Osc 1 and 2 sound different

Interesting find I have to admit.

The reason they are not the same is because OSC 2 is slightly detuned by default. Set detuning to 0 OR sync the oscillators to make them sound the same.

If you insist that one osc is missing some frequencies make sure you are not using the filters in split configuration.

I am pretty sure they are identical when setup exactly the same (detune!). I verified that with an oscilloscope when I started out with the Virus in order to better understand all the available options.

Zitat

Try putting the 'phase init' to 1 in the oscillator common section. This way the waves start from the same place every time

Zitat

The Common pages contain parameters which affect more than one oscillator at the same time.
Initial Phase

Value	Meaning
Off	The phase angle (see glossary) of each oscillator is reset to a random value at the start of each note. This closely emulates the “free-running” oscillators found in real analogue synthesizers.
1 to 127	The phase angle of oscillator 1 is xed at 0°, oscillator 2 phase is shifted forward, oscillator 3 phase is shifted backwards by the same amount. Because it ensures that the “click”at the start of notes always sound the same, setting the Initial Phase to a value other than Off is particularly useful for percussive sounds.

Modulation destination “Osc Initial Phase”.

Alles anzeigen

Zitat von syqmusic

set the unison voices to 8

This is new information, and changes the game. Unison introduces a good deal of random changes to the sound for each note. It's quite possible that Access decided to use different pitch randomization for OSC 1 and 2 which may lead to the impression of a duller sound. However, I have to try myself.

Syqmusic, thank you for your comments and clarifications.
There is a difference indeed between all 3 OSCs soloed, noted throughout the FQ spectrum.

Using init patch:
-setting OSC1, OSC 2 and OSC 3 to pulse
-setting unison to 8 voices
-setting unison detune to 58
-setting OSC balance all the way to OSC1
Result is OSC1. png

-setting OSC balance all the way to OSC2
Result is OSC2. png

-setting OSC balance all the way to OSC1 back again
-setting OSC1 PW to 100% (muted OSC effect)
-setting OSC 3 volume to 100
Result is OSC3 .png

All images taken using the freeze mode of the analyser after holding the same C key for 10 seconds. This means that the frequencies displayed are the loudest that occured in that 10 seconds interval.

More evidently the curve post 10khz on osc1 is "outbound" convex, while on OSC 2 its more "inbound" convex; indeed OSC1 has richer higher frequencies. OSC 3 looks a bit "clumsy" but with overall similar qualities to OSC1.

There is also a very minor difference noted between OSC1 and OSC2 without unison voices but it really is non audible - you literally have to see it to believe it.

I hope I got this right! Please do not ask "why" this is happening, I do not know

I just tried that myself and come to the following conclusion: Unison detune introduces some randomness in a way that if you hit a key ten times in a row, each note will sound slightly different. If you keep hitting the same note with Unison 8 detune 58 you'll hear that sometimes it will sound considerably thinner and sometimes darker. This is due to the initial phase of the voices. If you set Phase Init to a fixed value, every note sounds the same, also if blending between OSC 1 and 2. And there is no difference at all if Unison is off. At least not one I can see or hear.

psylence: what spectrum analyzer did you use? Looks nice!

If OSC1 and 2 are indeed not exactly the same, then maybe with unison applied this difference/randomness is emphasized. Even "randomness" is coded, which may be applied differently to the OSCs. Naturally I could be wrong!

Regarding initial phase, the above spectra displayed are the loudest that occurred in that 10 seconds interval. By the end of the 10 seconds, thousands of cycles have finished; I don't see how initial phase can affect the end/displayed result. I would like to understand though. Even with initial phase employed, OSC 1 has still louder higher FQs than OSC 2, at least here.

Btw, if it matters, phase initialization is not the same for OSC 1 and 2. For OSC 1, 0=random, any other value it forces the cycle to always start at the middle from the beginning of the cycle. So its more like an on/off button. For OSC 2, 0=random and any other value is the point you would expect a wave cycle to start (like the virus' LFOs).

I didn't want to "advertise" the plugin, although I'm very very happy with it, but since you are asking, it is Fabfilter Pro Q2

Regarding Initial Phase: Play around with the two OSC with identical settings and change the Init Phase value. I think it will become immediately obvious. Now if you think about doing the same with 8 voices you'll probably also come to the conclusion that two start different sets of initial phases might never produce the same spectrum regardless of how long you wait. Naturally, my logic may be flawed...

However, by now I'd really like to hear/see an example of the OPs "problem", because I don't have the means to do exact spectrum analysis, and I'd really like to know whether I simply don't hear the difference.

Alright I see your pov. Yes the difference in phase init is obviously there, i'm just not sure how (if) it affects the higher frequencies per OSC. I agree that the spectra will never be the same, especially if randomness exists, but the big picture is there: the OSCs sound slightly different.

Here's how it sounds here:
https://soundcloud.com/user-37…vs-osc2-access-virus-init

Using init patch:
-setting unison to 8 voices
-setting unison detune to 58
-setting phase init to 1 (so both osc start their wavecycles from the start)

First 4 plays OSC 1 solo
Then 4 times OSC 2 solo
Then setting detune from 58 to 0 (unison still = 8 voices)
First 4 plays OSC 1 solo
Then 4 times OSC 2

btw its funny how op is not sharing more thoughts but I quite enjoy this conversation

Wow thats crazy. The difference is subtle (to my ears), but it's there. Especially clear with the second example, where OSC 2 sounds definitely darker.