KVR Audio

Uncle E · Post by **Uncle E** » Wed Apr 23, 2014 5:21 pm

robotmonkey wrote:It would be even sweeter if someone like DSF would revive E-MU/Ensoniq legacy and made a synth that would combine the sampling of Emulator with the transwave synthesis of Ensoniq + add all the E-MU Z-plane filters.

I respectfully disagree. We have lots of great sample manglers but no samplers with the amazing sound of DIVA. Even a basic sample loading oscillator with nothing more than the ability to define start, end, and loop points would be awesome.

fmr · Post by **fmr** » Wed Apr 23, 2014 5:30 pm

robotmonkey wrote:
Uncle E wrote: A sample loading oscillator and Z-Plane filter bank for DIVA sure would be sweet.
It would be even sweeter if someone like DSF would revive E-MU/Ensoniq legacy and made a synth that would combine the sampling of Emulator with the transwave synthesis of Ensoniq + add all the E-MU Z-plane filters.

Keep dreaming

It would be sweet indeed, though

fmr · Post by **fmr** » Wed Apr 23, 2014 5:36 pm

Uncle E wrote:
robotmonkey wrote:It would be even sweeter if someone like DSF would revive E-MU/Ensoniq legacy and made a synth that would combine the sampling of Emulator with the transwave synthesis of Ensoniq + add all the E-MU Z-plane filters.
I respectfully disagree. We have lots of great sample manglers but no samplers with the amazing sound of DIVA. Even a basic sample loading oscillator with nothing more than the ability to define start, end, and loop points would be awesome.

I respectfully disagree with your disagree. DIVA filters are great for subtractive synthesis, but for sampling I tend to appreciate other kind of filters, more in the vein of Z-Plane (formant filters, comb filters, complex filters of any kind). This because samples tend to be complex spectra sounds, and a subtractive filter usually just takes things away, without adding nothing. Besides, a sampler has to be more than just a map of sounds that replaces an oscillator.

Anyway, we already have great samplers in the market, tough to beat: HALion, Mach Five and Kontak are three really great samplers. Pity that none of them have anything like Z-Plane filters.

Uncle E · Post by **Uncle E** » Wed Apr 23, 2014 5:38 pm

fmr wrote:I respectfully disagree with your disagree. DIVA filters are great for subtractive synthesis, but for sampling I tend to appreciate other kind of filters, more in the vein of Z-Plane (formant filters, comb filters, complex filters of any kind). This because samples tend to be complex spectra sounds, and a subtractive filter usually just takes things away, without adding nothing.

I humbly and regretfully disagree with your disagree with my disagree. Case in point is the Emulator III that defined the sound of Daft Punk's "Homework".

Urs · Post by **Urs** » Thu Apr 24, 2014 8:40 am

robotmonkey wrote:Filterscape though does not sound even close to E-MU Z-plane filters in Emulator X3. Completely different characteristics.

Well, if the X3 uses the same principle as devised in the original ARMAdillo patent, then the filters in Filterscape should sound exactly the same - provided one finds a combination of filter cutoff + EQ sweep with the same rations of frequency, gain and q.

However, no idea what other kind of processing is involved before and after the signal chain.

robotmonkey · Post by **robotmonkey** » Thu Apr 24, 2014 10:25 am

It's probably due to several factor like different algorithms and programming used. Also Emulator X3 does not modulate only between different set of EQ points but also between a set of other filter parameters. Some special filters like Dual EQ Morph/Expression modulates allows modulation between two sets of different filter parameters by different modulation sources. Also Emulator obviously has an advantage in usability as you can use any of the 55 filters per voice while Filterscape can only be used on the combined output of the synth.

Urs · Post by **Urs** » Thu Apr 24, 2014 10:39 am

robotmonkey wrote:Also Emulator obviously has an advantage in usability as you can use any of the 55 filters per voice while Filterscape can only be used on the combined output of the synth.

Hehehe, in FilterscapeVA the EQ is per voice too, and any type of morph possible with what's described in the patent should be possible with FilterscapeVA too. Naturally, FilterscapeVA has more degrees of freedom and thus is more complex to use.

However, Filterscape/FilterscapeVA is a completely different product than Emulator X. I wouldn't even think about comparing them. But as far as "Soft synth with similar filters to Emulator X?" goes, Filterscape was based on the idea of doing something like that (not copy Emulators, but do morphable filters)

Trancit · Post by **Trancit** » Thu Apr 24, 2014 10:47 am

It´s quite a long time ago, since I last used Emulator X2, I think I had installed, but what I can remember:

I was never very impressed by those hyped Z-Plane filters and I was far away from considering to use them...

Most of these 50+ filters sounded to me:
1. very similar and very digital
2. as they took of a lot of punchness and pressure out of the sound
3. often just like phasing
4. as they could easily replaced by one effect device on the output instead of letting my CPU calculate one for each voice...

I don´t miss them at all and I think these EMU plugins were the worst and "buggiest" VST I ever used...

robotmonkey · Post by **robotmonkey** » Thu Apr 24, 2014 11:06 am

Trancit wrote: Most of these 50+ filters sounded to me:
1. very similar and very digital
2. as they took of a lot of punchness and pressure out of the sound
3. often just like phasing
4. as they could easily replaced by one effect device on the output instead of letting my CPU calculate one for each voice...

I don´t miss them at all and I think these EMU plugins were the worst and "buggiest" VST I ever used...

Some of the filters indeed sound similar (phasers, vowels etc). And they certainly do sound very digital no doubt but that's what I actually like. You can make those digital sounding long evolving pads like the E-MU Morpheus could do. It's absolutely sweet.

All filters do not sound good at all source material so one needs to find the right combinations. If I remember correctly some of the E-MU synths like Morpheus had close to 60 page chapters that explained all the filters and how to use them. There were a lot of emphasis on how to use them for specific acoustic modelling tasks.

I know that there were a lot of complaints about the X2 and previous E-MU VSTi's and that in the end E-MU did not get to fixing it all. That said Emulator X3, the stupid copy protection excluded, has been working nicely. Recently I have been again working with it and must say it's definitely a great instrument. Workflow wise it blows all the other VST samplers out of the water and it has plenty of programming potential. It kind of has that nostalgic feeling that it still belongs to hardware paradigm closely resembling E-MU synths (and it comes with a 500 page manual!).

Ingonator · Post by **Ingonator** » Thu Apr 24, 2014 11:20 am

Just tried to get what Z-Plane filters are about as i never had an E-Mu synth myself so far.
Looks that besides some others it is possible to do some morphing between diferent filter types.

Not sure if it is the same but here are some experiences with other synths:

In e.g. Tone2 ElectraX there are 32 filter types including some more "exotic" ones like e.g. Comb or Fractal filters.
For each oscillator (4 layers with 3 oscillators each) you got a Mix knob to route the oscillator to one of the two multimode filters. The mix parameter could be modulate using the mod matrix. This way it is possible to morph between two filter modes.

Also in Largo for each oscillator you got a "Balance" parameter (inside the mixer section) to adjust the routing of that oscillator to both multimode filters. The Balance parameter (e.g. "O1 Balance") could be modulated in the mod matrix.

In both ElectraX and Largo the filters could be set to either parallel or serial mode.
Also in both there is a panning control for both multimode filters.

Somehow this kind of morphing should be also possible in Diversion but i have not checked that yet. Each oscillator could have it's own filter and there are two additional ones in "Bus 1" and "Bus 2".
Maybe with using the "Bus" parameter within the oscillator section.

There also seem to be other kinds of morphable filters like e.g. the Uhbie filter in Diva or the "Multi" and "Dual" filter modes in Synthmaster 2.6 (in the "Dual" filter mode of Synthmaster the parameters Mode1, Mode2 and Mix 1/2 could be modulated, besides other parameters).

Anyway based on those comments which i found in the web this is maybe not really the same as Z-Plane filters:

A z-plane filter is just a discrete-time (aka digital) implementation of any filter. Analog filters have poles and zeros in the s-plane, which is represented in rectangular coordinates. Digital filters have their poles and zeros in the z-plane, which is represented in polar coordinates, and maps to the s-plane through the bilinear transform. My understanding of EMUs implementation is that the poles and zeros on the plane can moved around in realtime, creating a morphing of the filter type over time. ie lowpass to bandpass of different cutoff frequencies and resonances. Not quite the same as fading between two different filter types, as any given filter between the starting and ending filters is likely to be entirely different depending on the path of the poles.

Yeah, thats exactly right.

For instance, the peak/shelf z-plane - you set 2 groups - each has two freqs - the top and bottom, and a gain paramter. The morph function moves top1 to top2, bottom 1 to bottom 2 and gain 1 to gain 2.

If your frame 1 freqs are 60hz to 1000hz, and the frame 2 freqs are 10khz to 20khz, then you are morphing from a low pass to a high pass filter. In reality, you have ONE bandpass filter, and the morph function is just shifting the freqeuncy ranges and changing the gain.

BUT, if you crossfade between a highpass and a low pass, all you are doing is running your noise thru 2 filters, and changing the output level of each - theres a big differance since it isn't interpolating between the two - if your lowpass is set to 200hz, and the highpass is set to 3khz, you'll miss everything inbetween.

The GRM tools has a bandpass filter which can replicate the emu shelf filter quite well. However, the lowpass+res+eq morph filter is a bit more tricky - you'd need a GRM like bandpass folowed by a lowpass, and then you'd have to set it up so you can control the lp cutoff and the bp bandwidths with one midi cc (although this is actually a piece of piss in logic, just use multiple transformers in the environment to convert, say CC1 (the mod wheel) to all of the applicable CCs) Also, the emus use a limiter, so you can really overdrive them.

Ingo

Dan_E10 · Post by **Dan_E10** » Thu Apr 24, 2014 1:29 pm

Trancit wrote:It´s quite a long time ago, since I last used Emulator X2, I think I had installed, but what I can remember:

I was never very impressed by those hyped Z-Plane filters and I was far away from considering to use them...

Most of these 50+ filters sounded to me:
1. very similar and very digital
2. as they took of a lot of punchness and pressure out of the sound
3. often just like phasing
4. as they could easily replaced by one effect device on the output instead of letting my CPU calculate one for each voice...

I don´t miss them at all and I think these EMU plugins were the worst and "buggiest" VST I ever used...

I haven't had issues with bugs in Emulator X or X2 but it sounds like I've been lucky. I may be leaning in your direction on how Emu's filters sound. They definitely have a lot of character. I'm still trying to decide if that's a good thing or a bad thing. Some of the Morpheus demos on youtube do sound pretty good though.

I've been playing around with X2 again trying to see if I can work up the courage to buy a copy of X3 so that I can transition everything over to the Win7 64bit. There is a place that still distributes it pretty cheap along with a bunch of Emu's sound sets. Even given the low prices, I'm still trying to decide if it's just throwing money away.

Filterscape sounds very intriguing. I may pick up a copy of that even if I get X3.
Dan

Uncle E · Post by **Uncle E** » Thu Apr 24, 2014 4:20 pm

Ingonator wrote:Also in Largo for each oscillator you got a "Balance" parameter (inside the mixer section) to adjust the routing of that oscillator to both multimode filters.

Do you think Waldorf will ever add sample support to Largo? It seems like an obvious move to me, they could simply sell it as an add-on pack just like they do with the Blofeld SL.

EvilDragon · Post by **EvilDragon** » Thu Apr 24, 2014 4:32 pm

I don't think they will. Blofeld has to have some upper hand to keep selling. Besides, Largo is already quite a hefty plugin as it is (no shared memory between plugin instances, so each instance eats about 300 MB of RAM...)

Ingonator · Post by **Ingonator** » Thu Apr 24, 2014 4:57 pm

EvilDragon wrote:I don't think they will. Blofeld has to have some upper hand to keep selling. Besides, Largo is already quite a hefty plugin as it is (no shared memory between plugin instances, so each instance eats about 300 MB of RAM...)

Not with the 64-bit update as far as i tested it...
In the 64-bit Largo at Windows 7 64-bit and in Live 9.1.2 64-bit a Largo instance is around 115 MB. The high RAM use could be a bit-bridge related problem.

Also freeing up RAM when Largo instances are deleted seems to be fixed.

Ingo

Uncle E · Post by **Uncle E** » Thu Apr 24, 2014 6:08 pm

Ingonator wrote:Not with the 64-bit update as far as i tested it...
In the 64-bit Largo at Windows 7 64-bit and in Live 9.1.2 64-bit a Largo instance is around 115 MB. The high RAM use could be a bit-bridge related problem.

So about that Largo SL...