Hi,
I am in the process of imagining a pressure oriented controller,
as I like how my Mc Millen QuNexus reacts to pressure, I had a look at pressure sensitive fabrics
such as velostat or eeontex (which seems to be the one used by Mc Millen/Bebop).
Does anyone experimented with these materials ?
There are many shapes and sizes of these available. Sparkfun.com sells a tone of them, both pressure, bend/flex, and positional sensors. You find these sold often in Robotic supply stores as FSRs were originally developed as grip force sensors for robotic hands. The inventor of this technology lives just not of me here in Seattle. His company makes most of the FSR sensors in the music industry. The sensor is made by silkscreening metal layers between layers of conductive plastic that changes resistance when compressed. This means that almost any shape or contour can be made to order.
You can also make somewhat crude versions by using the black conductive foam that many ICs are shipped in. It also changes resistance when compressed.
My company, Synthwerks (www.synthwerks.com) makes a whole line of Eurorack modules based on FSR technology. They are not super expensive (around $9-$15 depending on quantities and shape) and like I said, typically come in a number of sizes and shapes. Almost every pad oriented controller, and almost all the aftertouch on keyboards use FSR tech to do it. VERY responsive stuff.
Commercial FSR from interlink (and others) are really great I already use them, they are sensitive, responsive, accurate and reliable.
The controller I imagine will be shaped for human hands natural positions.. It will not look like an electronic instrument but rather have something organic and almost alive (haptic vibrations maybe). I picture a pressure sensitive leather surface instrument.
The pressure sensors I am imagining would not be flat... so, the rigidity of FSR is a problem.
The other problem is their cost... 120 $ per octave 
velostat and eeontex (which are not plastic foam) can be an alternative but I doubt their sensitivity and accuracy.
eeontex seems to be a better choice than velostat.
I've just received a 30cmx30cm patch of eeontex.
I cut a 1cm square piece of this one for a quick experiment of volume resistivity pressure sensing.
I used 4 mm copper tape strips : one horizontal at the bottom side, the other vertical on the top side.
The result is a resistance varying from 8k ohms to 200 ohms under the pressure of my index.
With a 20*10mm rectangle crossed by a 15mm and a 5mm copper tape (the size of the sensor i want) I have a usable pressure range from 500 Ohms to 20 Ohms.
That's not as good as a genuine FSR, but much better than I expected.
I also tried surface resistivity sensing with to parallel strips on the same side of the fabric (1cm wide 2cm long).
It needed a higher pressure to get resistance variations (from 40kOhms to 5kOhms)... not good for finger sensors, but maybe good for a pedal sensor.
Compared to FSRs, this fabric need some experimentation but it cost a fraction of the price of FSRs.
Have you thought of maybe using soft pots instead? Spark fun has a 500mm one for $17. Like a ribbon controller it changes resistance with position 100 ohms to 10K ohms. You could quantize it to get discrete notes. If 19" is too long they also make an 8"($10) and 2"($5) version - https://www.sparkfun.com/categories/143
I've ordered (and received today) a 20cm softpot, It's smaller than the one of @timeorspace...
First i planned to use it as a ribbon controller (layered with a long FSR for pressure) but you're right @Ersatzplanet , it can be used with quantifications too 
Even better, it can be used "quantified" at stroke + relative glide in case of a continuous sweep (i think that the Linnstrument and/or the seaboard work this way).
I think this ribbon controller will be for my left hand. It will control the "Reference pitch"
For the right hand, i think that i'll use a set of eeontex FSRs, one or two for each finger, they will control notes relative to the Reference pitch and will extend the pitch range (thumb controlled octave ?... ) allow arpeggios and chords
FYI This article may be of some interest...
more here... http://www.plusea.at/?page_id=2259
Wow, i did not saw this one.
It would be great to control drum sounds with such an "e-sock"
@SmashedTransistors came across this by chance tonight and thought of this discussion. 
It's quite fun but the demo sounds somewhat repetitive, I guess that these gloves allow more complex melodies and expression.
oh for sure, certainly not a 'serious' type of instrument in my opinion. just found it interesting.
If any is interested, I found some interesting articles about home made conductive ink / paint / glue etc, I tried the paint, and it really works, you can vary resistance in many different way. Oh and I should explain what it is, it is using graphite powder and mixing it into glue / paint etc..
Here is a cool link to help get you started..
Another thing I saw in a you tube video, was someone made a silicon like molding of his figure and it was conductive, I tried adding graphite powder to silicon, but the result was a failure, non conductive, maybe I didn't add enough etc, not sure, but I wasn't able to find the ingredients online, if someone can find it, please please let me know. Pressure on one of these molds might be enough to register a change in resistance.
Maybe it was not silicon but another kind of rubber.
These pressure sensitive touch surfaces seem to be Arduino tweakable:
I hadn't posted it until now, but the other day I found a comment on the WWW that you need a ratio of at least 2 parts graphite to 1 part silicon before it starts conducting, they were recommending 3 to 1.
I bought some equipment to test it, hopefully will do in the next day or two.
I'm very interested in the results you'll obtain.
I've found another textile pressure sensitive surface here:
It comes with some Arduino/teensy code
Graphite and Silicon definately works, but may require a quite bit of experimenting to get the correct resistance and flexibility you require.
great to hear that, I should experiment with it some day.
Here is a link to the Linnstrument patent.
http://www.rogerlinndesign.com/assets/us-patent-9779709.pdf
It describes how the scan works to get pressure AND x and y position for each sensor. That's quite a smart design with lots of analog multiplexers !...