[ANN] Repro-1

[aaron aardvark]

Earlier I thought my 6+ year old iMac was 32 bit, but after doing some research and finding this about my iMac:

Model Identifier: iMac9,1
Processor Name: Intel Core 2 Duo

I found information that seems to show my iMac is 64 bit from the above information and the link directly below:

https://support.apple.com/en-us/HT201948



But in my System Software Overview it says:

64-bit Kernel and Extensions: No

Now I am confused if my computer is going to run Z3/Re-Pro-1 (in the future) or not, assuming Z3/Re-Pro-1 will require 64 bit. Any people out there that can tell me?

[Mac Hine]

My number 1 wish has come true! Pro-1 emulation and from U-he. 2016 is looking good! Can't wait!

[Urs]

So...are we looking at a Synthex module for Diva then?

Depends. I'm certainly tempted to model the filter and the oscillator section. But as long as Diva doesn't do splits and layers she won't be a full emulation. And then of course the parameters and levels are matched between the modules.

[Urs]

Intel Core 2 Duo

That's a 64 bit one, no worries. We have one too to test our stuff on.

[david.beholder]

So...are we looking at a Synthex module for Diva then?

Depends. I'm certainly tempted to model the filter and the oscillator section.

But as long as Diva doesn't do splits and layers she won't be a full emulation.

Both ableton and reaper have amazing layering capabilities and except DCO phase align everything else is doable on daw level.

[david.beholder]

Anyhow... it's a long way to go. We've received first requests to be included in the beta team... that's not going to happen for another few months.

Hm did I missed beta-testers recruiting? Isn't it too late? I want to try.

[knowix]

Well, getting the sound right is not really that big of a challenge. The actual challenge is making it fast.

That reminds me of something I've always wondered. No-one seems all that interested in using non-realtime circuit emulations, to prove that software can emulate hardware 100% perfectly.

Is it because it's not that financially useful? Though there's plenty of totally financially unuseful audio applications out there.

Or is it because it would be bad for plugin PR? If a 100% accurate emulation of device took 2 minutes to render on a current generation i7 chip, someone could (correctly or incorrectly) estimate how many years it would take before 100% realtime emulations were available.

[Sendy]

/\ Interesting question.

I think there'd be a market for a synth that sounds good in realtime, but is flawlessly rendered when given arbitrary time to do so. It wouldn't have the instant inspiration feedback loop that a good sounding realtime synth has, but it might woo a lot of analog purists.

It would also give them something to do for the next two years, as they analyze and dissect each cycle of each waveform and compare it to their stash of analogs

[bafonso]

I just went to uhe's website looking for the alpha... there is no alpha, right? this was just a pre pre pre pre announcement of a plugin in 2018? </I kid>

[db3]

So...are we looking at a Synthex module for Diva then?

Depends. I'm certainly tempted to model the filter and the oscillator section. But as long as Diva doesn't do splits and layers she won't be a full emulation. And then of course the parameters and levels are matched between the modules.

I think of Diva as a toolbox anyway...sounds delicious

[aaron aardvark]

Intel Core 2 Duo

That's a 64 bit one, no worries. We have one too to test our stuff on.

Urs,
Thank you for replying!

[News From The Sky]

So...are we looking at a Synthex module for Diva then?

Depends. I'm certainly tempted to model the filter and the oscillator section. But as long as Diva doesn't do splits and layers she won't be a full emulation.

Unless the splits and layers interact somehow I don't see any difference from just using a DAW to run two instances, and do the split or layer via track settings. The Jupiters do splits and layers but they're in Diva. Or am I overlooking something?

[david.beholder]

[I don't see any difference from just using a DAW to run two instances, and do the split or layer via track settings.

There are some technical difference:
1. dcos have the same phase on both layers
2. lfos are shared instead of independent
3. on board efx (if any) are shared too

and all 3 are probably not that important pr easy to mimick if needed

[Urs]

Well, getting the sound right is not really that big of a challenge. The actual challenge is making it fast.

That reminds me of something I've always wondered. No-one seems all that interested in using non-realtime circuit emulations, to prove that software can emulate hardware 100% perfectly.

Is it because it's not that financially useful? Though there's plenty of totally financially unuseful audio applications out there.

Or is it because it would be bad for plugin PR? If a 100% accurate emulation of device took 2 minutes to render on a current generation i7 chip, someone could (correctly or incorrectly) estimate how many years it would take before 100% realtime emulations were available.

Interesting thought... What makes a circuit simulation slow is two things: They use a big matrix to cater for arbitrary connection between the parts and they use variable characteristics for each part.

But if one models a given circuit with given parts one can strip the process down to solving a few handy equations. There is not necessarily a difference between the two. In fact, we nowadays start out with a set of equations that give identical results.

However, even if one models that circuit perfectly in Spice it won't sound like the original hardware. There is an aspect of reality that is not present in theoretical models. We use our ears and creative intervention to make it sound right.

Nevertheless, I would fancy trying this out... Can't be too difficult to compare the optimized model to the circuit simulation and plot the difference in some graphs...

[Urs]

[I don't see any difference from just using a DAW to run two instances, and do the split or layer via track settings.

There are some technical difference:
1. dcos have the same phase on both layers
2. lfos are shared instead of independent
3. on board efx (if any) are shared too

and all 3 are probably not that important pr easy to mimick if needed

Pssst: DCOs typically have a much smaller common denominator in their phase counter than divide down oscillators found in string machines. They are not really audibly phase locked. For example the Juno 60 has beating between two voices playing octaves.