Arturia V Collection 6

[VariKusBrainZ]

Did a quick comparison between DX7v and Sytrus importing sysex and the DX7v did better job, it was quite obvious for many of the imported sysex patches that Sytrus had converted them wrong

[VariKusBrainZ]

Maybe, but, 25 voices also isn't exactly much compared to what you have with most VSTi's. I'd guess the minimum processing power in hardware synths is always due to cost efficiency. And, of course, because it's doable, when you don't have to process a complex OS, and all sorts of background tasks.

But often these 'DSP chip synths' are running at a higher sample rate than most people run their soundcards

[ghettosynth]

I never disagreed with this. It is just that many emulations today use more advanced methods and are technologically better emulations. But a very important aspect in replication is measuring too. So there are different aspects that can be objectively evaluated, but it is not always that easy.

It is still not subjective though. You measure to get an objective range and objective behavior to model. If it is not right within that range of variation, it is wrong, simple as that. Now, whether you like it or not is subjective, you may like aliasing filters, you may like a waveform that is not accurate better than the original, but it's not a matter of subjective opinion if it is accurate or not.

I have had hardware of the same type with different sound. I am not talking about broken chips. When one emulation sounds different than another it does not have to be wrong. Objectiveness isn't that easy achievable as stated. You can have 10 analog chips and all sound different.

You did not understand what I said. That difference between those 10 different synths is the range, it still has to be within that range. I'm not going to argue this any more though, it's getting silly.

Yes, it is silly. This is the developer perspective, the user at home does not measure this and so it isn't easy for him to judge objectively. Or would you get three TB-303 before you decide which emulation you want to purchase? (which does not mean it would be easy for the developer either).
That is also the problem with some comparisons on YouTube. When emulations are compared to one original hardware there, that does not necessarily mean one of those emulations is more authentic.

Again, that doesn't make it NOT OBJECTIVE. The old Arturia models are objectively less accurate than the newer models and the models of some other vendors with respect to the analog circuits that they purport to be emulating. Whether a user knows this has fuckall to do with objectivity.

There's no wrapper that you can dress this up in that will make bad models of analog subjectively bad. They are bad because the models fail to capture important aspects of circuit behavior.

[el-bo (formerly ebow)]

Thanks to this thread, I have demoed these products for the first time, ever. And while I think most of them sound pretty good, I've kinda fallen for the CS-80V

And while I was originally tempted to just go for the Analog Lab (Not much of a sound designer), the CS interface offers so many little tweaks that even a novice can use to effect positive changes to the sounds.

Whether it is a 100% accurate emulation is not important to me. I have never played a real CS, but I have heard many presets that throw back to the sounds it seems to offer. Suffice to say, this thing sounds exactly like I would expect a CS to sound. That's all I would ask

[Ingonator]

After i played a bit with Piano V2 and DX7 V and figured out how to do certain stuff there (FWIW i had posted about that on the previous pages...) i now start to ckeck CMI V in more detail.
Obviously this is much more complex than just being a sample playback source.

I now tried to find out what exactly is done in the "Time Synth" after a sample was imported by analysis/resynthesis.
So far at the Edit page of the Time Synth it looks like there are 32 additive harmonics where each harmonic/partial has it's own multstage amp envelope.
So opposing to having a static sound based on a single additive waveform more or less this is "complex additive synthesis" comparable to what was possible with the resyntheis in e.g. Alchemy with the additive engine there.
In theory this kind of resyntheis could replicate any possible sound. In practical use especially in the way it was used in the Fairlight of course this method has it's limitations concerning properly replicating a sample/sound.

If you edited the imported sample/sound in the "time synth" it s also possible to convert it back to a sample with the "Compute !" button at the edit page.

Of course you could also program the harmonics via their corresponding envelopes from scratch. At the time synth edit page you could select "Reset All" at the "Wave Envelope" section to reset all 32 envelopes for the 32 harmonics. You coudl also create a "static" waveform if you create a kind of "Gate envelope" or you could use the same envelope shape for all harmonics you want to use. There is also a copy/paste feature for the envelopes to be able to use the same with multiple harmonics/partials.

Just nocticed another cool feature in the "Time Synth" edit page. By default the 32 harmonics each use a Sine wave like all additive synths. Anyway with the "Wavetable Voice" drop-down menu you could choose between 6 single waveforms that are then used for all 32 harmonics or you could choose one of 4 wavetables which replaces the 32 harmonics with a waveform in a wavetable. Still each of the 32 waveforms then has it's own amp envelope so it is not a usual wavetable morph.

At the "control" tab you could have a 2D or 3D display of how teh resulting sound looks and/or changes over time.

[chk071]

Maybe, but, 25 voices also isn't exactly much compared to what you have with most VSTi's. I'd guess the minimum processing power in hardware synths is always due to cost efficiency. And, of course, because it's doable, when you don't have to process a complex OS, and all sorts of background tasks.

But often these 'DSP chip synths' are running at a higher sample rate than most people run their soundcards

I'd still think the same amount of optimizing flows into software synth development. It's probably easier to optimize something for a hardware DSP though, because you specifically develop for that one "platform". On PC or Mac, you have several processor specific features you can use, or not use.

[ghettosynth]

Whether it is a 100% accurate emulation is not important to me. I have never played a real CS, but I have heard many presets that throw back to the sounds it seems to offer. Suffice to say, this thing sounds exactly like I would expect a CS to sound. That's all I would ask

I think that there's some mixing of ideas in this conversation about emulation. I don't care at all if an emulation of a synth is completely accurate with respect to all of the details of the synth. What I care about is that the analog modeling is accurate.

So, the CS80 is not an accurate filter model. I still use the CS80v, it's one of those synths that can sound ok if you don't push the filter. I don't really use the 2600 much though and that's because I think that ACE does a better job of capturing the spirit of the 2600 because the modeling is better. I don't care that it sounds just like a 2600, I care that the filter behaves well when you push it. I care that audio rate modulation works as well as it can in a digital environment.

[foosnark]

The easel GUI is beautiful but honestly it is a chore to work with, Aalto is just so much better to get a glance of what is happenning and parameter values.

I don't find BEV difficult at all, aside from still thinking upside-down envelope sliders are weird and not liking having the extra modulation options hidden behind another panel.

I'm not fond of the pie-chart dials in Aalto, and the control scaling is kind of weird. Dialing in smaller amounts for FM index, timbre etc. isn't as easy as it should be.

After getting used to BEV I miss some of the things it can do when I use Aalto -- like letting the sequencer be clocked by keyboard notes instead of internal or external clocks. And I really wish the modulator could be routed to other things besides FM; I haven't been able to get the LFO or looping envelope to track pitch well enough to substitute.

Overall though, those two synths are a modular (and a specific, weird flavor of modular) vs. semi-modular approach.

[egbert101]

<delete>

[Ingonator]

Based on what i just posted about the "Time Synth" additive synth in CMI V i got a little "warning" for those who had used additive style waveform editors (like in a few wavetable synths) to create single and "static" waveforms.

The way how it works in CMI v is NOT the same. Opposing to drwaing the partials for a single waveform in a simple editor as already mentioned in my previous posts in teh CMI V "Time Synth" you got 32 harmonics where each partial has it's own multistage envelope that represents the amplitude/volume of that partial/harmonic changing over time. So the envelopes represent both an amp envelope and a change of the timbre over time.
Those envelopes could have just a few stages but go up to a few hundred stages if you like (which is often the case if you do resynthesis of a sample).

[wagtunes]

As far as emulations go, everybody has a different priority. For those who want accuracy, especially in regard to how the filters respond, the Arturia synths probably won't appeal to them. For those who just want a synth with a decent sound, even if not 100% accurate in any aspect, Arturia might be sufficient.

Personally, I like their synths. I'm not in love with them, but I like them and I do use them. I have certainly heard synths that sound worse.

I mean let's face it. There are some people who are so picky in regards to authenticity that they don't own any soft synth emulations because they don't feel anything can replace hardware.

That doesn't make any of us wrong or stupid or lacking in taste or any of those things that get constantly implied here (just read some of the comments if you think I'm kidding). It just means we all care about different things.

Me? I'm happy if the GUI is usable and the synth doesn't crash my system.

[el-bo (formerly ebow)]

Thanks to this thread, I have demoed these products for the first time, ever. And while I think most of them sound pretty good, I've kinda fallen for the CS-80V

And while I was originally tempted to just go for the Analog Lab (Not much of a sound designer), the CS interface offers so many little tweaks that even a novice can use to effect positive changes to the sounds.

Whether it is a 100% accurate emulation is not important to me. I have never played a real CS, but I have heard many presets that throw back to the sounds it seems to offer. Suffice to say, this thing sounds exactly like I would expect a CS to sound. That's all I would ask

Someone else is coverted to the church of CS-80V Welcome brother. Enjoy time travelling back to the seventies.

Haha! Thanks. It's really a beautiful thing. Great pad/ambience machine

Unfortunately, I've also been demoing the Analog Lab. It seems that I'm finding very few dud presets, from any of the collection. Considering that I went through most of the Repro-5 presets, these last few days, I'm surprised that this is where I'm finding the most inspiration

[aMUSEd]

The ARP is also great - much prefer it to the Timewarp one which is dissapointing (might be technically more 'authentic' but this just goes to show how authenticity isn't everything)

[hyperscientist]

Considering that I went through most of the Repro-5 presets, these last few days, I'm surprised that this is where I'm finding the most inspiration

Why? One could maybe argue that U-He's engines have more finesse, but Arturia's presets are excellent and always were! Very musical and usable.

[noiseboyuk]

I'm kinda surprised how many MAHOOSIVE loop clicks there are in CMI. Presumably that's period accurate, but sheesh.

Overall, based on the Analog Lab 3 noodling today, I don't think AVC6 will be high up my own list of priorities. If I really want that DX7 or Fairlight sound I have FM8 and Darklight. I'm not massively into either sound, so making use of the programming isn't so appealing. The Buchla might be fun, but it seemed very sfx heavy and I didn't especially warm to that sound either. The pianos sound pretty good to my ears, but I have a ton of those so.... death by over-consumption on my part perhaps.

At the end of the session I just spent 15 minutes going through the release candidate of Repro-5, and it thrilled me 50x more than anything in AVC6. It just sounds gorgeous.