HUROLURA wrote:That's just what I meant: as it is now, each of the DSP is statically loaded with modules to perform 2 voices. But the Midi stuff doesn't have to be loaded on the 6th DSP. From what I understood, it is handled by the Blackfin which is the DSP master and tell them what to do.
MIDI stuff doesn´t need much DSP, so where this goes is eventually 2nd row.
We know a lot about XITE but not Solaris hardware design,- so to me it appears, the Blackfin is the "communication chip" of Solaris,- similar to the DSPs # 3-6 in XITE,- each one controlling 1 of the DSP slots.
Can be, the MIDI would work on the Blackfin,- or on the 6th SHARC.
John knows.
HUROLURA wrote:
The Solaris voice handling is quite different from the Xite but as you said it is much more predictable which is a good point. I would see a Solaris multimode being more like the one used on the Noah. Noah provided 2 old DSP for each of its slot and 3 extra used for other stuff: one was used to do what the Blackfin is doing in the Solaris and the 2 remaining were used for mixing the slots including an external signal path and provide resources to compute FX.
For that reason, the upgrade from the 2 slots standard Noah to he 4 slots Noah Ex is a board featuring 4 old DSP.
There is a chance, Solaris has similarities w/ NOAH.
I don´t own a NOAH, but my friend, (service-) tech and sounddesigner/mixer has a NOAH EX.
I´ll meet him next week and will investigate more.
HUROLURA wrote:
My guess is also that the Solaris is already organized hardware wise with 4 slots as are the Xite-1 and Xite-1D.
In fact, I thought about that too, because we have a "4-parts" architecture in Solaris already,- 4 OSCs, 4 mixers, 4 filters etc. and the block diagram indicates 4 slots too.
My "5-slot assumption" came only by the "factor of 5" mentioned,- so it´s a question of how the hardware design of Solaris really is in the end.
It could be 4 slots, but then, there must be 2 SHARC chips in 2 of the slots and only 1 in each of the 3rd and 4th slot.
I´d like to know how it´s been done.
But if it has 4 slots, I wonder where´s the factor of 5 for polyphony coming from,- mathematically and physically.
HUROLURA wrote:
As John already told us getting more polyphony would mean getting 3 voices out of each of the 5 DSP by mean of optimization or code improvement.
That´s what I said, mentioning, they squeeze out 1 voice more per chip and then it´s the limit,- overall 15 voices.
But in former posts in different forums, John also mentioned 15 -20 voices,- and it seems, this isn´t possible now.
And the truth is, we´re stuck w/ 10 voices until the overall improvements for SCOPE will go into SCOPE SDK related developements.
HUROLURA wrote:
I guess the DSP load for 2 voices currently reach more than 2/3 of the available DSP power and memory so this is the reason why the polyphony for each DSP is limited to 2 voices.
Yep, and that margin meets the experiences w/ XITE and it´s 60% DSP load at max. for the time being.
In fact, there´s a lot of up to now unusable DSP power hidden in these designs.
It can be, the very next updates for both, Solaris and XITE won´t bring everything we would like to see,- but sooner or later we´ll see more, depending on the evolution of developement of the software.
HUROLURA wrote:
Actually, several paths are possible on the way to increase polyphony.
We´ll see.
It´s not just only a matter of possibilities, it´s also a matter of time investment, money and more.
S|C and J.B. Synth Design are small and enthusiastic companies which need more time and breaks to make progress.
We´re not talking about companies like UAD here, using a similar DSP design, but have more funds and coders.
And if you look at the price for a UAD Apollo incl. all 53 plugins, you´ll get pale.
It´s ~ EUR 5.000.- here and these devices have 4 SHARCs only and just only deliver FX plugins for a DAW host.
Compared to these,- Solaris and XITE are a steal,- even not being perfect.
I´m pretty sure we see improvements by the time and I wish John as also S|C the very best and lots of luck.
It´s also not too difficult to create a machine offering up to 80 voices and MIDI multi mode w/ only 2 of the new SHARCs,- that´s what a Access Virus TI is,- but I think the sonic quality of SCOPE/XITE and Solaris is top notch and has it´s price,- DSP power.
Doing compromizes in sonic quality would bring more voices for sure,- but wasn´t the target constructing XITE and/or Solaris.
HUROLURA wrote:
Adding an external 1U unit should be more easy, and also good to bring founds to John team to go on development. And anybody acquiring a 1U unit on top of the keyboard would benefit from these later enhancements.
That´s a important factor of a rack module,- they will probably sell good and make additional money for developement.
What these small companies need is sales,- that´s it.
I´d buy a rack also w/ 10 voices only and wait for the improvements,- but if I have to spend 4K$ for the keyboard, I´m eventually not that patient.
HUROLURA wrote:
Having this multi mode would make sense for people already having the keyboard to buy a 1U expansion.
Well,- the question is "who is willing to buy a rack unit next future,- let´s say after Winter NAMM or Musikmesse 2013.
I´d do,- with or without MIDI multimode.
If we´re talking about all the features a rack device should have in addition and compared to Solaris keyboard, we don´t see the rack at all soon.
HUROLURA wrote:
I wouldn't see much people buying a 3 1U units to get the 4 slots each with 10 notes...
Confirmed !
HUROLURA wrote:
No multi mode is seen more and more in the current hardware synth available on the market. For example, I do not know what the Novation Ultranova 18 notes polyphony while being mono timbral is useful for...
Well, for the gigging keyboard player not only doing sequencing and pushing the play button, high voicecount is an advantage always.
Hein