Improvements!
-
- u-he
- Topic Starter
- 30188 posts since 8 Aug, 2002 from Berlin
-
- u-he
- Topic Starter
- 30188 posts since 8 Aug, 2002 from Berlin
Yes, but I can't figure out how to download them as ttf...
-
- KVRAF
- 4197 posts since 23 May, 2004 from Bad Vilbel, Germany
"In www.google.com/webfonts you can add a specific font family to your collection, and then in the top right click 'download collection' to download the binary TTF files."
-
- KVRist
- 402 posts since 16 Sep, 2005
I like the League open-source fonts http://www.theleagueofmoveabletype.com/
-
- KVRian
- 1243 posts since 24 Oct, 2003 from Maine
Hello there Urs and co!
I have a suggestion relating to Zebra's implementation of Unison/multi/supersaw whatever you want to call it. I wouldn't expect something like this to appear in z2, however with z3 in the works I hope it's something you'll think about.
Seeing as Diva will have a JP-8000 osc soon, I'm sure you guys realize there's more to it than a bunch of saws detuned. In fact, each hardware VA had their own way of doing it and each sounded a little different. Some had movement, some huge stereo, some little movement but fat sound. Specifically, there were a couple tricks you could do with the Nord Lead 3 and JP-8080 that are basically impossible with 1 instance of Z2.
The NL3 has 5 voice unison with variable detune and full stereo spread, only requiring 1 voice in unison mode. It also has a "Stack" feature which would layer voices until your run out of polyphony, this was also detunable but didn't have stereo spread. So you have 20 saws, 3 in each pan position, 4 sets of 5 pitches, each slightly offset. The result is this huge sound that has very little movement but a lot of harmonics. This does NOT sound the same as 20 saws with linear spread and pitch variation. Seriously. Try it. Sylenth 1 can do a decent impression; put each osc on 5 voices, full spread, the same detune for each, 2 at +1 octave, and offset the pitch with the fine tune control. In z2 you can't select 5 voices. 4 doesn't quite do it and 11 sounds like mush.
Now for the JP-8080. The tricky thing here as I'm sure you know by now is the detune curve is non linear, IIRC it's exponential-ish. Again this sounds different, there's tons of movement unlike the nord. Also the relative volume level of the voices could be tweaked with a knob. The 8080 could layer two patches, and layering the same supersaw patch twice, while panning them in opposite directions would give you that characteristic massive JP8080 sound. z2 has a linear detune curve so the only real way to get a JP8080 supersaw is by using all 4 Oscs with 3 set to double and the pitch and volume adjusted manually. So it's like a 1 OSC JP, with no patch layering, that takes 4 knobs instead of 1 for detune and mix.
So this has a lot of variables, what to do if you want something similar in software? Synthmaster had a decent idea but I think it's kind of cumbersome. Basically you can manually control each voice in the multiosc, volume, pan and pitch. It works, but you need to tweak each voice manually. Not exactly fun to program.
So I propose this:
Get rid of the detune, width, and voice selector, replace them with a small sub panel. It would consist of 6 knobs/options: voices (1-??), detune, detune curve, width, width curve, and mix. The "curve" knobs would interpolate between a concave and convex curve for detuning and stereo width. "Mix" would be analogous to the JP-8080 "mix" knob.
So if I want a JP supersaw I set voices to 7, stereo to 0, and turn the detune curve up. Two of them for the dual layeres. If I want a Nordic sounding one, I bring up all 4 OSCs with 5 voices, full spread, and linear pitch curve.
So basically we've added only 3 knobs, and can now emulate the unison sound of two famous hardware synths and increase the potential for experimentation quite a bit as adjusting the curve parameters would give you a slightly different texture each time. And in zebra fashion, you could even modulate them!
Just something to consider.
I have a suggestion relating to Zebra's implementation of Unison/multi/supersaw whatever you want to call it. I wouldn't expect something like this to appear in z2, however with z3 in the works I hope it's something you'll think about.
Seeing as Diva will have a JP-8000 osc soon, I'm sure you guys realize there's more to it than a bunch of saws detuned. In fact, each hardware VA had their own way of doing it and each sounded a little different. Some had movement, some huge stereo, some little movement but fat sound. Specifically, there were a couple tricks you could do with the Nord Lead 3 and JP-8080 that are basically impossible with 1 instance of Z2.
The NL3 has 5 voice unison with variable detune and full stereo spread, only requiring 1 voice in unison mode. It also has a "Stack" feature which would layer voices until your run out of polyphony, this was also detunable but didn't have stereo spread. So you have 20 saws, 3 in each pan position, 4 sets of 5 pitches, each slightly offset. The result is this huge sound that has very little movement but a lot of harmonics. This does NOT sound the same as 20 saws with linear spread and pitch variation. Seriously. Try it. Sylenth 1 can do a decent impression; put each osc on 5 voices, full spread, the same detune for each, 2 at +1 octave, and offset the pitch with the fine tune control. In z2 you can't select 5 voices. 4 doesn't quite do it and 11 sounds like mush.
Now for the JP-8080. The tricky thing here as I'm sure you know by now is the detune curve is non linear, IIRC it's exponential-ish. Again this sounds different, there's tons of movement unlike the nord. Also the relative volume level of the voices could be tweaked with a knob. The 8080 could layer two patches, and layering the same supersaw patch twice, while panning them in opposite directions would give you that characteristic massive JP8080 sound. z2 has a linear detune curve so the only real way to get a JP8080 supersaw is by using all 4 Oscs with 3 set to double and the pitch and volume adjusted manually. So it's like a 1 OSC JP, with no patch layering, that takes 4 knobs instead of 1 for detune and mix.
So this has a lot of variables, what to do if you want something similar in software? Synthmaster had a decent idea but I think it's kind of cumbersome. Basically you can manually control each voice in the multiosc, volume, pan and pitch. It works, but you need to tweak each voice manually. Not exactly fun to program.
So I propose this:
Get rid of the detune, width, and voice selector, replace them with a small sub panel. It would consist of 6 knobs/options: voices (1-??), detune, detune curve, width, width curve, and mix. The "curve" knobs would interpolate between a concave and convex curve for detuning and stereo width. "Mix" would be analogous to the JP-8080 "mix" knob.
So if I want a JP supersaw I set voices to 7, stereo to 0, and turn the detune curve up. Two of them for the dual layeres. If I want a Nordic sounding one, I bring up all 4 OSCs with 5 voices, full spread, and linear pitch curve.
So basically we've added only 3 knobs, and can now emulate the unison sound of two famous hardware synths and increase the potential for experimentation quite a bit as adjusting the curve parameters would give you a slightly different texture each time. And in zebra fashion, you could even modulate them!
Just something to consider.
♫♪♫♫♪♫
-
- KVRAF
- 12522 posts since 21 Mar, 2008 from Hannover, Germany
I'm quite sure Urs is aware of this. BTW this is mentioned in the Bachelor thesis of Adam Szabo:3*s wrote:
Now for the JP-8080. The tricky thing here as I'm sure you know by now is the detune curve is non linear, IIRC it's exponential-ish. Again this sounds different, there's tons of movement unlike the nord. Also the relative volume level of the voices could be tweaked with a knob.
http://www.nada.kth.se/utbildning/grukt ... _10131.pdf
He built his JP6K synth based on that.
Recently Alex Shore made a Reaktor ensemble based on this too:
http://alex-shore.com/jp-4c/#.Ugs0FNtXvAM
http://www.kvraudio.com/forum/viewtopic ... light=jp4c
If you check the Supersaw macro you see that Alex added a look-up table for the detune curve.
Alex made a thesis about this too. It's included with the download of JP-4c.
He's a KVR member too and posted in the related threads.
Some modifications by myself (including using the NI Monark filter):
http://www.kvraudio.com/forum/viewtopic ... light=jp4c
There were modifications by another member to add stereo spread. The problem is that the Supersaw no longer sounded like a Superswaw. Best is to keep it like it is.
Ingo
Last edited by Ingonator on Wed Aug 14, 2013 8:14 am, edited 1 time in total.
Ingo Weidner
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1
-
- u-he
- Topic Starter
- 30188 posts since 8 Aug, 2002 from Berlin
Dunno. Not my cup of tea.pdxindy wrote:In the meanwhile, there is Spire...
I've meanwhile perfected my "weekend project" to a supersaw algorithm that takes up about half as much CPU than Sylenth1 at comparable quality (which is excellent IMHO) while maintaining fully functional PWM. It can play back arbitrary waveform, so it could be built into Zebra's oscillators.
I also experimented with various detune laws. While Sylenth1 seems to have slightly different offsets for each voice I think I found my personal sweet spot.
I think indeed that Z3 might deploy some sort of advanced unison detune control. Not sure as of yet.
-
- KVRAF
- 2258 posts since 25 Jun, 2008 from Montreal, Canada
Spire is not doing anything for me and I was feeling like something was wrong with me. Now I know I was right!Urs wrote:Dunno. Not my cup of tea.pdxindy wrote:In the meanwhile, there is Spire...
-
- u-he
- Topic Starter
- 30188 posts since 8 Aug, 2002 from Berlin
The detune curve is based on the same polynominal function they used for waveshaping ("normal" Triangle and "normal" Saw). That's how we matched it. - This however is about the way the knob is scaled, it isn't about the way the sawtooths are related to each other. While the main sawtooth remains at its main pitch position, the 6 additional sawtooths are almost perfectly detuned in even pitch distance to each other.Ingonator wrote:Recently Alex Shore made a Reaktor ensemble based on this too:
http://alex-shore.com/jp-4c/#.Ugs0FNtXvAM
-
- u-he
- Topic Starter
- 30188 posts since 8 Aug, 2002 from Berlin
I wouldn't go as far as to say "something is wrong with you" if you don't like something.xx JPRacer xx wrote:Spire is not doing anything for me and I was feeling like something was wrong with me. Now I know I was right!Urs wrote:Dunno. Not my cup of tea.pdxindy wrote:In the meanwhile, there is Spire...
I'm sure Spire is "just the thing" for many people, just like Sylenth1 is for many others. If we wanted to we could have done a similar synth years ago - yet we didn't. Even though the idea of a u-he style EDM synth sounds increasingly tempting.
-
- KVRAF
- 2258 posts since 25 Jun, 2008 from Montreal, Canada
^ Urs, I was kidding, hence the . Spire is a good synth, just not for me. I wasn't really feeling bad! But it's always nice seeing you're not alone thinking something and it's especially nice when that someone is an authority like you!
. Spire is a good synth, just not for me. I wasn't really feeling bad! But it's always nice seeing you're not alone thinking something and it's especially nice when that someone is an authority like you!-
- KVRian
- 743 posts since 14 Apr, 2004
I'm the only one thinking that EDM isn't supersaw and hating anything related to this unending supersaws blurb?!
I think that the more rewarding thing making EDM is to spend hours finding "that unusual sound" than using the mo*th*r*uck*r supersaw...
Bazille and ACE, from my point of view of course, are already the u-he EDM synths, with spaghetti's you can experiment a lot and make something more original than this ****** supersaw sound.
I think that the more rewarding thing making EDM is to spend hours finding "that unusual sound" than using the mo*th*r*uck*r supersaw...
Bazille and ACE, from my point of view of course, are already the u-he EDM synths, with spaghetti's you can experiment a lot and make something more original than this ****** supersaw sound.