ACSensorizer Questions

[audiocommander]

Hi Sasha,

thanks for your feedback -

I love them all! ;D

well, there are sensors that look amazing when being used, but the level of control depends on the purpose. That means, I love to let others operate with IR-Sensors, but when it comes to make music, I prefer the unspectacular ones with a bit finer haptic control, such as linear pots, pressure pads, lower range IRs.

But with the new features like the fine output range tuning and the optimized auto-calibration, it's really fun to operate any sensor type. It can be even funny to make music with an analogue pot! I rather find that the quality of movement or interaction changes with the different types; so to say a cybernetic feedback mechanism :)

Cheers,

Michael

ps: if I find some time, I will upload some tutorial videos, too

[Sasha]

well, there are sensors that look amazing when being used, but the level of control depends on the purpose. That means, I love to let others operate with IR-Sensors, but when it comes to make music, I prefer the unspectacular ones with a bit finer haptic control, such as linear pots, pressure pads, lower range IRs.

I am boiling an idea for some time to make some kind of interactive sound/video/mechanical installation so conventional making music or performance is not my main idea. I`m more interested in experimental non-conventional interaction with that art installation. The problem for ealization is that sensors are pretty expensive and hard to get locally. I should search the web a bit so maybe I could find some cheaper to order. Any source recommendations?

But with the new features like the fine output range tuning and the optimized auto-calibration, it's really fun to operate any sensor type.

Yes, these are great features. I really find very easy how to configure min/max value of sensors. :)

ps: if I find some time, I will upload some tutorial videos, too

That would be great. I love your videos. If you put aside content which is fun itself, the videos are very well done.

Thanks for your answers

[audiocommander]

Hi Sasha,

thanx again for your kind words  :D

The problem for ealization is that sensors are pretty expensive and hard to get locally

well, that depends...

The problem gets bigger the bigger the interaction area gets. Normally sensors must not be expensive. For example my skin-resistance sensor circuits: I started with buying premade circuits for 10 EURs, now I'm making my own circuit that's 1000% better and costs about 3 Cents  ;D

A nice source for sensors are homebrewed LEGO Minstorm sensors, you could do a search on google; also: our wiki where I placed some nice links (esp. the sensor-wiki: http://www.sensorwiki.org/index.php/Sensors )

you can always go and spend 100$ on a super hitech tilt-sensor, but at the same time you could (just as an example) experiment with salted water and two electrodes... or some mechanical solutions (see the famous midi-sofa from seppoman)... possibilities are endless!

especially for someone with your recycling skills :)

best,

Michael

[bosone]

I finally finished my AC Sensorizer and it's working! (this is really incredible, since i used a protoboard and you should look at the mess on it!)

now it's time for some questions...

1) i cannot handle the use of the control CC64 as a pedal. i want to use the sustain pedal of my masterkeyboard as the pedal of the ACSens. but for me it donesnt work (or i didnt understood how to make it work)

i think the ACSens. is receiving MIDI, since if in MIDIOX i send the sysex string

F0 00 00 7E 40 00 01 F7

i immediately receive a series of F0 00 00 7E 40 00 0E 0C 00 F7 as "answer", so midi is sent and received.

is there any other way to test midi input?

i cannot understand how to use control changes to control the AC sens. (via CC98,99...etc) , neither use CC 64 as pedal, or use the external pedal to trigger autosense max/min

do i miss something?

2) i didnt measured the output of my distance sensor GP2D12. my question is: did you ever tried to amplify a little the signal with an OP amp like the LM324? i connected my pressure sensor and, since it outpou was really low (i think 0.3 V max) i succesfully managed to use a ~10 gain to have a better output. moreover, i had to use a "bridge amplifier" (see LM324 schematics) to hook the sensor outs.

this is because the sensors seems not so stable... can the gain of the LM324 be better of the PIC's one?

3) regarding distance sensors, which setting are you using and which sense values work best for you? did you built something to "cover" the sides of the sensors? mine are just fixed at the top of the box, and they are completely "naked" in the air...

4) if i play notes and set up a scale (for ex. C MAJ), and i change the root key (D MAJ) nothing seems to happen

anyway, after an entire afternoon of soldering and a night of rest, i'm still quite exhaust... but  i look forward to play with the new toy! :-)

thank you! :-)

PS: i have to admit i don't have the standard 6n138 as optocoupler, but an equivalend one named HCPL 2730 which has two inputs, but i think the wiring is correct (different from 6n138 one)... moreover sysex seems to be received....

unfortunately i cannot find the 6n138 here... :(

PPS: i'm using ACsens ver. 0.4.4

additional edit: during buildup stage i uploaded MIOS code using another midibox. in order to avoid complication and to proceed with work without excessive debug  but even if now sysex are displayed in MIDIOX when i do the "ping" using this optocoupler, i cannot upload mios code in this core using midibox studio. i receive the message "MIDI IN FRAME ERROR"

[audiocommander]

Hi Bosone,

congrats :)

From what I read, I think you don't have no HUI yet, and probably no LCD either, right?

An LCD is really required for this; except maybe youre swapping Banksticks with presetted patches.

The pedal only works with certain pedal modes; so if no special pedal mode is set, you won't hear any difference. The modes are:

pedal modes:

filter ? only forward if pedal down;

panic ? send panic on release pedal;

combinations of all pedal mode options are possible

Also see the "Pedal Modes" on the Onscreen-LCD-Menu Overview picture:

http://www.midibox.org/dokuwiki/doku.php?id=acsensorizer_04#manuallcd-menu

---

You can only setup the sensorizer with NRPN Messages' date= these are ordered sequences of Controller Messages. I think it's documented quite well in the Wiki; though I must admit this method is a pain in the arse, but it's just an emergency setup possibility in case no Encoders are available:

Turning the Menu Wheel (virtually):

1. Send NRPN MSB (Controller# 99) with value 0×61 to select control type

2. Send NRPN LSB (Controller# 98) with value 0 to 10 to set the control parameter

3. No NRPN Data MSB (Controller# 6) and NRPN Data LSB (Controller# 38) required

Setting the CC to send for (already selected) sensor 1:

1. Send NRPN MSB (Controller# 99) with value 0×00 to select the setting for sensor 1 (or 0×01 for sensor 2)

2. Send NRPN LSB (Controller# 98) with value 0×71 to set the control parameter “CC†for the selected sensor

3. Send NRPN Data MSB (Controller# 6) with a value from 0 to 127 to set the CC Number. NRPN Data LSB (Controller# 38) is not required (only for values greater than 127)

A remote setup program emulating the HUI is included for Mac OS X. If you're on windows or linux, there are some MiniAudicle Scripts attached to make this process easier; though you have to adapt these scripts to your requirements and system. Therefor I definitely recommend building a Huser Interface Input with Encoders!

So, to sum it up:

- connect a LCD

- connect a HUI module with Encoders and Buttons to be able to calibrate the sensors

---

2) i didnt measured the output of my distance sensor GP2D12. my question is: did you ever tried to amplify a little the signal with an OP amp like the LM324?

Yes, I tried many times, but always failed. The thing that annoyed me most is the requirement for a higher voltage. But anyway, that's exactly the reason why there's the sensorizer! You don't need no amplification circuits with the ACSensorizer, just connect the sensor to the AIN, calibrate: done!

---

3) is it possible to receive midiclock and use the ACsensor. as slave?

Yes.

Just send a clock and it's automatically slave.

Stop the clock and it switches to master.

syncronizer: master/slave mode (autodetect clock input switches to slave) and selectable BPM

---

4) what about the possibility to arpeggiate external chords?

Agreed' date=' this would be nice, but this is no arpeggiator/sequencer, it's a sensorizer. I share the believe that a device/application should concentrate on it's main purpose and do this perfect, instead of the ability to do everything "a little bit".

Afaik there are one or more MB-Arpeggiators available, just build one and hook it after the sensorizer.

Besides, there is really nearly no more space available for additional features and I don't plan to swap the PIC type atm.

Just connect your HUI and some more sensors and you see that you get arpeggiator/chord-like structures if you vary the quantisation values of the sensors!

---

5) regarding distance sensors, which setting are you using and which sense values work best for you? did you built something to "cover" the sides of the sensors? mine are just fixed at the top of the box, and they are completely "naked" in the air...

See the last step, "Sensor Calibration" in the step-by-step tutorial:

http://www.midibox.org/dokuwiki/doku.php?id=acsensorizer_04#step-by-step_building_instructions

I would have to look up some values first, but in v.0.4.4. the autosense-mode is really working beautiful, you should give that a try! (but you can only access this feature with a working LCD and a pedal). I'm using the sensors in quite different ways, but most times I have them covered with a red perspex (plexi) plate, like the small LED windows that can be found on the front of remote controls. You should experiment first, so no reflections can occur if you place something in front of the Sensors.

keep me updated with your progress :)

Best,

Michael

ps: I must admit that I have no clue about Optocupplers and such; but this is also more related to building a Core Module than something to be covered here in the ACSensorizer topic. Maybe you should open up a new topic for help on the Core and then return to specific questions on the sensorizer  :-\

[bosone]

thanks for the reply!

for now, i have the complete thing working

including the breath controller!

you may wonder why i have only three encoders... it's because i had only three left spare at home and i cannot buy them in my city (that one came from the midibox 16e order!)... so instead of ordering just one encoder from VOTI i preferred this solution...

i used a two-pole deviator in order to "double" one encoder to two! :D i use only one encoder + the deviator for sensor select and menu navigation. it's quite easy also in this way!

the only thing that gives me problem now is the midi IN... and i don't undestand if it's the optocoupler or the circuit design...

i hope to find info... my last resort would be to buy some optocoupler from another midiboxer... ;)

i have only three button since i planned to use the external CC64 as pedal... the third button is "panic" and now i'm desoldering it to use it as a pedal one.

do you fell "panic" is useful? (sorry, i didnt tested the car on the road, yet :D

the last thing will be to add a softpot XY!

as i said in another thread.... the only matter here is how to connect that damn plastic connector to a wire!!!  ???

thanks for your help (and for the project! ;))

[Sasha]

the only matter here is how to connect that damn plastic connector to a wire!!!

Well, this is not a problem at all. Just make one PCB that will go under the foil contact, and put it in a sandwich made out of 2 pieces of wood, metal or anything with holes at the end for screws. Put a piece of rubber on top of the foil and tighten the sandwich. Pretty easy if you understand me. Thin the PCB contacts to prevent oxidation.

here is the pic

[bosone]

now that i solved the midi-in matter, everything is OK!!  ;D

i even understood how to use NRPN events!

just some  more questions...

1)  if i play notes and set up a scale (for ex. C MAJ), and i change the root key (to D MAJ for example) nothing seems to happen... do i miss something? instead, the different scales are working OK!

2) i think the slow-down factor could be more "fine"... i mean, it has a great range (the maximum value is 127). but if i set up for examlpe 5 or 6 as a slow down factor, the time-distance between events is very very high, and practically is unusable (i think ). what about to dither it a bit? so that for example at a slow-down of 127 the temporal distance between events is something like 500 msec?

3) the features list  says: "harmonizer with 20 scales (minor, major, blues, spanish...) ? harmonizes generated or received (by MIDI!) note-signals"

acutally, i didnt understood what it should do... i send midi note events (now they are working cause i can autosense with my masterkeyboard sustain pedal!) but nothing seems to happen or change if i set up "note" (CC0) as event and i play a chord on the keyboard... i still miss something?

i'm using distance sensor to do this, maybe i don't have a so-stable-hand and the signal is flickering, so i don't appreciate the difference!! :-)

one addendum to the manual: i don't think i read anywhere that the DIN module is to be connected to J9 of core module. if i miss that excuse (again!) me, but while i was building the box, i initially connected the DIN to core J6 and it wasnt working! it took me a while to debug it!  :P

anyway, it's a great project!  ;D

thanks again

soon i will post photos... for videos (if i have time to record them) you have to wait until end of august, when i will return to office and i will have access again to wideband connection!  ;D

PS: i finally use the XY touchpad as a linear sensor... not XY but only X. it works very well but i miss the XY control... i hope to find something else to connect.. maybe a laptop touchpad???

PPS: take a look at this for ribbon controller!

http://asmidius.googlepages.com/home

[audiocommander]

1) if i play notes and set up a scale (for ex. C MAJ), and i change the root key (to D MAJ for example) nothing seems to happen... do i miss something? instead, the different scales are working OK!

You're right, I have to take a look at this; but it seems that you can just ignore the base note for now.

2) i think the slow-down factor could be more "fine"... i mean, it has a great range (the maximum value is 127). but if i set up for examlpe 5 or 6 as a slow down factor, the time-distance between events is very very high, and practically is unusable (i think ). what about to dither it a bit? so that for example at a slow-down of 127 the temporal distance between events is something like 500 msec?

the slowdown is from the time where there was no quantizer. it is true that values above 25 are somewhat unuseable, but I see no reason why to waste some bytes to restrict this; maybe it's useful for someone.

I recommend not to use the slowdown at all, until you have a reason to. (also see below)

3) the features list says: "harmonizer with 20 scales (minor, major, blues, spanish...) ? harmonizes generated or received (by MIDI!) note-signals"

acutally, i didnt understood what it should do... i send midi note events (now they are working cause i can autosense with my masterkeyboard sustain pedal!) but nothing seems to happen or change if i set up "note" (CC0) as event and i play a chord on the keyboard... i still miss something?

Not sure what you mean?

it says: "assignable (...) Controller-Number or Note_On generation"

That means if "CC" is saying "NTE" (0) you are sending NOTEs; if "CC" is showing a number between 1 and 127 you are sending ControllerChanges! I chose this, because it's very unlikely that anyone wants to send CC#0 which is a bankSelect, and I really had to watch over the available space.

Just connect two sensors, set NTE for the first one and CC#7 for the second one and see what happens :)

If you're sending Notes from a connected Keyboard, the Sensorizer harmonizes the incoming events too and sends them to the output, regardless of sensor-specific settings. If you hear nothing, it may be probably connected wrong (?)

i'm using distance sensor to do this, maybe i don't have a so-stable-hand and the signal is flickering, so i don't appreciate the difference!! :-)

Just increase the quantisation! I assume you have your signal sent with 32. or 48. or without any quantisation! Set it to 1/8. notes!

After you have done that you could increase the gaps between the notes further with the Slowdown; that could be useful to trigger more random events.

anyway, it's a great project! ;D

(...)

soon i will post photos... for videos (if i have time to record them) you have to wait until end of august, when i will return to office and i will have access again to wideband connection! ;D

Thanks :)

I plan to upload some vids too, but haven't also found no time yet.

Cheers!

Michael

[bosone]

i began to play with the now toy! :-)

beside the fun,. i experienced two problems, both related to midi-clock and maset/slave issue

1) AC Sensorizer set to master

the AC Sensorizer seems to correclty sending midi-clock, but i think it lacks a "song start"/"stop" signal.... i will try to explain:

i'm using sonar6 as sequencer. if sonar is set as slave, when i start "play" or "rec" it is waiting for a sync in order to hook to the master.

for example, if i use the midibox sequencer as master, sonar as slave start to play/record ONLY if i hit the "start" button of the midibox sequencer.

when i'm using the AC Sensorizer, this signal never arriver, so that sonar has never the possiblity to "hook" to the master. i even tried to change the BPM value in AC Sensorizer to see if this send the start/stop, but it doesnt happen. are you experiencing the same? (if this is really a bug, maybe you can send the "start" when the BPM value is changing - just an idea, since the box lacks a start/signal)

2) AC Sensorizer set to slave, sonar 6 as master

in this case, the BPM are correcly received but i found that, when i play notes (CC0), the arpeggiator is not downbeat-synced to the master. so for example i load a drum loop at 100 bmp, set the AC Sensorizer to slave and play notes: the first note that i'm playing are always delayed with respect to the downbeat. i found that there is always an appreciable delay between the hand that arrives on the distance sensor (in order to trigger the note) and the note played. when you are not playing synced to anything is not appreciable, but if you play along a loop or try to "stay in time" with something else you notice it...

i dont' know if it is a problem of my  midi-interface (i'm using a mini-DIN midi of a doughter board of a soundblaster live, since i don't have any more MIDI in port available!) or of the box itself.... did you noticed this?

moreover, i already deleted two saved patches because i think that when you press the save button you immediately overwriet the patch. what about a two-three seconds delay (like in the midibox sequencer saving pattern) before actually writing the patch? so that one is warned that it's writing a patch...  :P

[audiocommander]

Hi Bosone,

good point ;D

Would it help to send a CONTINUE signal each bar when in MASTER mode? I don't know if your sequencer will behave correctly this way; I can't see how to implement a START/STOP handling because it's no sequencer.

You're right with the slave-mode; I forgot to implement a start/stop receiver, that's because you notice a delay, because it simply does not recognize the right downbeat; I will correct this.

Regarding the load/write buttons: I would really recommend to use green/red buttons and keep the recommended HUI-layout, else it's surely a bit more complicated to handle.

thanks for beta-testing :)

best,

Michael

Edit: I attached a quick fix with the sent continue (master & slave) and the recognition of a start signal, should work now...

ACSensorizer_0.4.5_encType2.syx.zip

ACSensorizer_0.4.5_encType2.syx.zip

[bosone]

hi!

i just tried the new version.

now, when the sequencer is set as master and the ac sensorizer as slave, there is no more delay and everything is perfectly in sync.

i still have some problem when i try to sync sonar as slave with ac sensorizer as master. now the sequence correclty starts, but there is still some appreciable delay between the down-beats

for me it's not so critical, since i can use sonar as master and sync the box, but maybe other people could experience problem, expecially if they don't connect a midi-in input...

anyway, thanks for fixing the problem

PS: i just painted my two load/save button gren and red! good suggestion! ;)

[stryd_one]

I can't see code so I can only guess but... Remember that SPP Song Start does not actually start playing when received - it waits for the next F8 before that.

So you need to send Start first, then the next 0xF8 tick makes it play.

[audiocommander]

yeah, that's right. Sending a CONTINUE works with my Electribe, but it might not work with every sequencer. I just added a START signal whenever a patch is loaded. I don't think it's useful to send a START with every bpm change; I guess the best way for this is the load button.

I see this is a bit tricky, maybe I get some ideas how to handle this later on; for the moment I think it's tolerateable...

As the sensorizer is receiving START messages in both modes (and the forwarding is enabled anyway); it might work to send a START message from another connected device via ACSensorizer?

Cheers,

Michael

[audiocommander]

Hi,

I just released version 0.4.5:

* sync start signal is now recognized (SLAVE & MASTER),

* a continue signal is sent each bar (MASTER)

* a start signal is sent on patch load (MASTER)

* the harmonizer is now working as expected with base notes

http://www.midibox.org/dokuwiki/doku.php?id=acsensorizer_04#application_software

thanks again for beta testing; I haven't recognized these things as I was running this version mostly stand-alone without any sequencers ;)

best regards,

Michael

[sneakthief]

Crazy question: what about using the AC sensorizer as a CV quantizer?

This would of course involve connecting it to an AOUT module.

Here's what I propose:

1. Standard 10V p-p CV signals would be halved on input (see this thread)

2. Internally calculated midi-notes would then be converted to voltages using the AOUT module.

3. Output voltages would conform to the 1V/Octave standard (see MIDIbox CV output table)

Options:

4. Inputs from one channel could transpose (offset) inputs from others

(imagine an LFO CV on one input, then a keyboard CV from a second one: instant arpeggiator!)

5. CV input from a channel could select from the various scales

6. Harmonies could be simultaneously output to multiple CV outputs

Is there anything I can do to help you implement this?

Background:

As modular synthesis enjoys an amazing renaissance, there's a high demand for CV quantizers. This would put all of them to shame!

Here are some of the current ones - you can see how simple they are and how much room for improvement there is:

(120 euro)

($435 US)

($260 US)

($570 US)

[sneakthief]

OK - for the basic CV out functionality, stryd_one and I discussed these updates to main.c:

 #include "aout.h"

#if SENSORIZER_INTERFACE_SYNC
	ACSYNC_Init();									// Init Syncronizer
	ACSYNC_BPMSet(120);
#endif

#if SENSORIZER_AOUT
	AOUT_Init();									// New: Init AOUT Board(s)
#endif

void Timer(void) __wparam
{
#if SENSORIZER_INTERFACE_SYNC
	ACSYNC_Timer();
#endif /* SENSORIZER_INTERFACE_SYNC */


#if SENSORIZER_AOUT
	AOUT_Update();									// New: Update AOUT Voltages
#endif

...and in ACSensorizerDefines.h:

#define SENSORIZER_AOUT	1 	// if AOUT conncected

Any thoughts on this?

[audiocommander]

Hi sneakthief,

thanks for your input!

I have two thoughts on that:

1. I have never worked with AOUTs; I don't even own an analogue synth, nor do I have any AOUT board. This means I do not have the suitable environment to develop this.

2. I have a severe memory space problem regarding new features. The current program grew the last two years and now I reached the maximum available space for the PIC18F452. That means, I will still be able to add one or two lines, but more would make necessary a PIC change... and I guess this would mean a lot more, because it would require to add some menu items as well. And changing the PIC type is a bit of a hassle which I am definitely not going to do if the gain of the additional features isn't really worth all the action.

Also: I'd like to concentrate on inputs and sensory readings improvements for this box. If I add AOUT probabilites, the question arises: why not DOUTs, too?

So, I get the impression, that what you're describing would be very nice for a seperate box! I don't know if there's a Midibox converting MIDI-signals to AOUTs, but I guess there is. This way you could easily hook them together.

Another argument pointing towards this solution is the code integration of the AOUT controls:

I would not add the AOUT functions in the Syncronizer Classes like you proposed, I'd add it to the Midi-Classes, where the final messages are being sent. That's another hint speaking clearly for the realisation as a seperate Box, 'cause it shows that the functionality has nothing to do with the code going on inside the box, it would be just glued on the output side.

If in this case, any adaption on the midi-output would be required, I'd be very happy to help out and integrate some small additional lines :)

Hope this helps,

best,

Michael

[sneakthief]

AC - thanks for the reply :)

The MIDIbox CV converts midi to cv, but it's written in ASM.

I suppose that the best solution would be to import your scaler & harmonizer code into the Analog Toolbox.

I'll give it a shot!

[bosone]

i just uploaded this video on youtube of me using acsensorizer with a solo violin of garritan personal orchestra

and added some photos on my website

http://www.alchemystudio.it/Strumenti/ACSensorizer/ACSensorizer_en.html

i also have a quick question...

when you set the mix-max sensor value (for example with autosense) what happens to the values below and above the min / max ? i sometimes have the feeling that they are somewhat interpolated depending on the sense factor.... because it happens sometimes to me that if the sensor output is below the minimum, it behaves strangely...

i mean, i would have written some lines of code like:

if AIN_value > max_sense_factor then AIN_value=max_sense_factor

if AIN_value < min_sense_factor then AIN_value=min_sense_factor

where AIN_value is the output from the sensor min_sense_factor / max_sense_factor are the values set with autosense (for example)

i didnt look at the code so i don't know how are the variables and how they are handled...

[audiocommander]

bravo !!! :D

about your question:

I'm quite sure you're using the latest 0.4.5 version; there's been a bug in the previous versions that might explain this.

Normally all events below min and above max are ignored.

It may be though, that if the factor is not well chosen, there might be a loop within the scaling function (the last menu page -> from...to). If you used the Autocalibrate function, this should normally not happen.

Maybe it helps if you describe the "strange behaviour" better.

I'm on holiday for about 12 days starting wednesday, so please don't wonder if I'm not answering at once the next days...

Best ;)

Michael

[bosone]

why it does flicker??

i built a ribbon controller, and it behaves good, but it flicker with an high range when i don't touch it (so when it should be at minimum value).

look at my project here:

http://www.alchemystudio.it/Strumenti/ACSensorizer/ACSensorizer_en.html

if i don't touch the wire (so the "circuit is open") and i measure the voltage, it reads 0.35 V and it's pretty stable. it maybe reads +/- 0.01 V

but on the ACSensorizer, when i use autosense, the values read flicker in the 0-70 range (out of 255 scale), very fastly

if i touch the the ribbon, the voltage remains stable as before, and the ACSensorizer values become stable, and the minimum value that i can achieve is around 35.

of course i could set the minimum value to 80 or 90, but i'd lose a significant part of the ribbon...

what should be the reason of this flickering at minimum values on the ACSEnsorizer??

i tried to change PIC AIN channel, but nothing happens.

and why the voltage read seems stable and the acsensorizer input is not?

beside this, the two distance sensors doesnt show this problem.

an idea i had: if i short circuit one of the two side of the guitar string  with the tape, i would obtain a definite and fixed value (for example the minimum) when i dont touch the string.

but, even if this is working and i could use the "release detect" function, this changes the behaviour of the ribbon from linear to something logaritmic (or parabolic or so on...),

so it's not so usable... you have for example the first half of the ribbon which changes from 1 to 30 (in MIDI CC steps) and the other half from 31 to 127 (and if you consider the half of this half the behaviour is practically the same!)... so this is not option i would like to consider...

[bosone]

yes!

i solved the flickering problem!!

preamble:

i checked now with a tester and between ground and the string wire (signal) at rest (so when the string is not touched) i read 0.21V.

i added a diode with threshold voltage of about  0.15V (i had it a home, i was lucky). the diode reads: 1N5180 between ground and signal.

in this way flickering was reduce but not deleted.

then i had the idea of adding a small cap. it was a cap spared from the sid filter or from the midibox power supply i think, so it should be in the 1-100 nF range, i don't know...

anyway, by doing these two operations, the flickering was totally eliminated!! :-)

edit: i feel the diode is completely useless... but now since it's not broken i will not fix it!! :-)

[audiocommander]

Hi,

hehe... it's good to see that some things solve themselve ;D

I was away, so I just read your postings. Indeed, I would have suggested a small cap as well, so that the signal is grounded if you don't touch it. If it's not grounded (floating) there are some random values like as you would expect from an unconnected AIN pin. This behaviour is related to the MBHP Hardware concept and not specific to the ACSensorizer program.

Best,

Michael

[audiocommander]

There's an update available!  :D

http://www.midibox.org/dokuwiki/acsensorizer_04#application_software

v0.4.6 / 2007-11:

- fixed Master/SLV autodetection bug, now behaves correctly. Overworked areas were Clock-Detection, MASTER/SLAVE auto-switch, feedback-loop protection.

- The PANIC button now also sends STOP in MASTER mode (LOAD sends START since 0.4.5)

- The Global MIDI Channel now defaults to CH16, messages are sent/received on all channels except for PRG_CH (Global CH only). Now ACS patches can be loaded and saved without interfering with connected instruments.

- fixed some minor bugs

I also added a MIDI Implementation Table to the Wiki-page.

Cheers,

Michael

ps: of course I'm still interested in reports  ;)