Small sound test - ModelD vs. Obsidian
Just for fun tried to match Model D patch in Obsidian as close as possible..
I think result is very interesting if we take in accout that Obsidian can handle 16 poly voices, with fraction of CPU load of Model
D (it means you can easy layer 4 instances of Obsidian for huge polyfonic sounds)
There is few sweeps and short sequences, always played Obsidian and Model D..
https://www.dropbox.com/s/s1mp8lsr7zvs9yq/ModelD_vs_Obsidian.wav?dl=0
even more interestig it is if some FX are applied (which is usually the case when sound is used withing track)
https://www.dropbox.com/s/58xxv4o5jsdghbj/ModelD_vs_Obsidian(fx).wav?dl=0
Comments
That’s interesting. Thanks.
Would you be able to share the Obsidian preset(s)?
I’ve been trying to get it to sound as Moogy as possible, but I think you’ve got further than me!
here patch
https://www.dropbox.com/s/z84ounw1s6k2uje/AYER BASS.zip?dl=0
will try to reproduce more of them later..
Brilliant. Thanks so much @dendy
Doing something like this has been becoming a fun learning experience lately. While composing I’ll often grab external sounds I’m already familiar with because it’s easier to keep the creative flow going. But as things start to gel, or I get temporarily stuck, I’ll pick one of those sounds and do my best to create it in Obsidian. Generally I’ll start oscillator by oscillator, filter by filter, modulator by modulator ... until I get as close to the other patch as I can. Once there I remove the other app.
I’ve been pretty successful and had a lot of fun so far.
Yeah same motivation here :-) Of course you never get 1:1, that's obviously not possible, but if it works in mix after applying fxs together with lot of other tracks, why not ...
Plus it means saving a LOT of CPU to recreate it in Obsidian
I just wish to have time machine :-) I would then return back it time and use some techniques in Obsidian which indiscovered just lately during such challeges for factory presets i made
That’s a pretty nice, thick analog patch you shared. Thanks again. Particularly nice with some delay added, as you say.
@Cinebient
Yeah of course. My goal was not tomproof thst Obsidian can replace Model D. Just that it can sound lot closer than some people think ;-)
But on other side wondering nobody noticed that issue in Model D on biggest resonance in some speciffic filter position.. in my example aroud 0:10 .... obviously some bug in implementation, lots of aliasing and shit there :-)
Disagree with this, exact opposite in my opinion ;-) Obsidian have lot more sharp snappy envelopes... i think actually more than most of iOS synths which are not primary aimed for drum synthesis...
Yeah, challenge accepted. But if possible put also screenshot with patch settings from Model D app ;-)
I'm also experimenting with using multisamples of oscillator waves from my analogue synths, loaded to obsidian sample oscillator - and some very interesting results are going out ...
Now i got idea to sample MD's oscillators with various settings and create multisample based patches in Obsidian... lot of fun :-)
btw i was never big fan of Model D, even HW... there are other things which i'm
missing on iOS .. for example Microwave XT - literally nothing is even remotely comparable to it's FM modulation applied on wavetables.... magic orange box :-)
Oh just realized i didn't mentioned, although it is probably obvious, in my example it starts with model d, then obsidian, then again model D, then again obsidian, and so on...
pianoroll range is from C-2 to G8 ... but of course because Obsidian does not use audio rate modulation in extreme settings you can reach some boundaries
you cannot use OSC as mod source
btw. did another “model d emulation”
https://www.dropbox.com/s/l034tiqazmv1ob8/ModelD_vs_Obsidian2.wav?dl=0
@Cinebient
ok looks like this is lowest and hoghest possible
tone of Obsidian :-)
https://www.dropbox.com/s/joew6eweqgjxo7h/high_low.zip?dl=0
To my ears, I knew the first one was Model D and the second Obsidian. I mean, I've been a synth head for 15 years, so of course that'll be an unfair advantage in determining which is which given Model D has a slightly meatier sound. However, Obsidian is at the very least 99% close to it, and if you were to put both synths in your standard 8-bar loop (one playing for four bars, and the other playing for the other four bars) with other instruments/drums, even I may not be able to tell the difference in sonic quality! Really, the only major difference would be how much CPU is used.
Well done Dendy! Looking forward to what else you come up with here.
@Cinebient
ask whatever you want, i'm happy to answer your questions if i know answer :-)
regarding pitchbend - theoretically you can map
it to +/- 120 semitones, but of course there is limit what really engine can play (so even if you set max range to pitchbend, obviously if you play note c-2 you cannot pitch it down to -120 semitones)
however, higher ranges sounds like there is some kind of "stepping" - i guess this is result of some cpu optimalisation - it is obvious that this brutal CPU efficiency costs some price in more extreme settings ..
here i recorded various pitchbend variants (+/- 12 semitones, then 24, 36, 48,60)
https://www.dropbox.com/s/tqbdshqh80k57ok/pitchbend.zip?dl=0
There’s just no way to get the full sound of the Moog apps from Obsidian.
Obsidian isn’t a Moog emulator. It’s like tofu. It’s a base upon which to build a tasty sound. Moog Model D and Model 15 are meant to emulate those specific analog machines.
That said, Obsidian can get almost close enough to fool you in a mix, and it can do some other stuff that the Moog apps cannot.
I don’t prefer Obsidian to the Moog synth apps. I’m very glad that both are available in iOS.
Yah yah nice nice, so can we get a cool preset pack to buy please?
that's meee :-))
plus wife&son are already sleeping, so what i should
so ? Playing with NS :-))
plus i'm my personal feelings related to these holidays are near zero
that is 12 years old bottle ! I bought it at my wedding, now it's too old to open it :-)))
Comparisons are by nature invidious and this is no different. I think Obsidian’s strengths are more as a hybrid synth/9sampler than solely as a virtual analog synth - it’s too clinical, and the filter does not have a pronounced character. I suspect that this is intentional and that’s fine.
Just need finish my minibrute/sh02/se02/volca/electribe based patch bank and then i will jump on this :-)
Hi @dendy
Would you be able to step me through how I can assign a +-48 pitch bend range to midi notes coming into Obsidean? I need this to enable my Roli Lightblock’s pitch bend to work properly. And so far I haven’t worked out how to do this (pitchbend in Obsidean seems stuck at +-3). So any help much appreciated.
Matt
@MattFletcher2000
it’s +/- 2 by default (init patch)
just go to mod matrix page, locate “pitch wheel” source, it’s by default routed to “Osc All: Coarse Tune” with amout “2”. Change it to 48
Brilliant. Thanks!
Sorry but obsidian isn’t going to replace Model D for those kind of sounds and feeling ‘alive’ unless my iPad is puking in agony and I need more poly Model d is vibez. Obsidian is tick boxes. Nothing wrong with that but even the dev said it was coded to not have any type of character. I like obsidian, don’t get me wrong, but it is what it is.
My experiments are continuing This time 4 instances of obsidian, loaded with multisampled short loops of Model D oscillators (just flat oscillators, no env/filters applied in Model D)
https://www.dropbox.com/s/djdwzyhm89ifef6/Modelized_demo01.wav?dl=0
You rock Dendy! I love the Moog sounds and anything Lofi like. So this will help a lot. Plus I hate how much CPU model d uses vs Obsidian so If I can do it all in Obsidian then I’m down
Funny u mention this. I loaded up @dendy Ayer Bass and then swapped it with the Model D one and went back to Dendys preset. Just more of a full sound in regards to that preset right out the gate without tweaking. Kudos to you. I hope to up my obsidian game like that.
I just came late here, trying to listen to the examples but they seem to be gone from Dropbox.
Is that really the case? Would like to listen to those experiments, since I love Model D.
@sequence
Aah sorry, i did some cleaning yesterday on my dropbox, needed to free some space so i deleted it all .. but i will record some examples later and put them here again
But i do not have anymore that initial AYER patch which i created just from pure obsidian oscillators (maybe somebody can share it, @nkdvkng i think you downloaded it ?)
here example.. but this is not comparation obsidian vs. model d from first post - i do not have those source files anymore ...
here is more creative approach - 5 instances of obsidian playing various patches created from multi sampled single-cycle (or to be exact "few cycles") Model D oscillators...
https://www.dropbox.com/s/hrlqx1s1atc08eq/modelized.wav?dl=0
One thing I noticed is that Obsidian sounds so amazing inside NS2 compared to Auv3s. The in the box instruments just gel so well together. The mix engine sounds so good i find myself just sampling my other apps.