My Movies Wiki

Page History: My Movies XBMC Add-on

Compare Page Revisions



« Older Revision - Back to Page History - Newer Revision »


Page Revision: 2010/01/14 21:49


Introduction

Using the My Movies API, you can now access the high quality My Movies meta-data from within XBMC using your My Movies account. The service can be accessed either by purchasing an API access ($20 + VAT %25 in EU), by contributing data to our service up to where you have 2500 points on your account, or you can create a 21 day trial for the 2500 points, to evaluate the scraper.

Notice that the "scraper" is a partly commercial option, as you can choose to purchase access or contribute data to access the service. The scraper is created for users who do not want other programs than XBMC itself to maintain the meta-data. You can as an alternative download and install My Movies Server/Client or My Movies for Windows Home Server to freely store My Movies meta-data for XBMC into your movie folders, or you can combine the usage of the scraper and the software.

Notice! Because of changes in XBMC scraping functionality, XBMC build 21882 or later is required. The latest nightly build can be downloaded here. My Movies "scraper" is now part of XBMC builds but might not be latest version (awaiting acceptance of updates). You can always download the latest version of the "scraper" here.

Scraper Definition

The "scraper" is actually not a scraper in the traditional way, as XBMC does not "scrape" a website, but instead access the My Movies API service directly, and received easy to process XML data, which is imported into XBMC's library. Along with the speed and stability of our servers, this process makes the My Movies "scraper" much faster and reliable than any other scraper for XBMC, at the same time as providing you with the best available meta-data.

Account Information

Before you start using the scraper, you must create a My Movies user account here. You now have three different options to enable the API access for your account:

  • Purchase a $20 (+ VAT %25 in EU) API access here. This gives you a not time limited access to the My Movies API. You will be able to download up to 2.500 titles from our service, and at 2.500 titles, our service will temporary hold your account from downloading more titles, to avoid data harvesting. Should you reach the 2.500 title limit, you can contact to lift the limitation and receive access for another 2.500 titles. Unless we find that your account have been used in ways that does not comply with the usage terms of the API, lifting the limit can be done indefinately, making the access unlimited.

  • Gain 2.500 points for your account. Users who have 2.500 points or more on their account will automatically be given access to the API without the need to purchase API access. This means that users who already use our software, and either have purchased points for their account, contributed data to our service or users who would like to purchase points to unlock features in our software can use the 2.500 points to use the service. New users can start to contribute data to reach the 2.500 points. Accounts accessing the service using points will also be limited to loading 2.500 titles from the service at a time, just as the purchased API option.

  • Create a 21 day 2500 point trial here. During the trial period, your account will have access to the My Movies API as if you had 2.500 points, however to avoid data harvesting using trial accounts, during a trial period, your account is limited to loading maximum 250 titles from our service. During the trial period, if you purchase an API access, the 250 title limit will be lifted.

Even if you have purchased API access to our service, your data contributions are much welcomed, to increase the quality of the service for both yourself and others.

Configuring Scraper in XBMC

Copy the "scaper" files (mymoviesdk.xml + mymoviesdk.png) from the downloaded zip file to the "System\Scrapers\Video" folder in your XBMC installation directory.

In XBMC it is then possible to select the My Movies.DK video scraper when setting content for your video sources in file mode.

XBMC Configuration 1. Click for full screen.

You are required to enter your user credentials on the settings page. The user account must have purchased API access or have 2500 points.

NOTE! XBMC 9.12 Camelot, and My Movies scraper versions prior to 1.10, does not support URL encoding of settings. So you either need to avoid characters which need encoding or you will have to do this yourself. The chars needing encoding can be viewed here: w3schools. For example, if your password contains a space like this 'My Movies' you need to write this as 'My%20Movies' in XBMC.

For proper support of URL encoded username, password and country, please use a SVN build later than r26777 and My Movies XBMC Scraper version 1.10.

XBMC Configuration 2. Click for full screen.

This is enough to use the scraper with it's default settings.

Optional Settings

Disc data: This option enables searching by disc title rather than in the movie database. This makes it possible to find DVD's which does not have IMDB id's.

Enable Fanart: Use this option to enable or disable the addition of fanart/backdrops to your library.

Preferred trailer bitrate: Choose the highest trailer bitrate your internet connection can stream. If you have pauses, you need to lower this value and update movie information.

Preferred language: Used when Disc data is not enabled. Text, trailers and posters fetched for your movies will be returned in the preferred language if available. If preferred language is not available, text and posters will be fetched in English.

Country: Used when Disc data is enabled instead of preferred language.

Include adult movies: Adult titles will be included in title searches, and will be added as normal titles.

Example

Below is shown an example folder and file structure for use with the "scraper". It is recommended to have each movie placed in it's own folder:

Image

Optionally, you can place an NFO file containing the IMDB url of the movie. This forces the My Movies scraper to locate this specific movie to get a 100% safe match.

Image Image

For titles that does not have an NFO file, the folder name will be used for searching for the movie.

Image

Having added a source to this structure and scanned for new content we end up with a library showing this:

XBMC Example 1. Click for full screen.

If a trailer is available, it can be reached via the Movie Information.

XBMC Example 2. Click for full screen.

Finally, for each movie it's possible to change fanart/backdrop and poster to your liking if available:

XBMC Example 3. Click for full screen. XBMC Example 4. Click for full screen.

My Movies vs the Competition

Comparing My Movies with the most popular scrapers; TMDB and IMDB looking at speed and quality of metadata and trying to be as objective as possible.
For comparison we use IMDB's top 50 movies which are voted by a lot of their users. This might be to an advantage for the IMDB scraper. I choose to not have any IMDB urls in NFO files to test how the scrapers manage to find correct titles (especially remakes).
A folder structure is created with movies in each folder. A movie is a 0 byte MKV file called movie.mkv. XBMC's database etc is deleted before each run. SVN build of rev 22180 is used (latest at the time of testing).
Trailers are not fetched as it's not fair to TMDB and IMDB since My Movies scrape these from an other site (no impact on speed since links are in My Movies database).

TMDB

Time: 20 minutes 23 seconds.
Incorrect titles: 5: Anna M (M), Psycho (wrong version), Push (Up), Rear Window (wrong version), Star Wars Collection (Star Wars).
Missing titles: 3: Léon, Le fabuleux destin d'Amélie Poulain, WALL·E.
Missing poster/cover: 5.
Missing fanart: 4.
Missing description: 4: Rear Window. Plus the missing titles.


IMDB

Time: 18 minutes 2 seconds.
Incorrect titles: 0.
Missing titles: 0.
Missing poster/cover: 0.
Missing fanart: 0.
Missing description: 0.


My Movies

Time: 2 minutes 1 second.
Incorrect titles: 0.
Missing titles: 0.
Missing poster/cover: 0.
Missing fanart: 1. M.
Missing description: 1: UP.


Conclusion

TMDB scraper is no match for My Movies. It makes too many mistakes and is very slow.

IMDB scraper does a very good job as it makes no mistakes (Would also have been bad since the list of movies comes from their top 50 movies). However it is also very slow and for large movie collections you need to set aside many hours for updating your library.

MyMovies.dk scraper does a very good job. It is 9 times faster than the IMDB scraper. It makes no mistakes when finding titles. It is missing some fanart and posters for older movies. But in it's defense, this option was only added recently and the numbers of contributions grow every day and will explode once they are used in the My Movies software. Currently My Movies does not have plots for new movies which have not been released on media yet. This is why UP has no description. This is something which will change in the near future with introduction of new options on the film info website.

IMDB fanart is fetched from TMDB. TMDB does not have a report or rating system like My Movies does which is why I will rate My Movies fanart and posters/covers as higher quality.

If you want trailers scraped, you will see IMDB and TMDB use even more time to finish since another website must be searched and scraped all together (Dtrailer.com). In the result above MyMovies.DK scraper already includes trailers (filmtrailer.com). IMDB does too (from own links).


Result

This is how I would rate the result of the comparison:

Speed: MyMovies.dk
Matches: MyMovies.dk + IMDB
Text: MyMovies.dk + IMDB
Posters+Fanart: IMDB for count. MyMovies.dk for quality


Final Words

With the My Movies.DK "scraper" you are able to get high quality meta-data and graphics into your XBMC library at a very high speed.

The data retrieved is film data. This is because XBMC currently do not support disc id's (which My Movies do), so we are unable to determine the media type when loading data from the My Movies WebService. This might change some day in XBMC and if it does then the My Movies scraper can easily be extended to support this and then optionally correct DVD, HD-DVD or Blu-ray front and meta-data will be shown in XBMC, rather than movie posters and movie meta-data.

The XBMC skin used for the screenshots in this guide can be found here.