Waves Element - How to eliminate attack pops? (patches and audio included)

[lfm]

This is obviously unwanted thing. I asked Waves tech support too, and they said they are aware of the issue and looking into it. They promised to come back to me when they get to a conclusion. Hopefully this gets sorted out fast.

Excellent, I haven't heard back yet - other than it was forwarded to people to investigate.

Really sad to avoid such a good synth because of this.

[fluffy_little_something]

[*]Retrologue - No way that I could see to adjust OSC phase like Massive or Sylenth1. Although the pops were not as extreme as you get in Element, they were audible- but mostly addressable with a 2-3ms attack.[/list]

Maybe via the mod matrix, constant to phase in case there are such items...

[Andy Gone]

I have to say that generally a real generic synth would behave as Element, because the oscillator is not starting zero phase at each note. I would say, to reach good attack, you will need some other tools. However, let's see what Waves tech support is gonna tell.

[lfm]

I still think, since O2 has a sync option - O1 could always start at zero crossing. It could even vary the phase it start coming out, but always at zero crossing - and no problem would be, I think, no pop in audio.

And DCO version should surely start at zero each note, but with the proper wave. Problem now was that is makes a wrong turn, somehow.

It's been since 70's I ran a truly analog synth, and don't remember if those started as low as 1ms attack anyway. I started with Casio CZ in the 80's as my own synths and very digital that was.

Roland SH-32 had this setting, to emulate analog, you could choose how much to have the osc wobbling to create the phasing feel of it. But never got any pops as I remember - but Roland have the knowhow of synths, don't they.

Element was the first from Waves, or did they buy it and branded it only?

Anyway, we wait and see...

[Andy Gone]

After playing a bit with it, I have to say that usually a generic analog synth would do the same. There is no generally a function in an analog synth to start main OSC in zero phase with the key stroke and the VCA attack (which would cure the 'pop' issue), otherwise you get the pop with very short attack of course.

Anyway, it would be nice to eliminate this effect of course, but from what UltimateOutsider is saying ^ there is no so far all good approach on that.

I've got intrigued and will dig more into the issue

[xoxos]

you could probably dig up several threads like this for other synths if you look, they're out there.

why you are unhappy is because your synth is too powerful. i don't have any bias towards waves - they stole my vst name, and they sell shitty shit to people who aren't learning anything.

but for reals take it from a synthesis divinity - there's nothing wrong with what you have, except that the attack time is too short.

resettable osc on gate event is a nice feature - i try to include recallability on all of my synths, yet people still buy synths from waves?! huh, how about that.

but for the functionality it provides, with non-recallable oscs, it's working just fine, you just need to move the attack setting up some.

i'm not saying this to belittle your experience or get into a fight, because i know all about this stuff and everyone knows it - let's consider the other case, if you could reset the oscillator to 0 phase on gate events - then, a mono note that begins before the previous note has faded will cause another arbitrary interruption and click/pop.

there are solutions for this as well but how you handle it depends on how much cpu you want to use among other tolerances.

but definately, definately, believe me, and turn up your attack some, and you'll be having the best time. you are *not* going to find a better solution with a continuous oscillator, no matter how much their tech supports "look into it"

[UltimateOutsider]

but definately, definately, believe me, and turn up your attack some, and you'll be having the best time. you are *not* going to find a better solution with a continuous oscillator, no matter how much their tech supports "look into it"

I don't know how serious you were intending to be, but for the record I'll note that with the patch I was talking about in the OP I had to dial back the attack to 27ms to (mostly) eliminate the pop. On other synths when I have similar attack pops, usually just increasing it from 0 to 1-5ms is sufficient.

27ms significantly changes the character of the sound. There are plenty of other plugins that can handle fast, popless attacks without compromising attack time to that extent.

[xoxos]

wrote this before your last post..

a helpful way to think of it is to consider the wavelength in ms - eg. if you're playing A1 at 110 Hz, then a single wavecycle is about 9ms. i doubt you'd hear too much of a click if the attack time is at least 1/4 the wavelength.. still plenty fast enough for the patch. in fact, if the envelope is any faster, it would be more about "shaping the transient," and you can chuck that out the window as far as being intentful if the osc is non-resettable..

after reading your post.. 27ms is rather long to affect it, unless you're patching bass..

..back in the h/w days, ppl used to say cheap fm synths pwned for bass.. these of course do reset the osc at each note..

well.. given that, as a developer, my next step would be to look for any discontinuity in the attack performance between 2ms and 27ms.. because that's real weird.. you're right, 1ms-5ms should be generally sufficient..

also.. not that i'm real interested in picking on this.. that mini little attack contour looks more like a pickup contour to me... (like the actual attack time is pure 0 and the contour is produced by a declicking circuit in the envelope) .....the attack contour of an exponential envelope should "bow" upward, not downward..

[lfm]

but definately, definately, believe me, and turn up your attack some, and you'll be having the best time. you are *not* going to find a better solution with a continuous oscillator, no matter how much their tech supports "look into it"

Increasing attack time did not solve it for me, at least.
If having long enough release, next key down is creating a new osc instance - and something happends doing that. I could 150 ms attack and still get pops(tried different hosts to be sure, also).

Especially the DCO which is said to be phase synced, and is, have an erratic first halfwave(as OP image shows). Poor implementation in my view. If that worked as described, no issue with pops.

As I see it, they have to emulate a free running osc - since it's software doing nothing until you press a key.

The way to do it would be to emulate a free running osc at random phase in waveform, but always release it at zero crossing, so no pops is generated. Rather delay start of ADSR a bit, even if 27ms is rather long I admit. But you also have keyboard adjust(how each part of keyboard get different speed on ADSR) so you could adjust that to be acceptable response. I would prefer that to sudden pops in sound - but that's me. So using subosc you have to adjust other parameters also to make it work.

This problem was not imminent in analog synths, due to that voltage of the key down should get the osc to the proper pitch first - and you probably never had 1 ms attack times in the first place.

I have a setting in daw at about 5-10ms fade in/out when starting/stopping transport - not to get pops in audio doing that. So range where you sense pops is probably around that.

Waves certainly did something wrong for DCO and I think they should look over that part at least. The waveform should always be a sine and not that quirk OP had in image. If any quirk in sine it would be from zero crossing - having this phase synced osc.

[xoxos]

thank you for explaining, sorry i can't help.

[realmarco]

i remmeber a time where analogue synths didnt make any clicks...why u ask ?

because the envelopes werent fast enough

now with digital you can, if u don't just tweak the attack knob.

I wouldnt be suprised if the "update" just tweaks the knob by default and any faster will be unavailable