to fix incorrect matches.

That might be the primary intention but you also have to use it to get the extra info, which is my primary intention

Again, I'm not seeing the downside.  Someone is going to lose, but why should the 80/90% case have to?

Any way we can get a consensus from the users?

[ebr 14947]

32 minutes ago, justinrh said:

but you also have to use it to get the extra info,

What extra info?

"Identify" is designed to help you identify items properly, not gather additional metadata.

We  used to pre-populate the name but in the 80-90% case this was incorrect because the whole reason someone was trying to identify the item is because it was mis-identified.

[justinrh 175]

ebr, read the post

If that's the case then do we have a bug?

Edited June 6, 2022 by justinrh

[Luke 37153]

Let's go back to my earlier post:

Quote

then let's look at why and get to the bottom of that.

 

[justinrh 175]

Hey @Luke just to be sure (because I deal with this problem often and must be sure ), you do have this on your list of things to fix, right?

[Luke 37153]

Yes but first you'll need to provide a specific example in detail:

Thanks.

[Happy2Play 8324]

On 6/6/2022 at 10:29 AM, justinrh said:

Emby does a match (I get the movie poster), but I don't get the ratings or actors.  You have to Identify to get everything.

Do you have a specific example as this does not sound write unless there is a provider issue?

But an example with metadata/image provider order is needed.  And potentially a log to see the actual queries.

[justinrh 175]

Fellas, really?  Just record a movie from live TV.

OK, if I must go thru this, tell me exactly what you need for the report.

AND you waited all this time to say you were waiting on me?  I had the impression you got the idea and needed to investigate.  Good thing I asked...

Edited June 12, 2022 by justinrh

[Luke 37153]

53 minutes ago, justinrh said:

Fellas, really?  Just record a movie from live TV.

OK, if I must go thru this, tell me exactly what you need for the report.

AND you waited all this time to say you were waiting on me?  I had the impression you got the idea and needed to investigate.  Good thing I asked...

That would be great thanks. You simply need to provide everything requested in how to report a problem. Thanks !

[justinrh 175]

After recording is done:

 

After Identify is done:

 

embyserver.txt

[Happy2Play 8324]

Isn't the first one from EPG/Guide data and the second from actual online metadata providers?

So at what point should one be Refreshed ie ignore existing data?

 

But Emby use to populate filename and created more issue then it was worth so it was removed.  But 99.9% of the time filename is irrelevant as folder name is used.  But recordings are a totally different beast.

Edited June 13, 2022 by Happy2Play

[justinrh 175]

6 minutes ago, Happy2Play said:

Isn't the first one from EPG/Guide data and the second from actual online metadata providers?

@Happy2Play  You already knew this, right?

Both of you have acted like Emby was not working properly.  Now you are saying it is as-designed?

We are getting nowhere here.  I"m asking for an improvement to the software somehow.

See?  I go thru all this effort to provide info for naught!  (My prediction was correct.  )  And I already knew answer, thus the enhancement request.

[Happy2Play 8324]

Just now, justinrh said:

You already knew this, right?

No as I don't uses Live TV at all so it was a guess.

But your request "Identify to default to video filename" has nothing to do with Recordings needing to be Refreshed to get full online provider metadata.

[justinrh 175]

37 minutes ago, Happy2Play said:

"Identify to default to video filename" has nothing to do with Recordings needing to be Refreshed to get full online provider metadata

Yes, it does.  That is the use-case.  Even Luke understand this.

Can you explain to me how the filename is irrelevant?  Don't your own naming rules dictate a suitable filename, e.g.,

\Movies\Reservoir Dogs (1992)\Reservoir Dogs (1992).mp4

Recorded movies have the same filename syntax.

[Happy2Play 8324]

10 minutes ago, justinrh said:

Yes, it does.  That is the use-case.  Even Luke understand this.

Can you explain to me how the filename is irrelevant?  Don't your own naming rules dictate a suitable filename, e.g.,

\Movies\Reservoir Dogs (1992)\Reservoir Dogs (1992).mp4

Recorded movies have the same filename syntax.

Folder name is used for any automated identification process.  Only time filename is used is in a Flat structure ie no item folders.

\Movies\Reservoir Dogs (1992)\ABDC.mp4 ie filename is irrelevant as folder name is used.  So file ABCD will be queried as Reservoir Dogs.

 

But in the end your issue it Re-Identifying already Idendified items, so in there a Refresh w/replace images should already do that.

Edited June 13, 2022 by Happy2Play

[justinrh 175]

But recorded movies also have the correct syntax:

 

@Luke intimated that the identification was not correct to begin with.

Can someone in Emby land make a definitive judgment on right/wrong/WAD and propose a solution to this ER?  That's all I'm asking, is to provide a solution.

[justinrh 175]

On 6/6/2022 at 4:44 PM, ebr said:

the whole reason someone was trying to identify the item is because it was mis-identified

And why was it misidentified?  If it was named correctly, it might be wrong, but it would be close and the filename would still give you the needed results to correctly identify it.

[Happy2Play 8324]

From a Identify standpoint it should be manual User required input information as that is the entire purpose of the Identify feature.

Now if item is already correctly identified then Refresh options are the route to update/find missing.

[Luke 37153]

OK what happens is, the nfo written by the recording process appears complete enough that the internet metadata providers are not used. That's normal behavior when a pre-existing nfo is used, although we should find a way to override that for recordings I guess.

[justinrh 175]

Okay, so I gotta ask again:  just to be sure (because I deal with this problem often and must be sure ), you do have this on your list of things to fix, right?

[Luke 37153]

Actually I'm not so sure. The problem is we don't get movie db ids from the guide data, so looking up by name the random movies you might get over the air will have a high potential for incorrect matches . That's actually why we write the nfo in the first place, to avoid internet metadata on recordings.