Also, depending on the design of the system, sensors may or may not be able to be replaced individually, and so one silent or full-volume key can require replacement of at least an octave or the entire keyboard which is another cost to consider.

aandrmusic wrote:

I think two sensors SHOULD be enough to derive key speed.
Key speed measured at two different intervals, i.e. between sensors 1 & 2, then again between 2 & 3 could represent acceleration, but I think final speed is the only thing that matters.

I'm thinking more parts usually means more failures and more co$t, so if the benefit is only theoretical - why would you want 3 ?

Because you want it to repeat notes like a grand piano.

Yes, and again that CAN be done with 2 sensors - - SPEED matters, acceleration doesn't.
PLACEMENT of sensors matters more than NUMBER of sensors.

"More" parts (for the sake of spec bragging) usually loses out to better (simpler, more elegant, better executed) design.

Judging from the video, the opposite end of the key stick is shaped very differently (perhaps to simulate more weight and action movement in a grand) and the sensors appear to be differently positioned and perhaps a different design of sensor altogether.  Unfortunately, the video doesn't get close enough to the action model to read what it says.

From the Kawai CA99 US Product Page: https://kawaius.com/product/ca99/

To provide outstanding performance, the upgraded Grand Feel III keyboard action with Ivory & Ebony Touch key surfaces is combined with an improved Shigeru Kawai SK-EX Rendering piano sound engine (featuring multi-channel sampling and resonance modeling).

NEW: Grand Feel III wooden key keyboard action with Ivory & Ebony Touch key surfaces

Grand Feel III (GFIII) action

- 88 wooden keys
- Grade-weighted hammers
- Ivory Touch white key surfaces
- Ebony Touch black key surfaces
- Let-off simulation
- Triple sensor key detection
- 88-key Graded counterweights

From https://www.kawai-global.com/news/kawai...al-pianos/

Articulating this beautiful SK-EX sound is the brand new Grand Feel III wooden-key keyboard action, which draws upon Kawai’s 90+ years of acoustic piano craftsmanship to provide an exceptionally realistic playing experience.  As with a grand piano, all of the Grand Feel III action’s eighty-eight black and white keys are crafted entirely from long pieces of wood, and pivot on a central balance pin with a smooth, seesaw-like motion.  This natural movement throws grade-weighted metal hammers located at the back of the keyboard, with their individual velocities precisely measured using triple-sensor technology for improved responsiveness and realism.

The Grand Feel III action’s extended wooden keys also boast an extra-long pivot length – matching that of a Kawai acoustic grand piano.  This provides a more consistent touch weight when playing towards the rear of the keyboard, while additional counterweights placed below each key help to further balance the weight of the action during pianissimo passages.  Finally, the Grand Feel III keyboard action also features Kawai’s Ebony & Ivory Touch key surfaces that gently absorb moisture to assist playing control, and possess a natural, matte finish that is smooth, but not slippery.

2020 NAMM Video: https://www.youtube.com/watch?v=3hYY7rz-b8g

It's also possible that James Pavel Shawcross did a video of Kawai's booth at NAMM that will show up on Youtube soon.  He usually will get into more technical details than the promotional videos.

DonSmith wrote:

Last week a review triggered a thought about three sensors. I think it might have been on the Nord Grand, but I saw quite a few. I wondered if Keyboards would get to the stage where the senors development had more grades than the three.

A couple of years ago when I bought my keyboard I wasn't even aware of the sensor technology until I saw a review on it, luckily I seem to have the tree as I don't have to let the keys come all the way up.

https://www.youtube.com/watch?v=LmD2_9APBaE
https://www.youtube.com/watch?time_cont...NxacJZXPic

FYI the Nord Grand uses an old Kawai RM3 action. For what it costs, I'd just get a Kawai CA99 which has the new Grand Feel III action... three iterations newer and improved over the RM3. RM3 --> GF --> GFII --> GFIII --> Millenium III (only in Novus NV10)

Are you sure Grand Feel 3 exists? I haven't found any info about it. Only GF Compact and it isn't next generation of Grand Feel.

Last week a review triggered a thought about three sensors. I think it might have been on the Nord Grand, but I saw quite a few. I wondered if Keyboards would get to the stage where the senors development had more grades than the three.

A couple of years ago when I bought my keyboard I wasn't even aware of the sensor technology until I saw a review on it, luckily I seem to have the tree as I don't have to let the keys come all the way up.

https://www.youtube.com/watch?v=LmD2_9APBaE
https://www.youtube.com/watch?time_cont...NxacJZXPic

FYI the Nord Grand uses an old Kawai RM3 action. For what it costs, I'd just get a Kawai CA99 which has the new Grand Feel III action... three iterations newer and improved over the RM3. RM3 --> GF --> GFII --> GFIII --> Millenium III (only in Novus NV10)

I think two sensors SHOULD be enough to derive key speed.
Key speed measured at two different intervals, i.e. between sensors 1 & 2, then again between 2 & 3 could represent acceleration, but I think final speed is the only thing that matters.

I'm thinking more parts usually means more failures and more co$t, so if the benefit is only theoretical - why would you want 3 ?

Because you want it to repeat notes like a grand piano.

A couple of years ago when I bought my keyboard I wasn't even aware of the sensor technology until I saw a review on it, luckily I seem to have the tree as I don't have to let the keys come all the way up.

https://www.youtube.com/watch?v=LmD2_9APBaE
https://www.youtube.com/watch?time_cont...NxacJZXPic

I read in one of the reviews that it [PX-S1000/3000] always knows the position of the keys, part of their new smart features.

If the Casio "always knows the position of the keys", why using dual-sensors then? (would be redundant information).

Is there any more detailed information, how this smart sensing of the position works? (if it exists).

I don't think it would be at all difficult to determine this and to get it "right".
Where "right" is good enough for all but maybe the world's top three concert pianists.
The rest of us... three vs two sensors may be the least of our problems.

OTOH, the marketing appeal of "more is better and three is more than two"
is always hard to argue against - the masses will usually opt for "More".

If the top sensor is located such that it senses at the right point (maybe ~30% of full key depress ?) then it will "catch" rapid reps.

Maybe "~30% of full key depress" is too optimistic. In another thread I measured the trigger-points of my Kawai ES3:
------------------------------
0 mm (0 %)

5 mm (42 %) note-off

9 mm (75 %) note-on

12 mm (100 %) full key depress
------------------------------

In my opinion this AHA3-action is not good enough for fast repetitions. And this is with 42 % full key depress. So my estimation more tends to >50 % than "~30% of full key depress".

But this is an old Kawai action from 2003, maybe newer double-sensor actions have evolved in the last 16 years.

aandrmusic wrote:

I think two sensors SHOULD be enough to derive key speed.
Key speed measured at two different intervals, i.e. between sensors 1 & 2, then again between 2 & 3 could represent acceleration, but I think final speed is the only thing that matters.

I'm thinking more parts usually means more failures and more co$t, so if the benefit is only theoretical - why would you want 3 ?

I have a Yamaha CP4 Stage which has a triple sensor wooden key action. The benefit of the 3rd sensor is for repetitions of a partly pressed key, similar to how it is on an acoustic grand piano.

I understand that - and the "ASSUMPTION" that the third sensor is "needed" is based on another assumption, which is that the first (or "top") sensor is placed too high.

If the top sensor is located such that it senses at the right point (maybe ~30% of full key depress ?) then it will "catch" rapid reps.

I think two sensors SHOULD be enough to derive key speed.
Key speed measured at two different intervals, i.e. between sensors 1 & 2, then again between 2 & 3 could represent acceleration, but I think final speed is the only thing that matters.

I'm thinking more parts usually means more failures and more co$t, so if the benefit is only theoretical - why would you want 3 ?

I have a Yamaha CP4 Stage which has a triple sensor wooden key action. The benefit of the 3rd sensor is for repetitions of a partly pressed key, similar to how it is on an acoustic grand piano.

I'm thinking more parts usually means more failures and more co$t, so if the benefit is only theoretical - why would you want 3 ?

Since the PX-S3000 has a pitch bend wheel and two rotaries that can send CC, I am now looking at it as a new controller.  Nobody in my area has the PX-S3000 in stock, but I did find a PX-S1000 in stock.  Same keybed in both, S3000 has 700 internal sounds and the pitch bend and two rotaries that S1000 does not, plus a better menu system/user interface on the S3000, still a little lacking in that aspect though.  So I went and auditioned the PX-S1000 to see if I would like it.  I was absolutely amazed with the action.  I also tried repeating notes (fast and slow) without letting the key come all the way up.  It played the repeated note each and every time and as many times as I wanted it to, so retriggering was no problem with it. I read in one of the reviews that it always knows the position of the keys, part of their new smart features. For myself, I think I have found the new controller that I am going to buy.  I just found it interesting that they went from three sensors back to two and are able to achieve better results.

I don't think it makes much difference, sonically. The main thing is that the action allows relatively shallow repeats - i.e - the keys don't have to be released very far, before they can be played again.

However, some two-sensor actions have the top sensor rather low down (at almost the same vertical position as the middle sensor of a tri-sensor action), so the whole tri-sensor thing is very subtle indeed I suspect.

Note that even if consecutive Note-Ons ARE transmitted, you may find that your DAW will convert that sequence into Note-On/Note-Off pairs.  The MIDI recorder in Pianoteq is raw, and will not behave like that, as far as I can recall.

Greg.