Notes held with filter adsr amount

What You are trying to do is to scale trigger gain value.
Scaling like that works only if set before trigger takes place,
sending another value restarts the trigger.

Check legato mode if that is what You are after.
You'll find that in retrigger section of help file.
In legato mode ADSR will not ramp to 0 before retriggering,
but will just restart the envelope.
In that case the Dial will generate new attacks, which would jump to
set value according to attack time.
And - You don't need that {> 0.}
if You multiply the incomming velocity by any amount,
zero will remain zero.
By the way that dial "Amount to ADSR" is set to range -18000. 18000.
what do You do with that amount of gain ?

Hi there,

Firstly thank-you for your reply I really appreciate your time, I have tried legato mode as you suggested on the adsr but it doesn't seem to make any difference but I'm probably doing something wrong.

The dial is on that amount because these values allow for the sustain of the envelope, for example, to take the cut-off frequency up to it's maximum when the amount dial is fully turned..although there's probably a better way of doing it, as there will be with most of the patch no doubt.

It's my first device so I have a lot to learn but it would be great to get this one problem solved, here is the full poly subpatch :

and the main patch :

You'll see if you play it that the filter amount dial is heard when notes are released and pressed again, but not with notes held. Thank-you so much again for your time.

I'm sorry, but I can't play anything, the connections to poly are all lost.
That's because when one makes new from clipboard, the patch does not see the
subpatch.
But anyway adsr is just adsr and is not going to behave differently whatever You try to do.
If You update trigger value it is going to retrigger the whole envelope,
and there is no way to avoid it that I am aware of, legato mode just does not zero
the output before retriggering.
Output of adsr is nothing else but control signal, if You want to
scale it's output, then You could do it at the adsr output, rather than input.
That dial with -18000. 18000. :
negative values amplify and invert phase, so -100 is as loud as + 100,
just phase gets inverted. You need something else for that.
If You want to modify just the part of the running envelope,
like gain of sustain (which works only before release phase starts)
or release time then it would be ok.
But maybe You could try to control that EQ frequency
with some other control signal, like curve, line or whatever else.
or maybe using output of the first adsr, but scaled, inverted or I don't know what...
It all depends what sounds You are after.
Auto wah could be combo of sharp Q and low freq ramped
into lower freq and smoother Q with adjustable ramp time.
I hope that helps You a bit.
--------
I see You use a lot of cycle, saw,tri ,rect and switches to select waveforms,
would 1 cycle object with selectable waveforms be more effective ?
And also waveforms choice could be more interesting ?

There's the answer..scale it at the output, thank-you!! This has solved it, excellent.

Yes there's room for much improvement in the device, it's my first so early days for me at the moment. The negative amounts on the filter amount dial are just to give a wider range of results when altering the adsr amounts, I've narrowed it down to -10000 +10000 which works for now but yes I understand it's not entirely correct for the negative amounts in that it doesn't need to go back that far, so thanks for the advice on how to do it a better way..I shall take note of this.

Your suggestion of the cycle object for waveform choice I'm not sure I know what you mean? I know about using phasor as I have in the LFO's but that's about it.

Thanks again for your help so glad the dial works properly now.

If you did want to play around on it, or to see what I mean about the dial..it's here :

http://maxforlive.com/library/device/4666/altun-update

It does use an outrageous amount of CPU at the moment though, my next step is to learn how to improve this.

I had a look into poly patch and had to rearange it a bit to understand what is going on.
maybe You could use it .

Have a look at help file of cycle~ object. It can use a buffer preloaded with
waveform in any form.
So You could load 4 buffers with sine, saw, square or any other waveform
and simply tell cycle to refer this or that buffer.
Let's calculate - You use 16 oscilators with cycle, saw ,tri and square, makes
16 x 16 voices = 256 osilators running all the time.
if You used cycle method then it would be 4 x 16 = 64,
And You would not need all that selectors for choosing waveform,
which would also add to lighter cpu load.
Of course all that routing would have to be rearanged,
but I am sure it would be worth it.
-------------------
I have tons of 512 sample waveforms If You wish to try,
here just 4 as replacement for the ones You use :

Wow the poly patch is much better, thanks!!

The question about why are those velocities being sent to the lfo phasor phase inlet, they are for the option to retrigger the phase for new notes (unless you were referring to the '>0' box, in which case I now know they are not necessary )

The cycle and buffer method obviously sounds much better and I am looking into this now and will use your waveforms..thanks a lot!
If I succeed maybe I'll try out some of your other waveforms too, I'm just looking now at how the user can specify which wave to use from the cycle / buffer combination.

As for all those gates, again I know there must be a better way of doing it and I aim to get this solved one day too.

It should be just simple message to the cycle object "set Sine" if buffer is named so.
You can leave selection menus as is, or maybe extend them
for additional waveforms.
If buffers would have exact names as in menu den simple prepend set
between menu and cycle would do.
And put buffers into main patcher, not into poly.
And random selection - didn't think about it, but there surely is a simple
solution...

512 sample buffers are really small, so if You prefer not to load files from hd,
one could embed sample values in coll objects and just dump to buffers on patch load.
I can send You exampl if You want.
I'll send You some waveforms to try, there are too many of em,
most of it sounds quit similar, but You can see yourself .