|
Welcome to the Piano World Piano Forums Over 3 million posts about pianos, digital pianos, and all types of keyboard instruments. Over 100,000 members from around the world.
Join the World's Largest Community of Piano Lovers
(it's free)
It's Fun to Play the Piano ... Please Pass It On!
|
|
63 members (BWV846, Anglagard44, clothearednincompo, brdwyguy, amc252, Bellyman, 16 invisible),
2,265
guests, and
375
robots. |
Key:
Admin,
Global Mod,
Mod
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
OP
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
If you've followed some of my endeavors on this forum, you probably know about CyberGene's DIY hybrid MIDI controller. Sleutelbos suggested the name Cybrid and I liked it Basically I created a MIDI controller from an old grand piano action by putting optical sensors that detect velocity of the hammer shanks. After some considerations I decided to make it open-source and free for people to contribute. Here it is: Cybrid on GitHubI started publishing it 3 months ago but I am always lazy and slow writing technical docs, so at the suggestion of others, I decided to just publish the current state of the project with some section of the main description still in progress. The readme.md that contains the description is too rushed and probably inconsistent and full of typos or even wrong information. But the code and the PCB design is there. I can answer questions and will allow for interested smart-heads to be contributors.
|
|
|
|
Joined: Aug 2017
Posts: 801
500 Post Club Member
|
500 Post Club Member
Joined: Aug 2017
Posts: 801 |
Thanks. Now i need to find the parts to make a Tybrid.
Kawai: NV5 | Yamaha: CLP-745R Pianoteq 7 Pro | Garritan CFX Full | Vienna Imperial | Keyscape Full | Ravenscroft 275 - Modern U - Model D - AG | Minigrand | The Oeser
|
|
|
|
Joined: Aug 2017
Posts: 1,756
1000 Post Club Member
|
1000 Post Club Member
Joined: Aug 2017
Posts: 1,756 |
CG, very good, even for a work in progress page. Kudos to you.
Kawai ES8, Roland RD2000, Yamaha AG06 mixer, Presonus Eris E5 monitors, Sennheiser HD598SR phones.
|
|
|
|
Joined: Jul 2009
Posts: 508
500 Post Club Member
|
500 Post Club Member
Joined: Jul 2009
Posts: 508 |
well done! congratulations from us!
|
|
|
|
Joined: Aug 2019
Posts: 1,395
1000 Post Club Member
|
1000 Post Club Member
Joined: Aug 2019
Posts: 1,395 |
Cybrid is great achievement for sure. Well above what I would ever try. Let's hope somebody gets their feet wet in there; this would be the best way to find ommissions or ambiguities in the project docs.
Wise men speak because they have something to say; Fools because they have to say something. (falsely attributed to Plato) Vlad, Adult beginner
|
|
|
|
Joined: Nov 2018
Posts: 1,797
1000 Post Club Member
|
1000 Post Club Member
Joined: Nov 2018
Posts: 1,797 |
Impressive work and nice write up. I enjoyed the pics!
I tried to implement Cybrid based on your template and using your code, but the high 3 keys after A7 are not working on mine. Any idea what may be wrong?
😅
|
|
|
|
Joined: Jan 2016
Posts: 3,375
3000 Post Club Member
|
3000 Post Club Member
Joined: Jan 2016
Posts: 3,375 |
Thanks CyberGene - this was a fun ride!
|
|
|
|
Joined: Oct 2006
Posts: 287
Full Member
|
Full Member
Joined: Oct 2006
Posts: 287 |
Wow. Gonna have a lot to review.... Thanks for this.
|
|
|
|
Joined: May 2020
Posts: 682
500 Post Club Member
|
500 Post Club Member
Joined: May 2020
Posts: 682 |
I was very impressed before, and I am even more now that I've seen the additional details. My jaws dropped to the floor, broke it and fell at the bottom of the crawling space. This is way more hardware design work than I had anticipated, and I am astonished by how much you have accomplished, starting from zero in such a short time, while having a full time job and a family!
Thanks for sharing!
|
|
|
|
Joined: Sep 2015
Posts: 5,820
5000 Post Club Member
|
5000 Post Club Member
Joined: Sep 2015
Posts: 5,820 |
This is amazing. Thank you so much for this.
I am thinking of having a go. I can see some problems adapting this to my piano, but such problems are insignificant compared to making the impossible possible, as you have done.
👍👍👍
|
|
|
|
Joined: Sep 2007
Posts: 19,099
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2007
Posts: 19,099 |
Kudos CG, excellent work!
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
I see that you use reflective optical sensors. What characteristic of the moving hammers are you detecting with those sensors?
Your readme in git says that "the phototransistor will change its collector current depending on the proximity of subjects to the sensor".
So is it proximity? Or reflected light? If the latter, how is it working properly when the action is open to varying ambient light (as seen your video), rather than tucked away inside the dark innards of a piano?
|
|
|
|
Joined: Jun 2020
Posts: 646
500 Post Club Member
|
500 Post Club Member
Joined: Jun 2020
Posts: 646 |
CG: Thank you for your generosity. I antecipate that many clones will appear in the immediate future (and many old actions will get a second fruitful life...).
|
|
|
|
Joined: Jun 2013
Posts: 383
Full Member
|
Full Member
Joined: Jun 2013
Posts: 383 |
I see that you use reflective optical sensors. What characteristic of the moving hammers are you detecting with those sensors?
Your readme in git says that "the phototransistor will change its collector current depending on the proximity of subjects to the sensor".
So is it proximity? Or reflected light? If the latter, how is it working properly when the action is open to varying ambient light (as seen your video), rather than tucked away inside the dark innards of a piano? CyberGene will be able to answer this more properly, but as far as I know, those sensors have a some kind of light filter which works quite well, and are not afected by ambient light (I don't know if direct sunnlight can bother them...)
Pianoteq / Kawai CL 35 & MP11 / Old 1920's Upright Zimmerman
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
OP
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
The CNY70 uses its own infrared emitter and the phototransistor is sensitive to 950nm light, i.e. to its emitter and the daylight won't affect it. So, it's a proximity sensor that uses its own reflected light to detect proximity. I've been using my controller without any cabinet in daylight and it hasn't been affected
Last edited by CyberGene; 07/07/20 03:24 AM.
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
OK, that explains it. One more question: How are the sensors situated with respect to the action? Do the sensors "see" the hammer heads? The shanks? Or something else?
It was not clear from the photos I looked at.
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
OP
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
OK, that explains it. One more question: How are the sensors situated with respect to the action? Do the sensors "see" the hammer heads? The shanks? Or something else?
It was not clear from the photos I looked at. Sensors are above the bases of the hammer shanks, close to the pin.
|
|
|
|
Joined: Jun 2020
Posts: 646
500 Post Club Member
|
500 Post Club Member
Joined: Jun 2020
Posts: 646 |
Hi CG... Couple of technical questions regarding Cybrid (and small bump to the thread 😊) :
1. Any particular reason for not using Teensy 4.1? (probably not yet available when you started the project). It has a MUCH faster clock, although I am not completely sure if the faster speed affects only internal computations or if it also increases the IO cycle rate (it should...). The overall pinout remains the same (it even has 2 more pins easily accessible).
2. I cannot see in your code an implemmentation for key-off speed release. I assume that this cannot be calculated using only 3 trimpots (note off is the time at which sensor 3 changes state, irrelevant of time between 2 and 3, correct?). If so, then note off requires an additional sensor 4 well above escapement in order to calculate a speed. Which brings me back to the first point: teensy 4.1 may provide the speed necessary for an additional trimpot in the scanning cycle. I have not checked what are the hardware consequences of this change (I assume that 42 IO pins are enough for everything, but the voltage comparators and multiplexes need to change from 3:1 to 4:1... And the IDC probably needs to jump from IDC34 to IDC50 with very wide ribbon cabling...)
Am I thinking correctly? Is speed of note-off overkill? Maybe only Pianoteq will be able to fully use the note-off speed? Moreover, you don't need this to signal silent undamped keys, although again I don't think silent undampening is implemented in MIDI references...
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
OP
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
^ To answer your questions:
At the time Teensy 3.6 was the latest one. 4.0 was later announced but it was not pin compatible. However 4.1 is pin compatible and I received one on Friday, however it turned out it has no pins soldered. I didn’t have long enough pin socket remaining to solder both sides and the nearest electronics store was closed due to COVID and I’m currently on vacation, so will finish it next week and post an update on the scanning speed improvements.
You are correct, note off velocity is not compatible with the three sensor point implementation. This is the same as NV10 which is why the Kawai doesn’t support note off velocity either. If I’m to put four points, I’d prefer an approach with two points at hammer and two points at keys for much more consistent damper detection behavior.
|
|
|
|
Joined: Jun 2020
Posts: 646
500 Post Club Member
|
500 Post Club Member
Joined: Jun 2020
Posts: 646 |
Thanks for the quick reply.
You are always one step ahead of everybody else... Let us know if the teensy 4.1 improves cycle efficacy. And I fully agree with simultaneous detection of hammer movement and initial key displacement (you may even use the key sensor information to "jump over" some note groups and not scan the hammer sensors if in that group of 5 keys all of them are at resting position, and shave a few cycles - unless the "if" commands generate more lag than straightforward dumps).
Wow: the NV10 does not have note-off velocity???? A real grand action with real dampening hardware, but without its digital implemmentation? Are all Kawai keyboards the same? From ES110, MP11 through NV10 (the only models for which I checked the manuals) the MIDI chart state that there is velocity in both note-on and note-off.
|
|
|
Forums43
Topics223,403
Posts3,349,419
Members111,636
|
Most Online15,252 Mar 21st, 2010
|
|
|
|
|
|