i've been trying to resize a table (using resize) that I'm about to record audio to, for example set the table to a specified number of bars worth of samples at a given tempo, but the table doesn't seem to change.

Instead, I created a new array in place of the table, and this works. however, having an array that's 44103 long in the sub-patch slows the whole patch down horrendously to make it unusable!? why does this happen with arrays but not when you have a table that is at least long as this?

hmm... can't imagine why the table isn't resizing. wanna post a patch to demonstrate?

as for arrays and tables, in general pd is not so bad at maths, but terribly slow at drawing...and tables just store numbers, while graphical arrays store those numbers and draw them too.

tables will always be heaps faster.

so when you see the waveform in a table is this just an approximate representation of the numbers, as opposed to each sample being drawn?

as far as i know, the data only gets drawn if you actually open the table..so if you don't open the table, there is no graphical representation of the data - just numbers.

if you make a gui array it will always get graphically represented, and use heaps of cpu.

..these are just observations i have made from experience though, i don't actually understand how pd works behind the scenes.

Mmm... very strange. I'm using quite some gui arrays in some of my patches, but as long they're inside a subpatch and not being displayed, i've not encountered any problems with them so far (i've usually about 30 arrays loaded at the same time). I just don't understand the actual difference between a table and an array...

my guess is that the array in brett's subpatch is being refreshed too quickly for the cpu to handle. hard to tell without seeing the patch though.

domien, i think in your case if you just load data into arrays and read it , then you will be fine. it shouldn't make much difference to cpu usage..

BUT, if you want to continually load data into arrays while the patch is running, then the gui refresh of these arrays is going to slow things down a lot. in that case you would use a table.

just thinking about the same thing...
if write table in an object it contains an array called table1. if i put an array it contains itself and is called array1.
there is no difference besides that the object "table" only shows the array if openend.
or am i wrong?
y

yeah i think that is basically correct. also, with an 'array' you can resize the graph to suit your needs.