No need to apologize for noobish questions!
"wave/play ..." objects are for streaming sample playback from sdcard. That means data is read from the card in real time. Sdcard can only be accessed through buffered reading, the "wave/play ..." objects do some juggling to hide this, but this is not entirely flexible: for example samples can not be transposed. The advantage of streaming is that there is no length limit except the file size on sdcard (2 Gigabyte I believe, perhaps 4). Several "wave/play..." objects should be fine, but I expect you'll bump into limitations if you 'd attempt to play 16 samples simultaneously.
The other approach is using the on-board SDRAM memory. Besides the "SD" characters in the name there is no similarity. Sdcard is flash memory (keeps data without power). SDRAM is dynamic volatile memory - contents are gone when powered off. SDRAM has better performance and does not need buffering, but on-board SDRAM is limited to 8MB capacity. This is entirely available for use in patches, and corresponds to 86 seconds of mono sound at 48kHz/16bit, for delay lines or tables. To play a sample from SDRAM you need to:
- allocate a table though "table/alloc 16b sdram"
- load a sample (from sdCARD) into that table with "table/load", both allocation and loading at startup are integrated in the "table/alloc 16b sdram load" object
- play it with the "table/play" object, referencing the relevant "table/alloc..." object.
For multiple notes with different samples, use several "midi/in/keyb note" objects, each triggering a "table/play..." object.
If you're using the last "stable" 1.0.6 release, significant improvements are in test release concerning automatic upload of sample files and error reporting. Converging improvements in development into a stable release is one of my highest priorities.