Scriptar JS Permutation Generator

ColdFusion CFSCRIPT version:

<cfscript>
permArr = [];
usedChars = [];

function permute(input) {
  var i = 1;
  var ch = '';
  var chars = ListToArray(input);

  for (i = 1; i < ArrayLen(chars) + 1; i++) {

    /* get and remove character at index "i" from char array */    
    ch = chars[i];
    ArrayDeleteAt(chars,i);

    /* add removed character to the end of used characters */   
    ArrayAppend(usedChars,ch);

    /* when there are no more characters left in char array to add, add used chars to list of permutations */
    if (ArrayLen(chars) eq 0){
        ArrayAppend(permArr,ArrayToList(usedChars));
    }

    /* send characters (minus the removed one from above) from char array to be permuted */
    permute(ArrayToList(chars));

    /* add removed character back into char array in original position */
    if (ArrayLen(chars) lt i){
      ArrayAppend(chars,ch);
    }else{
      ArrayInsertAt(chars,i,ch);
    }

    /* remove the last character used off the end of used characters array */
    ArrayDeleteAt(usedChars,ArrayLen(usedChars));  
  }
}

mylist = 'A,B,C';
permute(mylist);

WriteDump(permArr);

</cfscript>

Example Output

4/4 time is the de-facto standard for popular music. Sure 3/4 was popular back in the day (the day being oh some 100+ years ago) but 4/4 remains strong because it allows our minds to effortless glide through a piece of music on our base instincts. Left foot – right foot – left foot – right foot, tick – tock – tick – tock, wax on – wax off, etc. For the most part our western world is guided by the on-off switch and it takes quite a bit of effort to break this vicious cycle so we can explore other musical worlds.

When first delving into odd signatures… a phrase which I despise because it immediately casts the subject matter as less than acceptable… most people struggle with the overall mechanics of the count. It’s important to understand the counts and basic mechanics, but far more important to move beyond that as quickly as possible and develop a sense of ‘feel’ to the meter. Just like with the practice of scales, one must make the practice musical in order to avoid rigidity (and being a bore.) Nobody wants to listen to you sing “One and Two and Three and Four and Five and” over and over.

So before we get into counts we should take five…

Play this piece for anyone you know, including non-musicians, and they will almost immediately recognize it from TV or film… and be able to follow it. Well it’s 5/4 time. If odd signatures are generally considered too complex then how is this possible? Simply put, it’s not projected as a musical theory project, it’s feel and that is what connects with the listener. Now let’s have Brubeck hurt us a little…

When I first listened to Unsquare Dance I got angry because I couldn’t keep up… even with the intro primer handclaps I just got lost. It’s 7/4. I can follow it on feel, but deconstructing it seemed to take a Herculean effort on my part. I needed a way to increase my knowledge of odd time signatures which led me to Bounce Metronome.

Bounce Metronome is an awesome application for hearing and seeing complex time signatures including polyrhythms.

Check out this 5/4 visualization

Now how about a 5/4 over 4/4 polyrhythm?

The visual and sound approach provided by this app is by far the best I have found for quickly learning the feel of time signatures. Within a short timeframe I was able to get over Unsquare Dance and even start developing that sense of anticipation in the alignment of polyrhythms. Go on and check it out… make the time.

For whatever reason (probably my reluctance to part with cash) I decided to look and see if anybody had virtualized MIDI over LAN. I remember Steinberg made some early attempts with this type of thing long ago for net jams and MIDI has skirted around a CAT5 type connection so surely somebody just got smart and decided to ‘packetize’ the messages in a client/server app. Yes.. they did.

rtpMIDI is a really great little utility to have in your studio toolbox.

In all fairness, Mac has had this for awhile but I’m still in PC world. It’s rare I find a PC based MIDI utility that actually works… that’s why I’m posting this. In a matter of minutes I was able to setup my DAW as a server and connect to it from the laptop client. Running Finale on the client machine I was generating live MIDI over the network to my Midex8 hooked up to a Yamaha P-60. The best part? The laptop was on a wireless connection so I can freely pop into the next room to sit on the couch and still listen to performances or even edit without having to switch back to localized SoftSynth sounds or go back in the studio.

On the project front I’ve started plans for two new boards which will augment my guitar rig. I don’t want to get into the details of one because it is just a little too awesome and ear-popping to give away in words alone… so more to come there. The second of these new boards is pretty simple. I’ve been playing around with Mobius and live looping with Albeton Lite. It’s quite a different animal than studio recording but I’m having some fun so far. I’ll be building a special controller for Mobius based on what feels right once I get more familiar with the system.

Now on to Conjure… my other pet project. I’ve briefly talked about this with a few people on Facebook but the Conjure project is a music composition system I’ve been working on since about Feb of 2010 but is finally getting close to a useful beta. It’s a rather complicated long-term development project but essentially Conjure is a preference engine for producing MusicXML scores. The idea is that given enough information about music theory with a combination of personal music preferences the system can create root song structures. The intent here isn’t to create complete scores as much as serving as a virtual muse. Going through the system has given me a great opportunity to relearn alot of music theory knowledge that’s been lost over the years. At this point Conjure has awareness of about 85 scales and some 60 odd chord structures and how they relate to all the standard western music keys. There is a title generation system, ensemble assembly function, GM MIDI database, drumkit constructor… and hordes of other tables cobbled together for one reason or another. Development has primary settled in with ColdFusion as the web services provider and a mix of XSL/XML/DHTML and possibly Flex for an alternative interface. The past two weeks have been consumed with providing Conjure information about song structures while I’ve just recently started refining the music key/scale selection criteria to introduce harmony and chord progressions (which will then be coupled with the structures.)

Here is an early twiddling from Conjure months ago when it was locked down to a single key but allowed to roam freely among orchestra instruments from IK Multimedia’s Miroslav Philharmonik VST plug in.

One aspect of my pedalboard are these two oddly placed switches kind of lower and between what would normally be bypass switches for the pedals.

Preset Browser Switches

My intention was to make these preset browsers but never tried to do so until this writeup. I have spent a lot of time in the MIDI Control Preset window where I never saw the option and obviously never really looked at the Global options. Getting to the Global options is simply a matter of clicking the MIDI button in AT3.5 then selecting the Global button in the top left corner.

Global MIDI Control Window

I’m glad to see that AT3.5 has two specific pre-configured entires for preset changes. If you recall from my previous post “Quick Guide To Using External MIDI Controllers with AmpliTube 3” all I need to do is highlight the Preset Next entry, click Learn then tap the next button on my controller. Repeat for the previous button and we’re ready to change presets at the simple tap of a foot.

Other Fixed Options in Global MIDI Control

So next my eye is drawn to the Volume, Wah, and Wharmonator entries. Probably because I want to try and exploit them in unintended ways (it’s a personal tic). I can’t help but wonder what happens if I use two volume pedals in series? Ok sorta dumb I know, but what if I run a parallel config like signal chain 2 and have a pedal on each, what happens then? If I use two different Wahs, does it alter both? Let’s find out what’s going on here.

So after training the volume in signal chain one, anytime I drop a volume pedal, and regardless of which slot I move it to, the volume control just works. Pretty nifty. A huge time saver for a commonly used pedal. Dropping a second volume in-line defaults the control to the first volume in the series. That makes sense. Setting up a parallel config and placing a volume in stomp A and stomp B results in only the pedal in Stomp A responding to the controller. This makes a little less sense to me but I get it. Ok Wah time. I set the Wah row to learn in Global MIDI control and tap on a pedal. Good to go. Returning to the stomps I run the same tests as the volume and get the same results. I’m not terribly surprised to get the same results.

This all may sound like a bunch of overkill but I was just curious to learn the primary intent of the global ‘default’ controllers and now know it’s just for your first matching pedal in the chain (and behind the scenes AT3.5 scans Stomps A through B in slot order order making no distinctions between serial or parallel configs.) Obviously any deviations from this single pedal approach are best handled by MIDI Preset control setups as covered in the previous post but part of me wouldn’t mind a ‘cascading’ approach to control the first ‘non-bypassed’ Wah in the chain should I want to use two flavors in the same preset. Just a thought.

So if you’ve never seen AmpliTube, crawl out from under your rock and take a peek.

Amplitube 3

Pop into the device setup (for standalone instances) and setup your interface and MIDI device parameters.

AmpliTube Settings Screen

Now flip over to the STOMP A section of the signal chain and let’s drop a Wah 46 on the floor.

Wah 46 (my fav)

IK Multimedia couldn’t have made the MIDI setup for this any easier or more straightforward. Simply right mouse click the aspect of the effect you wish to alter to reveal the Assign Midi -> Learn function.

The Learn Function

AmpliTube will now enter the Learn mode waiting for you to engage your physical pedal that you want to use to control the effect.

MIDI Learn Waiting For Your Input

Now for the hard part… drag your lazy foot off the floor and push the pedal you want to use. The interface will dismiss the Learn message and you should be controlling the pedal. Good times!

Now let’s try a pushbutton because there are a few ways those little gizmos work within the interface.

Right mouse click the mode (off, on , auto) above the wah. You might notice there is no ‘bypass’ option. This is one thing I don’t completely understand about the way IK implemented the MIDI Learn functionality. I personally would like to also see the bypass option available for any pedal that has a multi-mode like this, but that’s just a personal tic. Not a huge deal because IK has given us more than one way to assign control. In the lower right area of AmpliTube click the MIDI button.

MIDI Button

The MIDI settings screen will appear. In this area you will already see the Wah MIDI control we setup. Now we’re going to manually configure a controller.

MIDI Window

Click the Add button at the bottom of the window and a menu will appear. Navigate into the Stomp A signal chain -> Wah 46 then select Bypass. A new row will be added to the parameters list.

Adding A Parameter To Control

Now with the new row selected, click the Learn button and AmpliTube will enter the MIDI Learn mode.

Waiting For Button Input

Press your ‘real world’ button and AmpliTube should change the CC# to match the message sent by your controller button.

There may be one more step depending on your physical pedal construction. If your pedal snaps/latches like an on/off switch then the Latch Mode off would be preferred (and is the default) however if your switch is momentary then pressing your button would enable the pedal while letting the switch go would disable it. To alter this behavior simply click the word OFF in the parameter row to change it to ON and all is well.

Altering Latch Mode

So there you have it… two ways to get immediate satisfaction in AmpliTube 3 with your external MIDI controller.

Lather, Rinse, Repeat.

Pedalboard Controls

Analog Functionality

• 1/4 inch input
• 1/4 inch TRS insert
• 1/4 inch output
• SPDT switch for selecting to enable the insert or not
• SPDT switch for defeating the input
• SPST momentary switch for temporarily stopping input until switch is released
• Audio taper potentiometer for volume pedal

Issues – I don’t know alot about passive analog circuitry but now know there is a huge difference in the selection of the potentiometer rating for the volume control. It’s currently a 10K pot when it should be a 250K instead… which is closer to what’s actually inside a guitar to begin with. An incorrect selection here degrades the power of the signal.

Potential Fix – Just replace the pot but I also like the degraded effect and will find a way to mimic the circuit path even with the new pot in place.

MIDI Functionality

• 2 SPST Momentary switches intended for patch changes
• 6 SPST Momentary switches intended for bypass/defeat of the MIDI pedals
• 6 10K linear potentiometers for the MIDI pedals
• 1 Pitch bend/MOD wheel from a discarded Roland Keyboard controller
• 1 SPST Momentary switch for MIDI Panic/Reset

Issues - The only issue is the pitch bend wheel. It doesn’t allow for a full throw (not a physical limitation) and as such won’t transmit the full 0-127 MIDI data range. I can only assume the wheel had accessory circuitry to translate this properly.

Potential Fix - At some point I’m going to have to pull apart the wheel assembly and see if I can replace the potentiometer to the proper spec.

Other

• 5 Pin MIDI In/Out
• 7-12V / 100mA DC Input
• SPDT switch for power enable/defeat
• LED for MIDI activity

Issues - None

Completed Guitar Pedalboard

Keep in mind that I never really intended to post a How-To/DIY on this project and because of that I’m missing various details on how some things were done. At a certain point I just started photographing everything as I progressed. Overall there are plenty of photos and descriptive info to foster ideas if this is something you want to take on. My only advice is to take your dear sweet time. This was just about the most challenging project I’ve taken on in the past few years and is evidenced by the 6-7 months of development it took to get to where I am today. That said… it passed the all important “smoke test” the first time I plugged it in, so taking your time on this stuff is a good thing.

The purpose of the pedalboard is to send MIDI messages to my Presonus Studio One digital audio workstation running AmpliTube 3 from IK Multimedia. AmpliTube can receive MIDI commands to turn on and off guitar pedals and alter parameters like volume, gain, wah-wah pedal filters, etc.  There are plenty of midi controller pedals out there (IK even makes a line) but with what I wanted I knew I would end up in a wiring mess and alot deeper in the hole cash-wise for this adventure. Since I only play in my studio I wasn’t overly concerned with building the pedalboard like a tank, I just focused on good design and general ruggedness.

My first stop was trying to figure out how to even tackle the electronics aspect of this. I asked a horde of music hardware vendors if anyone made some kind of black box you just plug a bunch of expression pedals into and then it all runs to a simple MIDI output. To my knowledge, nobody (commercially) makes such a simple little beasty. Pressing on I got into the idea, and alluring underworld,  of DIY MIDI… USB controllers using joysticks, programming your own joystick to MIDI drivers, etc. All way too much for something I just wanted to have fun with.

After alot of searching I finally discovered Doepfer Electronics in Germany, producers of MIDI keyboards/controllers galore and one great lineup of OEM/DIY controllers. They have plenty of them to choose from but ultimately I selected the Pocket Electronic as it offered up to 16 buttons or Potentiometers (sliders/pedals) running a MIDI in/out pair (which could be daisy chained)… just enough for the design I had in mind. I got the board from Doepfer’s US Distributor, Analogue Haven, based in California.

Pocket Electronic

This is where I started design-wise in November of 2010:

Guitar Pedalboard Design

The original design (sorry, not alot of light near the whiteboard) was to include 1 analog pedal with an insert, 5 midi pedals, and an array of switches to send signals, reverse pedal polarity (up becomes down), etc. The pedals were shaped like feet and all of the wiring was to run out of the back of the unit.

Early Prototype

This early prototype of a pedal was tossed together by myself and my good friend Jeff. I traced my Cons and we used whatever was in the garage to toss together the initial concept. If I remember correctly the first pedal was the rack and pinion from a telescope, curtain rod, potentiometer, a bracket from those western-style swinging doors, and some screws/plastic tubes and brackets. Crazy… but it worked long enough to give me hope, then it broke. Good enough.

The first thing I struggled with were the mechanics of the pedal itself.. specifically the issue was driving the potentiometer via the pedal. I went through a few designs then ultimately opened up my old Roland volume pedal (uses a sliding mechanism) and then the Crybaby Wah pedal (uses a rack and gear assembly). Preferring the Crybaby engineering because it seemed a little simpler to engineer, I was set. Finding a source for the Crybaby parts took some time but I was very happy with the service from NewOldSounds.com. I ordered a set of racks and gears then played the waiting on shipping game.

Pedal Design

The pedals were the first thing I built. Ignore the nails and gizmos all around the above image… we’ll get to that later. The pedals are simply just 3 inch wide oak from Lowe’s. I cut two three foot sections into foot long pieces then routed off the sharp edges with a 45% bit.

Near the top of the pedal you can see the pedal racks. I cut then filed pieces of aluminum channel (again Lowe’s), added a hole for the rack pin, then screwed them to the board and mounted the rack via the pin insert.

The pivot point of the pedal is a really crazy design but this is what happens when you start with one thing, get too far in, and have to make it work. The base of the assembly is aluminum channel which is screwed to the board with countersunk nuts on the top of the pedal. Within the channel (at the ends) you can see nylon grommets which have been cut to fit inside the channel. Through the grommets is a length of 1/4 aluminum rod. The rod has a setscrew which keeps it from pivoting in the grommets. Working our way out you will find a 1/4 washer, 1/4 clear tubing to help pressure fit the pedals together, another 1/4 washer, then finally a 1 inch nylon grommet. I can probably imagine nine better ways to do this now … live and learn.

Mounting Layout

Back to the nails. The blocks are oak which has had holes (the same size and the nylon bushings) drilled through them at a 90 degree angle then four mounting holes at the corners of the blocks. They were all placed on a curtain rod then run through the table saw to give them the angled edges. The pedals were then placed on the pedalboard at even intervals and the blocks spaced between the pedals. The nails were simply tapped in to provide marks ensuring a proper lineup for each block when I drilled the mounting holes.

Mounting Holes Revealed

Holes were then drilled in the board then working left to right, each block was temporarily mounted followed by a pedal until all blocks were in place. No photos exist of this, but obviously I put some non-skid tape on the pedal faces. This is just the standard black no-skid stuff you can get at Lowe’s.

Pedals Temporarily Mounted

Finally, the end blocks were capped with an oak dowel and set screw.

End Block

Now to backtrack for a moment. You may remember that oddball pedal design and mention of some clear tubing. The reason for the tubing is to allow for adjustments in the placement of the pedals so they glide effortlessly between the blocks without touching. There was alot of troubleshooting to get that to work. Basically if the left side of a pedal was rubbing the block, I needed to pull everything and put a longer spacer. Then that may change the pedal next to it, and so on. So this was a tedious chore to finalize and one day I’ll change the design, but it works now and that’s enough for me.

After the pedals were working well I marked the location of each rack gear, labeled each pedal and block, then pulled the whole thing apart (again… this design will never be repeated!)

Marked Locations for Gear Area

Ok so on to the gear and potentiometer. The Crybaby Wah pedal has the rack gear as shown earlier which marries up at a 90 degree angle to a gear on a potentiometer. I needed to find a way to mount the potentiometer, keep it from rotating, and also take a little abuse from the constant up-down action of the gear. So here is what I came up with:

Potentiometer Assembly

The frame is 90 degree aluminum stock with a mounting hole and (to the right) a ‘stop hole’ for the little tab that sticks out on potentiometers. This holds the pot in place without rotation. The potentiometers were 2 inches long and here is where I got fancy. It’s hard to see in these photos but there are two problems with standard potentiometers and the Dunlop replacement parts. First, the gear has a very shallow D shape that is not common and second, there is a snap ring on the end. I don’t  know how I decided to do it this way, but before assembly I put the potentiometer in my drill press, just like it was a bit then threw the switch. Using a hacksaw I was able to cut a thin notch for the snap ring. I then raised the blade about 1/8th of an inch and cut a little deeper to score a cutoff for the post.

Once the post was cut off I then filed the rough end and while in a vise, carefully filed just enough of the rounded post off to form the shallow D to mate with the gear. Then I did that whole process over and over for the rest of the pots.

Nothing Quite Like Garage Machining

It was time to mount the potentiometer assembly. Using my marks from before I measured the space needed for the gear and 90 degree angles then knocked this out with a jigsaw:

Potentiometer Mounting Hole

The left side is for the pot body, the slots are for the 90 angle facing and the open area is for the gear. Using the 90 degree stock and this ‘slot’ allowed for a good bit of contact surface area and added strength.

The Potentiometer Assembly Just Slides Right In

Underside Shot Of The Rack Mated To The Gear

All Pots In Place

After cutting then staining the board and pedals, the assemblies were screwed into place followed by remounting the pedals. Starting to look like something, isn’t it?

Assembled Pedals

In order to hold the rack against the gear, and to keep the pedal positioned when your foot is taken off, Dunlop uses a greased wire grommet pushing up against the back of the rack. Ok… good enough for me.

Wire Grommet Mounted to Stock Aluminum

Only three major tasks left in the construction phase.

Angled Housing Box Built Around Board

I found an old modwheel while cleaning up… couldn’t resist. The handle is off an old sink. Why not? Looks great!

Mod Wheel

Footswitches Added

I added a series of footswitches for various purposes that I’ll outline later. Some momentary, some on/off. I got all the switches from AllElectronics.com. For whatever reason these footswitches vary wildly on price and All Electronics seemed to have the best deal and their site gave me the most confidence. I also got the 5 PIN din plugs and 1/4 inch jacks through them.

This is where things in this post will speed up. Wiring this beast took about 2 days and if you want to learn about that, opt for reading the manual with the board. Instead of soldering  to the ribbon cables that connected to the board I opted to use a series of terminal strips. My thinking was that I’ve never done this before and I didn’t want to have to undo alot of stuff if I made mistakes. This turned out to be a wise choice as I made about a half dozen mistakes that were easily corrected by unscrewing a few terminals and switching wires.

Wireapalooza In Effect

Some of the things to point out in my wiring:

Audio Wiring

The first pedal is an audio pedal and serves as a passive volume control. There are three audio jacks for in, out and insert in the chain. there is an a/b switch for the insert along with a master kill switch and a momentary kill switch for stutter effects or just temporarily muting the line. The exposed copper wire feeds through all the pot mounts and is grounded to the MIDI board as well. I’m currently using a pretty mismatched pot for the audio. It is degrading the signal but in a way I like. It almost gives my guitar a semi-weak vintage pickup sound. I trying to figure out a way to allow that with an upgrade potentiometer before I change it. If I need to hot-rod the guitar I currently just run it right to the audio input of the DAW.

Midi And Power Hookups

Another early design decision was to put all the inputs on top of the pedalboard. I decided to also run two cropped MIDI cables to 5 Pin DIN connectors instead of mounting the board in such a way to allow direct plug in of the cables. I did this for two reasons. First, I didn’t want to break the board with general wear and tear plugging in cables and second, since multiple OEM units from the vendor can be daisy chained, this design allowed room for a second board and easy connectivity. In this show you may also be able to see the MIDI panic button near the DIN plugs. There is also a power switch just to the opt out of view.

The Final Board

And there you have it. One 6 month long, but worth it, ordeal. I’ll be posting some follow up videos on how the board interacts with AmpliTube 3 along with an overview of the various board functions.

Go forth and create!

“Duo” – 48×48
Congrats to Jim Nichols who won Duo in the “name my studio” contest on Facebook.

“Solo” – 48×48

“One” – 27×38
Congrats to Trey Lominack who won One in the follow-up “Name in the hat” giveaway
on Facebook.