Update on Cybrid

Hi @CyberGene

In the last message you sent on this topic you mentioned something like "the N1 is a better piano". Do you care to elaborate more on that and perhaps posting an update on the Cybrid project?

Thanks!

Cybrid has not been at my apartment for months. I miss it from time to time.

Thinking about it in retrospective, I would say it all boils down to the action. It has a pretty old and beaten up action that needs a complete restoration including difficult repairs such as fixing the gapped balance holes... It's not worth it to go in that way though. It is very playable at this stage considering its age but is a no match for the perfect Yamaha N1X action that is also perfectly regulated from the factory (I guess I may be a bit lucky in that though).

However in terms of sensing and electronics, I'd say they are on par which is kind of amazing considering it's just something I made up without prior experience and no, I'm not trying to boast of myself 🙂 On the contrary, I think if something that I made up in our bedroom at night can be so good, then maybe this entire task is not rocket science at all and it's all about how the action feels, not about the sensors. Of course the N1X is also a piano with great sound engine.

CyberGene something that I made up in our bedroom at night

I realize that can sound very odd 🤣 Luckily my wife would often sleep with our daughter in the kid's room at the time. I can't make another Cybrid in the bedroom anymore though 😀

CyberGene if something that I made up in our bedroom at night can be so good, then maybe this entire task is not rocket science at all and it's all about how the action feels

Well, I mean it sounds like things are fine in that department, but this is really OT.

CyberGene I can't make another Cybrid in the bedroom anymore

There must be other rooms, maybe a closet, or a quiet wooded area. Anyway, I had no idea Cybrid was a Bulgarian girl's name; it's very pretty.

LOL, I deserved it, didn't I... 😳

On more serious note, it seems non-rocket science or not, you have developed knowledge and skills that can turn worn acoustics into hybrids, and apparently for much less money than the Japanese hybrids go for. There must be a business model in there somewhere.

QuasiUnaFantasia On more serious note, it seems non-rocket science or not, you have developed knowledge and skills that can turn worn acoustics into hybrids, and apparently for much less money than the Japanese hybrids go for. There must be a business model in there somewhere.

Actually I seriously thought about it (I'm making my own too), because in the USA there is super-abundance of the raw material (old worn acoustic) at a very good price. It can be, but only for a piano store, IMHO. Mostly because of the "accessory" needs of moving around large/heavy boxes (both for the raw material and for the finished product) which the piano stores already have as incidentals and with all the logistics figured out. Starting from zero feels like a big deal

QuasiUnaFantasia maybe a closet

A closet? Come on Quasi… is that a danish thing? Large closets? You need to fit 88 keys, don't you? 😉

vagfilm I'll have you know my first Yamaha keyboard would have fit nicely into a closet. Harry Potter might even have been able to use it in his "room" in Privet Drive. 😄

Thanks for sharing, but you can keep the memories of your first yamaha all to yourself… 😉

QuasiUnaFantasia The business model is there and is already explored. But this is a niche market. People interested in buying a new piano will likely get a model with a silent system installed at the factory instead of retrofitting a silent system. The hardware and software package for a retrofit sells for around 1000-1500 EUR (upright pianos) and 1500-2500 EUR (grand pianos), excluding installation costs (which can double the price). A factory installed system on a grand will be even more expensive. This means you can buy a top range digital piano for the cost of a silent system - for example, you can buy an N1X for roughly the same price as a factory-installed Yamaha SH system on a grand 😀

The main difference between Cybrid and the majority of silent systems on the market is the sensor technology: Cybrid relies on hammer sensing whereas the majority of silent systems rely on (optical) key sensors. Afaik, only the Yamaha SH (available in grand pianos only) uses hammer and key sensors. If the actual results of hammer sensing are somehow "better" than key sensing, then surely there is an opportunity to enter this niche market: a silent system manufacturer could provide the same controller with two technologies: key sensing and hammer sensing.

Anyway, the market for retrofit silent system is quite crowded, especially given the small number of potential customers. Systems from AdSilent, Genio, MagicStar and Korg have been around for years. And there are plenty of newcomers boasting several GBs of samples in their sound engines, such as Motus and Kioshi. Yamaha has also been selling retrofit silent systems (such as the RSC2) for a long time, but I believe they are only available in Japan. Moreover, brands continue to actively develop products and introduce new features. For example, AdSilent recently announced a mechanism for grand pianos that intends to achieve the same results as Yamaha's Quick Escape (this keeps the regulation of the action in acoustic mode completely unchanged; with other silent systems the regulation of the action in acoustic mode needs to adjusted to accommodate the silent system). But I am not sure if AdSilent can retrofit such mechanism without messing up with the existing action of the piano.

A majority of piano manufacturers do not develop their own silent systems but rebrand existing systems and update the controller software with a new user interface and sometimes a different sound engine. For instance, Schimmel used the hardware from the old Yamaha SG/SG2 systems in their silent pianos until quite recently. But the current Schimmel TwinTone on grand pianos is similar to Yamaha's in the sense that it keeps the acoustic regulation unchanged (this is done by moving the rail where the set-off buttons are mounted when the silent system in enabled, which changes the distance between the jacks and the rail and allows keeping the regulation in acoustic mode unchanged). There were rumours that Kawai introduced a similar system in their silent grands with the new ATX4, but afaik this was s not confirmed nor announced by Kawai.

If I were @CyberGene, I would certainly contact one of the big names, like AdSilent, as well as some of the smaller players such as Motus and Kioshi, to try pitching the idea of including hammer sensing in their product line. Nothing to be lost 😀

arc to try pitching the idea of including hammer sensing in their product line. Nothing to be lost

Those companies don't need pitching, because they only need 5 minutes of RD to implement a hammer sensor tech… Not trying to downplay CG development (which I have extensively praised privately and publically) but those companies prefer the key approach because it is invisible to both the owners and the guests. Hammer sensors will always be visible in your 50k grand, and no customer would want that if they have an invisible alternative that is 99% as accurate as with hammer sensors. The programming needed for key sensing is more complex and error prone than the hammer sensing (remember the nu1 issue with silent notes?).

vagfilm Those companies don't need pitching

True. Piano makers are not interested in outside inventors. And they've already tried hammer sensors.

MacMacMac Which silent systems designed to be retrofitted include hammer sensors instead of key sensors? I am only aware of hammer sensors in factory-installed systems.

vagfilm The programming needed for key sensing is more complex and error prone than the hammer sensing (remember the nu1 issue with silent notes?).

The software handling key sensing is not more complex than hammer sensing. It is just different because the events captured by the sensors are different. The limitations of key sensing in silent systems mostly derive from using contactless (optical) proximity sensors instead of contact sensors, like the ones used in DPs. Hammer sensors also have limitations, such as the inability to detect silent key presses (which affects resonance), key release, and key-off velocity (which affects resonance as well as the proper simulation of the sustain and release envelope of the note). This is why Yamaha combines hammer sensors with key sensors.

But I couldn't agree more that the difference is minimal. Moreover, the silent system retrofit market is ridiculously small. Anyone planning to buy a acoustic piano with silent system will surely buy one with the system installed and regulated at the factory. In my view, retrofitting such a system on a piano makes very little sense unless the goal is to create a MIDI controlled like the Cybrid - but then the market is limited to a handful of enthusiasts…

arc But I couldn't agree more that the difference is minimal.

Depends on what you mean by minimal. For example, a hammer sensor will never produce the "loud note" issue like the NU1 had. A key sensor will always do unless software algorithms are utilized to detect the problem and correct it, leading to possible false positive (actual notes which are supposed to play loud being muffled).

So hammer and key sensors are very different, in my view. A real piano has both: hammers and dampers, and a hybrid can easily to have both too (I am attempting to do so by Cybrid-inspired https://github.com/jkominek/piano-conversion/ )

The problem with hammer sensor is that it's extremely hard (impossible?) to make it a retrofit system because the strings are in the way!

FWIW, I am not sure how the discussion switched to retrofit system. As stated by most of you in that digression, there's fierce competition for the small market in retrofit system, not a good business. Where the "potentially good" business lies is the digital-only (hybrid) models, to compete with the N1 and the NV10 (in a moment a discussion about the NU1 and NV5). The market for those is much larger than retrofit systems. The action and cabinets can be found (in the USA) at flea market prices. The needed electronics is also inexpensive. It's "just labor cost" to refurbish the action, modify the cabinet (you want only the front of the grand piano) and assemble the electronics. In my opinion (I am doing it), the final cost to the public could be 50-70% of the street price of the N1 and NV10 competition, if you can make "economy of scale" work (of course as a DIY project I am both doing "more" -- in labor without economy of scale -- and less -- since I don't pay myself to do it, but use spare time). The most expensive aspects are (as I wrote earlier) the moving heavy/large items on short notice. An existing piano store with such a logistics already in place can do it easily, someone starting from scratch would have big troubles.

The upright NU1 and NV5 like is also a possibility, however it's not so appealing: the final price must be less because so are the prices by Yamaha and Kawai. The involved work is more because you do need to remove the plate from the acoustic instrument, which is more complicated and time consuming than in a grand. There is a part of the work that is less because you could use the cabinet "as is" without major woodworking that you need in grands (chopping the front and dumping the tail), but that work is not that much of labor. So for uprights the total labor cost is more. You end up with instruments which are too large (and heavy), so less appealing to consumers (unless you go with spinets, but then the action is crap, so it becomes a moot point). Hence you spend more, have a less appealing product and the competition has lower price: no business.

In the long term, if everything goes well with my project (I will post some pictures one of these days) I may contact one of my favorite piano stores and see if they are interested.

RIP Where the "potentially good" business lies is the digital-only (hybrid) models, to compete with the N1 and the NV10 (in a moment a discussion about the NU1 and NV5). The market for those is much larger than retrofit systems. The action and cabinets can be found (in the USA) at flea market prices.

Sure, this is certainly a better market. But I think you are severely underestimating the costs and effort of adapting the cabinet of a grand. If the goal is "competing" with hybrids and top-range digitals , then you need to add a cabinet that can be placed in a living room. Or are we talking about an action with exposed wires that is supposed to be placed on top of a workbench? You need to cut the cabinet, remove beams, soundboard, strings, and plate , then add a new wooden back cover, new lid, new legs, and then paint/lacquer the whole thing. This process will be more expensive in terms of labour and materials than all the sensor hardware put together. Not sure how you can say that this is not much of labour and how adapting an upright is simpler - that does not seem to be case since you just need to remove the plate and strings while keeping everything else as is.

RIP Depends on what you mean by minimal. For example, a hammer sensor will never produce the "loud note" issue like the NU1 had. A key sensor will always do unless software algorithms are utilized to detect the problem and correct it, leading to possible false positive (actual notes which are supposed to play loud being muffled).

But that is because how key sensors are implemented. Digital pianos do not have this issue. The majority of key sensors in use rely exclusivity in proximity sensing and with that single variable there is not enough information to discriminate all possible events. It can be easily sorted out by adding an additional sensor, but no one does that because this is a microscopic market and this would increase costs.

RIP The problem with hammer sensor is that it's extremely hard (impossible?) to make it a retrofit system because the strings are in the way!

Why are the strings in the way? Even when using key sensors the system needs a (moveable) hammer stop rail placed between the hammers and strings. The stop rail actually needs more vertical space than the sensor rail. Pictures below show Yamaha's SH and Kawai's IHSS hammer sensor rail. There is a significant gap between the rails and the strings. So, what is the actual problem?

Yes, there is space for optical sensors of the hammers, just like in Cybrid (although it was easier in cybrid because it is devoid of strings).

Arc: if you actually spend some time thinking about this, you realize that the computation of hammers vs key sensors is different because you have to calculate if the key movement actually translates into real hammer movement and what value of hammer movement (the same key displacent leads to different hammer velocities in bass or treble because the hammers heads have different weights and thus different momentum; moreover, fast repetitions use kinetic movement of the hammer, which you can interpolate very roughly frkm the key velocity. Not easy nonlinear calculations that are done with a large margin of error.

When sensing the hammer, the actual movement is the actual movement: no correlation is needed. All the nonlinearities are done by the action itself and you just observe and measure…

arc You need to cut the cabinet, remove beams, soundboard, strings, and plate , then add a new wooden back cover, new lid, new legs, and then paint/lacquer the whole thing.

I've done that part already. Almost completed.

arc But I think you are severely underestimating the costs and effort of adapting the cabinet of a grand. If the goal is "competing" with hybrids and top-range digitals , then you need to add a cabinet that can be placed in a living room.

Exactly. I am doing quite that. I am not underestimating, I know exactly what it takes. Be patient and you will see the pictures.

arc his process will be more expensive in terms of labour and materials than all the sensor hardware put together.

That is exactly what I said. IIRC CyberGene said he's spent a few hundreds dollars in material, let's assume a few more hundreds in labor (probably less once you have economy of scale) and I've said flea market cost for the instrument itself. So at most we are at 2000 dollars (typing the dollar sign makes the editor crazy ☹️)
I then said that such an instrument will sell at 50-70% of an N1 or NV10, and you see the huge gap here, which is all in this work, if the logistic of "moving boxes" is there already (as it is for an existing piano store). Not free, mind you, but cheaper than for a new biz. So we agree here.

arc Not sure how you can say that this is not much of labour and how adapting an upright is simpler - that does not seem to be case since you just need to remove the plate and strings while keeping everything else as is.

Your sentence is not clear to me. If you are saying that

• removing the plate and strings from an upright is easier/cheaper than
• removing the plate and strings from a grand and then "chopping off" the tail of the grand

that is incorrect. The grand is easier. I know because I've done both. Moreover, both will need new surface refinish which will be easier on the grand (because it's easier to disassemble), but I am not including this aspect in my comparison above

arc Why are the strings in the way?

I've been sloppy in my sentence, but the problem of a retrofit system is that it needs to work on variety of different instruments which are all different in scale design and hammer placement. Plus what you describe is a hammer-shank sensor, not the hammer sensor. Very similar but needs better accuracy and resolution, hence it's more difficult to assemble than a proper hammer sensor. I could elaborate more, but the bottom line is that I think there is no business model there. If you think there is, go for it, I wish you good luck 😀

RIP The grand is easier. I know because I've done both.

Hi Delvento. Can you ellucidate since you have the experience with uprights. When CG presented his prototype, I was plannig to adapt it to an upright (there are so few grands in Portugal, that there is no second hand market - in fact there is: people ask 5k for a non functioning grand…). Decent second hand uprights are easier to find.

My idea was to keep harp and strings in place, just asking a technician to release the tension of the strings and dampen and secure them with felt on the back of the piano (soundboard). Optical sensors would be placed like in cybrid at about the level where the strings were previously.

Do you think this is doable?

I abandoned the idea when I realized that an acoustic action is a CLACKety mechanism that would drive me to divorce (or homeless or both)… of course, I could better isolate the cabinet with foam or some other dampening material because the soundboard would not be necessary, but even then…