Auto Save !

[T3t_one]

One thing i could do actually is using the load/save sampleblock functions when loading/replacing samples. This might fix the fragmentation issue since all is backed and cleaned up onto the sd card. But it is not seemless anymore to just replace one sample as i would have to do this everytime a sample is being loaded/replaced. And i am not 100% sure yet if that would work.

Usually in playback, replacing samples is only from LSMP. In fact, I only really replace samples when I am 'experimenting' with noises. I'm heavily utilising the LSMP to overcome the limitations of memory when playing back a vast array of breakbeats or glitch one shots. LSMP is not something I merely use to load a kit upon starting a new session. As stated above, the binary pop sound sound is overcome by killing the playback of the sample before loading a new kit usually with NOTE OFF. It would be neat a sample could be killed after a certain amount of ticks instead of a whole line, utilising every line without NOTE OFF. 

Never replacing samples on the fly in a live situation so autosave would only be for safety precautions in composition due to the knowledge of it being temperamental around near max memory. We would for sure turn it off in playback. 


I guess the summarise, manual saving isn't a problem if you do it every time you think to replenish the sample window, so not 'top of the list' stuff but an error message would be better than total melt down.

Regards <333

Year, sorry, the meltdown is out of my hands. If the errorcode appears then it's a detected error and i maybe might be able to make
a crash-backup and maybe be save.
The meltdown is something else i can't trap and do something with it.

And i don't mean autosave when loading a sample. I am talking about a full cleanup everytime when loading a sample. So the memory is never fragmented, but this at costs of a glitch is replacing while playing. Doesn't have anything to do with the LSMP command, but i utilize the same functions.
And it is not possible to detect when the memory is going at max, because this can be already with one short sample and loading another big one...or not.
So it must always be used.

But lets see, maybe it's not too bad at all do do it. And maybe it doesn't work at all....

Okay, the way this is understood is to be cautious in the sample screen properties and respect the memory bytes available. As you said NerdSEQ is not a sampler per se and its primary intended functions thus far work functionally smooth [channel 1-6].

How will you go about working on the NerdSynth sampler section? I understand your background somewhat so I don't expect bells and whistles on the sampler field but you have already opened the floodgates for a hardware tracker/sampler so you might as well go balls deep and make the best 8bit hardware tracker/sampler ever in modular format! one for after NerdSynth in 10 years maybe?. . . NerdSAMP? Who knows...

Dank u wel, Thomas < 333