hi all

[gigatable~] is an approach to overcome the 16777216 sample accuracy problem by grouping a number of tables to a bigger unit. see also http://lists.puredata.info/pipermail/pd-list/2007-09/053479.html

internally this prototype has 3 hard-coded tables and is able to record and playback as well as preload a sample. the 3 tables could be extended.

additionally a gigaread4~ , which is unfortunately quite expensive, allows to scratch or play faster/slower as you know it from your tabread4~ pleasures. to overcome the maximum accurate number of 16777216 it takes milliseconds instead of samples and allow to use 44.1 times longer samples.

normal table + tabread4~ 16777216 samples = ~ 6.34 minutes

gigatable~ + gigaread4~ 16777216 * 44.1 = 4.66 hours (not in the prototype yet)

(if i do the math correctly)

the prototype itself is just an example of how to group tables. for testing reasons each table has only 44100 sample atm.

have fun!

eni