adsr vco as would apply to Leslie

stefan · 7 Jan 2006

Try using the envelope follower from the toolbox! Just hook it up on your "speed" thingy.

infuzion · 7 Jan 2006

I think there already is an organ & lesle speaker emulation in skOre's mega pack

oddson · 7 Jan 2006

infuzion wrote: I think there already is an organ & lesle speaker emulation in skOre's mega pack

It's not as ambitious as Doug's plan.

stefan wrote: Try using the envelope follower from the toolbox! Just hook it up on your "speed" thingy.

I'm not sure why an envelope follower wouuld be needed. 😕

stefan · 7 Jan 2006

oddson wrote: infuzion wrote: I think there already is an organ & lesle speaker emulation in skOre's mega pack

It's not as ambitious as Doug's plan.

stefan wrote: Try using the envelope follower from the toolbox! Just hook it up on your "speed" thingy.

I'm not sure why an envelope follower wouuld be needed. 😕

The envelope follower has attack / release controls. Which in turn gives you the ability to control how fast the leslie motor can speed up, and how fast it can slow down. You put the envelope follower on the speed signal, set the envelope follower to the desired "motor behaviour" and then just set the "goal" speed for the engine on the envelope followers input.

Ofcourse, you can do it in other ways, but they'd be similar 🙂

oddson · 7 Jan 2006

How about this? [url=http://members.shaw.ca/oddson/leslieSpinUp.osm]

OSM[/url] It ramps up over the number of 'Spin-Up Ticks' and ramps back down the same way. I've attached it to a basic OSC/VCA to demonstrate.

It's not very elegent but I think it might be a start.

oddson · 7 Jan 2006

Rethinking this... it won't give you a gradual speed change if you have two set speeds and you change between them. Stefan's method would handle that better but I haven't figured out how to do it yet.

On another issue, shouldn't there be some sort of doppler effect? Unless the speaker's diaphram is in the centre of rotation there should be some pitch change. Sambean's fractional delay line will do the trick to mimic this effect (integer delays don't have the resolution because at the sub-sonic frequency of the rotation you're only dealing with a very small number of samples of delay.

[edit] -- here's an amended schematic with dopler shifting OSM. It also stops with the speaker facing the right way this time.[/edit]

Hope I'm not too far out to lunch on this stuff... 😕

Does anybody have a physical description of an actual Leslie (how is the speaker mounted. How fast are the rotation speeds etc.)?

doug4350 · 7 Jan 2006

Thanks guys for the info...

As part of the learning curve, I want to avoid the envelope follower so I can get a grip on the underneath components. The scale and timeinfo components haven't turned any graymatter lights on yet.

I like oddson's example. Short & sweet & not too busy with unused links (the end result that is). I think I can take that with some glue logic and get the comb's of on/off-slow/fast to set the up/down and min/max links of the counter.

That's kind of out there with the doppler effect. I'll keep that in the back of my mind though. I'm already getting pop/bump creep as I add complexity.

Anyway, got housework today, so I'll have to come back to this later... Aloha, Doug

oddson · 8 Jan 2006

doug4350 wrote: ...That's kind of out there with the doppler effect.

see wikipedia article

wikipedia wrote: The Leslie speaker consists of two driver units - a treble unit with horns, and a bass unit. The key feature is that the horns of the treble unit (actually only one working horn, but a dummy horn is used to counter-balance it) and a sound baffle for the bass unit are rapidly rotated using electric motors to create 'Doppler effect based' vibrato, tremolo and chorus effects. It can be switched between its two speeds, and the transition between the two produces the most characteristic effects.

I don't think the dopler shift is minor element of the sound. But if you're building the organ sound from scratch you can mimic the pitch shift prior to the oscilators instead of replicating it with a delay.

The next problem is simulating the amp and speaker's effect on the timbre of the sound. I once asked the community if anyone had emulated a tube amp sound. Closest I got was some basic staturation effects.

Hey sk - did you ever get your B3 going? [See threads on OSC code and drawbar controls - ingore some of my stupid comments in the former 😊 ]

Edit - BTW - there was a mistake link off the '1-x' module doubling up the signal to the tremolo part of the effect. It's fixed in the current posted version.

doug4350 · 8 Jan 2006

Thanks for the ref's oddson. I actually did use a bunch (9 that is) of sine osc's multiplied based on the footage, rather than a code component for the drawbars. They aren't the traditional 8-step tone wheel style, but rather respond to the midi cc range 0 - 127. I know all the stuff I'm doing can be refined down the road, just gotta learn to walk b-4 running. I'm trying to think like a programmer when connecting the dots, but some stuff is darker than the point of singularity when cpu efficiency is concerned. I am concerned about efficiency for obvious reasons.

I do have a vibrato effect upstream b-4 the osc's with a rate and depth control. I can probably sync that in with the Leslie speed somewhere. I also have a tremelo effect in between the osc's and the leslie. The tremolo is not by definition -- that is synonymous with vibrato. It's more of a lfo mono a.m. ripple. Unfortunately, I'm not in posession of a classic B3 or Leslie unit for reference, so am going off of past experiences I remember from repairing many church/home organs (Hammond, Wurlitzer, Baldwin, Viscount, etc.) and sort of adapting concepts from those designs. I also may be playing mono (for smaller venues), so flexibility, not replicability wins out here. The goal is to be convincable for the average live band enthusiast (not to mention the rest of the band.) It's all part of the creative process.

Looks like I've come to the right place. Seems everyone has their own way of skinning a cat 🙂 .

Aloha, Doug

AWright418 · 21 Jan 2006

Hi All,

I'm new to synthmaker. I love it. I'm very interested in Leslie simulation. Here is some interesting links that might be helpful.

Leslie Rotation Speeds:

http://www.dairiki.org/HammondWiki/Lesl ... ?version=2

Some Doppler related math and Leslie info:

http://ccrma.stanford.edu/~jos/pasp/Dop ... ation.html

http://ccrma.stanford.edu/~jos/pasp/Leslie.html

doug4350 · 21 Jan 2006

Thanks Awright418,

I'm still (time permitting) plugging away at it here and there. I got the wind up/down working on the slow/fast speeds (many thanks [and credit] to oddson for his feedback) and was fiddling with it this a.m. to get the slow-fast wind up/down working b-4 I had to go in to work 🙁 .

Looking forward to checking those links out when I get home tonite. I've pretty much been doing the leslie params by ear. I'm curious how close it'll be to the real thing.

Part of my repertoire consists of blues with some really sweet hammond accompaniment/solos that just melts the soul when the sound is right 😍 😍 😍

Aloha, Doug

oddson · 21 Jan 2006

Welcome to the forum AWright... Thanks for the link on Leslie rotation speed... I hadn't been able to find anything on that...

On the JOS links (Julius Orion Smith)... that's my favourite source of DSP knowledge right now. A lot of it is over my head but I still go there first when I want the full story on how something works (...that or I post here and Sam or Stefan sort me out 😉 )

oddson · 21 Jan 2006

From this http://www.dairiki.org/HammondWiki/LeslieRotationSpeed

When developing the PRO-3, ?JohnFisher measured his Leslie147. He found that the top rotor had a rotation speed of 400 RPM on tremolo and 48 RPM on chorale. This was with the belt in the middle pulley position and with a normal belt tension. The lower drum rotated at about 342 RPM on tremolo and 40 RPM on chorale.

SalAzz reports the upper rotor speed (middle pulley) to be 409 rpm, and lower rotor speed to be 396 rpm. Chorale speed is 48-49 rpm, for both upper and lower rotors.

That would put the tremolo at around 5-7 hz and chorale around 0.7hz. I didn't figure on the upper and lower rotors being at different rates... that could be tricky to emulate. You'd have to emulate the cross-over filter and then two different rotor emulation modules at the different rates... but given how close they are I'm not sure it's worth it or whether you'd really notice any difference. Then there's whether to do the doppler at all and if so whether to fake it by adding vibrato pre-OSC on using a delay.

Then there's the fact that the woofer doesn't really rotate but a horn below it does...

http://www.dairiki.org/HammondWiki/Leslie The primary effect of the bass rotor is to impart AM to the signal. There is very little phase shift of frequencies below 200 Hz due to their wavelength, though some phase shift may occur up around the crossover point of 800 Hz. The result is a low-frequency pulsation or throb that is very effective when used at the slow or chorale speed.

So it all get's a little complicated as to what parts of the physical reality of a Leslie you're going to emulate and how.

doug4350 · 22 Jan 2006

So it all get's a little complicated as to what parts of the physical reality of a Leslie you're going to emulate and how.

That's what makes it fun 🙂 .

I don't know if I have enough horsepower for all that math. Version 1 (or 0.9.5 😈 ) will cover the basics. There's a vibrato and tremolo module for mono. The leslie is intended for stereo or > 2 (multiples of 2 speakers -- maybe ver 2.0). I haven't split the horn and woofer rpm's. It's just a stereo stream at the moment. The effect emulates looking topdown, the speaker rotating clockwise. There's 2 lfo sine waves (one 90 deg. out of phase) to give the effect of the leslie rotation from right to left and then the volume fades as the speaker cruises around the back. So, I'm considering in a studio or production environment, with 4/8 speakers, you could "catch" the rotation around the back and route it appropriately. Ideally, you'd mount the 4/8 speakers on a pole, but I suppose having them in the room corners & against the edges would give an interesting effect too.

I've also added a leslie lfo sync out to the module to feed the vibrato and fudge the doppler a bit.

If there's sufficient cpu head room at that point, I'll tinker with the realities of the physical unit.

Hopefully (stressed & no promises), I can get it done this week. I'll probably be working late Mon & Tue, so not much spare time then. The rest of the week should be home free.

Aloha, Doug

doug4350 · 5 Feb 2006

Hi gang,

Sheesh am I hard core or what?!?! Workin' SM on Superbowl Sunday! I used to live in Seattle, so I'm a Seahawk fan, but I'm also a Steelers fan too. What a great excuse to drink beer and eat loooottttts of nachos 😀

Ok, question/problem. I've gotten pretty much of what I want of this puppy working (less some things Malc is checking up on for me .. Thanks my friend). While that's happening, I started fooling with the front panel/gui stuff and tripped out of the starting gate 😊 . I'm having a bit of trouble getting child module gui's to propogate to the topmost parent. I figured out the MGUI component is the key (in the children), but all the buttons/knobs/etc. of the heirarchy are stacked on top of each other in the parent. I tried poking around in the example.osm, but I'm about as clear as mud in the fog 🙂 . Can one of you gui voodoo masters point me to the light 🙂.

Aloha, Doug

stefan · 6 Feb 2006

If you select your "front panel module" and click on that "lock" icon on the bottom right (the right-most icon in the row) you enable the GUI editing. You can now move around different GUI objects on the front panel, resize them etc. Give it a try and let us know how it goes 🙂

infuzion · 6 Feb 2006

stefan wrote: ...click on that "lock" icon on the bottom right...

I was about to reply w/a pic before, but my inet went down...

stefan · 6 Feb 2006

Good addition!

doug4350 · 6 Feb 2006

Ok, I'll give it another shot when I get home tonite.

Thanks for the feedback... Aloha, Doug

doug4350 · 7 Feb 2006

Ok, I tried the unlock trick. Cool... things can be moved around now.

New problem... It seems that all the gui modules across all modules appear in the parent module. Take the "Vibrato" module for instance. It has two on/off buttons and a knob. Adding a wireless in & out to the mgui inside the vibrato module carries all the other gui controls from the drawbar, tremelo and leslie modules (these modules are at the same level in the heirarchy as the vibrato). I tried naming the wireless connectors and mgui in the "Vibrato" module to "Vibrato" (remembering something about names of wireless connectors in a recent thread), but then everything dissapeared. I proceded down into one of the vibrato's button components and named some things "Vibrato" there thinking what's the worst that could happen?.?. Still nothing.

I've been looking at the example osm and what I'm missing is not clicking. Do I need to add the mgui & wireless do-dads b-4 adding gui'ed submodules, or just pay a visit to the SM local balloon dealer 🙂 ???

Aloha, Doug

infuzion · 7 Feb 2006

Try looking at the example.osm. It uses sub-GUIs to arrange the controls w/in the module (need the G button on the module on), then when you just move around the groups in the topmost.

doug4350 · 7 Feb 2006

Ok, got it!

Thanks Infuzion. I was missing the outer vsti shell component.

Aloha, Doug

doug4350 · 11 Feb 2006

Aloha folks,

Hope you're enjoying your weekend. Mine is looking better.. la de da de daaa.

Ok, down to biz. This puppy is sounding pretty sweet now imo and I'm about ready to unleash it. It's not earth shattering, so don't get too excited & step on your tongues 🙂 . I'm patiently weathering the 'j' release & trying to tweak some subtleties. I'm having some confusion with the knob prop's. I'd like to have the rate & depth knobs all self contained. By that I mean the min/max/step values all driven by the prop's w/out external maths (in the interest of the cpu). However, because the values are minute, the knobs responsiveness is minute. For instance, if I want to have a min/max range from 0 - .01 with a step rate of .005, I can hear the audio effect, but I need an electron microscope to see the knob move. I'd like it to sweep the full gamut based on the min/max prop val's. On the other hand, the "Amp" knob performs perfectly 😕 .?.? The gui respects the min/max/step values acceptably.

I've searched the threads and my eyes are starting to glaze... Some help and visene please 🙂

Aloha, Doug

oddson · 11 Feb 2006

...if I want to have a min/max range from 0 - .01 with a step rate of .005, I can hear the audio effect, but I need an electron microscope to see the knob move.

I'm not getting that result... when I set the properties of a knob to those suggested I get the behaviour I'd expect including the visual response of the knob. Am I misunderstanding what you're saying?

doug4350 · 11 Feb 2006

Wow Oddson, you're on it this a.m. (in my timezone) 🙂

Hmmm, interesting... I think we're on the same page. Maybe I have some artifacts tripping it up. Lemme go do some fooling around with a virgin knob ... Ok, get your mind out of the gutter guys 🙂 ... and see what I come up with.

Aloha, Doug

doug4350 · 11 Feb 2006

More interesting...

Here's the virgin knob... http://home.hawaii.rr.com/doug4350/CorruptVirginKnob.osm The responsiveness of the gui is very sluggish compared to when I 1st dragged it on the slate. I'm wondering if my screen resolution or other graphics settings may be affecting this (a long shot).

Aloha, Doug

oddson · 11 Feb 2006

You've got the sensitivity turned way 'up' (very low number). That impacts how many pixels you have to move to get results....

see this threadfor details

doug4350 · 12 Feb 2006

Thanx Oddson,

I understand that now. It seems I was trying to relate it to everything else but the graphics (along with the step).

I owe you a beer if we ever run into each other 🙂

Aloha, Doug