Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Dynamic levels of Kawai CL36

No, I don't see how it could be. Such a piano would play horribly. Let's imagine that there are only two sensors and they are super-close together at the bottom of the stroke. Push the key down, everything works just as you would like it to. Now lift the key up. The note cuts out after a certain delay.

The problem is, we don't always lift the key at the same rate. And in general, we may not hold the key all the way down, but we don't want the note to cut off until the key is half way up and the felt comes into contact with the strings. Careful releases of the notes would be essentially impossible because the notes would die a preprogrammed amount of time after the key is lifted off the bottom. And how would such a piano sound when played stacatto?

Some players may release keys slowly so that there is essentially the same effect as a half pedal (the dampers are partially in contact with the key and kill the sound, but not immediately). How would your design know when and in what manner these dampers had been released?

What if we push the key down and then inadvertently lift it just a hair (but enough to pass the second sensor) and then push it back down. Would this be an unwanted repeated note? Would the damper silence the first note after the second is played (because the damper is on a timer)? Any way you cut it, the behavior will not match that of an acoustic.

No, the problems associated with putting both sensors at the bottom of the stroke are very much larger than the problem of having the first sensor be half way through the stroke. The sensor associated with the damper is critical to a functional piano.

Agree. Didn't think of holding a key. But then the third sensor is necessary.
But still, I'd prefer to have the top (or the second - in three-sensored keyboards) sensor not higher than 30% from the bottom. Otherwise, correct dynamic in repeats and thrills is unachievable - they will be loud, 'cause you are to play quickly, and there is no way to get pianissimo, if the velocity is average.

Yeah, you are right. We need three-sensor actions (with the sensors in the right place) in all our pianos.

Could anybody make the measurements? Please!!!

I have a kawai CA63, and from what I can see, there are just one sensor for each key, continuously sensing the hammer position .... I think that this is vastly better than N sensors for each key.

How would it sense hammer action if it isn't measuring time events (which would imply more than one sensor)? That would have to be a very elaborate system - or else just pressure measurements, as far as I can see.

I don't really know ... as you suggest, I think it is some sort of pressure sensor. Theoretically, at any time the system could know directly the position and estimate its velocity, not calculating the position/velocity from just 2 or 3 time events. Another question is what is being measured, the key travel or the hammer travel.
What I wanted to said, is that maybe the sensors in others systems also are more complex. It is not just the number of sensors, what is being measured is even more important.
I don't think that your testing DP procedure would render meaningful results on my CA63. If I press a key slowly it will not produce any sound, and if I press a key to middle and stop, then press hard, it will sound ...

That would be nice, but it's not the case. Not for any major digital piano. What you have is a rubber thing with two prongs that trip a mechanism similar to the one inside a computer keyboard. Since one prong trips its switch earlier, the DP calculates the velocity from the time between.

All Kawai pianos without exception have two sensors. All major digital pianos without exception have either two or three. The only sensors I know of that are "better" are on the AvantGrand. They are also three discrete on/off sensors, but they are optical, rather than mechanical. Also, two are on the hammer and one on the key.

Continuous position monitoring is a pipe dream in the world of digital pianos.

(the possible exception to the above is the PNO scan midi strip that can be installed in acoustics. We don't really know how it works in detail, but since it is unable to measure hammer velocity as well as key velocity, it probably does not work well--all major hammer-action digital pianos have the sensor tripped by the hammer, not the key.)

Also not correct. No major DP has a pressure sensor. The sensor setup of essentially all digitals is shockingly similar to each other. The three sensor actions simply have three prongs instead of two.

As I mentioned in my last post, all major hammer-action pianos have the sensors tripped by the hammers, not the key.

Also the testing procedure you suggest will not work absolutely does work, including on your model.

These speculations you are making are not really in the realm of things that are subject to speculation because we already know how DP's work, including the CA63, which has one of the most looked-at and studied actions of all digitals.

If you want to know how you sound, imagine going to an automotive forum and saying "My Cadillac has a jet engine and shoots fire out the back. It's a much simpler and better design than the internal combustion engines in other cars." It sort of stinks up the forum with its wrongness and makes you seem very silly to anyone who does know how it works, even a little.

In short, it's better if we restrict our posts to either questions, opinions about subjective things, and true statements about matters of fact. Yours in this thread have been none of the above.

Sorry. I said that "I don't really know", it was my opinion, and maybe you are right and I am not. Thank you very much for the information. My conclusions comes after inspecting my own keyboard, but I did not remove the rubber thing, so maybe it is true what you said. Maybe you can give some references.



Ok, I'll delete my post then.

That sounds a bit harsh, gvfarns. It sounds to me as if Hardwired was only trying to get to the truth after doing a bit of investigation on his own keyboard.

I have tested as Tesir suggest, (with organ sound now, I miss that part).
The key height is 2cm, at 1.1 sound start and stop at 1.5.

Yeah, you are right, it was a bit harsh. Sorry, Hardwired.

To explain what happened, at first I just wanted to nip the wrongness in the bud, but also, to me at least, it seemed that there was already plenty of information in this thread to preclude what he was asserting so I felt that he wasn't reading sufficiently before posting. After the first post telling him he was wrong, he posted again continuing his wrong line of thinking. It turns out we were just writing at the same time, so he did not see my first post. I do think he would have done better to have asked rather than stated how his DP worked, but that doesn't mean he needs too much public embarrassment.

Thank you, Hardwired! This means, they are 0.5 and 0.9 from the top of the key. I guess, the whole travel of the key is 1 cm, isn't it?
But I also suppose, the way of measuring I suggested, may be incorrect for RM3 Grand of CA63. Its sensors are related to the hammer to more extend, than in PHA. This fact may influence dynamic and let the RM3 be more acoustic-like. But I don't know exactly.
Be more precise, as I see, sensors in GH, GH3, PHA II and III, RHA answer the keys action, not the action of hammers, as it should be. That is why their dynamic differs grately from APs, and what should be measured is the velocity at the very end of a key travel.

And what happens, if you press a key slightly bellow the level of the top sensor (at about 6 mm from the top of the key) and then stop, after somme seconds - press hard?

I cannot agree. If the hammers were alike those on APs (or, at least, if they could move separately), then this statement could have sense. But till the hammers move with the keys, it doesn't matter, what trips the sensor - the hammer or the key.
It seems, the hammer in RM3 can move separately, but, I'm afraid, this separate movement is too insignificant to have value for velocity measurement.

I don't know what you are disagreeing with here. I didn't say acoustic and digital hammers were alike, only that the sensors are associated with the hammers in all major digitals. That is an indisputable fact, so there's not really any room for disagreement.

However, I should point out, while we are on the subject, that the insignificance of the separate movement is not clear. In fact, it seems likely that it is just like an acoustic in this respect. Digital actions differ from their acoustic counterparts in that there is no true letoff, yes. That's not particularly relevant to the issue we are discussing here, though. You notice the letoff in an acoustic because the key does not press up as hard after it has completed its swing, but that happens after the velocity of the note has been fully determined. That is, the difference between acoustics and digitals affects how they feel, but not how they perform.

Let's be more specific: The difference between acoustics and digitals that you have suggested (if I read you right) would happen when you hit a note hard but stop the key before it bottoms out. Then the momentum would carry the hammer up to hit the string anyway. You have suggested that this does not occur meaningfully with RM3, but to show this you would need to demonstrate that the same strike that leads to a note on an acoustic would not lead to a note on RM3. Since most of the RM3 momentum is also in the hammer (the key itself is mostly balanced on the fulcrum as it is on an acoustic and certainly isn't any heavier), I am skeptical that there is a meaningful difference between RM3 and an acoustic in this respect. I think it's more likely that you are overestimating the ease with which this happens in an acoustic, underestimating the ease with which this happens in a digital, or both. Possibly you are confusing the letoff feel with something that actually affects the note velocity.

I've made plenty of hard strikes with no follow through that cause a note to fire on my Kawai, even though the key bottom out is small or not present. For this reason, I think you are probably wrong.

Kawai's wood action allows this kind of key-hammer separation, as you also point out, but I don't know for a fact about the other actions around. I could easily believe that many or most of them do.

The lack of letoff does lead to a significant difference in the feelof an acoustic and a digital, but as far as I know, the hammer-key separation issue you reference is not part of that difference, at least in RM3.

Was there another way in which you believe velocity measurement to be impeded by the form of the separation mechanism in RM3? I would hate to mischaracterize what you are saying, as you have done with what I have said.

Am I reading this right?
"... the sensors are associated with the hammers in all major digitals ..."
The sensors on mine are associated with the keys, not the hammers. I think that's true of all the Yamahas (GH/GH3/NW).

We looked at this a while ago and I recall looking at the actual teardown and diagrams and the Yamaha sensors were indeed associated with the hammer. At the time I also found an older thread in which people took it apart with the purpose of determining this very thing (as well as the distance between sensors, etc). They also concluded that the sensor was hammer-triggered.

I'll dig around and see if I can find the thread. It's also possible that we made a mistake. If you can find clear evidence that it's other than as I said I will be grateful at having learned something.

Of course, it's all stuck under the key so it's harder to tell than it is on Kawai's wood action. And as has been pointed out, I didn't check whether you can separately move the hammer, or whether the fact that it's the hammer that triggers the sensor is irrelevant to the way it plays.

Yes, a bit more, because the key bottoms softly and you can press a little more.



I don't know how GH, GH3 or PHA works, but I think that it is more realistic to measure the velocity of the hammer, because in a real piano the hammer is what hits the string.


Yes, it is difficult to measure, but it seems that there is a point, below that it is not possible to trigger the sound, but I can not see now what happened with the hammer.
On the other hand, I can block the key with one finger, and knock the key with a knuckle of the other hand, such that the key do not travel almost nothing, but the hit is sufficient to make the hammer jump and make the sound. The key and the hammer are disconnected. Maybe you can try this on other types of keyboard to see, if I am not wrong I hope, only if the hammer is disconnected from the key, and the sensors measured the hammer travel the hit will make sound.
See this vid http://youtu.be/yxVGDdvkXhs at 2:30