Arturia V Collection 6

[Harry_HH]

I haven't had time to demo this Arturia's clavinet. Waves launched a year ago series of 'little old keys' in very cheap price, including clavinet.
I think Waves did a great job, relating to the expectations. That clavinet is one benchmark to the Arturia, too, although Arturias approach is different.
I did a small demo 'a la Herbie' of the Waves clavinet, when it was published.

https://soundcloud.com/harry_mccloud/watermelon-man

[ghettosynth]

To get back to the point: HOW can you calculate the accuracy of an emulation mathematically?


I'd really like to know, because i wonder how you want to accomplish that. If you mean that you can judge how accurate the circuit modelling is to the real deal, then, yeah, there's surely a point to that, but, mathematically measuring something you have no idea about, code wise? How's that going to work?

Maybe i misunderstood you there, and that's not how you meant it. Not sure.

Yes, it was stated in context of refuting the universal claim of objectivity via statistical methods. As I said, some emulations can be shown mathematically to be more accurate than others, I did not say all, nor did I imply that I meant anything in specific. It was a refutation of the given statement, not a universal claim. Of course if we want to say something about a specific instance of an emulation, e.g., the 2600v, we either have to know something about its model, or we have to measure it to determine where its model fails. However, it's not unreasonable to assume, based on several factors, that we can infer some details of the modeling techniques used.

When you compare without measurements, just pure mathematically, then you just compare the emulation against another model. This says nothing about the real accuracy. You may match your model, but how the real hardware behaves is a completely different thing.

I think that you need to take your own advice here, you can't "play the nerd." Your assumptions are of greater magnitude than the implicit assumptions in my claim and you are so caught up in this that you can't see the forest for the trees.

We can absolutely draw bounds about the expected behavior of SOME circuits and then show that SOME emulations are more accurate than others with respect to those bounds. Your absolutist stance betrays your ignorance.

Beyond that, nobody cares, really. You should just just "be done" because no amount of your prattling on in an attempt to save face is going to change that I don't respect your knowledge of science and if you aren't trying to convince me, then who do you think is listening?

[braj]

So, the current incarnation of the Moog Modular, is it decent? How do the filters sound?

[ghettosynth]

I haven't had time to demo this Arturia's clavinet. Waves launched a year ago series of 'little old keys' in very cheap price, including clavinet.
I think Waves did a great job, relating to the expectations. That clavinet is one benchmark to the Arturia, too, although Arturias approach is different.
I did a small demo 'a la Herbie' of the Waves clavinet, when it was published.

https://soundcloud.com/harry_mccloud/watermelon-man

I like the Waves piano instruments, however, they are sample based instruments. That said, I also passed on the clavinet because I don't care enough about the instrument to have more sample libraries of it than I already have.

The Arturia is physically modeled, no? I'd probably be more interested in the Arturia PM models if they would embrace that aspect and give you a lot more control over the parameters of the model. I'm less interested in a clavient than I am an alien clavinet.

I guess I'm agreeing with you. I think that sampled instruments done well give a great experience that captures enough of the detail and physicality of an instrument that PM instruments don't necessarily improve upon, especially for those not that sensitive to the nuances of the instrument.

[ghettosynth]

So, the current incarnation of the Moog Modular, is it decent? How do the filters sound?

I don't think that they've updated the Moog Modular since V3/V4 in terms of the filter models, but I could be wrong.

[Spitfire31]

Sorry to disturb the current agenda around with an actual a question about one of the Arturia V 6 instruments.

I suppose it's mainly directed at Ingo, who is actually using the CMI, not just listening to the videos.

Last night, I was playing around with the CMI demo and noticed some rather ugly loop clicks in several of the presets. Maybe part of the desired legacy sound? Anyway, my question is, did Arturia provide any loop crossfade option? And if I move the loop points around, do they stick to zero crossings?

Maybe not true to the original, but would be very handy…

The sound of the CMI does make me feel a bit homesick for my old Emax, long since gone. I think I'm going to go for it.

/Joachim

[jme-audio]

To get back to the point: HOW can you calculate the accuracy of an emulation mathematically?


I'd really like to know, because i wonder how you want to accomplish that. If you mean that you can judge how accurate the circuit modelling is to the real deal, then, yeah, there's surely a point to that, but, mathematically measuring something you have no idea about, code wise? How's that going to work?

Maybe i misunderstood you there, and that's not how you meant it. Not sure.

Yes, it was stated in context of refuting the universal claim of objectivity via statistical methods. As I said, some emulations can be shown mathematically to be more accurate than others, I did not say all, nor did I imply that I meant anything in specific. It was a refutation of the given statement, not a universal claim. Of course if we want to say something about a specific instance of an emulation, e.g., the 2600v, we either have to know something about its model, or we have to measure it to determine where its model fails. However, it's not unreasonable to assume, based on several factors, that we can infer some details of the modeling techniques used.

When you compare without measurements, just pure mathematically, then you just compare the emulation against another model. This says nothing about the real accuracy. You may match your model, but how the real hardware behaves is a completely different thing.

I think that you need to take your own advice here, you can't "play the nerd." Your assumptions are of greater magnitude than the implicit assumptions in my claim and you are so caught up in this that you can't see the forest for the trees.

We can absolutely draw bounds about the expected behavior of SOME circuits and then show that SOME emulations are more accurate than others with respect to those bounds. Your absolutist stances betrays your ignorance.

Beyond that, nobody cares, really. You should just just "be done" because no amount of your prattling on in an attempt to save face is going to change that I don't respect your knowledge of science and if you aren't trying to convince me, then who do you think is listening?

Did you ever notice that some people at KVR expect authentic synth emulations?
That you can use a method SOMETIMES is not of help, when we want all our emulations authentic.
You didn't prove it useless to compare an emulation to the original hardware. It is absolutely necessary.

[sprnva]

I watched the CMI V videos at launch and I think they said that if you hold down shift when moving the loop points you have smoother control of their position. This might help with the clicky samples.

[ghettosynth]

That you can use a method SOMETIMES is not of a help, when we want all our emulations authentic.

LOL! You don't even understand the parameters of the discussion.

[elassi]

Last night, I was playing around with the CMI demo and noticed some rather ugly loop clicks in several of the presets. Maybe part of the desired legacy sound? Anyway, my question is, did Arturia provide any loop crossfade option? And if I move the loop points around, do they stick to zero crossings?

Noticed the clicks, too. After zooming in you can see that the loop points sometimes indeed aren't located at zero crossings (ex. preset "Art of Voice").

You can fine tune the loop point by holding shift, though, to correct it.

edit: Sorry, sprnva, wrote it already.

[hellomrbike]

So, the current incarnation of the Moog Modular, is it decent? How do the filters sound?

I don't think that they've updated the Moog Modular since V3/V4 in terms of the filter models, but I could be wrong.

Yeah, I can confirm going back and forth between the versions in V3/V4 and V5 that the sound is still the same. One can try some of the presets common to each version to validate this, for example (the presets designed by Jean-Michel Blanchet tend to be carried over between versions, I notice).

In fact, I found that pretty much all of the pre-V5 synths (CS-80, ARP, Modular, Prophet, Jupiter, SEM) have retained the same relative sound as before, with the exception of the Mini V, which I believe got a new filter model. From the patch notes, it seems they haven't modified the code bases of any of these synths, aside from squashing bugs and updating the GUIs.

[waltercruz]

In fact, I found that pretty much all of the pre-V5 synths (CS-80, ARP, Modular, Prophet, Jupiter, SEM) have retained the same relative sound as before, with the exception of the Mini V, which I believe got a new filter model. From the patch notes, it seems they haven't modified the code base at all, in fact, aside from squashing bugs and updating the GUIs.

last time that I checked the changelogs, the only synth that seemed to have some slightly improvement on code besides Mini-V filter was Prophet V - "The square waveform of the Prophet 5 sounds closer to the original instrument". The rest seems to be all framework based. On 3.0 release there's a "Improved sound engines" - but with no further explanation of it.

[Spitfire31]

Last night, I was playing around with the CMI demo and noticed some rather ugly loop clicks in several of the presets. Maybe part of the desired legacy sound? Anyway, my question is, did Arturia provide any loop crossfade option? And if I move the loop points around, do they stick to zero crossings?

Noticed the clicks, too. After zooming in you can see that the loop points sometimes indeed aren't located at zero crossings (ex. preset "Art of Voice").

You can fine tune the loop point by holding shift, though, to correct it.

edit: Sorry, sprnva, wrote it already.

Thank you, elassi and sprnva!
Now at least I can put the loop points where I want them manually.

/Joachim

[cowby]

Guys, I just demoed DX7 v but I couldn't find the FM Matrix session (I am from FM8 vst). Is there something I've overlooked?

Please advise

[foosnark]

If I run 1 instance of Buchla with the opening patch "volt-aged piano", I get a few pops and clicks, CPU runs around 35-50%. If I load the patch "pad me up", it's unplayable 50-60%.

If I load 4 instances of Buchla, all with the patch "volt-aged piano" playing simultaneously, there are zero clicks and CPU goes down to 30-35%

On my machine I get 8.7% CPU with one "Volt-aged Piano", or 6% CPU with the GUI closed.

Setting it to 4x polyphony, I get 16.4% (GUI open) or 14.2% (GUI closed).

With 4 monophonic instances and the GUI closed, I get 14%.

"Pad Me Up" is about 11.4% with the GUI closed. But things got brutal when I had two instances of it running and I didn't even measure CPU time because Maschine got unbearably sluggish. No pops and clicks though.

FWIW, the magazineware version of Aalto is about 4% for one instance, or 3.4% with the GUI closed... and Maschine itself is 2.2% with no plugins or anything else going on.

(Maschine 2.6.10 32-bit, on a 2011 Core i7 960 @ 3.2GHz, with 8GB of RAM. Focusrite Saffire Pro 40, at 44.1kHz and a 256 sample buffer, with 14.1ms latency total reported by Maschine.)