Skip Navigation.
Section 0

ChromaTalk Archives: April 2001

2 Chromas to be sold

Papa Templeton · Sun, 1 Apr 2001 18:49:08 -0700

I've had the very good fortune of recently acquiring 2 Chromas: 1 Rhodes in very good shape for its age, which was used in the Rhodes factory for testing computer interfaces (retofitted with legs from a Rhodes 4-voice at the Rhodes plant), and one nonworking machine which is rough but marked as ARPCHROMA...it's complete minus one dual channel board (a diamond in the rough?). This package also included 2 original flight cases (which need foam), expression pedal, volume/effect pedal, computer interface, Chroma to MIDI, service manual, and some extranneous Chroma parts (yes, I was the one who got these parts from ebay awhile back....see last month's notices). I intend to sell this complete package for the best offer I receive over the next few weeks. I am located in New Hampshire but am willing to ship these items. For more information, please contact Pete at [e-mail address removed]. and thank you!!

JLCooper ChromaFace Manual

Chris Ryan [21030691] · Mon, 9 Apr 2001 20:46:04 -0700

I've posted the manual for the JLCooper ChromaFace MIDI retrofit.

Thanks to Dave Clarke for providing a copy.

Chromas on Ebay

David Clarke [21030085++] · Tue, 17 Apr 2001 19:11:46 -0400

I wouldn't normally re-post Ebay listings, but since traffic has been quite light as of late, I thought these two would be OK:

RHODES CHROMA CLASSIC SYNTH.
Item #1424673627
Ends Apr-24-01 04:34:52 PDT
Currently at $1,100.00
Located in Belgium

RHODES CHROMA SYNTHESIZER/W ANVIL CASE
Item #1423319306
Ends Apr-21-01 03:19:23 PDT
Currently at $510.00 (reserve not met)
Located in USA (SF)

Sample "Subroutine" Patch (AKA, how to have your Chroma/Expander say "HELLO")

David Clarke [21030085++] · Sun, 22 Apr 2001 20:19:46 -0400

The designers of the Chroma included an ability to load and run code via the cassette interface. Specifically, there is a special type of cassette packet that can place external code into the Chroma. The "branch subroutine" option (Set Split 25) then allows that code to be run.

While this functionality is present, we really don't have any examples of it being used. That being the case, I've put together a short sample/demo.

The attached .WAV file [hello.wav, 12K] contains a single subroutine packet that will print the word "HELLO" to the small LED display on the Chroma. (A duplicate copy of this file is included in 22050 Hz, mono, signed raw format as hello.snd [20K]).

To run the subroutine:

  1. Press the "Load One" button the Chroma.
  2. Send the attached .WAV (or .SND) file to the Cassette input on the Chroma. The LEDs should blink, and then you'll see "Subr" in the LED readout. (This will tell you that a Subroutine packet has just been read.)
  3. Press "Set Split 25" (This will run the Subroutine you just loaded, printing HELLO in the Data Readout window).

Note: In the above, you can also push "Load All" The subroutine will still load, but you don't get the 'Subr' message in the display.

For those of you interested, the following is the actual assembly code which is loaded: 

34 02           pshs A     ; Save scratch variable on stack
86 00           LDA #$0    ; Code for a blank
97 20           STA $0020  ; Clear first LED spot
97 26           STA $0026  ; Clear last two spots
97 27           STA $0027
86 37           LDA #$37   ; Letter "H"
97 21           STA $0021
86 4F           LDA #$4F   ; Letter "E"
97 22           STA $0022
86 0E           LDA #$0E   ; Letter "L"
97 23           STA $0023
97 24           STA $0024
86 7E           LDA #$7E
97 25           STA $0025  ; Letter "O"
35 82           puls A, PC ; Restore original A value and return

If you want to confirm this is all that gets loaded, you can run the "TAPE2TXT" utility on the .SND file (with the DEBUG option turned on). When you do, you'll see the decimal versions of the hex opcodes above.

If you're so inclined, you could insert different code - maybe some code to create a "random patch" feature for instance.

Enjoy!

Dave

Eirikur Hallgrimsson [21030467+] · Sun, 22 Apr 2001 23:50:18 -0400

Yikes, the reprogrammable Chroma!

Very interesting, Dave. The hacker in me wants to know where this code gets loaded, and how much space there is there. Obviously it's in the main processor's memory, but just how big a routine could one fit in there? I suspect there's just enough memory left over to make sure that the machine doesn't run out of buffer space-rounded up to an even number of chips :-)

I don't know about a random patch generator. They are non-trivial programs--at least if you want more than a small fraction of the results to produce audible sound. Perhaps a morph routine (ala Galaxy) that took two patches as inputs and convolved (averaged?) the parameter values to get a new patch would be small enough. Then we need some way to designate the parents, which would be adding code just to do the UI.

Duh. I just remembered the brain-dead way to do a random patch generator. For each parameter, take it from a randomly selected patch in the current bank. That's pretty darn small, codewise, given a source of randomness.

Oh, am I having ugly thoughts now! Hideous distorted sounds caused by audio-rate modulation of parameters. I can't think of a way to get a digitized waveform "played" but one could definitely do things that would not sound like a Chroma.

Jeffrey D. McEachin [21030073+] · Mon, 23 Apr 2001 00:15:27 -0700

At 11:50 PM 4/22/01 -0400, Eirikur Hallgrimsson wrote:

I don't know about a random patch generator. They are non-trivial programs--at least if you want more than a small fraction of the results to produce audible sound.

I wrote one for the JP6, in 8051 machine code, no less, and I didn't think it was that big a deal. The trick was applying a random number generated w/ an exponential decay distribution to certain patch parameters, such as VCA LFO modulation depth, that you usually want to be a low value, but occasionally might want to be high. Though it might sound frighteningly complex on a chip w/ limited math capabilities, this exponential decay distribution was done w/ simple math. I generated an 8 bit random number, then randomly shifted it right (dividing it by 2) 6 times. I did it 6 instead of 7 times (which would have yielded a truely exponential distribution), because it sounded better. :) The end result is, the ms bit has a 1 in 128 chance of being set, the next bit 1 in 64, 1 in 32, 1 in 16, 1 in 8, 1 in 4, 1 in 2, and 1 in 2 for the ls bit.

JDM

David Clarke [21030085++] · Mon, 23 Apr 2001 18:08:02 -0400

Eirikur Hallgrimsson wrote:

Very interesting, Dave. The hacker in me wants to know where this code gets loaded, and how much space there is there. Obviously it's in the main processor's memory, but just how big a routine could one fit in there?

The "subroutine" packet gets stored in the cassette buffer, so in Rev. 14 FW that would mean it sits starting at $0141 and extends upwards to $0240 (the safe buffer sits at $0241). That would give you about 255 bytes. Through some judicious use of existing functions in the FW, you could do quite reasonable things ...

While I'm sure it's more trouble than it's worth, if you really wanted/needed more space, then could write a little code-loader which would say take up 20 bytes or so. The purpose of this code could be to copy the following 235 bytes elsewhere in RAM.

Using this technique (and multiple subroutines) you could assemble a larger piece of code in RAM (there's another 1000 bytes of RAM avail. up around 0x0c5a, if you're willing to limit yourself to only a few "sequencer" notes). Your last subroutine could then branch to/from this code, as required.

...Perhaps a morph routine (ala Galaxy) that took two patches as inputs and convolved (averaged?) the parameter values to get a new patch would be small enough.

This would be quite easy, and considering you'd effectively be doing the same operation to all parts of the patch (get two values, average, store), it wouldn't take up much code at all.

Then we need some way to designate the parents, which would be adding code just to do the UI.

Actually - we could make use of pre-existing program defines in the Chroma. For instance, Patch 1 could be the "current" patch (the one shown in the display) and Patch 2 could be the link program, the 'sequence' patch, the one in the safe buffer, etc.

Oh, am I having ugly thoughts now! Hideous distorted sounds caused by audio-rate modulation of parameters. I can't think of a way to get a digitized waveform "played" but one could definitely do things that would not sound like a Chroma.

If anyone out there really wanted to hack some 6809 assembler code for the Chroma, I could probably provide a mini utility which would convert a 6809 binary file (assembled code) into a properly formatted Subroutine packet (in .SND format) ready for loading into the Chroma ...

Chroma Soundconverters on Mac?!...

Fredrik Beckmann · Mon, 23 Apr 2001 08:32:59 +0200

Hi List!...

I dont know if this subject has allready been handled, but is there any aplications to convert Chroma Soundbanks for Mac like those for PC?!...

:-) Fredrik

Chris Ryan [21030691] · Sun, 22 Apr 2001 23:55:26 -0700

I have a friend who's working on it ... he's pretty busy, though, so it may take a while.

My friend never got to the project; I hope to include Dave's patch conversion routines in my Chroma app for Mac OS X.

Screen Shot of DOS Editors?

Chris Ryan [21030691] · Sat, 28 Apr 2001 23:30:22 -0700

Can anyone send me a screen shot of Tim Godfrey's Chroma Editor/Librarian, and/or James Moyer's DOS Chroma Programmer for PC?

Both can be downloaded from the [Editors & Librarians: MS-DOS] page.

And, I did receive screen shots for both, which are at the above page.

Chroma on eBay

Chris Ryan [21030691] · Sun, 29 Apr 2001 00:27:26 -0700

"Power supply and eight sound cards: in good shape. Slight damage on 9 volt battery backup board, otherwise sound. Voice cards in good shape. A good buy for parts or swapping parts if you own a Chroma."

Item #1425605892

[URL removed]

Current bid: $425 (reserve not yet met); ends May 1.