John Bowen Synth Design

Sorry Marzzz, but use a sweep generator to check the frequency response envelops of the anti-aliasing filters, and you'll see that there was aliasing indeed.

The stepping issue was also an issue.

The Andromeda was an analog synth, it did not alias. Do you mean the stepping/resolution problems of the controllers and knobs? I owned one for a decade, and sold it primarily because there were several issues about it I didn't like (no response to PolyAT over Midi, resolution and range of the controls, etc.) and the fact that someone offered me more than I paid for it(!). A lot of the sonic disappointment was a design flaw, to get it to sound smooth and full, total OSC or filter levels had to stay below 30/100.

I had the A6 for a while and sold it as I was very annoyed with the aliasing and the lack of smooth filters. I know two other users in France who did the same as I did and sold their Andromeda for the same reasons. Nice synth the A6, but sonically very disappointing.

As for using a Solaris to play the piano, each to their own, but I would bet no many people who buy a Solaris do so for its piano sound!

I always have my ob12, solaris seems like hdtv near it.
It's just ob12 has a "crazy/(buggy?)" side, where solaris is neutral.
I don't say one is better than other, it's just a different approach.

For obie filter, I was refering the a6 state variable filter that has a lot of color.

You should sell me your Solaris for $500 to get this OB-12 you say you miss so much...

[quote="kalaspace"].... -the philosophy seems different on the 2 synths : solaris has proportional volume (when filter cut it reduce the volume), andromeda has gain compensation, it's why andromeda low end is stronger...[/quote]
Gain compensation in the Solaris depends on the filter model. The MultiMode 1 (MM1) Filter does use gain compensation, but not with SSM or Mini.

I published this review 3 weeks ago

http://fr.audiofanzine.com/synthe-numerique/john-bowen-synth-design/Solaris/avis/r.127420.html

primary due to the following comparison

http://46.105.122.202/zic/moog_filter_solaris_vs_andromeda.wav
(for each note, solaris first)
in this one, I used solaris moog filter and andromeda moog filter with 50% resonance each.
I was focused on harmonic excitation generated by moving filter on analog side, particularly around 3:20

http://46.105.122.202/zic/solaris_andromeda_piano.flac
(solaris first, andromeda at 1:40)
in this one, I tried to do the best possible piano
For me it is obvious that Solaris is less alive than Andromeda

http://46.105.122.202/zic/solaris_vs_andromeda_overdrive.flac
(solaris first)
in this one, it is my first solaris overdrive patch vs my best andromeda overdrive patch
it is a little "early" but it gives a good idea of solaris "versatility less alive" vs andromeda


secondary, because i tried to do some stabs and other classic patches on solaris, but it always sound so "modern/digital",
it's like an impossible desperate battle to obtain the "vintage" kind of stabs the andromeda can naturaly do.
I had a lot of expectation on the analog emulation features ...
On this problem, I tried a lot of thing ... lots of attempts results in moody sound
(you can have an idea in my 54 first patches, all patches cat2=moody, i dont like them,
I try to be less wild on feedback now ...)


John asked me to redo the moog test
i try to put resonance real equal (and not parameter equal),
so i put solaris around 62% and for some low around 67%
(for each note solaris first)
http://46.105.122.202/zic/solaris_vs_andromeda_moog_filter_static.flac
http://46.105.122.202/zic/solaris_vs_andromeda_moog_filter_env3sec.flac
http://46.105.122.202/zic/solaris_vs_andromeda_moog_filter_env10sec.flac
http://46.105.122.202/zic/solaris_vs_andromeda_moog_filter_env10sec_ad.flac
http://46.105.122.202/zic/solaris_vs_andromeda_moog_filter_env10sec_ad_low.flac


My conclusions

for the moog test,
-it is very difficult to be objective,
analog filter react differently with osc volume,
andromeda volume and resonance is never proportionnally same as solaris on low/end part of keyboard (i tried to do the best compromise after 30 minutes calibration, before becoming crazy :))
-globally there is a good accuracy,
-the difference around 3:20 in the first test is partially generated by the detune of the octave notes.
-the philosophy seems different on the 2 synths : solaris has proportional volume (when filter cut it reduce the volume), andromeda has gain compensation, it's why andromeda low end is stronger

for the complex patch test,
-I partially maintain my point of view for the moment,
Solaris has better versatility than andromeda but sound less alive
-For overdrive type patch, it is not the same kind of sound but, i like each synth equally.
-I tried to do my lead opera patch on andromeda, but it seems very difficult (quasi impossible), due to the vocal solaris filter the andromeda dont have


For oberheim state variable filter, it sounds to different...
Maybe i will publish more later ...


Globally, the test are "parameter equal" oriented,
I try to find the best approach to have a "sound equal" result, but it seems difficult.
My best option now is
filter1 = comb.tube -> vca1 = 0%
filter1->filter2
filter2 = moog or obi -> vca2 = 7%
filter3 = moog or obi -> vca3 = 100%
all with post equalization (bass and treble boost)


I wait for best advice you can give for analog emulation and life giving to patches.