Connecting the Akai s950 and VX-90

Following up on my last post about connecting Akai samplers and synths with their proprietary 13-pin cable, I managed to get it working polyphonically with my s950 and VX-90. I expect this would also work with the AX-73 (which is virtually identical to the VX-90) and the AX-60 (which is similar) synths, and with the s900 and s612. Based on what worked for me and stuff I’ve read about people using the s612/AX-60 combo, I suspect all the units with the 13-pin connection work (somewhat) similarly.

How to get s950 voices to be processed polyphonically by the VX-90:

  1. Connect the voice out of the s950 to the sampler in of the VX-90. I found one on eBay described as “for Roland cable GKC-3 13-pin” — it’s a knockoff of a Roland guitar synth cable. If you get some other cable, make sure pins 1-7 are connected (1-9 to get all 8 voices).
  2. Connect the MIDI out of the VX-90 to the MIDI in of the s950. Not the thru! The VX-90 is going to transform the midi notes you send it before passing them on to the s950. If you want to be able to run the s950 without the VX-90, you’ll need to be able to reroute its MIDI input.
  3. Connect the outputs of the VX-90 to your mixer or amp.
  4. Set the VX-90 MIDI channel to 1 (edit, parameter 70). Set the s950 MIDI channel to 1. You can probably use other channels, but the two are actually going to communicate over multiple channels, so it’s easiest to just set them at 1 and let them do their thing.
  5. Create a patch on the VX-90 with sample input (parameter 06) set to ON and A/B balance (param 07) set to 100 (or at least 50). Set assignment mode (param 50) to POLY and portamento (param 51) to 0. For starters, open the filter up all the way (cutoff, param 10, to 100, and resonance, param 11, to 0).
  6. Create a program on the s950 with a single keygroup containing a simple synth sample, with the filter wide open (to 99). Duplicate this keygroup until you have 6 identical keygroups. Set the midi channel of KG1 to 1 and its output channel to mono 1. Set KG2 to channel 2 and mono 2, and so on.
  7. Select your sample patch on the VX-90 (the midi mode that communicates with the s950 is only enabled when you select a patch with sampler input set to ON).
  8. Send the VX-90 some midi notes. You should hear the s950 sample playing through the VX-90. Make sure when you run through a series of notes, all the notes are audible and that when you play a chord, you hear all the notes simultaneously. Play with the filter (param 10) to see it processing the sampler output. Assign some envelope modulation to the filter cutoff (param 13) to hear independent filter processing for each voice.
  9. To tweak your program on the sampler, edit the keygroups with KG set to “all”, which will change all 6 keygroups at once. Of course, you could give them different settings if you want a sort of random note effect.
  10. Play with the A/B balance (param 07) to mix the sampler with the internal oscillator. By turning the tune knob on the front of the VX-90 you can get a nice detune going between the two. You should also be able to put the VX-90 in DUAL or UNISON mode (param 50) to get extra stacking. If you just want to use UNISON, you actually don’t need all the keygroup hoo-ha on the sampler; that’s just necessary for polyphonic operation.

Why does this work? When the VX-90 is in sampler mode (i.e. when the current patch has sample input enabled, param 06) it spreads its midi notes across 6 channels, one corresponding to each voice. If you send it a series of notes on channel 1 and monitor the output (the  s950 has a midi monitor function), you’ll see the notes come through on channels 1-6. The s950 program is set up to receive on those 6 channels and route that note through the corresponding output in the 13-pin cable. The VX-90 and s950 have to be in sync this way to make sure the sound from the sampler goes through the corresponding voice that is being played at the same time.

Akai Samplers and Synths

For a while, Akai was making both samplers (s612, s900, s950) and analog synths (ax60, ax80, ax73, vx90) and they had this proprietary system for connecting them — a 13-pin cable that lets you run multiple channels of output out of the sampler and into the synth, allowing the sampler’s voices to be polyphonically processed by the synth’s analog filters and so on. This sounds like a great setup, so I was surprised when I recently scored an s950, vx90, and 13-pin cable to find very little info on this online, aside from the s612/ax60 combo. After a bunch of experimenting with my setup, I decided to record what I found for posterity.

  • The 13-pin cable is just a multi-channel audio connection
  • Each sampler voice corresponds to a single channel in the cable
  • The s612 has 6 voices, and the s900 and s950 have 8 (not sure about s700/x7000)
  • In the s950, a simple program with one keygroup assigned to “all outs” will get sent out the cable, apparently rotating through the 8 channels. Assigning the keygroup to a specific out will seemingly always send the audio out the corresponding cable channel.
  • In the s950, the individual channels are indeed monophonic (i.e. the do correspond exactly to the 8 voices). A later event will seemingly always cut off an earlier one.
  • The vx90 has six voices and apparently only listens to the first 6 of the 8 cable channels. So if it’s hooked up to an 8-voice sampler like the s950 that’s rotating through its voices, every 7th and 8th note will be dropped by the vx90
  • The s950 seems to have no way to assign a keygroup to a set of outs (i.e. outs 1-6). You can assign to a single out or to all (or to left/right, even/odd, but that doesn’t help).
  • If 2 keygroups are assigned to outs 7 and 8 while a third is assigned to all outs, the s950 is not smart enough to prevent the all group from sometimes using voices 7 and 8 and therefore cutting off those notes. So you seemingly can’t trick it into just using 1-6.
  • Since individual keygroups can be assigned to specific outs, a multi-keygroup program (e.g. a drumkit) could be set up to only use outs 1-6.
  • However, even if you get the s950 to not use outs 7 and 8, there seems to be no way to ensure that, on receiving a midi note, the s950 and vx90 will choose the SAME voice channel to use. So the notes still won’t line up.
  • The s612 and ax60 combo apparently address this by having a midi mode where the voices are controlled on 6 separate midi channels so that the two units can be forced to allocate the same voices. Presumably if you can put the ax60 in this mode, you could do something similar with the s950 by assigning keygroups to multiple midi channels. I don’t have an ax60 and don’t know if this is true.
  • If the vx90 is in unison mode (i.e. all 6 voices stacked), it will pick up audio on any of the 6 cable channels, so that works. Of course, this is just using the vx90 as a single monophonic filter on the s950’s output, which could be accomplished with any external filter.
  • You could also presumably make a cable that would wire a single audio channel to all 6 voice inputs in the 13-pin cable, allowing the vx90 to be used as a monophonic filter when in poly or dual mode, for whatever that’s worth.
  • The upshot is that, as far as I can tell, the whole Akai-sampler-synth connection thing is mostly bogus and unusable, unless you have an ax60, or just want to use the vx90 unison mode.
  • I believe everything I say about the vx90 applies to the ax73. I don’t know about other synths that may have the 13-pin connection.


Time, a new song, with original lyrics, a rarity for me. The lyrics and music were heavily inspired by early New Order — the opening line is direct from “Denial”: “here I am in a house full of doors and no exits”, one of my favorite lines ever) and it just grew from there. It’s also by far my shortest song ever, at a radio-friendly 3:32. Writing lyrics was fun. In the past I’ve had trouble with that, tending to be too literal, resulting in awkward, heavy lyrics. This time I worried less about telling a coherent story and went with things that sounded like they fit the mood, and I think ended up with something that flows better and ended up meaning something after all.

Hungarian Rhapsody

Liszt’s Hungarian Rhapsody #2 is one of my favorite pieces (me and everyone else). I was reading about it last week and listening to some performances, including Horowitz (I like what he does starting at about 4:10) and Jung Lin (check out the section starting around 5:00). I also came across a reference to Friz Freleng’s Rhapsody in Rivets, a cartoon that sets the construction of the “umpire” state building to the melody, which I had to watch. Hungarian Rhapsody might be the #1 piece of classical music for cartoons, because Rhapsody in Rivets led me to (in chronological order), Mickey Mouse, an orchestra of animals, Bugs Bunny and a mouse, and Tom and Jerry — all of them playing Hungarian Rhapsody, and all sharing similar gags. Rhapsody in Rivets has some great visual interpretations of the music, like the guy running up and down the ladder (4:20, 4:50), the diggers (4:10), the hammerers (5:20), and the bricklayers (5:30, 6:15). The Mickey Mouse cartoon probably tops the list for surrealism, when the piano and stool boot Mickey off the stage and the piano starts playing itself with its front legs while the stool dances (starting around 1:35). And then of course there’s Victor Borge‘s take, which is sort of like a Tom and Jerry cartoon but with real people.


“It is necessary for you to understand the functions of the controls and selectors of the Juno-60 perfectly to fully enjoy the advantages of the unit. Some setting examples are shown in this manual to make it easier for you to master how to operate the Juno-60, but you are the one who creates the sounds. Please find out your own setting and new ways of playing.”

Univox SR-95 and Roland TR-55

The Univox is probably my all-time favorite preset drum machine. I had one a long time ago and more or less accidentally produced this track. But in the great “I have a computer now!” purge I foolishly sold it. Well, I managed to find another one in terrific shape from a retired guy who was the original owner and used to play shows with it back in the day. Actually, his machine was the Roland TR-55, and his father used the Univox. Nora and I had to drive out to the hidden enclaves of Bonney Lake in the pouring rain to get them, but they’re both mine now, and they will be handed down to the NLP when I am gone.

My Life in Song

A dumb “meme” that I couldn’t resist doing: using only song names from ONE ARTIST, cleverly answer these questions. Try not to repeat a song title.

1. Your Artist: new order
2. Are you male or female? the him
3. Describe yourself: someone like you
4. How do you feel about yourself: doubts even here
5. Name one thing you are not: true faith
6. What is the best advice you have to give: every little counts
7. The first thing you think of when you wake up: leave me alone
8. If you could go anywhere, where would you go: all the way
9. Your favorite form of transportation: 60 miles an hour
10. Your best friend is: mr disco
11. Your favorite color is: everything’s gone green
12. What’s the weather like: sunrise
13. If your life were a TV show, what would it be called: blue monday
14. What is life to you: regret
15. Describe where you currently live: rock the shack
16. If you could change your name, what would it be: krafty
17. Your favorite food is: temptation
18. How I would like to die: dream attack
19. My soul’s present condition: in a lonely place
20. How would you describe your love life: world (price of love)