Synapse Audio OBSESSION is now available!

[Echoes in the Attic]

No VST3?

No, but Synapse have said they have plans for VST3 of all their products.

[canadian_moose]

All I can say is it sounds fantastic to me. I now have 'The Legend', Repro 1/5 and Obsession. I could only dream of owning synths that sounded this good back in the day. I am equally as 'wowed' with this as I was when I first loaded up Repro 5.

[canadian_moose]

I would also like to thank the two members who posted the free sound patches. Much appreciated as with an 18 month old and 3 older kids, sound design is not always top of my 'to do' list

[Mutant]

BTW Here is a quote from that guy on GS i was talking about earlier:

Originally Posted by recnsci View Post
I've checked OBXA schematic, and you might be onto something. Filter envelope amount is controlled via OTA VCA, and this OTA is driven pretty hard by 3310, so you get smooth saturation/rounding of top of the envelope.

Did some research on it.
http://www.synthfool.com/docs/Oberheim/Oberheim_OBXa/
http://www.siliconbreakdown.com/AANewIm ... 2large.jpg
Maybe he was talking about that little chip just under the CEM3010 marked 3080N.
Here is its data sheet:
https://www.digchip.com/datasheets/part ... 0N-pdf.php

And a quote from a page written by someone who clearly knows what he is talking about:

The 3080 OTA is not exactly a high fidelity component and some distortion is introduced here, which lends the OB-X its dirty edgy sound.

(The chip is present in both OB-X and OB-Xa.)
http://www.analoguediehard.com/studio/k ... index.html

If the original hardware the Obsession is modelled after did not have that smooth transition, then it was most probably modified.

[e-crooner]

For me the differcne betwen activa and muted voices loks very clear, even if i have a greater distance from my display (which is a notebook) and the the medium sized GUI.

Btw, I forgot to reply to that. I think "active" does not refer to voices that are not muted, but to voices that are currently sounding (basically like the Midi LED per voice), so the red LED and the cross are for two completely different things.

[e-crooner]

As long as you understand that the longer you carry on like an idiot, the more it looks like you might actually be one.

What, did teacher make you Inconsequential Details Monitor for the month? Did they make you sign an oath of office and swear on the bible?

And yet you failed to notice all the misaligned labels or that the Mod Wheel isn't a wheel at all, it's a slider. Maybe it's time someone else got to be Inconsequential Details Monitor? You're obviously not up to it.

Nah, only idiots consider me an idiot

As I already said weeks ago, I don't like the GUI (the Arturia looks much more like the real thing), but I don't see any misaligned labels, which labels are you talking about?!

The mod wheel control is assigned to my mod wheel, so I don't have a problem with that, no. Does anyone even use the GUI pitch bend and mod wheels?!

[mholloway]

I'm very annoyed by the 'default bank' of patches that *always* shows up in the main-level of the preset menu. Every time I click the preset name to access the menu, I have to scroll up through this list of patches, just to get to the Bank area (where I have my User bank saved, plus the factory stuff, etc). It seems totally pointless to me that a bunch of presets are stuck at this 'root' level, all it does is cause a bunch of unwanted scrolling to get up to where the actual banks are.

EDIT: OK, so I did the obvious thing and just manually got rid of them (actually, I made a folder called 'Showcase' and stuck 'em all in there). So, no real problem, but still an awkward choice.

[e-crooner]

Yes, I agree that with some sounds the forced cycling thru the voices is not ideal. With long attack and release times one can hear the voice stealing when playing the same note repeatedly

[urosh]

BTW Here is a quote from that guy on GS i was talking about earlier:

Originally Posted by recnsci View Post
I've checked OBXA schematic, and you might be onto something. Filter envelope amount is controlled via OTA VCA, and this OTA is driven pretty hard by 3310, so you get smooth saturation/rounding of top of the envelope.

That was me. I didn't want to press matter further because I'm not 100% certain and effect is (melo)dramatic - I would hate to be source of unconfirmed rumor on ze internets. But since you've brought it up again, I feel compled to add few details.
Disclaimer:
1) I'm not 100% certain that this is happening as further described
2) Have not investigated did Synapse modeled this behavior

As far as service manuals for OB-Xa will tell, envelop modulation of filter goes like this: filter envelope is generated by CEM3310 chip (same type of chip is used for amplitude envelope), which generates envelope that goes from 0 to 5V and back again. This goes to VCA which controls level of envelope applied to filter cutoff (and VCO2 pitch when this is enabled), that is envelope modulation of VCF. This VCA is based on very simple and back then ubiquitous 3080 OTA, which provides warmth in bunch of VCFs and VCAs. This OTA does not like high signal levels at it's input because it saturates easily, so you always see resistor divider at it's input to bring signal level down to acceptable range.
This is where things start to look weird on OBXa schematic. Resistor divider is 1k/47k which is very low by large margin (for example, divider for main voice VCA is usual 330/100k). I've checked two different versions of service manual, and it's 1k/47k for all voices in both.
If envelope signal is 5V, that would mean that at attack peak signal at OTA input is around 100mV. For OTA, that means slamming it very hard. I was bewildered, because this still does not look right.
Here is simulation of output of envelope, output of Env Mod VCA, and two different emulations of this behavior:

OBXaEnv.PNG

Green line is envelope (this has same attack asymptote as 3310), red is circuit simulation of actual circuit, brown is (computationally) expensive simulation using tanh() function, and blue is cheap simulation based on env*(1-0.5*env) where env goes between 0 and 1.

This is so dramatic that it doesn't look right. And there is one practical issue, top half of sustain range would be severely compressed. There is workaround for this, limit maximum sustain control voltage to something around 1.5V, which could be accomplished in software on OBXa which generates all control signals.

I'm not sure at all this is going on in OB-Xa, only way to be certain is to open one and scope around.


BTW, that cheap simulation (env*(1-0.5*env)) gave me an idea to try this in mod matrix on Obsession. Select filter envelop as mod source, env mod as mod destination, and set negative level for mod amount. My ear fancies -33, and it looks like I will use this gimmick a lot. You can go all the way down to -100, at one point it becomes very obvious that attack peak becomes attack dip, which is useful for some stuff.

Again, I'm not certain that any of this is accurate, please don't take for granted, and I have no idea what and how Synapse emulated in Obsession.

[Chrisk-K]

I'm away from home and cannot try the demo. I'm curious about one thing. Can Obsession modulate the filter cutoff frequency with Osc 1? (a la the Tom Sawyer sweep sound)

[urosh]

Just to be sure, again, I'm not certain this is actually going on in OBXa. There are several possibilities, like, implementation of 3310 envelop in OBXa might not put out 5V at peak (OBXa has weird asymmetrical power, and this might affect 3310, details of it's internals are not known beyond high level functional schematic). Also, both service manuals might be wrong, and intentionally, manufacturers did indulge in such affairs on more than one occasion. I might have made some mistake, but I double checked three times and run sim, because it looks bloody wrong, especially because of effect on sustain level of VCF env.

[exmatproton]

I'm away from home and cannot try the demo. I'm curious about one thing. Can Obsession modulate the filter cutoff frequency with Osc 1? (a la the Tom Sawyer sweep sound)

That isn't possible

[Ingonator]

For me the difference between active and muted voices loks very clear, even if i have a greater distance from my display (which is a notebook) and the the medium sized GUI.

Btw, I forgot to reply to that. I think "active" does not refer to voices that are not muted, but to voices that are currently sounding (basically like the Midi LED per voice), so the red LED and the cross are for two completely different things.

Sorry i meant voices that are activated, not those currently playing.

[Chrisk-K]

I'm away from home and cannot try the demo. I'm curious about one thing. Can Obsession modulate the filter cutoff frequency with Osc 1? (a la the Tom Sawyer sweep sound)

That isn't possible

Too bad but that is not a dealer breaker for me.

[Ingonator]

EDIT: OK, so I did the obvious thing and just manually got rid of them (actually, I made a folder called 'Showcase' and stuck 'em all in there). So, no real problem, but still an awkward choice.

All the "showcase" presets in the root folder are also found in the category folders but without the category in the patch names.