Multi stage modulation envelope MOVED TO PROJECTS

sunsynth · 25 Dec 2007

Many thanks 4 your Present - Exo! 🙂

I don't have 1 this christmas...

Will search 4 a use in my always grown 'SUNglower' 😉

Merry Christmas from me 2 and i really will try not 2 get 2 much drunk 😉

Happy new YEAR...!!! Walter

infuzion · 25 Dec 2007

Really cool Exo!

Of course, you know I have to report bugs: 1) In the right-click menu, for the "Number of Sections" & "Polarity" sub-menus, nothing is pre-selected by default. 2) Snapping seems to work, but the dots & dot-lines are mis-aligned are visually off of the bar dividers. Even the first dot is off.

BTW, would help if your OSM came pre-setup with the basic keyboard, OSC, Poly2Mono, Direct Sound out 🙂 BBTW, I wish I knew about you doing this sooner; I've already edited SM's multi-env to change to Bus system in some parts, & some minor cleanup. I PM'd the link, the the current (& I think that had a bug, so I sent a last-known working copy) older version, in 7zip.

I still have a bugin SM's version of the multi env 🙁

It is snowing hard so I have to shovel; perhaps be back later.

sunsynth · 25 Dec 2007

Infuzion please send me your Snow... 😉

In northern Germani there is no... until now 🙁

Walter

exonerate · 25 Dec 2007

infuzion wrote: Really cool Exo!

Of course, you know I have to report bugs: 1) In the right-click menu, for the "Number of Sections" & "Polarity" sub-menus, nothing is pre-selected by default. 2) Snapping seems to work, but the dots & dot-lines are mis-aligned are visually off of the bar dividers. Even the first dot is off.

BTW, would help if your OSM came pre-setup with the basic keyboard, OSC, Poly2Mono, Direct Sound out 🙂 BBTW, I wish I knew about you doing this sooner; I've already edited SM's multi-env to change to Bus system in some parts, & some minor cleanup. I PM'd the link, the the current (& I think that had a bug, so I sent a last-known working copy) older version, in 7zip.

I still have a bugin SM's version of the multi env 🙁

It is snowing hard so I have to shovel; perhaps be back later.

1) I guess you mean when you look in the sub menus nothing is visually selected? there's nothing i can do about that at the moment, currently you can only ever have the last selected item ticked even if you have lots of sub menus, so you end up with menus that look as though nothing is selected. I made a feature request about this a while back...extra ticks for sub menus it's low priority but IMO really needed.

2) I guess your referring to the fact that theres a gap at the start, if you look closely you'll see that the first yellow line is lined up directly with the center of the first point and from there they all line up to the center of the points, the snapping is exactly right, i just need to move the outline across a bit to line up with the first yellow line, means the first point will over hang the outline slightly.

Regarding your bus implementation, i might implement that at a later stage when there are no known bugs ect... I guess this makes it clearer for though that don't use straight links?, it does look a mess other wise.

cheers 😉

Exo

exonerate · 27 Dec 2007

Just updated the multi env to version 0.02, you can get it at the link in my first post.

Changes.....

clipped the outline area of the env,

I was previous scaling the scroll speed by the zoom level this was causing the scroll bar to move very fast to the right if you don't scroll perfectly straight. Now scaling with static values to prevent this.

Moved 'reset env' to the bottom of the right click menu.

Remaining minor issues....

when changing the 'number of sections' one of the points always seem to start at the top,I haven't found the cause but I'm thinking it could be a link order issue.

Snapping, snaps from the previous point so the snapping will not always line up with the grid but it will be the snapping distance from the previous point ( not sure if or how to fix at the moment ).

------------------

Also with this version I've included it in a test schematic.

The multi env is hooked up to a self oscillating filter i made, it's set to 'self oscillate' already so the env is modulating the pitch of the oscillation.

I've set the filter up so that when the volume of the osc is zero and the filter resonance is set to one i feed the filter input with an impulse causing it to 'self oscillate', its fairly quiet but watch out with turning up the volume of the osc this will cause the filter to oscillate very loudly, especially on saw or square waves.

Also with the multi env I'm noticing slow down on redraws ect.. I'm going to have to go through and optimized as much as possible. I'm going to see if i can come up with a way of generating the background grid from only one 'V loop' primitive, I'm current using 6 just for the background grid, so i think this is where the bottle neck is.

cheers 😉

Exo

infuzion · 27 Dec 2007

Thanks for the updates!

It unfortunately have a few issues: The env start & end can be moved from 0 up & down (which I prefer, but might weird out some... should have some sort of sign that the start & end are locked to 0 though).

I added an extra section, & it is locked horizontally to the next-to-last section. I can't move it L-R, only up-down.

Scroll bar too sensitive, a slight move makes the area jump.

I still don't know what line is actually a measure, beat, etc.

Prefer: All colors & line-types have front-preferences. For me, there are too many different colors for each measure, sub-measure, sub-sub-measure...

Turn off/on extra fills... there is some sort of foreground fill that makes everything harder to read. I can't even find it in the schematic, or else I'd just delete it myself.

I prefer SM's use of the square & labeled ASR points over your non-labeling, because I at first thought your labels "start" & "end" meant the start & end of the entire env rather than the loop points. & SM's current method is easier to read faster.

Insert the new section into a location arbitrarily. So, if I want to put it in the middle, I can.

Snapping should take place when I move the dot +/-40% of the snap-point. Eg right now, if I want to move a dot from the 3rd measure to the 4th, I have to move it R all the way to the 4th line, but to move it from 4th to 2nd I only have to slide it very slightly to the L... hope that makes sense? == Sorry if it alot. That's why I tweaked SM's MSE to use a Bus system, & organized it better; to make the MSE easier to hack.

exonerate · 28 Dec 2007

Thanks for the feedback infuzion 😀

I've took your points into consideration and i think it's coming on nicely now 🙂

New version ( now 0.4 ) http://www.box.net/shared/qsjttwdk4c

Changes....

Brought the properties up to the properties panel, you can also turn off extra frills like the cross hair ect...

removed the start and end labels for the sustain section,now using the 'S' points instead. (does look better i think)

Fixed the snapping, now snaps 50/50 between each snapping point.

When turning the release point into the sustain end it now keeps it's y position instead of jumping to zero.

Made the scrolling less sensitive, (you can hold 'SHIFT' if you need more control)

The grid now only shows bars and 1/4 (beats) , makes things easier to make out although the smaller measures are useful when zoomed in, so i might make them show when zoomed in very far.

The cross hair was getting calculated when just moving the curve points, now only gets calculated when turned on and your moving a control point.

cheers 😉

Exo

infuzion · 29 Dec 2007

exonerate wrote: Thanks for the feedback infuzion I've took your points into consideration and i think it's coming on nicely now

Welcome; I hope I wasn't too picky. Please don't make changes ATT; I'm editing your 0.4.

infuzion · 29 Dec 2007

http://www.mediafire.com/?6yxek0s2ii0 v0.4.7 Changes: ALOT of cleanup; I removed like 110 primitives & about that many links. Hopefully the parts I worked on are easier to read; they are for me 🙂

+ the "0" line will highlight no matter the polarity. The refresh in that module is a full area now, but shouldn't really matter.

I wanted to add vertical snapping, at least to the quarter steps, but I couldn't figure it out. Maybe next time.

edit: Thanks for the line removals... it was getting too busy for me. Though I do like the zoom idea... so I left in those sub-modules for you.

infuzion · 29 Dec 2007

0.4.8 http://www.mediafire.com/?abcdzhymo1t Nothing like hand-hacking some ASM to help one to fall asleep... Only a small % of CPU reduction, but measurable.

I do notice a bug (likely in the SM version also) where if there is no Sustain, the completion of the entire ENV doesn't release the process. The coded ADSR had the same problem, but they fixed it. I forgot the exact fix action, sorry. So I kept in the Code version for reference.

exonerate · 29 Dec 2007

Thanks for the updates infuzion 😀

I do notice a bug (likely in the SM version also) where if there is no Sustain, the completion of the entire ENV doesn't release the process. The coded ADSR had the same problem, but they fixed it. I forgot the exact fix action, sorry. So I kept in the Code version for reference.

Yeah i think the multi env needs the 'fast release'(kill voice) thing added, like the other env's with releases. I give it a quick mention here....ADSR in VSTI module needs updating

In the other env's the fast release is added to the release part of the env but with the multi env there is no separate release section (that i can see) so I've not found a place to put it. I find the section player code to be quite cryptic i can't get my head around what half of it is doing, sambean or malc should know where to put it though.

Maybe a separate thread requesting this is needed? I think it should be there by default because it does affect the toolbox version also.

cheers 😉

Exo

infuzion · 29 Dec 2007

exonerate wrote: Yeah i think the multi env needs the 'fast release'(kill voice) thing added, like the other env's with releases. I give it a quick mention here....ADSR in VSTI module needs updating

I'm not sure what you are talking about; both the SM MSE & other Code env have a "Retrigger" sub-module. I do notice the MSE has this line that the others don't: out = env&(env>=4) + out&(env<4);

The MSE's status coding is such a mess; it has 4 separate Code sections just dealing with each stage/finishing, not including the main Code.

To Do perhaps: * Right-click over dot to force locking X &/or Y directions (good for live situations) * Right-click over dots to have 0,25%,50%,75%,100% jumping (well 0 & 100% aren't needed, but when the range is -1..1 then a middle-0 jump is needed) * Right-click over curve for similar defaults, or atleast remove curve. * Zoom with sub-divisions showing *Look at using Properties or user-prefs in a seperate page for Polarity & other right-click menu items that won't change that much.

Let me know if you plan to attack these, or you want me to. I don't want to work on it & have you make changes at the same time 😇 I'm on GMail & Myspace the most for IM...

exonerate · 31 Dec 2007

infuzion wrote: exonerate wrote: Yeah i think the multi env needs the 'fast release'(kill voice) thing added, like the other env's with releases. I give it a quick mention here....ADSR in VSTI module needs updating

I'm not sure what you are talking about; both the SM MSE & other Code env have a "Retrigger" sub-module. I do notice the MSE has this line that the others don't: out = env&(env>=4) + out&(env<4);

The MSE's status coding is such a mess; it has 4 separate Code sections just dealing with each stage/finishing, not including the main Code.

To Do perhaps: * Right-click over dot to force locking X &/or Y directions (good for live situations) * Right-click over dots to have 0,25%,50%,75%,100% jumping (well 0 & 100% aren't needed, but when the range is -1..1 then a middle-0 jump is needed) * Right-click over curve for similar defaults, or atleast remove curve. * Zoom with sub-divisions showing *Look at using Properties or user-prefs in a seperate page for Polarity & other right-click menu items that won't change that much.

Let me know if you plan to attack these, or you want me to. I don't want to work on it & have you make changes at the same time 😇 I'm on GMail & Myspace the most for IM...

Your OK leave these with me 😀

1) do you mean lock the point so it cannot be moved? hmm.. yeah i could see the advantage that. So you could move a point across without moving the one next to it or the rest of the points.

2) So snapping for the y axis.

3)I could add that but I'm not really convinced it's a useful feature.

4)Definetley adding this, well the subdivision are already there it's just blending them in nicely at a certain zoom level.

5)I prefer to have polarity as a right click option instead of a properties option, everything is easily accessible then.

cheers 😉

Exo

infuzion · 31 Dec 2007

exonerate wrote: 1) do you mean lock the point so it cannot be moved? hmm.. yeah i could see the advantage that. So you could move a point across without moving the one next to it or the rest of the points.

Yes, that scenario is a good one. But please also on one axis only, so for example I can adjust the time of an attack without changing the "height" amount. Or adjust the amount without adjusting the time.

2) So snapping for the y axis.

Yes... really I'm more worried about getting the dot exactly on the 0 when in bipolar mode. So if you're not into Y-snapping, but can get a 0-reset in that would be fine please.

3)I could add that but I'm not really convinced it's a useful feature.

Atleast a "remove curve" or some sort of indicator that there is no curve; sometimes I want a MSE patch to be a clone of a ADSR, which has no curves. Once I move the slope, I can't really tell if I returned it to 0% slope, or 0.01%.

Thanks for your time & thoughts!

infuzion · 20 Jan 2008

How's this going Exo?

Acrobat · 21 Jan 2008

wow thanks Exo, a bright idea! One question: do you think it can be ...ehm... looped?

It's cool how you mixed up two very different things like the multi and your loop player!

exonerate · 21 Jan 2008

infuzion wrote: How's this going Exo?

I did a little update a bit back to 0.5 but forgot to post, you can get that at the same link.

I added a right click option to straighten the curves (right click the curve point)

Added the smaller measures when zoomed in.

I think i did some other things but i can't rememeber what. 😊

I didn't implement locking points, or snap for the y axis, The snapping for the y axis blagged my head a bit it seems the lines other than the center are off, i had problems lining them up with the snap.

I think i had issues with the locking points as well.

If you want to implement these your self then go ahead 🙂 , I'm a bit busy at the moment with some other things so I'm going to leave the env as it is for now, If someone reports a bug or something then I'll get that fixed though no problem.

cheers 😉

Exo

infuzion · 4 Feb 2008

exonerate wrote: infuzion wrote: How's this going Exo?

I did a little update a bit back to 0.5 but forgot to post, you can get that at the same link. I added a right click option to straighten the curves (right click the curve point) Added the smaller measures when zoomed in.

Thanks, that right-click to flatten the curves is handy!

FYI: I'm still working on my updates. So far: * code reduction; removed alot of extra primitives & modules & links, not sur how much K yet but likely 20kb or so uncompressed. Should be a tad less CPU when zooming. + Can zoom out all the way to about 11 minutes (for those long ambient songs 😉 ) + Alpha to some prefs * changed the inner outer box

I like how you added the smaller measures when you zoomed in; I finally understand how the measure lines work I think. I'm going to try to consolidate all the measure lines' modules into one module; should save Kb & CPU. I'm going to also try (time permitting) to add "Insert Section Here" to place a point in the middle easier.

infuzion · 8 Feb 2008

[old link] I added the variable beats per measure in the sub-module muti-env/grid/vertical. There are "sections" which is group (set at 8) of measures, beats, (each line becomes dimmer), & dotted beat-dividers seen only if you zoom close enough. Help needed: I think the beat lines, env dots, or both are off; can someone please test to see which one?

You can also set any number of horizontal lines.

Right-click over points will allow "set to 0", setting the Sustain start & Release start (Sustain loop end) if Sustain Loop is on, & neither if you are over a S or R point.

The very first point is an unmarked square, locked to the left but can have a non-0 value. The very end is marked with F (=fini), & can be moved anywhere.

Some modules are messy, but I have to to keep the size down. The env alone is like 170kb uncompressed.

bug reports welcome, as is setting up the WiKi page better with full menu instructions, etc. cheers

exonerate · 8 Feb 2008

Very nice infuzion!

I like the grid better, you made it so it's not too pronounced but stands out well 🙂

Quite a few little bugs though ( I think i fixed most of them)

the grid wasn't lined up properly as you mentioned, i fixed that.

when scrolling along you could no longer move the points, fixed.

Reset to default wasn't working for the y axis or sustain markers, fixed.

You'd removed the snapping for the points? put that back.

I think that was it, I've added version 0.6.1 to the wiki just below 0.6.

--------------

You've added a 'zoom boost' to the zooming at certain levels i found it quite confusing because the zoom jumps at one level and the grid changes ,(some lines dissapear) and i kinda lost by bearings with the grid layout kinda looked like I'd zoomed in rather than out, judging from the grid.

Its great how your using one V loop for the vertical grid, although trying to modify that module was a pain in the arse, SM became extremely unresponsive, but at least the one V loop should use less CPU when using the env. 😀

Cheers 😉

Exo

infuzion · 9 Feb 2008

Thanks for fixing my mistakes Exo; I rolled back some modules because of a pref error (one pref changed another), but eliminated some improvements you made by accident.

I didn't see where you fixed the line width; seems you left a Float Mult hanging without an output. So I connected it where I think you meant to put it.

I also don't like the XY scroll/zoom bar. It is too awkward to pan & zoom where I want quickly if I want towards the middle, & it isn't clear what direction is zoom. Think my split bars solution is better? I'm not too sure about the ratio for the zoom, but I think the scroll is OK. Looks cleaner anyhow, though the smaller slide-handles can be harder to use? I can see how zooming in very close could use more vertical bars...

Speaking of which, the vertical Draw Loop routine is faster, thanks to a well-placed Select.

0.6.6 here Maybe we should use a separate file host while it is still in beta?

aliasant · 9 Feb 2008

This is one incredibly nice little envelope you guys are making.

But

I had a weird bug in the 0.66 version.

First it loaded just fine and I played it from my midikeyboard. The I clicked the arrow to change presets and everything went to very sluggish behaviour. I checked the performance and I could see that SM had jumped from 2% to 25% cpu usage.

Closed it, loaded it again and repeated it with the same result. Oddisish behaviour.

infuzion · 9 Feb 2008

aliasant wrote: First it loaded just fine and I played it from my midikeyboard. Then I clicked the arrow to change presets and everything went to very sluggish behaviour. I checked the performance and I could see that SM had jumped from 2% to 25% cpu usage.

It is the empty presets. In those empty presets, if you right-click-menu/Reset to Default then everything is fine.

Thanks for testing & the bug report! Sorry for the inconvenience. Anyone else?

infuzion · 12 Feb 2008

I uploaded fixed file... any other issues? I'm planning to fork a version of this for myself, so I want to get this version finished. cheers

infuzion · 13 Feb 2008

Uploaded 0.7.0: fixed bugs: wrong curve set to 0, sustain loop start & end when set to 0

Lots of reduction by pulling sub-modules up a level, some math reduction. Looks uglier & a tad harder to understand, but I got the file size from 154Kb in v0.6.1 to 146Kb, about 5% smaller.

I'm making a fork, so I don't think I'll upload any more changes right now unless they are minor bug fixes. If aliasant or exonerate or someone else wants to share improvements that is cool

TomC · 13 Feb 2008

infuzion wrote: Lots of reduction by pulling sub-modules up a level, some math reduction. Looks uglier & a tad harder to understand, but I got the file size from 154Kb in v0.6.1 to 146Kb, about 5% smaller.

Wait: you've made a schematic harder to understand to reduce the file size by 5%?

Dude, we have terrabyte drives going for 200 bucks featuring transfer rates of 100 MB/sec and you optimize for, let me repeat, 8 kb?

Only my good manners prevent me from writing HOW brain dead I think this is.

Sorry infuzion, no cookies this time...

Tom

aliasant · 13 Feb 2008

Well.

Uploaded 0.7.0: fixed bugs: wrong curve set to 0, sustain loop start & end when set to 0

At the wikisite there is now only one file that looks like the 0.6.osm I downloaded it but am currently on my mac so I cant check if it is 0.7 miss named or 0.6...

About the optimization..

I used to not give so much thought about optimizations but when I released Inerthia I got a lot of bad talk about the loading times being, for some, astronomical. So I decided to have a look at it and try to optimize it. After 6 hours I cut the filesize by more then 30% and loading times improved a LOT.

So I do understand (finally) the need for optimizations but maybe 8kb is a little bit to much ( little ? ) But then again.. If you have a scheme with ten envelopes in that will increase loading times quite a lot. Wouldnt you like it to be optimized then?

I bet this can be discussed for an eternity so maybe we should just let it go and be darned happy that some of us actually do optimize the stuff before sharing it. I dunno..

infuzion · 13 Feb 2008

Sorry, I took 0.7.0 down due to a bad array-points/default bug on my part. Working on it 🙂

8k isn't much, but if every other module is trimmed 8k it can add up. Especially if you use the same module 4 times. Also, often (not always) a smaller schematic = faster running VST, since you have fewer links and primitives processing data.

aliasant · 13 Feb 2008

infuzion wrote: Sorry, I took 0.7.0 down due to a bad array-points/default bug on my part. Working on it 🙂

8k isn't much, but if every other module is trimmed 8k it can add up. Especially if you use the same module 4 times. Also, often (not always) a smaller schematic = faster running VST, since you have fewer links and primitives processing data.

So true!!!

Happy Bug Hunting... I should do the same myself.....

infuzion · 13 Feb 2008

aliasant wrote: Happy Bug Hunting... I should do the same myself.....

Thanks! My problem was I went to far into changes without testing them... & would up with a broken module 3 hours in & not knowing where I want wrong. Fortunately I make incremental backups, but I still don't know were between v0.6.7 & 0.6.8 things went wrong 🙁

TomC · 13 Feb 2008

infuzion wrote: Sorry, I took 0.7.0 down due to a bad array-points/default bug on my part. Working on it 🙂

8k isn't much, but if every other module is trimmed 8k it can add up. Especially if you use the same module 4 times. Also, often (not always) a smaller schematic = faster running VST, since you have fewer links and primitives processing data.

Sure, but moving things up from a module doesn't save you a single link. The art(!) of programming IS having everything nicely laid out (makes it easy to understand AND to maintain, which is often the most important part) and have a small footprint. That's why modulization was invented at all.

Having cryptic but small&fast code or having extremely nice looking code which does NOT perform efficient is like brute force, it solves the problem at hand but doesn't do it nicely and costs you much more time/money in the long run.

Tom

.

infuzion · 15 Feb 2008

TomC wrote: The art(!) of programming IS having everything nicely laid out (makes it easy to understand AND to maintain, which is often the most important part) and have a small footprint. That's why modularization was invented at all.

I agree, but if the said OSM is too large & slow to edit, let alone open in 30 seconds, what good is all that beauty? 1.0.9: SM VERY LAGGY with large files So until auto-reduction happens, I'll reduce by hand. Also, when I unroll a sub-module, I often find I can combine routines.

So, thanks Tom for volunteering for the code beautification & documentation team. Feel free to have at the WIKI for this project.

==

I've improved the right-click menu for points so you can't 0 the sustain loop points (I can't get the GUI to update properly when you do this), put an Release before the Sustain start, etc.

I'm stuck on the snap; I envisioned the snap would snap to grid. In my vision, if you're snap-dragging the 3rd point, it will be on beat, while the others stay off beat. It will only do that all points before are snapped also. But if you snap the 3rd with out the others, then it will be off as much as the first 2. So I have to sum all previous points of a snapping point to get that particular point to snap on beat.

edit: I uploadedmy latest & exo's slightly older bug-fixed versions. I fixed a few more minor bugs, now lighter & faster. > file here< To Do/need help with: + Any-point to snap on beat (eg while previous stay off beat, & ones after if possible) + re-triggering reduction, so same routines (modules) don't get re-triggered unnecessarily ! finding bugs! I found a few that existed for a few versions Things I won't fix/someone else can - Properties panel layout (I just go inside the module(s) - "anchor" for the dashed sustain & release point-lines are rooted at the bottom, but should they be rooted at 0 when polarity is changed?

aliasant · 15 Feb 2008

Thanx Infuzion. Gonna play with it.

Im afraid I cant help with the stuff you mentioned but maybe I can figure out something else that is useful.

=)

infuzion · 15 Feb 2008

aliasant wrote: I'm afraid I cant help with the stuff you mentioned but maybe I can figure out something else that is useful.

That's cool... we can really use good bug testers & feedback. I'm decent at finding bugs, but when you are programming, you are in geek mode & not user mode.

Does anyone have a problem with MAX ENV POINTS being semi-hard-coded to 32? There is a wireless Out for it set at 64, but it isn't used everywhere it should. No one has complained so far that they tried to use more than 32 points, & found that bug. I'm thinking to re-code the prefs to handle up to 100 points, so if someone were to change the max points from 32, the saved prefs would be backward-compatible (they are not now it seems).

infuzion · 15 Feb 2008

infuzion wrote: > file here< To Do/need help with: + Any-point to snap on beat (eg while previous stay off beat, & ones after if possible)

I've noticed that the crosshairs will always snap the way I want, but the point sometimes (when previous points aren't snapped) won't....

infuzion · 17 Feb 2008

new file: 0.7.1 Smaller, faster, better (subjective) + proper snap points I had to move the snap to before the LeftClick Drag's Limit Range. What a pain; took 4 hours to figure out. + reduced CPU by reducing redraw from 25/second to 20/sec + improved refresh speed by limiting Sustain Markers's refresh. Let me know if there are problems. + sexier point-area background + moved/unpacked sub-sub-modules into their parent module, mostly to avoid wireless bugs, but helped me optimize for speed & size. + tweaked scroll & zoom bar ratio + made right-click over points more idiot proof for the sustain loop + a few extra labels for primitives & links just for TomC 😉

Please let me know if you find bugs, ideas, & issues. I think this where I'll fork.

infuzion · 18 Feb 2008

uploaded multi_env_beats_mod0.7.1.1 + fixes background error sorry 🙂 + 7zip Zip compressed

xenophea · 2 Mar 2008

I'd just like to say that this envelope is the dogs. Seriously, I've been trying to do something like this for a while now and you guys have just saved me shite loads of work 🙂

I'll be implementing it into my synth (with a few minor alterations) that i'm working on and rest assured I'll post the OSM up here when it's done.

Thanks again fellas!

exonerate · 2 Mar 2008

xenophea wrote: I'd just like to say that this envelope is the dogs. Seriously, I've been trying to do something like this for a while now and you guys have just saved me shite loads of work 🙂

I'll be implementing it into my synth (with a few minor alterations) that i'm working on and rest assured I'll post the OSM up here when it's done.

Thanks again fellas!

Cheers!

Look forward to seeing your synth 🙂

Acrobat · 3 Mar 2008

infuzion wrote: uploaded multi_env_beats_mod0.7.1.1 + fixes background error sorry 🙂 + 7zip Zip compressed

Hi Crazy Envelopers, I offer my help to re-organize Properties panel, tell me if it is appreciated, I'll put my hands on it!

A. 🙂

infuzion · 3 Mar 2008

Acrobat wrote: I offer my help to re-organize Properties panel, tell me if it is appreciated, I'll put my hands on it!

You can do as you wish: I don't use the Properties panel for the MultiEnv (too much for me), & my own fork is already v0.8.8 after anther 20 hours of work. cheers

infuzion · 15 Mar 2008

updated: * fixed a few smaller bugs thanks to malc * shrunk scroll & zoom ranges * moved “0 tiny db” code into Section Player’s hop (3% CPU reduction) * removed In/Out (use external Stream Mult if needed) * + Optimized Readout-Vector Knob & HardLimitor [old link]

exo: I moved your -140dB audio zeroed sub-module in Section Player into the hop. Is 0.0000000001 = -140db, or do you just want to curb de-normals? With that adjustment, are the DENORM adds with BlkStVal & BlkEndVal needed? -- If someone can please help out with adding new presets, disappearing line-color, Pref-panel layout, optimization, & what ever you want to do, that would be great. thanks!

exonerate · 26 Apr 2008

I've just updated the multi env to version 0.8.....

The main thing in this update is the huge CPU saving from the new addition of a 'hop' memory read. 😀

* Fixed a bug with the sustain looping after resetting the env, (The loop was including the release section) * Consolidated the audio code into two code components. * Now using the the new hop read component instead of the float array to poly, this is saving a huge amount of CPU, about 75 percent on my PC. This multi env now uses less CPU than the basic ADSR Env. * Tidied up the properties a bit.

[old link]

Cheers 😉

Exo

infuzion · 26 Apr 2008

exonerate wrote: * Fixed a bug with the sustain looping after resetting the env, (The loop was including the release section) * Consolidated the audio code into two code components. * Now using the the new hop read component instead of the float array to poly, this is saving a huge amount of CPU, about 75 percent on my PC. This multi env now uses less CPU than the basic ADSR Env.

Thanks Exo!!! I didn't see the new hop-read yet; cheers. I was about to give up on this thing since the CPU was so high, even after all my ASM-fu. Is there any audio difference?

Now I have to redo all my ASM. There goes another 4 hours 😕

Nu Audio Science · 26 Apr 2008

Nice one Exo its actually faster than the toolbox one now :0 I have avoided using this because here it was always way more cycles than the toolbox version Very very cool

NAS

Panoply · 5 Jun 2008

Firstly,

Damn! Nice job - I've been playing with this quite a bit, and its very, very good. Been getting some really satisfying squelch going on using it.

However (sorry, soooo so sorry I wish there wasn't a however)

I've been having issues with values vanishing from the right click menu. Kinda like a presets issue. I'm quite happy to bug hunt, but I'd be interested to know if others are suffering from this, or if I'm on my own, and whether anyone has any ideas what could be causing this behaviour?

Even when I'm using just one instance of it on a very basic synth this issue crops up... so quite confused!

However, the issues are worse when I'm running two together, and it was definitely getting confused about what was going on... polarity modes going astray with no visual prompt etc etc etc... I think.

Apart from that, this is incredible, and hats off to Exo for all the hard work!

I'm running the latest version from the Wiki on SM 1.1.1b

Would it be as simple as going through and arbitrairily adding explicity named 'physical' controls to all the right click options, instead of using the right click?

exonerate · 6 Jun 2008

Panoply wrote: Firstly,

Damn! Nice job - I've been playing with this quite a bit, and its very, very good. Been getting some really satisfying squelch going on using it.

However (sorry, soooo so sorry I wish there wasn't a however)

I've been having issues with values vanishing from the right click menu. Kinda like a presets issue. I'm quite happy to bug hunt, but I'd be interested to know if others are suffering from this, or if I'm on my own, and whether anyone has any ideas what could be causing this behaviour?

Even when I'm using just one instance of it on a very basic synth this issue crops up... so quite confused!

on but not the others, so you cannot see what is selected.However, the issues are worse when I'm running two together, and it was definitely getting confused about what was going on... polarity modes going astray with no visual prompt etc etc etc... I think.

Apart from that, this is incredible, and hats off to Exo for all the hard work!

I'm running the latest version from the Wiki on SM 1.1.1b

Would it be as simple as going through and arbitrairily adding explicity named 'physical' controls to all the right click options, instead of using the right click?

I think I know what you mean, in the right click menu things are not visually selected? with the tick? the problem there is that only one thing in a menu can be ticked at any one time, so this means that one of the sub menus will have a ticked option while the others will not.

I've made a request about this before.....extra ticks for sub menus

In fact, I think that thread needs a bump. 😀

Panoply · 6 Jun 2008

ah yes 🙂

Thats it exactly 🙂

NP, I can work around it 🙂

exonerate · 10 Jul 2008

Update to version 0.9 😀

Added some nice features I think....

version 0.9 changes....

Fixed the end sustain point not being set correctly after resetting the env. (graphically was ok, though the audio was looping from the release) **edit** hehe looking at the last version I thought I'd fixed that (must have been tripping off my avatar 😀 )

Now hiding the mouse while moving scroll or zoom. Added a "Shape menu" which includes saving,load, copy and paste of shapes, also flip shape vertically.

----------------------------------------------------

I think though the menu is getting seriously crowded now, I like having the menu because it simplifys the GUI, but it's getting rather big, what would be good is the breaklines I've requested before to break things up a bit, or maybe a little button somewhere for the shape menu rather than having it in the right click menu. Maybe I should move the scroll and zoom over slightly to the right and have a little button there? not sure, any ideas are welcome 🙂

I'll post this here for now.....

exonerate · 10 Jul 2008

Update to version 0.9.1....

Just added a little button for the new shape menu for easy access, I've left the shape Menu within the right click menu for now. 🙂

Cheers 😉

Exo

kirkagur · 11 Jul 2008

I have a suggestion.

when the env is in snap mode the time (in sec) will be the snap fraction

i hope you understand

exonerate · 11 Jul 2008

kirkagur wrote: I have a suggestion.

when the env is in snap mode the time (in sec) will be the snap fraction

i hope you understand

Sorry I'm not sure I do, when in snap mode you move a point and it snaps to the fraction, which is shown in seconds.

exonerate · 11 Jul 2008

Update to version 0.9.2...

changes...

Fixed a bug when copy and pasting with sustain loop off, turning sustain on the other env added lots of points to the end of the shape. Removed the menu for "straighten curve", now right clicking on the curve just straightens it.

Cheers 😉

Exo

Thorsten · 13 Jul 2008

exonerate wrote: Removed the menu for "straighten curve", now right clicking on the curve just straightens it.

Actually, it seems right clicking resets the last curve you selected with a left click, but I can see how you could miss that in testing 🙂

But wow, very nice stuff!

exonerate · 13 Jul 2008

Update to version 0.9.3.... 🙂

changes....

Added middle mouse button drag and drop for copy and paste. Straighten curve now straightens the right one! (thanks Thorsen 🙂 ) Removed the shape menu from the right click menu.

------------------------------------------

Inspired by kirkagurs Middle mouse drag I've added drag and drop with the middle mouse for copy and paste, check it out! now copy and paste is super quick and easy 😀

This env is really starting to feel complete now, I think the next things I'll do before going to version one is give it a good tidy up, it's getting a bit messy in there now!

Cheers 😉

Exo

kirkagur · 13 Jul 2008

smacking!!!

very nice idea indeed!!! 😀

I think that if you copy&paste the zoom also it will be more like a copy (for the eyes)

exonerate · 14 Jul 2008

kirkagur wrote: smacking!!!

very nice idea indeed!!! 😀

I think that if you copy&paste the zoom also it will be more like a copy (for the eyes)

Yes I think your right, I wasn't sure about copying the zoom also, but I guess this will be expected behaviour.

So Heres an update....

Version 0.9.4 changes....

Copy and paste now copys the zoom level also.

Cheers 😉

Exo

kirkagur · 19 Jul 2008

exonerate wrote: kirkagur wrote: I have a suggestion.

when the env is in snap mode the time (in sec) will be the snap fraction

i hope you understand

Sorry I'm not sure I do, when in snap mode you move a point and it snaps to the fraction, which is shown in seconds.

o.k

i will try to abstract this..

insted of time in seconds i want to see bars&beats like in cubase or pro tools.

1 | 4 | 3 | 96 bar - beat - 3\16 - Fine

hop this will be more understandable

exonerate · 19 Jul 2008

kirkagur wrote: exonerate wrote: kirkagur wrote: I have a suggestion.

when the env is in snap mode the time (in sec) will be the snap fraction

i hope you understand

Sorry I'm not sure I do, when in snap mode you move a point and it snaps to the fraction, which is shown in seconds.

o.k

i will try to abstract this..

insted of time in seconds i want to see bars&beats like in cubase or pro tools.

1 | 4 | 3 | 96 bar - beat - 3\16 - Fine

hop this will be more understandable

Ah ok I get you now 🙂 ,

I'll see if I can add that as an option, it's a good idea!

Cheers 😉

Exo

Large · 20 Jul 2008

Have you added this yet?

-Clicking on any curve will add a control point there.

-Right click / "Delete" will delete a control point.

This is much more efficient than the current method of adding/removing control points.

Thanks.

exonerate · 22 Jul 2008

Large wrote: Have you added this yet?

-Clicking on any curve will add a control point there.

-Right click / "Delete" will delete a control point.

This is much more efficient than the current method of adding/removing control points.

Thanks.

Yes I'm going to try and add such a thing for version 2, I want to get this to version one first with all the 'extra' features I want, The way the env will behave will be very different than it is now when adding control points like that, at the moment when you move a point all the others move along, a lot of multi envs behave this way zebra 2s do for instance, to have adding points with a click will mean changing the behaviour so moving the point doesn't affect the others.

Some people may prefer the current behaviour so I'm getting this one to a 'stable' version 1 release and then I can work on version two, so people will be able to choose between two different envs with different behaviours.(although same basic feature set).

I can't give a time estimate for version 2 though, It will depend on 'time available' and how hard it turns out to be.

Cheers 😉

Exo

exonerate · 23 Jul 2008

Update to Version 0.9.5 😀

Version 0.9.5 changes....

Added option to show section times in beat fractions or seconds.

Cheers 😉

Exo

Large · 23 Jul 2008

"so people will be able to choose between two different envs with different behaviours.(although same basic feature set)."

They don't need different behaviors; just add a "slide" on/off button. If it's on, all points to the right of the currently selected point will move. If it's off, only the currently selected point will move. 🙂

exonerate · 24 Jul 2008

Like a noob, I didn't test the beat fractions with other tempos, so they where wrong for tempos other than 120. 😊

Version 0.9.5.1 changes....

The beat fractions where not correct for tempos other than 120, fixed.

Cheers 😉

Exo

stw · 4 Sept 2008

I did some small changes to the envelope. I don't want to scamp exos nearly perfect work, but maybe you like it. changes: added dotted and ternary measures. Altered "snap" handling. Now it's always snapping to grid. Grid resolution is choosable and labeled on the grid. Additional bar-line for better overview on not matching grid settings Snap feature is now on the gui and not in the menu.

@ exonerate: I hope i don't rain on your parade. If you'd rather want your work untouched please tell me

philter5 · 4 Sept 2008

🙂 thx for sharing

exonerate · 4 Sept 2008

stw wrote: I did some small changes to the envelope. I don't want to scamp exos nearly perfect work, but maybe you like it. changes: added dotted and ternary measures. Altered "snap" handling. Now it's always snapping to grid. Grid resolution is choosable and labeled on the grid. Additional bar-line for better overview on not matching grid settings Snap feature is now on the gui and not in the menu.

@ exonerate: I hope i don't rain on your parade. If you'd rather want your work untouched please tell me

I like the change 😀 , while adding the snapping I was thinking it needed to be more accessible, It was a bit of a chore going into the right click menu and setting it. So your solution is a good one I think. changing the grid with the snap amount is a nice touch. 🙂

Oh and don't worry about raining on anyones parade here 🙂 , theres no parade in town, it's good for the community when people post improvements to posted modules, I do it all the time. This env AFAIR was started by shifrin and then sambean took up the reigns and turned it into the multi env you now see in the toolbox, and then from this I've turned it into a more flexible syncable modulation envelope, after posting it infuzion has helped with a lot of it. So this is more of a community project with different people bringing different talents and vision to the table. So I can't really lay claim to the env being mine, but I have spent plenty of time and effort to get it to this point, along with others.

BTW I'm currently working on version 2 of this env, it behaves very differently, I much prefer it actually, basically the points are insertable by left clicking a blank area and when you move a point it no longer moves all the others.

I think I'll most likely add your snapping mechanism because it keeps with the ease of use now offered by version 2.

Anyway It would be wrong of me to just tease, so here is a very alpha version.....

I'm going to start a new thread at some point for this new version most likely within the next week or so, it's not quite ready for testing or using just yet, but there it is anyway 🙂

Cheers 😉

Exo

Large · 6 Sept 2008

Exonerate, I'm happy to see version 2 (Sytrus-Style) envelope is coming out 🙂

Just make sure it doesn't hang the CPU like your current versions. Talk with MALC about that. We discovered CPU hanging after notes were played, as if the notes wouldn't shut off. Please make sure this is fixed in Version 2.

Also, can you add an on/off button that ALLOWS all the points to slide when you move a point? That would not be cool if that feature was lost.

Thank you very much 🙂

bootsy · 7 Sept 2008

looks great!

stw · 8 Sept 2008

the journey goes on... 🙂 until exo comes out with V2 i go on playing with the 0.9 incarnation. An optical feature that i missed with this envelope. Not only for viewing pleasure but also for easier distinction between +/- polarities i added the curve fill. Colours can be adjusted for both polarities. Bipolar has no colour add. It's not that easy to implement this, and even this version was much harder to do than i did imagine before. So i think that feature could be optimized. But for now i'm glad it works at all.

Have fun...

exonerate · 8 Sept 2008

Wow Excellent stuff! 🙂

I had an idea to do that a while back, I just turned on solidfill on the graphlines but it didn't work properly, I never really looked into doing correctly but I could tell it wouldn't be such a simple thing to do.

I'll look at it properly and see if it could be optimized when I come to add it to version 2. 😉

Cheers 😉

Exo

stw · 9 Sept 2008

exonerate wrote: Wow Excellent stuff! 🙂

I had an idea to do that a while back, I just turned on solidfill on the graphlines but it didn't work properly...

yeah, that was my first step as well. And the result made me eager... 😀

However, here's a small update with a gradient background approach.

Regards

Large · 9 Sept 2008

Hey Exonerate, did you read my last post about Version 2?

CPU hanging issues?

Optional control point slide on/off button?

Thanks.

exonerate · 9 Sept 2008

Large wrote: Hey Exonerate, did you read my last post about Version 2?

CPU hanging issues?

Optional control point slide on/off button?

Thanks.

So malc knows of the CPU hanging? was it an issue with this env, or an internal SM issue?.

I'm going to see if I can add the optional control point slide, I'm going to have to give that a good think through because I came to the conclusion that we couldn't have that AND inserting points with the left click, what I'm going to do though is get the env working perfectly as it is,( ie fix all bugs ect..) then I can have a good look at adding this feature but I'm not going to attempt it until I've got the env running nice smooth. So the first beta version most probably won't contain this feature.

Cheers 😉

Exo

infuzion · 6 Dec 2008

matti wrote: infuzion wrote: Strangely enough, if you remove the "movaps xmm1,F0; //removing +4 cycles?" line, then it will use about +3 more cycles, which is very strange.

Removing the line did nothing to the CPU usage on PentiumD 915, which is a 2.8GHz dualcore P4E. The mov gets masked by xmm3 with it's dependency from andps and cmpps. The latest ASM code crashes SM ...when I pasted it to an assembler primitive that was connected to an analyser. It uses 48 cycles over here. Original code version was 112 cycles.

Thanks for sharing your test results Matti! That is quite the cycle reduction; cool thing is the same ASM can be used in the other Coded envelopes... I think that paste-crash bug will be fixed soon.

matti · 11 Dec 2008

I heard there's problems with this envelope clicking. I'm not sure if this helps but here's what happens when i purposely steal a note.

matti · 11 Dec 2008

This is what the actual envelope output looks like in the same situation.

matti · 11 Dec 2008

And here is what my non clipping envelope looks like in a similar situation. This was just a quick test, so the sustain levels are a bit different, but the main intrest should be in the "stealing" point.

It's based on an older multi stage envelope code, optimized and faster than the new multi stage(tho there's great potential in the new one to be as fast or faster when slimmed down in asm).

So it seems the envelope should start from where the old one left? Hope this helps.

infuzion · 11 Dec 2008

matti wrote: I heard there's problems with this envelope clicking. I'm not sure if this helps but here's what happens when I purposely steal a note.

Good investigation! 😇 So we can repeat your tests, could you please post your OSM, or let us know what your settings in the MIDI to Poly are, & if you have "Hold" &/or "Gate" on in the MSE please? TIA

stefan · 11 Dec 2008

Isn't this a matter of hard vs soft retrig?

infuzion · 12 Dec 2008

stefan wrote: Isn't this a matter of hard vs soft retrig?

That is why I'm thinking we all need to see the entire OSM, including the patch to test more...

stefan · 12 Dec 2008

infuzion wrote: stefan wrote: Isn't this a matter of hard vs soft retrig?

That is why I'm thinking we all need to see the entire OSM, including the patch to test more...

Yes. I'm also curious about the completely flat lines at two different levels. Is it an AHDSR?

Large · 12 Aug 2009

Uh, never. The stock MSME and ASDR's click and pop also. BIG weakness in SM. That, and a lack of oversampling within oscillator's phase modulation.

Now, I tried an experiment using NO envelopes, and triggering voices really fast, and even THAT causes pops and clicks, and the CPU goes through the roof. Apparently SM can't handle quickly played notes. It may have to do with that "SSE" that is supposedly the best thing since sliced bread.

On a side note, SM, in general, is SLOW. The simple Triangle wave example that comes with FLSM eats 4-6% CPU. A similar setup in Sytrus eats 3-4% CPU.

MALC, you seriously need to look into this. Thanks

infuzion · 13 Aug 2009

Large wrote: Uh, never. The stock MSME and ASDR's click and pop also. BIG weakness in SM. That, and a lack of oversampling within oscillator's phase modulation.... MALC, you seriously need to look into this.

I thought this was fixed in an earlier SM1.1.3beta, with the hop() having an issue in the older versions? You'll have to upload an OSM to perfectly illustrate what you are claiming Large, since you may have the Attack or Decay or Release too sharp, or it just my be the way your OSC is. Also, bug reports need to go into the Bug forum, since Malc never reads this forum anymore.

I am still patiently waiting for a reply for the last working MSE & if help is needed for a new version please. 😇 cheers

Large · 13 Aug 2009

This post pretty much sums it up.

http://forum.image-line.com/viewtopic.php?t=37564

Again, multiple voices in succession seems for be a factor.

infuzion · 13 Aug 2009

Large wrote: This post pretty much sums it up.

http://forum.image-line.com/viewtopic.php?t=37564

Again, multiple voices in succession seems for be a factor.

& again, you are not using the latest Beta, which has the hop() issue fixed, which is affected by re-triggering new voices.

EDIT2: Related: Multi stage modulation envelope MOVED TO PROJECTS clicking because of bugged polystream init/shutdown? oddsons & infuzions adsr 1.7 bug fixed & asm optimized Multi stage modulation envelope MOVED TO PROJECTS I forgot the hop-click fix explanation: [old link]

Large · 13 Aug 2009

Using FLSM version 1.1.3.4 (the latest that comes with FL 8.5 beta 5) and still getting clicks with ADSR. Holding a single note from that patch is about 25%. When hitting just 3 notes quickly the CPU almost doubles to 45%. So apparently 1.1.3.4 is NOT the latest beta?

infuzion · 13 Aug 2009

Large wrote: Using FLSM version 1.1.3.4 (the latest that comes with FL 8.5 beta 5) and still getting clicks with ADSR. Holding a single note from that patch is about 25%. When hitting just 3 notes quickly the CPU almost doubles to 45%. So apparently 1.1.3.4 is NOT the latest beta?

Sorry man, you're way off topic about the MSE; please start a new thread in correct sub-forum. Plus, you'll need to post an OSM, & most ppl here don't have FL.

Large · 14 Aug 2009

I can't post an OSM of that patch, as it contains my synth's components and "I can't give away my secrets". Anyway, I can only assume I have the latest beta and the envelope / retrigger problem still hasn't been fixed.

So, AGAIN, in response to your upper post, there has never been a "working version" of any envelope.

MyCo · 14 Aug 2009

have you tried this one: Selfmade Exponential Envelope

Large · 14 Aug 2009

Unfortunately that one clicks also, even when I turn up the attack to like 0.33. Also has the same CPU issues with fast notes. I LOVE the "curve" feature though.

Now here's an OSM of a simpler patch, still with clicks. I'm using FLSM beta 1.1.3.4

Try hitting adjacent notes back and forth really fast to get clicks.

aliasant · 15 Aug 2009

Large. I cant get any clicks from that last example.

Sounds perfect to me. I noticed your not using an ASIO card? Personally I couldnt test the DS out because I dont get any sound out. ( tried the Example synth with DS out and that worked well)

Large · 15 Aug 2009

Is Creative Labs / Soundblaster considered ASIO? Should I use the "ASIO for All" drivers?

aliasant · 15 Aug 2009

Large wrote: Is Creative Labs / Soundblaster considered ASIO? Should I use the "ASIO for All" drivers?

Well. Honestly I dont really know but Soundblaster isnt ASIO all though you can use that ASIO4all to emulate it. Try that. Should give you better performance then running DS.

If you are serious about audio you should get yourself a proper audiocard.

I have my MOTU Traveler that souds fantastic and has a good built in midi+ uses firewire. I also have a Creative SB XFi Audio PCIe that I use for gaming mostly. Never for any real audio work.

sunsynth · 15 Aug 2009

Hi Large,

look if the kxdriver is compatible with your SB...! Awesome driver - believe me!even with routing, own effects/-slots etc.

Large · 18 Aug 2009

Thanks for the suggestions. I'll note this down for when I get a new system.

Which will be when this MSME is fixed 🙂

AlexT · 18 Aug 2009

Large wrote: Thanks for the suggestions. I'll note this down for when I get a new system.

Which will be when this MSME is fixed 🙂

Forgive me for being a bit ignorant but surely the most likely cause of your clicks is your soundcard/drivers and not SM. In fact, I would put money on it.Especially as people who have good quality soundcards with ASIO drivers don't have the problem. The reason ASIO drivers were invented was to stop this from happening. Using Direct Sound drivers the software accesses the soundcard via Windows own software, in particular the mixing kernel. ASIO bypasses this and therefore greatly reduces the latency time, ie the time between the software initiating the sound and the sound actually being produced by the soundcard. With non ASIO drivers try to make this latency anywhere near usable, ie less than about half a second, and the Windows software is unable to keep up, hence producing clicks instead of sound.

I suggest you go and do some homework and read this, it might help and good luck.

http://en.wikipedia.org/wiki/Audio_Stream_Input/Output

Large · 19 Aug 2009

That could be it. Or maybe not? Why is it that I don't get these clicks with these synths: Sytrus Toxic Biohazard Sawer Vanguard Wasp Simsynth etc...

infuzion · 19 Aug 2009

Large wrote: Why is it that I don't get these clicks with these synths...

Gone though this list? [old link]

AngularMomentum · 19 Aug 2009

I agree that the clicks might originate from a too short set latency. When I set a fairly large SM synth to 2ms on an ASIO driver, with 5 notes I can hear popping when the 5th voice starts...when I set it 1ms later the pops are gone. So it might not have to do at all with the envelopes, but with a voice starting that the audio system can't handle, or something else that is lagging in SM. I do not think you can compare code only VSTi's with SM exported VSTi's when it comes to CPU efficiency, specially when projects get large.

As Walter suggested the KX driver is great for sound blasters. I have used that driver for years on a Audigy 1. I would suggest this over an ASIO4ALL driver.

List of compatible boards in the following link.

http://kb.kxproject.lugosoft.com/list.html

infuzion · 20 Aug 2009

AngularMomentum wrote: I agree that the clicks might originate from a too short set latency.

Could mean SM VSTs just need the extra CPU cycles. Another trick to try, lowering the OS priority of the host: http://www.kvraudio.com/forum/viewtopic.php?p=3724802

MichaelBenjamin · 16 Nov 2009

i want to mod this envelope a bit and asm later on. i can post the result then, but the first thing i wanted to change i am already totally stuck with, so i need some help:

- only the dragged point should be moved, not the ones that come behind it. but i cant find the place where the later values get added up with the new difference. it seems the whole array gets updated with a loop each time a point is moved, so how do i change this?

edit: its ok, i found it. now needs a limit not to get higher than next point.

infuzion · 16 Nov 2009

MichaelBenjamin wrote: only the dragged point should be moved, not the ones that come behind it. but i cant find the place where the later values get added up with the new difference. it seems the whole array gets updated with a loop each time a point is moved, so how do i change this?

edit: its ok, i found it. now needs a limit not to get higher than next point.

Please please make this a configurable option, perhaps best with [CTRL+drag]. I'm thinking the limit-testing will eat alot of CPU, since you will run some compares & math for each trigger, at x times per second. But it should be a handy feature.

stw · 16 Nov 2009

MichaelBenjamin wrote: i want to mod this envelope a bit and asm later on. i can post the result then, but the first thing i wanted to change i am already totally stuck with, so i need some help:

- only the dragged point should be moved, not the ones that come behind it.

Which version are you talking about? The last one exo upped changed behaviour so that it only moves the selected point. I modded that one and added the "locked move" option which enables moving of all proceeding points. So what you need should alreday be there...?

MichaelBenjamin · 16 Nov 2009

V1.9.7.2 from first post, but that one shifts all points after the selected no matter in which mode - is there a newer version?

stw · 16 Nov 2009

MichaelBenjamin wrote: V1.9.7.2 from first post, but that one shifts all points after the selected no matter in which mode - is there a newer version?

maybe some nightly builds... 😀 Try right click menu and toggle between "Free move"/"Locked move".

MichaelBenjamin · 16 Nov 2009

oh yes, its already in. i thought that free move in the menu was the same like the button so have not used it.

Shalako · 17 Nov 2009

This is fantastic, nice work, can I make a suggestion, go on let me.....Consider an enlarge display toggle switch. I don't know how hard that would be to implment but having a larger display seems like it would be handy, think minimise maximise. Anyway just a suggestion, you are welcome to ignore it.

infuzion · 17 Nov 2009

Shalako wrote: This is fantastic, nice work, can I make a suggestion, go on let me.....Consider an enlarge display toggle switch. I don't know how hard that would be to implement but having a larger display seems like it would be handy, think minimise maximise. Anyway just a suggestion, you are welcome to ignore it.

Wayyy alot of work there; I think just a permanent size change was a pain for me.

ortys · 11 Dec 2009

Hello There, i am toying around with the FruityLoops Version of Synthmaker and building a bleep/clonk type of Synth and i integated the Multistage-env (last version) to it.... and it works! (more or less) 2 Problems are showing there ugly heads, the envelope is aaaaafully slow, means, if the time is not 0 it takes ages to do a step, very step transients go about 10 sec.... I think this could be because the envelope dont get any tempo-information from the host, i looked in the module but i didnt see a terminal where related info (bpm or something) could be piped in. i would be very glad If someone could give me a hint.... the Other Problem: Releasing alwas clicks, it doesnt matter how long the release segment is.

many thanks in advance ortys

infuzion · 11 Dec 2009

ortys wrote: ...the envelope is aaaaafully slow, means, if the time is not 0 it takes ages to do a step, very step transients go about 10 sec.... I think this could be because the envelope dont get any tempo-information from the host,... Releasing alwas clicks, it doesnt matter how long the release segment is.

You'll have to post your OSM; you may have things configured wrong etc. Also, how SM handles hop() in Code has changed, & the MSE may not have been updated. I do recall a few posts back on this thread others having click problems, but the demo patches for you synth will help isolate the issue.

ortys · 11 Dec 2009

many thanks for your reply, here my osm, 2 osc fm, the only special thing is that with the DeSyncer-code when the envelope is not 0 it prevents to sync the oscillators on keypress.... with the adsr it works.... many thanks for your help....

infuzion · 11 Dec 2009

ortys wrote: many thanks for your reply, here my osm, 2 osc fm, the only special thing is that with the DeSyncer-code when the envelope is not 0 it prevents to sync the oscillators on keypress.... with the adsr it works.... many thanks for your help....

Thanks. First, your patches have the sustain loop staring at the first point. MSE doesn't seem to like that, so if you start your patches on the 2nd point you're fine there. Turn on the Hold preference to get the release working (on the top of the Pref panel, or drill to {Multi-Envelope, Finish:Gate}. If you have the starting or release point any higher than 0, you are likely to hear clicks.

I tested your MSE just by using the Voices, Sine, Sawtooth, Multiplication, & DS Out module & primitives, hand-edited the MSE points & sustain looping, using on-board audio & cheap computer speakers. I heard no clicks, & seemed to work fine. Any more problems I'm guessing it is your OSM/patches.

ortys · 15 Dec 2009

Hello to check i made a supersimple thingy (just the init-Patch with stock adsr / MSE and a selector to choose... adsr doesn't click, MSE does (try'd many possibility's).... but the strange thing (that could also be the source of the trouble) is that the MSE is soooo slow in Fruityloops... could it be that the MSE needs some parameters (speed/signature/samplerate whatever, i am just guessing) that Fruityloops doesn't provide? is there any Patch that uses the MSE that i could test if works there?

many thanks for your effort all the best styro

MyCo · 15 Dec 2009

in the properties panel of the MSE toggle the "hold" option on. If this is off, the whole voice is getting killed by the internal voice management of SynthMaker when you release the corresponding key. So there won't be any "release state" at all. In the ADSR module from the toolbox, the hold option is on by default.

ortys · 16 Dec 2009

After finding out that in the Properties-Module (of the MSE) FruityLoops somehow links the wrong (upper) TimeSig module connector to the Beats/Measure wireless out and changing it to the lower one (that gives out 16) everthing works like a charm. Before the envelopes where so slow that i had to untick the hold, or the note never ended....

many thanks for everybodies help

styro

ortys · 23 Dec 2009

Hello, unfortunately i am still runnning in to some Problems /Question. When there is more than one MSE in a schematic and i want to save a preset to a file and reload the file the envelopes get garbeld. also i had the problem that sometimes the sound just stopped and there was 100% DC on the audiooutput (oscilloscope in FL). in this case the envelpoe out doesent go to 0, the poly readout show very small fluctuating values. when i changed to the default shape its ok again... also i have the impression that the MSE does not have much resolution in very fast transients, is this a hop(128) problem? is there a way to change this? and at last... i there a possibilty to change the default shape (to have a sharp attack?)

sorry if i am missing something and many thanks for any help

orty

infuzion · 23 Dec 2009

ortys wrote: When there is more than one MSE in a schematic and i want to save a preset to a file and reload the file the envelopes get garbeld. in this case the envelpoe out doesent go to 0, the poly readout show very small fluctuating values. when i changed to the default shape its ok again... also i have the impression that the MSE does not have much resolution in very fast transients, is this a hop(128) problem? is there a way to change this? and at last... i there a possibilty to change the default shape (to have a sharp attack?)

1) The "Preset" name is usually the issue, but it seems you have changed this. Try saving some patches with 2 Wave Draws, & if they are also garbled then the issue is with the SM primitive. If not, might be an issue with FL-SM, since I think my MSE versions had no problems. 2) That can be the last endpoint > 0, your OSM, or a bug in the MSE. 3) The hop can be the issue, though perhaps it needs better programming to catch these issues. Shrinking the hop requires altering all the maths that depend on the hop=128. 4) inside Multi-Env/Envelope Points/Presets/[all 3 modules]/__ array/__ array/__ array/unlabled/unlabled/__ array there is an In labeled reset. The Float Array Sample & Hold connected to that is what you need to feed the "default" you want.

ortys · 29 Dec 2009

Hello Now that the x-max-stress is over i could look a little deeper at the Preset-save to file-issue... this happens also in synthmaker 1.1.5 , meens it doesnt seem to be an FL-Problem when i replace the MSE with Wavedraws save a Preset to a file works flawless.

to post the Preset-Files i had to rename them to *.doc (*.txt seems to be not allowd by the forum...)

looking at the saved file (s2MSE-E.doc) of the simple2MSE.osm i saw after the Multi Env2-X Array Float Array (64) 1 0.000000 | 64 0.200000 Float 0.120000 <--------------Parameter without name Multi Env2-Curve Array Float Array (64) 1 0.000000 | 64 0.200000 Float 0.080000 <--------------Parameter without name

does this mean something? when i make a diff of this file with the saved (imported and garbeld) Preset (s2MSE-Ex.doc) these values are different...

also there in the save file there is the same 2 lines per MSE

Multi Env1-Sustain On Float 1.000000 Multi Env1-Retrigger Float 0.000000 <--------------- Multi Env1-Retrigger Float 0.000000 <--------------- Multi Env1-Sustain start X Float 2.000000

in the resaved Preset it changes to

Multi Env1-Sustain On Float 1.000000 Multi Env1-Retrigger Float 0.000000 <--------------- Multi Env1-Retrigger Float 1.000000 <-------------- Multi Env1-Sustain start X Float 2.000000

many thanks for any hints and your patience.....

styro

infuzion · 29 Dec 2009

Then it looks like there are missing & wrong preset names. You'll have to find & dig into each preset primitive, & check the Preset Name modules.

Large · 1 Mar 2010

Exonerate, have you made any progress on this? There is currently (2010) no decent envelope in SM, and you are our only hope 🙂

Please finish this 🙂

exonerate · 1 Mar 2010

Large wrote: Exonerate, have you made any progress on this? There is currently (2010) no decent envelope in SM, and you are our only hope 🙂

Please finish this 🙂

Thanks for the bump I needed it 🙂

I tried writing the audio code a couple of times to get rid of the biggest issue the env has, the envelope going out of sync when looping the sustain.

I wasn't successful, but in them attempts I was trying my damn hardest to keep the hop, to save CPU. But I'm going to have to do this without the hop this time so CPU use is going to increase, but I have a few optimizing tricks up me sleeves so hopefully it shouldn't increase to much in the long run.

Once I've done that I'll get through all the other remaining issues.

Cheers 😉

Exo

Large · 1 Mar 2010

Thanks for your reply.

I eagerly await the updated, fixed envelope 🙂

Acrobat · 2 Mar 2010

Great news! Strategically I think you are doing the right thing avoiding the hop, so you should see timing problems much clearer. 😀

exonerate · 2 Mar 2010

Ok done a quick update 🙂

changes in 1.9.7.6:

removed the hop Reduced the number of wave reads to two sustain looping is now in sync removed empty spaces at the end of drop down lists Added option to lock or unlock sustain points in point menu Given preset names to variables which didn't have them

This update is mostly about fixing the sustain sync issue, so if someone can test that. 🙂

The inside of this env is an absolute nightmare in places, really hard to understand some of it and make changes. But I'll get there 🙂

Cheers 😉

Exo

exonerate · 4 Mar 2010

This project has been moved to a new thread in projects HERE

Cheers 😉

Exo