Radias vs. Virus. Why the Radias lacks punch, explained.

Discussion relating to the Korg RADIAS, RADIAS-R and the R3

Moderators: Sharp, X-Trade, Pepperpotty, karmathanever

User avatar
Timo
Platinum Member
Posts: 3106
Joined: Thu Jan 24, 2002 8:53 am
Location: Kaoss central, England
Contact:

Radias vs. Virus. Why the Radias lacks punch, explained.

Post by Timo »

Scratched my head in the past, "no matter what I do, why can't the Radias sound full, punchy and controlled like the Virus?". So I finally did a few experiments tonight, comparing the Radias and an Indigo Virus side by side, and I found the answer - and it's very clear and simple - it is down to just one thing (filters aside). In my humble opinion, if Korg could perhaps address this issue (or add an option to somehow bypass it) the Radias has the potential to sound 10x better....

Short answer: The crucial difference relates to the way the phases of the two main oscillators are handled when triggered (when a key is pressed). The issue is much more prominent when both main oscillators are used and mixed together, causing random phase cancellations.

Answer in full: The phase positions of each of the individual oscillators in the Virus have memory. This means that when you set both oscillators to the same pitch (no detune), you can hear the phases of both oscillators are individually kept in sync, no matter how many times you press a key.

This results in a very focused and clean sound, each and every time you trigger the patch.

The phases of each oscillator in the Virus are therefore not freeform, they're still being calculated when you're not pressing a key. This is illustrated further when you detune the main oscillators slightly. The overall sound is still very focussed and punchy, as the Virus remembers (and continues calculating) the phases of the two oscillators even without pressing a key and the audible result is as you'd expect - subtle phasing, like an LFO, but which isn't interrupted inbetween key presses.

The Radias, on the other hand, selects a random phase position for each of the oscillators per every key press (or, the synthesis engine may just suffer from bad timing issues?) - so, every time you press a note, the phase between them has changed, which has a tendency to make them sound loose, unfocused and a tad messy when mixed, due to random phase cancellations and similar.

Simple example with audio:-

Process used:
* All patch settings are initialised and zero'd. On the Radias, analog tuning has been turned off. Virus pure-tuning has been set to "natural". Both "punch levels" (amplifier) have been turned off. All ADSRs are wide open.
* Both synths were then programmed with the same (very simple) patch settings. On each synth: Both the two main oscillators are enabled and set to sawtooth waveforms. Oscillator 1's pitch is at default, while Oscillator 2 is pitched up by 12 semitones (one octave) above Oscillator 1. Fine-detune is set to 0 (none).

[Virus - multiple keypresses]
[Radias - multiple keypresses]

Even though the patch settings are identical on both synths, you can clearly hear the phase anomalies between the Radias' two main oscillators each time you hit a note. Unfortunately this makes it virtually impossbile to make punchy and focused sounds (like you can on the Virus) as a result.

Below also illustrates the "memory" effect on the Virus, the way it still calculates the phases of each oscillator even in between pressing keys, resulting in an expected phaser type effect if you subtley detune the oscillators. yet it still retains punch and focus as a result.

[Virus - gradual detuning]
[Radias - detuning] (the Radias' fine detune control is a tad heavy handed, so it goes from subtle to heavy detuning much faster than the Virus! But hopefully you can hear that even at low fine detune values, you lose that gated "LFO phaser" type effect, due to the loss in memory of phases inbetween key presses, and the phase-cancellations also adversely affect the volumes of each key press).

I really hope Korg perhaps take this on board as the Radias is an amazing synth and deserves to do even better.

Thanks for listening.

Timo
Last edited by Timo on Tue Jul 29, 2008 2:18 pm, edited 1 time in total.
User avatar
slammah2012
Senior Member
Posts: 374
Joined: Tue Nov 28, 2006 6:22 am
Location: Canada
Contact:

Post by slammah2012 »

I think there is a parameter that, by default, syncronizes by timbre...but I believe I saw a setting in a sub menu that syncs oscillators to "note start'
Time Can be Mastered,
"NOW" is the Password....
Currently running, Hammond M3, Yamaha CS80 + MM6, Roland JX-3P + Mkb 1000 + Vk7 + D110, Alesis Quadrasynth plus, Haken Continuum fingerboard, Korg Radias + Lambda, Ensoniq SQ80, Waldorf Blofeld, Creamware Prodyssey + B4000, Use Audio Plugiator, Arp Axxe ....and some soft synths.
User avatar
X-Trade
Moderator
Posts: 6490
Joined: Tue Feb 14, 2006 9:47 pm
Location: Leeds, UK
Contact:

Post by X-Trade »

well, nt quite the same, but you can get some 'punch' in the unison osc1 mode, as all of the 5 virtual oscs all start on the same phase (which you adjust with knob #2).

I can see what you mean though - virus has constantly running oscillators, whereas the radias just sets a random phase point each time.

but i'm not sure if it could be a good idea on the radias. i've noticed laggy knob response and menus when i've got lots of notes in play. typically noticed when i've played something with a long release and immeadiately start to tweak afterwards for example.
but I can imagine it wouldn't be so bad if, as you say, korg left the phase accumulators running in the background constantly, ignoring the rest of the voice architecture.

much like the LFOs.
User avatar
Timo
Platinum Member
Posts: 3106
Joined: Thu Jan 24, 2002 8:53 am
Location: Kaoss central, England
Contact:

Post by Timo »

'ello :)
slammah2012 wrote:I think there is a parameter that, by default, syncronizes by timbre...but I believe I saw a setting in a sub menu that syncs oscillators to "note start'
Even if that setting was there (it's not, I've checked everywhere!), mearly resetting the oscillators back to 0º at 'note start' is still NOT the issue. The Virus doesn't reset both waveforms back to 0º (zero degrees) every time a key is pressed. It just remembers the phases of each oscillator and keeps calculating them even when a key is not pressed, so when you press a key the phase differences are still remembered and still synchronised. (Incidently, on the Virus you can also even specify what phase position Osc2 is to be reset to, on each keypress, as there's an sub-option that enables you to do so, but I digress).

For example, on the Virus, if Osc1 was triggered and the waveform started at 0º, and Oscillator 2 was triggered and it was at 45º, and both oscillators had no fine-detune. If you release the key, and re-press it, Oscillator 1 will still be at 0º and Osc2 at 45º. On the Radias, the phases of each of the oscillators would change and be completely random and non-sensical for each key press.
X-Trade wrote:well, nt quite the same, but you can get some 'punch' in the unison osc1 mode, as all of the 5 virtual oscs all start on the same phase (which you adjust with knob #2).
This is a completely different kind of punch, and not one I particularly like. It sounds very artificial as it resets all 5 oscillators to 0º each time, and is one of the reasons I rarely use the Unison oscillator. Tweaking the Unison's oscillator phase knob to the other extreme tends to make the 5 waveforms join along one by one in time - like one will start, then another one will join in, then the third - almost like a 'slew'- or fade-rate if you know what I mean. Again, it's not clean and lacks focus. The ideal issue, again, would be to have the phases of each of the 5 waves still being calculated in the background when not pressing a key, so when a key is pressed, all 5 waveforms can be launched immediately, regardless of their current state, and not reset each time. Would also be great if the volume was auto-compensated as you dial in the detune, but this is a separate issue. Does the Virus TI (Hypersaw) or JP80x0 (Supersaw) do this? I've never been able to try them. Again, I digress...

Timo ;)
Last edited by Timo on Sat Aug 09, 2008 2:57 pm, edited 2 times in total.
User avatar
slammah2012
Senior Member
Posts: 374
Joined: Tue Nov 28, 2006 6:22 am
Location: Canada
Contact:

Post by slammah2012 »

Does the radias still do this in Monophonic mode???mine is at the studio, so I can't test right now... it sounds like a polyphonic rotation of some sort...
Time Can be Mastered,
"NOW" is the Password....
Currently running, Hammond M3, Yamaha CS80 + MM6, Roland JX-3P + Mkb 1000 + Vk7 + D110, Alesis Quadrasynth plus, Haken Continuum fingerboard, Korg Radias + Lambda, Ensoniq SQ80, Waldorf Blofeld, Creamware Prodyssey + B4000, Use Audio Plugiator, Arp Axxe ....and some soft synths.
User avatar
Timo
Platinum Member
Posts: 3106
Joined: Thu Jan 24, 2002 8:53 am
Location: Kaoss central, England
Contact:

Post by Timo »

slammah2012 wrote:Does the radias still do this in Monophonic mode???
Actually, no it doesn't! It works properly, like the Virus in this regard! Just not in polyphonic mode. :( However, the Virus retains this focused character even in polyphonic mode, so it would be nice if the Radias could too. The sound quality for percussive stuff would rocket.
User avatar
meatballfulton
Senior Member
Posts: 351
Joined: Wed Jun 20, 2007 11:28 am

Post by meatballfulton »

Interesting that you think the phase coherence results in more punch because analog synths are generally considered punchy but of course the oscillators are always free running.
I sing the body electric
User avatar
X-Trade
Moderator
Posts: 6490
Joined: Tue Feb 14, 2006 9:47 pm
Location: Leeds, UK
Contact:

Post by X-Trade »

slammah2012 wrote:Does the radias still do this in Monophonic mode???mine is at the studio, so I can't test right now... it sounds like a polyphonic rotation of some sort...
I agree, Sounds like when no oscs are running, radias still progresses to the next voice instead of returning to the first. so, obviously, different oscillators are at different rates, the phase would appear to be random because all of the oscillators are at different phases. (okay, I know it but can't put it into words propperly!).

i'm sure you would find the same of some polyphonic analogues.
User avatar
slammah2012
Senior Member
Posts: 374
Joined: Tue Nov 28, 2006 6:22 am
Location: Canada
Contact:

Post by slammah2012 »

For example on my CS80, there is a rotation pattern that is followed by the KAS board,where it will trigger a new oscillator every time you repeat a note unless that osc is already taken...

now on the CS80V software version, you can change between rotation types( Random, Rotate, ReAssign,Reset,UniHigh,UniLow,UniLast......)
and allthough the real "CS80 KAS" rotation is not really available, these factors play into the way a repeated key gets doled its oscillator....

It would be interesting on your other synth TIMO, to give a third example with a chord cluster intersperced between each repeated single note....
to find out if the Virus is indeed sync'd across all oscillators, or if it is just using the "last note" priority on a repeated key........
Time Can be Mastered,
"NOW" is the Password....
Currently running, Hammond M3, Yamaha CS80 + MM6, Roland JX-3P + Mkb 1000 + Vk7 + D110, Alesis Quadrasynth plus, Haken Continuum fingerboard, Korg Radias + Lambda, Ensoniq SQ80, Waldorf Blofeld, Creamware Prodyssey + B4000, Use Audio Plugiator, Arp Axxe ....and some soft synths.
colinf
Junior Member
Posts: 50
Joined: Mon Feb 26, 2007 9:13 pm
Location: CHESHIRE UK

Post by colinf »

yes I agree, this in depth review between the two synths is really of use.
the Virus Ti Polar has a fat sound that the Radias cannot emulate.
its a pity because it should, either Korg release a software upgrade or over time the Radias will lose favour.
sadly I am considering selling mine and buying a Virus because it does produce the sound I want..
User avatar
X-Trade
Moderator
Posts: 6490
Joined: Tue Feb 14, 2006 9:47 pm
Location: Leeds, UK
Contact:

Post by X-Trade »

I don't really see how this can make that much difference either way.

I have recently aquired a Virus B albeit temporarily. I don't think it sounds anywhere near as clear as the Radias and I can easilly make the Radias sound much fatter and more powerful... although the continuous variaton on the waveforms and the more focused resolution on the detune control is much more useful...

But the Virus sounds so low-fi that I hardly use it.
fac
Junior Member
Posts: 62
Joined: Tue May 06, 2008 1:36 am
Contact:

Post by fac »

X-Trade wrote:I don't really see how this can make that much difference either way.

I have recently aquired a Virus B albeit temporarily. I don't think it sounds anywhere near as clear as the Radias and I can easilly make the Radias sound much fatter and more powerful... although the continuous variaton on the waveforms and the more focused resolution on the detune control is much more useful...

But the Virus sounds so low-fi that I hardly use it.
They simply sound different and have different features. Neither is better than the other. I just bought a Nord Lead 3 after people told me they were cold and sterile, but it sounds perfect for the kind of music I make. I also have an MS2000R, Virus A, Novation Nova, Nord G2, Waldorf XT, Ensoniq ESQ-1, and Yamaha FS1R, and they all have their own sound and cover their own territory. The Virus does better at dark, lo-fi sounds, whereas the Radias is probably good at crystal, hi-fi sounds (like the NL3 or G2). If you prefer one, then get it. If you want both, then get both. Just don't expect one to fully emulate the other.
Last edited by fac on Fri Aug 08, 2008 2:25 pm, edited 1 time in total.
http://cdbaby.com/all/fac

Nord G2 + Waldorf XT + Korg MS2000R + ER-1 + ES-1 + Boss DR-770 + Dotcom modular + Mobius + Ensoniq ESQ-1 + Atomo Mochika
User avatar
EJ2
Approved Merchant
Approved Merchant
Posts: 2291
Joined: Mon May 13, 2002 11:46 am
Location: Port Rowan, Ontario, Canada
Contact:

Post by EJ2 »

X-Trade wrote:I don't really see how this can make that much difference either way.
I have recently aquired a Virus B albeit temporarily. I don't think it sounds anywhere near as clear as the Radias and I can easilly make the Radias sound much fatter and more powerful... although the continuous variaton on the waveforms and the more focused resolution on the detune control is much more useful...
But the Virus sounds so low-fi that I hardly use it.
I recently acquired the RADIAS-R for my studio and from my perspective (once owned an Access Virus), saying the RADIAS lacks punch vs the Virus is a bit of hyperbole. I listened to the demos and to me, it's a matter of perception and taste. I'll take my RADIAS thank you.
Cheers,
Eric
Cheers,
Jim (aka EJ2) Karma-Lab Associate Combi Developer


CATALYST v 2 Blast of Inspiration for KRONOS & OASYS: http://www.karma-lab.com/sounds/catalyst2.html
CATALYST v 1 Combi Explosion for KRONOS, OASYS, M3, & K-M50: http://www.karma-lab.com/sounds/catalyst1.html
CHEMISTRY 3, a Groove Injection for Your Karma: http://www.karma-lab.com/sounds/chem3.html
SoundCloud MP3 Demoshttps://soundcloud.com/ej2-sc
gdh
Senior Member
Posts: 288
Joined: Sat Mar 04, 2006 9:27 pm
Location: NW Toronto, Ontario

Post by gdh »

Thanks for this info. Timo, I have never owned or played any of the Virus's and only listened to demos and ppls work. I like how you started each from a blank slate vs other reports that I have read that have fx on patches which tend to cloud the issue. The one thing that I like about the Radias is that for what I play it can fit in the mix very nicely, how is the access for this? I have only read from some that its darker nature makes it a little more difficult to fit in, then again I have no idea if they were simply using presets or had programmed their own patches. I recently traded my Voyager for a new Prophet 08 (and rec'd $) and have read many reports comparing it to the A6 just like the Radias to the Virus - I fully enjoy well thought out comparisons like the one you did!
nemmo
Senior Member
Posts: 309
Joined: Sun Aug 20, 2006 6:18 am

Post by nemmo »

Running the Radias through a couple of tubes (valves) should solve the problem.
Post Reply

Return to “Korg RADIAS / R3”