You have been given good advice here on how to create and load custom Voices into your T4. Looks like you are making it difficult for yourself trying to find some unconventional ways. Why not just follow the normal protocol?
Hi Joe H,
All that the OP is trying to do is to make full use of the very expensive flash memory on his expansion board!
His problem is that he has used up all of the 128 available slots in the custom voice table with the voices that he has currently loaded, but the wave samples associated with these voices only occupy around one third of the flash memory on the board. Other than the 20 additional slots available for custom drum voices, he can't add any more voices even though he has lots of empty flash memory.
However, he has realised that the wave samples from Yamaha Premium Packs are loaded onto the same expansion board, but they do not make use of slots in the custom voice table, so all that he is asking is if there is any way to 'package' some more of his own UVN voices to use T4's premium pack loading procedure and hence use the empty expansion memory.
I don't think that this is an unreasonable request. Its not as if he is trying to pirate copyrighted content.
Unfortunately I don't believe that he will be able to do this because Yamaha has the protection for their Premium Packs set up to be tighter than a duck's #### (enter your own word here!). There are also other limits relating to 'Wave Counts' for example, which once reached, prevent loading extra voices of any type, even Yamaha Premium Packs, regardless of how much empty expansion memory remains on the board! I ran into this problem myself using mainly Yamaha Premium Packs where I couldn't load another pack because I had reached the Wave Count limit even though almost one third of the memory was empty! Yamaha attempted a half hearted fix for this in firmware update 1.10 and although it helped, it was obviously what was easy for them to do rather than what was actually required! The introduction of YEM with the Tyros5 and PSR-S models resolved this problem.
The only thing that I think fantomx2019 is going to be able to do, is what has already been suggested by others, which is to make different 'libraries' and load them as required. Each one would make 128 'normal' voices and 20 drum voices available for use. A library file is nothing more than a list of voice files (the individual UVN/UVDs), where to find them, and the 'slot' number to load them into.
This can take time to load depending on the number of voices involved.
However there is another poorly designed 'feature' of the Tyros4 OS which he could possibly take advantage of to speed things up!
When you install/load a custom voice for the first time, its associated wave samples have to be loaded onto the expansion memory board and this is what takes up most of the time.
However when you delete expansion voices from slots in the custom wave table, this does not automatically delete their wave samples from the expansion memory, and they will sit there until you delete them manually. If you don't delete the waves and later reload one of these voices, the OS is clever enough to realise that the wave samples associated with the voice are already on the expansion board, so it does not try to load them again, and instead just loads the voice information, which is a much quicker process.
So, if you save all of the voices currently loaded in a library, then delete them from their slots, their waves will still be on the expansion memory board. The next time that you load them it should take only a short time because the waves will not need to be reloaded.
You can then load new voices to the slots to make another library. When you first do this it will take a while, because it has to load the wave samples for each new voice to the expansion memory. However as above if you save the voices from the slots as a library, you can delete them from the slots or simply overwrite them with other voices, and as long as you don't delete the wave samples from the expansion memory board, they can be loaded again very quickly.
You should be able to continue loading new voices (UVN/UVD) which will load their wave samples, and incorporate them into libraries if you want up to the limit of available space on the expansion memory board, or the Wave Count limit, whichever is reached sooner.
Although this is not ideal, once you have added the wave samples for all of your voices to the expansion memory board, it at least gives you relatively quick way of selecting and loading the 128 UVN and 20 UVD that you want to use on any given occasion. You can load individual ones as well as whole libraries.
Regards
Ian