INTERACT FORUM
More => Old Versions => Media Center 11 (Development Ended) => Topic started by: michel on October 20, 2002, 12:23:28 am
-
genre/style is a very important criteria to found audio tracks. Unhopefully it is very difficult to set. Even the very rich AMG genre-styles classification is not sufficient as it is an album classification when we would need a track classification.
So no solution ?
May be there will be one with Automatic Musical Genre Classification. The idea is to a analyse the audio signal to determinate the genre. A better idea is to found neighbour tracks based on similarities.
It is not for MJ9, for sure, but I think JRiver should keep an eye on these researches (and may be they are doing it) as Automatic Musical Genre Classification will be a killer feature IMHO.
It is funny: AMG (All Music Guide) classification could be called Automatic Musical Genre Classification...
-
Absolutely no disrespect to you, michel, but for my own purposes I say: forget that noise.
For me, genres are a flexible convenience for organization. I often make my own genres according to my impression of the music, not according to musical theorists' proposed definitions of the music.
After reading through some of the papers on the topic, I think that these folks have way too much free time. I also suspect that, like many decision-replacement software efforts, it would be something else for me to correct later :P
-
For me, genres are a flexible convenience for organization. I often make my own genres according to my impression of the music, not according to musical theorists' proposed definitions of the music.
I agree absolutely but I am not talking about that. My post was probably confuse (thanks to my nice english :( )
The objective is to classify by similarities, calculated from the audio signal using criterias to be defined. More than to found predefined subsets the objective is to found subset of audio tracks similar to a selected audio track.
These subsets don't have to map predefined categories, they just identify which audio tracks in your library are similars to one selected.
There would be many way to use it but, for example, you could select an audio track as your reference for a genre/style (that you could name as you want).
Thumb+ proposes the same concept for pictures: it is able to found similar pictures to a selected one; you are able to parameter the criterias used for the calculation.
-
Well that's more interesting, and different than the impression I got from some articles. Could you post links to information about this approach? I can see that being helpful in building thematic lists, such as mood, etc.
-
For example:
http://ismir2001.indiana.edu/pdf/tzanetakis.pdf
http://ismir2002.ircam.fr/papers.html
-
I don't like Genres.
They May limit the music you listen too.
I listen to what I like, without genre prejudice
-
I don't like Genres. They may limit the music you listen too.
Sometimes it is true. But that could be true also for 'artist ' field as you could not to listen a new album from an artist you didn't like the previous ones.
In fact, the feature I am talking about doesn't set any well-known genre but let group together "similar" tracks, on demand. A kind of auto-smartlist which you can give the name you want so there is no risk of prejudice.
-
Like the dance records with the number of BPM or BPS on the sleeve.
Kind of getting tracks on a moody base.
Would be a very neat feature, if the selection based on very specific criteria
-
They are calculated criterias based on audio analysis algorithms.