I'm new to max so please excuse the very basic question here. I'm hoping to use max to get analytic information about sound recordings, then output this as a data file, perhaps a .csv or something else that can become an excel file. I'm seeing a lot in the tutorials about things like data monitoring, and I see a lot of data-driven visualization and sonification, but what would be the best way to just get access to the data itself, as I'm describing? Any help much appreciated!

best-way-to-get-data-out-of-max

What sort of info are you looking to get out? I'm not nearly as much of a master of max-fu as some people here, but I think you could probably record a lot of info at a single sample level by using poke~ and buffer~, then peek~ and coll

As NOAH said, it depends what information you want to log. Maybe have a look at [dict] and [text]?

Have you looked into the jit.cellblock object? Very powerful matrix-like (think, excell) Ui object for viewing and editing collections, I believe this can output csv data, or if it doesn't, it wouldn't be too hard to diy. You should definitely look at dict as well, bit of a learning curve but well worth it.

Check out the dict and jit.cellblock help files, there's some great examples in there that should take you to the promised land!

Have you looked into the jit.cellblock object? Very powerful  matrix-like (think, excell) Ui object for viewing and editing collections, I believe this can output csv data, or if it doesn't, it wouldn't be too hard to diy. You should definitely look at dict as well, bit of a learning curve but well worth it.

Check out the dict and jit.cellblock help files, there's some great examples in there that should take you to the promised land!


Almost forgot, for the actual collection of data, check sf.info, metatag, external object (you'll have to download that one)and the ejies series of externals.

I have a similar project, where I use sf.info, metatag to to collect info from songs, id3 info, like artist, title etc. and technical data like length in ms, filenam etc.

Then I used ej.buffer to generate waveforms for each song - a collection of 1024 floating point calculations - which are stored into arrays contained in each song's , dataset, stored in a dict.

Then, when each entry is recalled, that array is then sent to a 'multislider', and blammo! Recallable, scrollable list of song info. With waveform view. And, with a slider overlay, it was possible to create a touchable (I'm building for iPad) song position playback head which updates in realtime as a song is playing, even creating in and out points, etc.

 I have a similar project, where I use sf.info, metatag to to collect info from songs, id3 info, like artist, title etc. and technical data like length in ms, filenam etc.

Then I used ej.buffer to generate waveforms for each song - a collection of 1024 floating point calculations - which are stored into arrays contained in each song's , dataset, stored in a dict. 

Then, when each entry is recalled, that array is then sent to a 'multislider',  and blammo! Recallable, scrollable list of song info.  With waveform view. And, with a slider overlay, it was possible to create a touchable (I'm building for iPad) song position playback head which updates in realtime as a song is playing, even creating in and out points, etc.

So a lot is possible, enjoy the journey!


Thanks for the responses everyone, this has given me some topics to start with!

best way to get data out of max