How about voice recognition? Possible?

3Dcore used for audio operations.

And yes, using the 3D-core for audio would be a great thing. It will be there anyways and it won't be used anyways if we're fooling around with music software.

Yes, pros would prefer a RME or Mark of the Unicorn interface, but:
PCI is dying (so only second hand interfaces could be used)
Pros would want MUCH more computing power and RAM

Let's face it: Pro audio recording can be done cheaper and more efficient on a intel-based PC. The Natami on the other hand could become a new standard for people with "tracker skills" and ADAT in/out could raise Trackers to a new level!

PCI is really dying. But with 16 ADAT I/Os onboard, most people won't need any third part audio interface. If needed, I would say pick a firewire one and write a driver for this.
 
  For the internal 16 channel (or even 8 channel) ADAT idea: still, someone would have to write good drivers with sub-5ms latency at 44kHz and also free signal routing. If you do not know what free signal routing means, take a look at LINUX JACK and also on a larger extent, the SCOPE FUSION PLATFORM (now Sonic Core XITE-1). This last one, in my opinion, is the Cream of the Crop of any audio system available today or yesterday. Too bad it is only Windows compatible. I mainly run windows today just because of Creamware cards. Have a look and try to understand the basic principles of free signal routing.
 
  If NATAMI can offer all this audio wise, even with only 8 channel ADAT I/Os, I am sure it will get attention from proaudio guys, even with modest processing speed and low RAM. Do you remember what we did with Pro Tools MIX and SCSI drives in a 64MB OS 8.6 Mac back in late 90's? Ahã!

Besides that 9 out of 10 times this would only be needed during audio creation and hence would not be required for play back since that can be in the default Amiga rank and file.

I figure there is enough power to mix 16 channels on the fly.

5ms latency is a lot, I have MCU's here that at Amiga speeds (8MHz) have a such latency.

Adding functions between point A to B or perhaps C adds latency this is unavoidable, questions is it acceptable?
 

Cheers to all and the team, I continue to follow discussion every day, I am sure you are on the good way for the future of amiga.

I agree too that ADAT would be very nice, but I really have no idea about the cost of the implementation...

One thing intrigues me :

  Does the 3D core will be powerful enough for an eventual combined use in sound and 3D production (eg, game)?

Thank you.

Wouldn't it be better to focus on Natami's sound chip Pauline before talking about external solutions? And why all this talking about drivers and stuff? "Drivers" what is that? I don't like the word. In the PC world drivers are required for everything, but on the Amiga? I don't need drivers to run OctaMED for instance.. ;) There's also no latency problems to discover. Please don't introduce typical PC-problems and or/ideas to the Natami. Keep it clean, fast and simple.

That's good to know :) I am interested in studio music but I'm not into MIDI and all that other stuff Rafael is talking about. Of course it's a great thing if the Natami can be used for more advanced stuff as well as demo music production. It just seems to me that people have too high expectations at times. Why don't people request a super-fast, optimized version of OctaMED SS or MilkyTracker? Something like that would be awesome, but perhaps I'm old fashioned... ;)

So in theory you think it's possible to create better low latency audio drivers for the Natami than for other computer systems out there? For Firewire and USB devices for example?

Rafael rafafredd wrote:

On the software side, we need good fast and stable low latency drivers.

The latecy problem was introduced with Windows. Hopefully this should be easier to overcome on the Natami and Amiga OS. Try FastTracker II or ImpulseTracker for DOS for example. No problems there. I still use DOS based tools on the PC due to "jerkiness" found in modern software.