Is there a workaround to 1000ms limit on modifier damp parameter?

[Dpoirier]

I'm in need of a special effect in one of my presets. I'm using a synth block to generate the tone I want, and I need it to slowly fade-in to the "mix", so I'm using a 500 ms damp setting on the level parameter. So far so good. But I need that tone to fade-out 2000 ms later, and I can't find a way to do it. If I assign a modifier to a downstream block's input level, for example, I'm still limited to a max of 1000ms, so I can use that to fade it out.

I tried mucking with the sequencer, but it cycles, and doesn't have a "fade out" setting.

I also tried fading in the tone and then killing it (using the synth block's mix parameter), again with a 1000ms damp, and then feed that signal into a multidelay to extend it to 2 seconds, but I've failed at making it work,

There's probably a way to do what I need, but I'm running out of ideas. Any thoughts? In short, some synth tone fades in (within 1/2 sec), stays strong for 2 seconds, and then starts fading out, down to nothing after another 1/2 second. If I had 3000ms as a limit on damp, I'd be fine, but I don't. Help!

 

[iaresee]

Use a continuous controller and do it with your foot? Use an external MIDI controller and have it alter the parameter over a longer period of time?

Otherwise, no, there's no way around this.

 

[Bakerman]

A slow LFO can do this, if you don't mind hitting the switch again to stop it after the fade out. Or try that with sequencer instead of LFO.

If avoiding a second switch press is important, it's likely possible using this technique:

http://forum.fractalaudio.com/threads/lfos-for-controller-transforming.48691/#post-634893

I haven't set up anything from that thread on the Axe II, but can try to get the synth fade thing working later today.

 

[iaresee]

I wonder if you could use a @MorningstarFX box for this? Set up a single switch to step through the values over a pre-programmed rate?

 

[h.c.e.]

Maybe try the sequencer controller? I've used it to generate a curve that fades from 100 down to 0 over the course of 30 seconds…

sequence with a rate of 1 Hz (lowest you can go) -- which will trigger 1 step per second -- and then set it to 32 stages (max you can create). If you set each step to be about 3.125 percent less than the previous one you should get a sequence that goes from 100 down to 0 in about 30seconds)…

Just need to sort the math for your particular timing, but between the number of steps and the rate you should be able to construct a sequence that rises up over .5 of a second, continues at 100 over 2 seconds, then fades off over however long you need (maximum of 30 seconds). Then just assign the run control to a wild card switch or the like. Might could do the trick maybe?

 

[iaresee]

Maybe try the sequencer controller? I've used it to generate a curve that fades from 100 down to 0 over the course of 30 seconds…

sequence with a rate of 1 Hz (lowest you can go) -- which will trigger 1 step per second -- and then set it to 32 stages (max you can create). If you set each step to be about 3.125 percent less than the previous one you should get a sequence that goes from 100 down to 0 in about 30seconds)…

Just need to sort the math for your particular timing, but between the number of steps and the rate you should be able to construct a sequence that rises up over .5 of a second, continues at 100 over 2 seconds, then fades off over however long you need (maximum of 30 seconds). Then just assign the run control to a wild card switch or the like. Might could do the trick maybe?

How smooth was it? Did you hear the steps when they changed?

(I guess I could just try it... )

 

[Moke]

Someone should put this stuff in the 'Wish List' section.

 

[h.c.e.]

How smooth was it? Did you hear the steps when they changed?

(I guess I could just try it... )

Really kind of depends on the parameter. good news is that with a shorter time requirement you could have more discreet steps and make it smoother probably.

 

[Dpoirier]

Maybe try the sequencer controller? I've used it to generate a curve that fades from 100 down to 0 over the course of 30 seconds…

sequence with a rate of 1 Hz (lowest you can go) -- which will trigger 1 step per second -- and then set it to 32 stages (max you can create). If you set each step to be about 3.125 percent less than the previous one you should get a sequence that goes from 100 down to 0 in about 30seconds)…

Just need to sort the math for your particular timing, but between the number of steps and the rate you should be able to construct a sequence that rises up over .5 of a second, continues at 100 over 2 seconds, then fades off over however long you need (maximum of 30 seconds). Then just assign the run control to a wild card switch or the like. Might could do the trick maybe?

This sounds like a great solution, but I'm confused... how do you set different levels for the different steps? I never saw that parameter in the sequencer page... I'm not near my Axe now, but I distinctly searched for some sort of attenuation control, without luck (may have been temporarily blind?)

But yeah, an improvement to controllers (to have longer damping times) would be phenomenal!

 

[mr_fender]

Yeah, sequencer can work. You will get discrete steps with default settings, but if you add modifier dampening you can smooth them out. For long fades, if you set your sequencer rate to 1 Hz as noted above, each step will take 1 second. Then pick the number of steps for how ever many seconds you want the fade to last. Divide 100 by one less than the number of steps used to determine how much to raise or lower each step to get equal value changes (one less because 5 steps would have only 4 value transitions between them). With one second between steps, if you then set the modifier dampening to 1000 ms, the transition between each value will take the full second and you'll end up with a nice smooth linear fade. You can also use a faster rate and more steps to make it smoother as well if needed, just shorten the dampening time accordingly.

If you can work out a way to control triggering it on demand, another option might be the ADSR controller. Each phase of the ADSR can be up to 10000 ms or 10 seconds in length. If you set the sustain level to 50% you can combine the decay and release times and create a slow 20 second fade with that. The sequencer way is probably easier though, since you can assign a switch to the run parameter and easily start the fade on demand.

 

[mr_fender]

The only issue with using dampening to smooth out the steps is that it's response is not linear in terms of time. Value changes start and stop slower than the middle of their transitions. The more stages you use in the sequencer, the smaller the value change between stages becomes and the less you notice the non-linear behavior of the dampening. Using all 32 stages will give the smoothest result, but you have to do a bit of math to calculate the exact rate you need for any given fade length.

 

[h.c.e.]

This sounds like a great solution, but I'm confused... how do you set different levels for the different steps? I never saw that parameter in the sequencer page... I'm not near my Axe now, but I distinctly searched for some sort of attenuation control, without luck (may have been temporarily blind?)

@mr_fender did a way better job of putting my suggestion into words than I could have

But yeah, here's a really quick patch I played around with to demonstrate what I was on about. It started out with some math, but math is hard, found you did need some decent damping to keep it from being steppy (as fender mentioned), and for the reasons he outlined that means you'd have to work out the timing for your event factoring that in. I kind of just futzed with the timing on the sequencer rate and the modifier curve scale and offset, so that the general idea was there of it fading up over the course of 8 steps then holding for 16 steps, then fading off for another 8, and having it take about 2-4 seconds to go through that maneuver... it's not gonna win any sound design awards or anything, but thought it might be handy to see the sequencer steps and the like... just flip on the run control on the modifier sequence manually to simulate a switch stomp or some other "kewl triggering mechanism"

 

[Trashface]

I have suggested this before. Extremely long dampening times and Extremely slow LFO's. I don't know anyone that wants to do math when trying to be creative with these work arounds.

Maybe add a switch that will allow the dampening past 1000ms to infinity? Think about Paulstretch... http://hypermammut.sourceforge.net/paulstretch/

I have used axe fx to create ambient beds of sound but it is limited in the sound design department.

 

[Dpoirier]

Thanks everyone, and particularly h.c.e. and mr_fendrer... but, I still can't get any of this to work for the following reasons:

- The sequencer: it keeps looping (I only want it to execute once, and then shut up). When I assign a modifier to the "Run" parameter of the sequencer, and attach an IA to it, then yes it will trigger the sequence, but it keeps looping... The sample preset graciously provided by h.c.e. (thanks!) does what I need except for the looping which I can't seem to be able to stop (i.e. there isn't a "run once" parameter to the sequencer).

- The ADSR: it also "almost" works, but I cannot find a way to trigger it simply by pressing a button on my floor controller. I seem to need to (a) enable it by pressing the button, and then (b) strike a chord or a note on my guitar (is ADSR only triggerable by some sort of audio signal?!?) ... the "Run" parameter is not assignable to a modifier, it seems to be hard-wired to playing envelope, with only a threshold parameter to play with.

Am I (still) missing something? Maybe our special effects expert, Simeon, has come across something like this in his never-ending explorations of the weird and wonderful?

EDIT: I'm trying to think of other additions I can make, for example to '"trigger" the ADSR with something other than my guitar's signal, or maybe another roundabout way to get the sequencer to be turned off after 3 seconds (how on earth will I achieve that?!?)... thing is, I'll be running out of cpu with all these additions and convoluted workarounds. Oh, what I'd give for a simple enhancement to allow longer Damp values...

 

[simeon]

sounds like you've exhausted almost all of the possibilities. the simplest way, which will also give you the most control, is to do what iaresee suggested and use a cc pedal. otherwise the sequencer (with a bit of damping) is the way to go. you'll need to switch "run" on and off again with a footswitch. you can have more stages than you need, with the extra ones all set to zero, to give you a bit of a "time buffer" so you don't have to switch it off immediately to prevent it looping. you know how many milliseconds to want the fade in and hold to be, so some simple maths and you're golden

[grumble mode on] this kind of thing would be so much easier if fractal would update the sequencer with some of my suggestions [grumble mode off]

 

[Dpoirier]

sounds like you've exhausted almost all of the possibilities. the simplest way, which will also give you the most control, is to do what iaresee suggested and use a cc pedal. otherwise the sequencer (with a bit of damping) is the way to go. you'll need to switch "run" on and off again with a footswitch. you can have more stages than you need, with the extra ones all set to zero, to give you a bit of a "time buffer" so you don't have to switch it off immediately to prevent it looping. you know how many milliseconds to want the fade in and hold to be, so some simple maths and you're golden

[grumble mode on] this kind of thing would be so much easier if fractal would update the sequencer with some of my suggestions [grumble mode off]

Thanks Simeon. So am I correct to conclude that there is no way to trigger the ADSR by a means other than the guitar's signal? If so, then I will probably use the sequencer, and force myself to turn it off (leaving several stages at zero to give me some buffer). Error-prone and convoluted, but that's the nature of workarounds.

 

[simeon]

no the only way to run the adsr is via envelope.

again...another wishlist item of mine is to be able to run the adsr via an internal or external controller...

 

[mr_fender]

You can turn off the run parameter of the sequencer and LFO's to stop them from cycling, but that will cause them to jump back to their starting value. For a ramp up/fade in, that will make it jump back to silence. For a ramp down/fade out that will make it jump back to full volume. One way around this is to use the same controller assigned to the run parameter to also bypass the Vol/Pan block. That will stop the sequencer and leave the signal full up at the same time. For ramp down/fade out, use a second vol/pan block that's turned all the way down and toggle between the two vol/pan blocks to leave the signal all the way off when you stop the sequencer.

If you use more sequencer stages and leave a handful at the end of the ramp at the desired end value, it will give you a buffer of time to turn the run parameter off before it cycles back to the start. For example, to make a slow fade in that stays at full volume, use all 32 stages, use about half for the ramp, and leave the other half at 100%. As long as you stop the run parameter during the stages that are at full volume, the vol/pan block then gets bypassed leaving the signal at full up as the sequencer resets back to zero.

Simeon is correct though, often the simplest solution is the best. Using an expression pedal avoids a lot of setup and ultimately has more control.

 

[GiRa]

For a fadeout I use the sequencer with more stages than needed and damping. Moreover, now I think that you can trigger a "mute" block (filter, volume) with the sequencer and have your chain completely silenced.

 

[markyd]

Long cross fades via the damping factor is one of my biggest wishes! It was on the wish list for a long time. I actually had a conversation with Cliff about this at an Amp Show several years ago, and he said that he would implement an "x10" switch to allow long cross fades by lengthening the damping factor by a factor of 10. Perhaps he will get around to it someday...