24 Synth Supersaw Comparison

VST, AU, AAX, etc. plug-in Virtual Instruments discussion
User avatar
Urs
u-he
26013 posts since 8 Aug, 2002 from Berlin

Post Tue Sep 21, 2021 7:33 am

I wouldn't call my findings scientific at all - just something that I heard someone else explain to me and it seemed so simple and logical at the time, I adopted the argument.

KVRAF
4979 posts since 17 Dec, 2009

Post Tue Sep 21, 2021 9:19 am

i'll shamelessly repost my post from another thread (because i thought i was posting it here anyway - so many supersaw threads lol).
Ploki wrote:
Tue Sep 21, 2021 4:50 am
I was bored so I made non-scientific test - ran few synths into one of their good sounding filters (fully open LP), into Pro-Q3 > Valhalla Plate
identical proQ3 and valhalla settings.
8bar loop :)
Zebra, ZebraCM, Diva, Hive, Alchemy and Bazille (not in that order), then last 4 are Zebra, Diva, Hive and Alchemy with as much as technically possible Unison. (Hive 2x16 + sub OSC, Diva 6x stack, Zebra 4x OSC with 11 unison, Alchemy with all 4 OSCs 16x unison)

Snap on envelope enabled where it exists.

https://we.tl/t-Pl6rf0QWap

Making a supersaw in bazille is a pain in the ass tho.
Hive is the fastest, diva is the 2nd fastest but the 6x stack makes your cpu go :party:

Wish i haven't sold all synths lol, so i could throw so more into the mix :lol:
Not that i need more than 5 flavours of supersaw in my life

edit: changed link, added Zebralette - surprisingly fast and good Supersaw, no filter tho. Last 4 the same.

edit2:
also here's a supersaw made with Microtonic:
https://we.tl/t-JrbU6Tr4hO
edit3:
also here is all 10 tracks (without microtonic) together :D
https://we.tl/t-j36hhHQTtF
Image

User avatar
KVRist
80 posts since 25 Dec, 2010

Post Tue Sep 21, 2021 8:05 pm

chk071 wrote:
Tue Sep 21, 2021 3:21 am

Can you clear up which one was A, and which one was B in your example earlier?
A = HIVE
B = JP-8080

User avatar
GRRRRRRR!
12572 posts since 14 Jun, 2001 from Somewhere else, on principle

Post Tue Sep 21, 2021 10:54 pm

chk071 wrote:
Sun Sep 12, 2021 7:41 am
I still thought Spire sounded best (unsurprisingly).
"Unsurprisingly", given that you are incapable of objective assessment of anything. I didn't think any of them necessarily sounded better or worse than any of the others, certainly not enough that any of them wouldn't work equally well in the same part in a mix, once you'd dialled in some filters, etc.
Biome_Digital wrote:
Sun Sep 12, 2021 9:34 am
Are you listening on a phone in mono??? :lol: They really don't all sound the same at all. There are some big differences.
Nothing significant I could hear on my $50 Sony headphones at work, which tends to tell me any actual differences are pretty inconsequential - they are all identifiable as stacked and detuned saw waves. I could hear the aliasing on the single waveform on JP6K but it wasn't discernable when you added in the extra waveforms. There is nothing there that would make me choose any one of those synths over any other if I was looking for a Supersaw sound.
Biome_Digital wrote:
Sun Sep 12, 2021 11:59 pm
I think you didn't read the page. Just get the individual wav files and do EXACTLY what you want. That's why I provided them too. :tu:
That's still just 24 wave files, one for each synth. It would take hours of editing to get what he was asking for. And he is right, it is way too hard to compare when you get 4 or 5 different lots of settings for each synth together.
NOVAkILL : Zenbook Duo, Core i7, 16GB RAM, Win10(64), Evo8 | Studio One | JP6K, Union, Hexeract, bx_oberhausen, Odyssey, TRK-01, Vacuum Pro, Invader, Concept, GR-8, Thorn, Equator, VG Iron 2 | Uno Pro Desktop, Uno, Rocket.

User avatar
KVRAF
3607 posts since 8 Jul, 2008 from UK

Post Tue Sep 21, 2021 11:14 pm

tbh I'd hate soft synths to make the same damn sound, I like the fact they all have their own nuances.

I like the supersaws that have a little drift and randomness to them, doesn't that emulate synths of old , heat expansion etc, I always wanted to make something in reaktor along those lines, but I simply didn't have the nouse .

I am pleasantly pleased with all the supersaws from the synths I own now, they all get used, layered, butchered, and manipulated in different ways depending on the way the synth works. I love the variety various plugins bring.

The tests were all much of a muchness, but interesting none the less. I could make use of all of them tbh, none were weak.

Just to touch a side note, i'd love to see Helix get a GUI update and VST3 updates, then id get it for sure, as I always loved the SS in that.
"Part Man, Part Machine, Part Bird, Part Drum - It's Robo-bongo-cuckoo-cop" - James Acaster

KVRian
957 posts since 29 Sep, 2004

Post Wed Sep 22, 2021 12:04 am

LeVzi wrote:
Tue Sep 21, 2021 11:14 pm
I like the supersaws that have a little drift and randomness to them, doesn't that emulate synths of old , heat expansion etc,
Super Saw is not really associated with that. The synth that made the original Super Saw the JP-8000, was a digital synth. It was just code running on Rolands own chips. You wont get different sound, old, heat expansion and shit that you get with analog synths ;)
http://www.adamszabo.com/ - Synths, soundsets and music

KVRAF
4979 posts since 17 Dec, 2009

Post Wed Sep 22, 2021 12:07 am

Well shit converters were a part of that sound
Image

KVRian
1180 posts since 29 Sep, 2003 from Karlshamn, Sweden

Post Wed Sep 22, 2021 12:28 am

Missing/too low values on the decoupling capacitors could have a bit of this impact on sound as well. I wouldn't be surprised if they cheaped out on such things.

User avatar
Urs
u-he
26013 posts since 8 Aug, 2002 from Berlin

Post Wed Sep 22, 2021 1:29 am

adamtrance wrote:
Wed Sep 22, 2021 12:04 am
LeVzi wrote:
Tue Sep 21, 2021 11:14 pm
I like the supersaws that have a little drift and randomness to them, doesn't that emulate synths of old , heat expansion etc,
Super Saw is not really associated with that. The synth that made the original Super Saw the JP-8000, was a digital synth. It was just code running on Rolands own chips. You wont get different sound, old, heat expansion and shit that you get with analog synths ;)
Well, if you look at the history of Roland, the Jupiter-8 (clearly an ancestor in name to the JP-8000) was rather famous for its unison sound. It did not have a detune knob though, which was added with the Jupiter-6. In order for the Jupiter-8 to sound really fat, one had to hope for slight temperature drift of oscillator tune to take effect. For instance the bass line of Nena's "99 red balloons" shows nicely how comparably little the detune of the Jupiter-8 was (it was also famous for its build quality and reliably stable sound).

But of course, there is no drift in then JP-8000 Supersaw, and if one looks at that alone, temperature drift has no role in it.

User avatar
KVRAF
3607 posts since 8 Jul, 2008 from UK

Post Wed Sep 22, 2021 5:20 am

adamtrance wrote:
Wed Sep 22, 2021 12:04 am
LeVzi wrote:
Tue Sep 21, 2021 11:14 pm
I like the supersaws that have a little drift and randomness to them, doesn't that emulate synths of old , heat expansion etc,
Super Saw is not really associated with that. The synth that made the original Super Saw the JP-8000, was a digital synth. It was just code running on Rolands own chips. You wont get different sound, old, heat expansion and shit that you get with analog synths ;)
Right ok, that was the valve ones then. My bad.
"Part Man, Part Machine, Part Bird, Part Drum - It's Robo-bongo-cuckoo-cop" - James Acaster

User avatar
GRRRRRRR!
12572 posts since 14 Jun, 2001 from Somewhere else, on principle

Post Wed Sep 22, 2021 5:44 am

LeVzi wrote:
Tue Sep 21, 2021 11:14 pm
tbh I'd hate soft synths to make the same damn sound, I like the fact they all have their own nuances.
Meh, I don't care one way or the other because I can make them all sound the same without even meaning to. Supersaw is just another way of doing unison, they are interchangeable as far as I'm concerned.
NOVAkILL : Zenbook Duo, Core i7, 16GB RAM, Win10(64), Evo8 | Studio One | JP6K, Union, Hexeract, bx_oberhausen, Odyssey, TRK-01, Vacuum Pro, Invader, Concept, GR-8, Thorn, Equator, VG Iron 2 | Uno Pro Desktop, Uno, Rocket.

KVRist
47 posts since 21 Sep, 2020

Post Sat Sep 25, 2021 8:59 am

Thanks Biome_Digital for creating this interesting comparison!
I learned a lot while comparing those wav files in DAW and recreating the sounds in various synths.
Biome_Digital wrote: Conclusions and observations:

It's no surprise that many of these top synths sound very good and very similar. However, there are some that sound quite a bit different. It's great that digital synths can have a unique character; whether that character is pleasing or not is up to you to decide.
The differences in your comaprison are not necessarily indication of unique character of those synths. It's more about different default settings, or lack of features and fine control of unison in some synths. The various settings affecting the unison sound were not properly matched even in synths that would have allowed such matching.
Like in Serum, you used the default Linear unison mode, but the Super unison mode (emulating JP-8000) would have been better for this comparison.

Various parameters affecting supersaw sound:
  • Full voice unison or per oscillator unison. Not really affecting sound in this comparison because only oscillators were compared. With filters and nonlinear processing in voice path there could be some differences in sound. And of course it would affect available polyphony.
  • Number of saw waveforms
  • Waveform shape, like ideal saws, or highpass filtered JP-8000 supersaw emulation
  • Oscillator/voice phase control, like random, free-running, etc
  • Oscillator drift
  • Detune amount
  • Volume level mix of unison voices
  • Frequency distribution of detuned saws. Linear, JP-8K supersaw emulation, asymmetric, etc.
  • Stereo spread amount, panning distribution of detuned saws, and volume changes of panned saws
  • Aliasing or other noise in the oscillators
All of those can be measured and analyzed objectively using oscilloscope, specrum analyzer, correlation meter and goniometer. Although measurement can be a bit difficult when synth is playing large chords like in this comparison. Much easier to analyze by playing the actual synth plugins than trying to analyze the wavs provided by Biome_Digital.
chk071 wrote:
Sun Sep 12, 2021 7:41 am
I still thought Spire sounded best (unsurprisingly).
I didn't find anything really special in the Spire sound. In the Biome_Digital's comparison Spire sounds just slightly different than the others, but it's mostly because of different settings. There is some aliasing noise and maybe a tiny high freq boost adding brightness to the raw saw sound. You can get pretty accurate emulation of Spire sound in Serum. Set Unison Mode to Super, 9 unison voices, Detune 0.24, Blend 100, NoteOn Rand1 modulating A Fine pitch set to 3, Unison Width 64, and maybe a tiny EQ boost in FX, something like high peak @ 16.7kHz, Q 25%, gain 0.8 dB.

That Spire supersaw sound can be reproduced even more accurately in synths where you have full control of detuning, panning and levels of unison oscillators. In this example I recreated the sound in Kontakt.
https://soundcloud.com/user-533108332-4 ... in-kontakt

KVRAF
5130 posts since 16 Nov, 2014

Post Sat Sep 25, 2021 10:55 am

My favourite supersaw synth was not included here. (Sunrizer). Just saying...:)

KVRAF
4979 posts since 17 Dec, 2009

Post Sat Sep 25, 2021 11:06 am

bluesawsq wrote:
Sat Sep 25, 2021 8:59 am
snip
Great post.
However as a fellow recreator (i often demo plugins and just snap presets i like into another thing i already have) - devices feature set and layout affects how you make sounds.
creating the same thing from scratch vs against a reference would likely not happen due to some wierd hoops you need to jump through that you simply wouldn’t if you were creating from scratch.

Just like the guy who copied Sonnox into MWaveshaper :)
Image

KVRAF
29968 posts since 11 Apr, 2010 from Germany

Post Sat Sep 25, 2021 11:19 am

bluesawsq wrote:
Sat Sep 25, 2021 8:59 am
In the Biome_Digital's comparison Spire sounds just slightly different than the others, but it's mostly because of different settings. There is some aliasing noise and maybe a tiny high freq boost adding brightness to the raw saw sound. You can get pretty accurate emulation of Spire sound in Serum. Set Unison Mode to Super, 9 unison voices, Detune 0.24, Blend 100, NoteOn Rand1 modulating A Fine pitch set to 3, Unison Width 64, and maybe a tiny EQ boost in FX, something like high peak @ 16.7kHz, Q 25%, gain 0.8 dB.
Or you could just... use Spire instead. ;)

Apart from that, Serum really sounds a lot different than Spire. I actually like Serum's sound, but, I doubt that I would use it for supersaws instead of Spire.
Plugins and a DAW. On an operating system. Which runs on a computer.

Return to “Instruments”