Print

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

[yannis]

If a pair of files have the same tags, they will always appear in a smartlist made to locate duplicates. Calibre has a nice feature to mark specific pairs as excluded. Is it possible in MC to exclude a specific pairs from appearing in a dupes smartlist while keeping the files available for other possible pairings?

[ferday]

There are a couple ways I can think of...what's the specific application?  Are any any tags different between the two?

[yannis]

I have some pairs of disks that are identical, but one is a remastered edition or whatever. To avoid having them appear as duplicates, I edit the album tag accordingly and/or use the new release date. But both of these take extra steps and I'm not happy with the complexity. In Calibre, you just mark the pair to be excluded, with no other change in the tags; that's what I'd like to have in MC.

[blgentry]

If I understand you correctly, you want to use a duplicate finder smartlist (or view) but make sure certain ones that you've marked don't show up as potential duplicates.

I have a Duplicate Finder view I designed that has this feature.  It's a combination of a few things:

1.  A pane selection that turns on or off the "not duplicates" feature.
2.  A check box next to every file in the view that lets you mark any file as "this is not a duplicate".
3.  A field in MC's database called "Not Duplicate" that holds the value of the check box.

See the attached screen shot of the latest version of this view.

Brian.

[8139david]

Thanks, Brian.
I've implemented your idea for finding duplicates of my movies, identified by IMDb ID (see attached pictures).

My Not duplicates correspond to movies which correspond to several files.

I also built a duplicate smartlist, identified by Name and Director, but in my case, it only found a subset of the first smartlist (since I may have one name in French, the other in English, for instance).
I could also have looked for duplicates of Director and Year: in theory, this might find a few missing cases. But it generates many false ones of course: all directors who made several movies the same year!

Then, seeing most of my movies already had an IMDb ID, I found the missing IDs for 20-30 movies.
Still a few are lacking, but it's unlikely I will ever have duplicates of those.

[yannis]

See the attached screen shot of the latest version of this view.

Brian.

Yes, this solution is exactly what I need, but for one thing: if you had a third version of Add It Up ("C"), it would create two new pairs, C+A and C+B. So how would you show them, so that you could mark the pair C+A but exclude C+B?

[blgentry]

Yes, this solution is exactly what I need, but for one thing: if you had a third version of Add It Up ("C"), it would create two new pairs, C+A and C+B. So how would you show them, so that you could mark the pair C+A but exclude C+B?

My solution is very simple.  It doesn't look at "pairs".  It looks at individual files.  If you mark any file as "not duplicate", it won't be shown in the results.

In your example, if you had my two versions of Add It Up marked as "not duplicate" and you added a third to the library, the third one would show up all by itself.  Because it matches the two others, but the two others are hidden.  I would probably be confused, and then I would turn the hide option off (in the second pane from the left).  Then I'd see the other two "add it up" songs, and probably mark the third one as a "not duplicate" also.

It's certainly not a perfect system, but I think it's pretty useful.

Brian.

[yannis]

Brian, thanks for your help; The function I proposed would exclude specific pairs, keeping the component files available for new comparisons/pairings. I was obviously not clear enough in defining this.

I'm still recovering from a crashed disk using a very unruly backup. My collection has hundreds of cover versions or new versions of the same song by the same artist, even in the same year, or sometimes with a missing year, or badly tagged and so on. So there's 100s of dupes that I need to keep. But in the long process of de-duping and retagging, I have to go again and again through these specific pairs. That's how I remembered this great function that exists in Calibre; I've no idea how easy it is to implement it in MC, I just thought it would be a good feature.

[blgentry]

I don't understand what functionality you would get with what you're asking for.  How would it be different than my method?  Neither method is at all automatic.  Both require you to make decisions and then mark files as non-duplicates.

Brian.

[yannis]

Here's a simple example. Suppose I use this smartlist to de-dupe mislabeled album titles.

~dup=[Artist],[File Type],[Name],[Track #]

https://www.dropbox.com/s/1rnip85gl003b55/dup%20MC.jpg?dl=0

The two cases of the first song are different versions. If I mark them as dupes to keep, they won't appear again in this list, right? But if at a later stage I removed the File Type to broaden the scope, your solution wouldn't allow me to find out that I have a third version rip in flac or whatever. And why would I change the smartlist? Maybe because I consider it more manageable to work in batches. Maybe because I may find out a new expression that suits me better. Or maybe because I didn't know better to begin with. Whatever the reason, the power of MC allows me to do it. The solution you suggest limits my possibilities, the one I proposed doesn't.

[blgentry]

You don't seem to understand my solution.  If you want to try it and see, I'd be happy to assist you in loading my view.  In fact, I'm attaching it here on the off chance that you want to try.

But honestly, we seem to not be communicating well between this and the last thread and I get the feeling that you don't think my attempts at helping you are very helpful.  So I won't post on this topic any more unless you ask for input from me.  I say this all intending to be polite and not confrontational.

Have a good day.

Brian.

[yannis]

Hey, no bad feelings here too  

I'll try this dvi as soon as I find out how to run it.

EDIT: Can't seem to find how - Would you care to show me how to load this view? Figured it out, testing.

[yannis]

OK, I needed to add the Not Duplicate field to my Library to be able to tick the boxes. I think I can see where you're getting at, but it's too late here and I'm sleepy. I'll get on with it tomorrow and report back.

[yannis]

Ok, as is obvious, I wasn't comfortable working with views, so this is certainly a jump for me. After fiddling with the Dupe Modes, I can now create lists that are meaningful to me and I can adjust to working with them. Plus, this opportunity opens for me an area of MC I hadn't messed with. So, thank you.

On the other hand, I hope you'll grant me that for a noob this is a rather steep learning curve; people don't always have the time or inclination to go with it.  So I still think the feature I proposed would simplify things greatly, if it were offered. And it wouldn't break anything.
 

[blgentry]

I'm glad you got the view to work for you.  Views are one of MC's big features.  That view is medium-advanced and fairly useful I think.  Though I haven't gotten a ton of feedback on it.

The last mode, ending with "Duration" uses a custom field called Duration Rounded .  It's a calculated field that rounds the duration of the song to the nearest 10 seconds.  Which helps you find songs that are almost identical in length, but might be off by a few seconds.  To use that you need to create the field.  It's a string type field with calculated data.  The calculated data part is:

math(formatnumber(math([duration,0]/10), 0) * 10)

Of course, feel free to lobby for your feature in one of the threads.  This is just one way of accomplishing this goal using the existing functionality.

Brian.

[yannis]

The time rounding was a valuable perk I didn't even know was possible; thanks for your time.