RipX=The Future of Sampling (Ai Sampling w/separate stems)

Definitely seems like something they will eventually introduce, would fit in well with the core MPC sampling ethos. I suspect it might arrive with the 'next generation' of hardware, whenever that may be, i.e. units with more CPU and RAM, or as someone suggested, it could be added as an online service where the audio is sent remotely for processing.

If the MPC could do this, even with non-real-time processing (e.g., just like we do for Normalizing and Time Stretching), it would be a game-changer, for sure.

Akai would likely have to limit the total time of the sample to process (maybe 15/20 seconds) to prevent overtaxing the CPU.

So imagine this process:

1) Choosing the start and end points of the sample for stemming.
2) The MPC does the offline processing to allow users to hear a preview of/audtion the stemming results.
3) After the processing, you can select which parts you want to hear in solo or as a group.
4) Then choose to export as individual stems or a mixdown of multiple stems, with the option to change the volume levels of the stems before mixing down to export.

Another cool feature would be allowing users to save samples as a stem file (e.g., "dope sample.STM"), which we can recall and later use/edit.

I don't think current mpc gen can handle it (I'm talking about non-realtime) ,unless Akai invests on developing using opencl (gpu)...
I guess if they decide to support stems it will be via online paid service
Tried Koala sampler on an ipad Air2 (which has a faster 3core cpu vs mpcs) . Certainly there are many factors (algorithm used, multi or singlecore processing) , but it took ages for a less than a minute sample

Elijah wrote:Serato Studio also allows stem splitting in the free plan. It is convenient to be able to chop the extracted stems directly

Are there any limitations on the free version ? (for the stem part specifically)
For freebies (standalone ,not online) there are currently , besides StemRoller , SpleeterGUI/SpleetSpace