Hi,
Just wanted to introduce myself as a new Pianoteq4 (Stage) user - very, very impressed with this so far, especially the new Blüthner model, which is nothing short of incredible!
Having been struggling with Ivory II Italian Grand and its infuriating velocity switching, I finally took the plunge a couple of weeks ago, and will not be looking back.
Congratulations to the Modartt team
-Ben
Hello Ben and welcome to this friendly forum!
As a former user of sample libraries from EWQL Pianos, Ivory (1), BDMO, and many others, I was also frustrated by the breaks in tone and volume with all of the velocity switched libraries. In addition, several of the tones were out of tune at one velocity level, and IN tune at the next higher velocity level, and then back out of tune in a still higher velocity level! Apparently the various pianos needed almost daily (hourly?!) retuning when they were being sampled so heavily.
I specifically remember having to record my own live performances into midi, using it as a kind of pseudo-tape recorder, and then post-edit those notes whose velocities were 85-90, back to velocity = 84. It does no good on my part to expose any particular vendor, because similar occurrences happened in all of the high-end piano libraries. A great deal of my time was spent adjusting, and then re-adjusting note-on velocities after the fact -- which tended to sap the life from the actual finished performances!
The problem of velocity "jolts" was not limited to software libraries, as similar (un)qualities were present in at least three earlier hardware boxes of pianos from Roland, Emu, Yamaha and others.
Once having acquired Pianoteq at the time that Version 3 became available, and discovered its immense playability, I removed all of my piano libraries from various hard drives on my computer. Yes, I have gone back and reinstalled two of the sample libraries -- only to completely de-install them for a final time. I have never looked back.
In my own experience, the true "playability" of Pianoteq's various instruments and versions have trumped any shortcomings of the sound, about which other people have complained in the past. The new Blüthner instrument is marvelous, both in terms of sound AND playability.
Welcome, again, to this forum.
Cheers,
Joe
Hello Ben, Joe (and Modartt),
I've been a long-time Pianoteq customer (since version 2) but have never posted anything on the Pianoteq forum. I recognize Ben's experiences with Ivory, so I thought I'd post something to share my enthusiasm. Like Ben, I was finally converted definitively by the new D4 and Bluthner models. Both are amazing technological achievements, I think, and especially the Bluthner is simply addictive to play.
With previous Pianoteq versions, I did feel there was a certain lack of realism to the basic piano tone, although I loved the responsiveness and playability and was convinced that physical modelling was eventually going to surpass sample-based methods. I think with the D4 and Bluthner Modartt have achieved that. My only gripe with Pianoteq is the microphone modelling. I find myself mostly preferring the Stereophonic mode to the Sound Recording mode, using the reverb to create a sense of space.
I bought Ivory 2 last year, mostly on the basis of the rave reviews it got, but have never been able to get along with it, not even with the Steinway (the Boesendorfer and Yamaha I think are not very good at all). It's OK to listen to as a recording but uninvolving to play; it feels like I'm constantly fighting the limitations of the software and the sample sets. Especially the velocity response doesn't feel right and Pianoteq is far smoother in this respect. I was interested (and a bit alarmed) to read about Ben's similarly experiences with the Ivory 2 Italian Grand.
The tone of Ivory 2 too lacks depth and warmth. Comparing Ivory II's Steinway to the D4 simply by playing middle C at various velocities, I was very surprised that the D4 actually has a warmer, fuller and richer tone than the sampled Ivory Steinway. The sympathetic resonance modelling in Ivory 2 is also thin compared to Pianoteq, and contains a weird metal zing that sounds unlike anything a real piano would do. The release samples are also less than convincing. There's audible reverb from the room in which the samples were recorded, and this gets cut off in an unnatural manner. In Pianoteq you hear the resonance of the piano body slowly fade away after you release the key, exactly as in a real instrument. This seems largely absent in Ivory 2, and as a result the notes don't always connect well. This is audible, for example, in the opening notes of the Chopin scherzo which Synthogy have as a demo on their website: the opening phrase sounds thin and unconnected to me. Even the new Steinway Concert D suffers from this, judging by the online demos.
Apologies for this anti-Ivory rant but it's a bit frustrating to buy a product that's been so well received and then find out it's not all that great.
Hats off to Modartt for creating the D4 and the Bluthner. Pianoteq will be my software piano from now on. I look forward very much to what else is in store. Considering the leaps which Pianoteq has made since 2007 (when I bought version 2), the future looks very bright indeed.
Happy playing to Ben, Joe and all other Pianoteq fans.
Jan
"Pianophile", you are welcome if do not abuse ony any baby grand.
But can take freedon complain about sampled libraries.
After get used to pianoteq response and prime velocities and sympathetic resonance, it's normal to create a dislike for most sampled libraries.
You are right, pianoteq basic tone is much better than previous versions, specially the new Bluthner add on.
"Pianophile", you are welcome if do not abuse ony any baby grand.
But can take freedon complain about sampled libraries.
After get used to pianoteq response and prime velocities and sympathetic resonance, it's normal to create a dislike for most sampled libraries.
You are right, pianoteq basic tone is much better than previous versions, specially the new Bluthner add on.
Hello Beto-Music, Pianophile, Ben, and all others,
It seems ironic that I would be one to complain about sampled piano libraries, because I also did make demos for the EWQL Boesendorfer and Steinway. Talk about "biting the hand that feeds you!" My irony is that I am not simply in the business of making demos in exchange for sample libraries -- I absolutely love the technology and mechanics behind real acoustic pianos, and am fortunate to have real acoustic Steinway and Kawai grands at my daily disposal.
After having acquired Pianoteq since Version 3, I simply have no desire to even try the Garritan Authorized Steinway or even the VSL Boesendorfer Imperial, much less making demos for those software creators.
In summary, I wish to paraphrase keyboardist Keith Emerson of ELP, who was once quoted about playing an electronic copy of a real pipe organ:
"In comparison to Pianoteq Version 4's D4 and Blüthner, every sampled piano library (to me) is the electronic piano equivalent of .... er, f***ing ... through a condom."
Enough said!
Cheers,
Joe
I would like to test Bluthner add-on with a controller like Yamaha Avantgrand or a real piano with PNOscan.
A natural sound with natural key touch probably induces our brain to believe it's closer to the real thing.
One question to everyone here:
Would you buy a real piano action midi controller (just controller and no onboard sounds) for something like of 1600 dollars?
I know low such price do not exist, but let's supose some manufacturer built one and sells for that.
Welcome Ben
A friendly welcome from me, too
@beto music:
Yes maybe I would have bought one, but most controllers on the market don't have much freedom when it comes to adjusting velocity sensitivity (ony three different presets) and many of them won't even work for the entire midirage (1-127), so maybe I would still have started my DIY project.
BTW: has anyone tested the avant grand with pianoteq and had a look at the midirange?
Currently I'm on a good way to acquire an old !grand! piano action, which hopefully will run smooth with my sensor bar.
The best way to achieve perfect response would be to measure hammer velocity rather than key velocity, but there is no standard for the grouping of the hammers in a piano, so I fear there will only be the high priced out of the box sollutions like the one from kawai, that will use hammer sensors.
finally I think something like the avant grand without internal sounds for around 2000 € would be really impressive!
I would like to test Bluthner add-on with a controller like Yamaha Avantgrand or a real piano with PNOscan.
A natural sound with natural key touch probably induces our brain to believe it's closer to the real thing.
Beto-Music,
This is certainly my experience. I don't have PNOscan but, like sebion, I designed and built a diy optical sensor strip and installed it in my acoustic piano a couple of years ago.
The piano itself is nothing special - a Rippen upright piano built in the late 1980s. I don't have a hammer stopper to prevent the hammers hitting the strings, but the piano does have a muting pedal which lowers a layer of felt between the hammers and strings. This mutes the sound sufficiently such that when I listen to Pianoteq through headphones it is the sound of Pianoteq that I hear rather than the sound of the acoustic piano.
When I play on this setup I certainly feel that the sound is actually emanating from the piano itself and it feels totally real. I'm sure that a large part of this effect is due to the fact that the acoustic piano keyboard feels more 'alive', due to a combination of the feedback from the real hammer action and the residual vibrational feedback from the muted instrument via the keyboard.
In fact, I actually prefer playing this setup through Pianoteq and headphones rather than listening to the un-muted acoustic piano (which, as I say, is nothing special). This is especially true since the new Bluethner add-on was released, and I take this opportunity to add my congratulations to Modartt on this superb instrument !
regards,
John
Hi Ben - welcome... and ya, the Blüthner is just great! Good enough to convert apparently
cheers
Hans
Thanks Johng
Thanks Sebion
I had a crazy idea, but perhaps possible if some people get interest.
But if the sensor strip project you did it's really good and not expansive to built, the idea get closer to realization.
My intention was to draw a simple but nice piano action case, in the style of Yamaha AvantgrandN1, but with less large wood pieces on sides and bellow (to make it light weight) and call some piano manufactuer here to propose the idea.
They would built the case and action, and some eletronic company would built the sensor strip project.
Initially I thought about PNOscan make a good price, in colaboration with the piano factory, but it is probably unlike to happen I supose. PNOscan costs € 856, or 20% less for pianoteq users.
But importation tax on Brazil it's 60%
May I see photos of you piano strip bar project?
Have it a quality closer to PNOscan?
Speaking for my build, it is documented online. Just Have a look at my signature
Thanks Johng
Thanks SebionI had a crazy idea, but perhaps possible if some people get interest.
But if the sensor strip project you did it's really good and not expansive to built, the idea get closer to realization.My intention was to draw a simple but nice piano action case, in the style of Yamaha AvantgrandN1, but with less large wood pieces on sides and bellow (to make it light weight) and call some piano manufactuer here to propose the idea.
They would built the case and action, and some eletronic company would built the sensor strip project.Initially I thought about PNOscan make a good price, in colaboration with the piano factory, but it is probably unlike to happen I supose. PNOscan costs € 856, or 20% less for pianoteq users.
But importation tax on Brazil it's 60%May I see photos of you piano strip bar project?
Have it a quality closer to PNOscan?
My own piano strip bar is hand soldered on strip board, so in terms of actual construction quality it doesn't look as neat as PNOscan. The actual performance may or may not match PNOscan - I can't judge because I don't have a PNOscan strip to compare with. I do have one or two photos taken during the development but can't see how to embed them in this message (it looks like they have to be uploaded to an external image hosting website first?). If you're interested it would be easier for me to send to you by private mail.
However, I must say that Sebion's project is much more suitable for your purpose, because he has gone to the trouble of designing and producing printed circuit boards which greatly simplifies the job of assembly.
regards,
John
Thanks again Sebion/Johng
How it's calibrated?
PNOscan requise intensive hours of calibration, as some buyers said, to get to the desirable response, taste.
I supose your strip sensor would need a calibration software.
Thanks again Sebion/Johng
How it's calibrated?
PNOscan requise intensive hours of calibration, as some buyers said, to get to the desirable response, taste.I supose your strip sensor would need a calibration software.
In my case, I wrote a java application which runs on the attached PC. It uses midi system exclusive messages via a midi usb interface to communicate with the microcontrollers that control the sensor strip. The application enables me to adjust the keyboard velocity curve, and also to set a velocity scaling factor for each note (to compensate for any differences between individual sensors that cause one note to generate a higher velocity value than other notes). I can also edit various other general parameters, for example how far down the key has to be pressed before a 'note on' message is triggered, and the point in the key travel at which the dampers contact the strings. All these parameters are stored in EEPROM memory on the strip microcontrollers, so this application only needs to be run if a modification to the stored parameters is required.
regards,
John
Thanks again Sebion/Johng
How it's calibrated?
PNOscan requise intensive hours of calibration, as some buyers said, to get to the desirable response, taste.I supose your strip sensor would need a calibration software.
Yep. For my setup the calibration is done by first showing the system the range of movement for each key. That is the basis for all further calibration.
I have to say my system differs from the sensor bars on the market in the way that it doesn't simply calculate the key velocity at the trigger point but tries to adapt a physical model of a piano action to that key movement and thus calculating the hammer movement and take its velocity when it collides with the virtual string. This introduces a bunch of physical paramers of the action that can be setup. These paramters can be setup by the user interface I've integrated in the mainboard, where you can change them with rotary encoders and see the values on an lcd. I also have a pc software, that can read/write these parameters and store them as a preset file, but the many developments refining my physical model step by step have made this program a little out of date, but finally I plan a software that will help the user to setup the parameters properly and give the ability of interpolating the parameters between two arbitrary keys as a ramp, like you can do in pianoteq via the note edit feature .
Just to give you an idea of it here comes a list of my current parameters:
- velocity sensitivity
- gravity
- Damping point
- Catcher position
- String position
- Repetition point
- relaxed Hammer position
- Trigger point
- Coupling constants describing the stiffness of the action (i.e. hammer shaft)
- friction / viscoelastic constants
- ... some more that are harder to explain
The thing that is really new in this model is that it can reproduce the timing behaviour of a piano action, i.e. on a real piano the hammer string collision does not always happen synchronously with the bottomning of the keys. It depends on the way you move the key and the strike velocity: for slow attacks you will hear the sound before the key reaches its bottom felt. for mezzo forte attacks both events will occur simultaneously and for forte attacks you can reach the bottom, before the hammer strikes the string. This is due to the elasticity of the hammer shaft and the felts in the action. and the timewindow for this delay is about 20 ms, so this is something that you will feel, when playing.
Another important effect of such a system is the ability to trigger a note without moving the key down to the trigger point. This way you can play staccatissimo like on a real piano.
Finally I have to admit that finding the right parameters can take a lot of time, and there is more then one possible setting that might fit your way of playing. I'm always experimenting with it to improve the responsibility. But you should never expect it to react perfectly like a real piano in every situation, because there will always be some deviation of the actual action to the virtual one. the only way to get this is to actually measure the hammer velocity. The KAWAY anytime does this afaik, but such systems are limited to the pianos of the manufacturer who sells it, due to the lack of a standard for hammer grouping.
BTW: John, did you follow my blog and read about my latest achievements? if not I'm sure you will like to read it and see the boards
The thing that is really new in this model is that it can reproduce the timing behaviour of a piano action, i.e. on a real piano the hammer string collision does not always happen synchronously with the bottomning of the keys. It depends on the way you move the key and the strike velocity: for slow attacks you will hear the sound before the key reaches its bottom felt. for mezzo forte attacks both events will occur simultaneously and for forte attacks you can reach the bottom, before the hammer strikes the string. This is due to the elasticity of the hammer shaft and the felts in the action. and the timewindow for this delay is about 20 ms, so this is something that you will feel, when playing.
Another important effect of such a system is the ability to trigger a note without moving the key down to the trigger point. This way you can play staccatissimo like on a real piano.
Finally I have to admit that finding the right parameters can take a lot of time, and there is more then one possible setting that might fit your way of playing. I'm always experimenting with it to improve the responsibility. But you should never expect it to react perfectly like a real piano in every situation, because there will always be some deviation of the actual action to the virtual one. the only way to get this is to actually measure the hammer velocity. The KAWAY anytime does this afaik, but such systems are limited to the pianos of the manufacturer who sells it, due to the lack of a standard for hammer grouping.BTW: John, did you follow my blog and read about my latest achievements? if not I'm sure you will like to read it and see the boards
Hi Sebastian,
Yes, I did look at your blog recently and saw that you are now getting your boards produced by a pcb fabrication facility. They look very good !
Re. your comment on simulating the timing differences between notes hit with different velocities - did you know there is a phenomenon called 'melody lead', which results directly from this effect ?
A pianist attempts to emphasise the melody relative to the accompaniment by playing the melody notes slightly louder. Assuming that the pianist is attempting to play the melody notes synchronously with the accompaniment, the melody notes will occur slightly ahead of the others, because the higher hammer velocity means that their hammers reach the strings first. There is an interesting paper on this at http://www.speech.kth.se/prod/publicati...s/3095.pdf
regards,
John
Thanks again guys.
There are low price piano actions from China, but I don't know the site and chinese actions are low quality and have low durability.
That's why I want to try approach a piano manufacturer with this idea.
If you could find a good way to produce a own piano midi sensor, in a single bar, easy to install in any piano or real action controller. I1, sure many people would get interest.
About calibration, one problem is the limite velocity, maximun velocity captured or computated. Sometimes And sometimes the maximun velocity captures do not match the maximum velocity strike of a piano software, as so,me softwares record a ultra high maximun velocity strike sound, and other use a not very high strike.
Sebion, I took a look, but not very long, but did not found a long bar sensor, just a compact sensor like 2 octaves.
A even more audacious idea I have, is to persuade a manufacturer built a real action controller with more than 88 keys, like for bosendorfer and pianoteq K1. Ohh yeah...
Now a "woodness" question...
Is that possible to apply a impermeabilization layer to the wood of the piano actions???
Moisture it'a a villain for real wooden actions. I always ask why not impermabilize the wood.
Thanks again guys.
There are low price piano actions from China, but I don't know the site and chinese actions are low quality and have low durability.
do they sell grand piano actions, too? would be interested to know the price, just for fun
About calibration, one problem is the limite velocity, maximun velocity captured or computated. Sometimes And sometimes the maximun velocity captures do not match the maximum velocity strike of a piano software, as so,me softwares record a ultra high maximun velocity strike sound, and other use a not very high strike.
are you talking about synthesizers, that only go up to a forte instead of a forte forissimo? I'm not sure if I get your point here.
Sebion, I took a look, but not very long, but did not found a long bar sensor, just a compact sensor like 2 octaves.
My sensor bar consists of 8 key modules. They can easily be chained up to any number of keys. For 88 keys you need 11 modules. My new manufactured SMD bar is currently running in a modern upright piano action. Just have a look at the photos. there is the entire bar:
http://sebion.wordpress.com/2012/07/18/...no-action/
Sebion, even than 3 piano softwares sampled up to FFF, some ones fit a more powerfull FFF than others.
Some early pianoteq models for example had FFF a bit lower in powerfull terms than some people wished.
How much costed to built such sensors?
I soldered the boards by hand, so the price for let someone produce them might differ.
I had to spend about <300 € for the entire build. Don't have the calculation at hand right now
Re. your comment on simulating the timing differences between notes hit with different velocities - did you know there is a phenomenon called 'melody lead', which results directly from this effect ?
A pianist attempts to emphasise the melody relative to the accompaniment by playing the melody notes slightly louder. Assuming that the pianist is attempting to play the melody notes synchronously with the accompaniment, the melody notes will occur slightly ahead of the others, because the higher hammer velocity means that their hammers reach the strings first. There is an interesting paper on this at http://www.speech.kth.se/prod/publicati...s/3095.pdf
regards,
John
Thats very interesting to hear. I always suspected such things to happen, as I realized that you can emphasize the notes in a trill just by the timing of the notes, when playing bach's first invention on a harpsichord. I always stress out the trill notes differently than intended when I play both hands.
PNOscan costs 850€ (or 20% less than that if you are pianoteq costume) and already have the connector plugs.
If you buy large quantity for the eletronic pieces, you may get a better price and could offer a competitive price.
I soldered the boards by hand, so the price for let someone produce them might differ.
I had to spend about <300 € for the entire build. Don't have the calculation at hand right now
what connector plugs are you referring to?
The jacks for midi in and out.
I have no picture of PNOscan pedals sensor. Do you have?
PNOscan
◦ General MIDI
◦ 32-note polyphony
◦ 88 optical sensors
◦ Velocity 0 - 127
◦ MIDI In - Out sockets
◦ Pedals, sustain, soft, on/off
◦ Control box measurement: 115 X 56 X 32 / mm
◦ Key optical sensors frame measurement: 1210 X 38 X 4 / mm
Just 32 not polyphony???? Isn't a limitation for sume performances???
Ah I see. I have designed a mainboard that controls the sensor bar and it provides midi in/out and usb-midi + virtual serial port for preset transfer and firmware update. So its pretty much the same as pno scan regarding the modules. That 300 € included the mainboard, of course, its the cost for the whole thing I built. Just have a closer look at my blog, I've made a lot of nice pictures and demos there, too
Where did you get that polyphony information? a just midi sensor bar doesn't have polyphony limits. A synthesizer may have, but 32 notes are quite nothing, nowadays.
I got from a site.
Need to see if find it again.
What is the best piano action after all, Renner or Kawai?
By the ray, the performance, for a video demo of Renner1s action, it's great.
http://www.youtube.com/watch?v=9dEOqU_Y-4A
Does anybody knows what music is that?
Steinway action http://www.youtube.com/watch?v=01DBtig_...re=related
Se the scapment in action.
By the way, I'm curious about the scapment in case of PNOscan or home made MIDI sensor. How to adjust the sensor to do note produce key sound (send message to software) before the scapment "clic" ?
In this video explaination about piano action, we get give a good idea of how large a action controller case would be.
http://www.youtube.com/watch?v=6qA_sHa9...re=related
What is the best piano action after all, Renner or Kawai?
By the ray, the performance, for a video demo of Renner1s action, it's great.
http://www.youtube.com/watch?v=9dEOqU_Y-4A
Does anybody knows what music is that?Steinway action http://www.youtube.com/watch?v=01DBtig_...re=related
Se the scapment in action.By the way, I'm curious about the scapment in case of PNOscan or home made MIDI sensor. How to adjust the sensor to do note produce key sound (send message to software) before the scapment "clic" ?
In this video explaination about piano action, we get give a good idea of how large a action controller case would be.
http://www.youtube.com/watch?v=6qA_sHa9...re=related
I think both actions are pretty good (when in good shape, of course). I suppose KAWAI action might need some less maintenance with their carbon parts.
regarding the escapement click or trigger point, yes this can be setup for in these sensor bars. You need well regulated action, so that the trigger point is at about 2/3 of the way. As the sensor bar knows what the entire way a key travels is it is setup to have the trigger point at 2/3 of that way.
New Hybrid digital piano from Yamaha, upright with real upright piano action.
http://www.production-room.com/yamaha/y...rid-piano/
Funny, in my idea of a real grand piano action MIDI controller, I was imagining some glass to show the hammers moving.
In this upright the glass make it quite nice.
I supose it is more adequate fur small spaces, as a upright action do not take space like a grand piano action.
But.... Isn't all upright piano action less sensible or playable than grand piano actions???
[...]
Isn't all upright piano action less sensible or playable than grand piano actions???
Hello Beto-Music,
You are correct: Upright actions are less sensitive and less playable than (quality name-brand) grand actions for one main reason -- grand piano actions utilize the force of the earth's gravity to reset the hammers and keys. Upright piano actions require extra linkages and springs to make up for gravity not being of much help in this regard.
Think of the string layout on a grand piano -- the strings are positioned horizontally, and the felt-tipped hammers strike the strings from below, and the earth's gravity helps pull them away from the strings, and reset the double escapement action.
Contrast the horizontal string layout of a grand piano with that of the VERTICALLY aligned strings in every upright piano. The upright's strings run parallel to the pull of gravity, so where does that leave the hammers? The hammers strike the strings essentially horizontally; gravity plays no part in retracting /resetting the hammers of an upright piano.
Upright pianos require an elaborate connection of springs and cloth strips to help pull the hammers away from the strings and reset the escapement, so the notes may be played again. Upright actions also require more linkages (and associated friction points and reduced sensitivity) than grand pianos' actions.
When playing repeated notes on a grand piano, one does not have to wait for the key to return completely to its most upright position. (That sounds like a statement heard by an airline stewardess!! I digress.) Unless you are playing a high-end, brand name upright piano, one cannot play quickly repeated notes, because the keys must return completely to their original positions.
I hope this helps clarify some of the differences between grand- and upright piano actions.
Cheers,
Joe
You are correct: Upright actions are less sensitive and less playable than (quality name-brand) grand actions for one main reason -- grand piano actions utilize the force of the earth's gravity to reset the hammers and keys. Upright piano actions require extra linkages and springs to make up for gravity not being of much help in this regard.
Think of the string layout on a grand piano -- the strings are positioned horizontally, and the felt-tipped hammers strike the strings from below, and the earth's gravity helps pull them away from the strings, and reset the double escapement action.
Contrast the horizontal string layout of a grand piano with that of the VERTICALLY aligned strings in every upright piano. The upright's strings run parallel to the pull of gravity, so where does that leave the hammers? The hammers strike the strings essentially horizontally; gravity plays no part in retracting /resetting the hammers of an upright piano.
Upright pianos require an elaborate connection of springs and cloth strips to help pull the hammers away from the strings and reset the escapement, so the notes may be played again. Upright actions also require more linkages (and associated friction points and reduced sensitivity) than grand pianos' actions.
When playing repeated notes on a grand piano, one does not have to wait for the key to return completely to its most upright position. (That sounds like a statement heard by an airline stewardess!! I digress.) Unless you are playing a high-end, brand name upright piano, one cannot play quickly repeated notes, because the keys must return completely to their original positions.
I hope this helps clarify some of the differences between grand- and upright piano actions.
Cheers,
Joe
I believe that is not entirely true:
- a grand piano action is 'reset' from escapement by a spring at the repetition lever that holds up the hammer as long as the jack is not under the 'hammer roll'(I don't know if it is called that way in english), but it is true, that the hammer as well as the entire mechanism is pulled down by gravity in a way that leads to a nearly constant reset force. That makes it less likely for the hammer to jump around and leads to the often called 'more direct' touch of a grand action. There are some rare upright actions that have an additional repetition spring that does the same for an upright action, but these springs make the action more time consuming to regulate.
- A grand piano action has far more parts as an upright one, and I don't think an upright action has more linkages that induce friction(but I'm always happy to learn): for an upright you only have the linkage capstan button - action lever and the linkage jack - hammer butt. Of course you have the friction of the axes, too, but the number of axes is speaks more for the upright action rather than the grand one
so lets come to the faults of an upright action:
due to the hammers nearly vertical position you get a hammer reset force that depends on the hammers position. It is proportional to the hammers weight multiplied by the sine of the angle between a line that goes through the hammers center of mass point and its fulcrum point and a vertical line going through the fulcrum point(just think of the angle of the hammer against its vertical position, that is a good approximation). Since the hammer has a way of motion of lets say less than 30 degrees you will find sin(x) ~ x a good approximation for the hammers reset force. That means, even when you take away all springs(as seen on many >100 years old upright actions) you get the opposite of a a-spring like reset force for the hammer. This means depressing the key gets easier the more you depress it, if you don't have any springs present. On modern piano actions you have springs at the hammer butt that will partly compensate for this behavior, but this is not entirely possible because a tighter spring will induce more friction.
You can easily observe this fact when sitting at an upright and weighting the keys with coins while the sustain pedal is depressed. just put an amount of coins on one key that is barely not enough to depress it. then you can depress it by about 1-2 mm and you will see the key move down to the trigger point. This non constant reset force of a hammer makes it more probable for the hammer to jump around, being reflected at its rest position when you play very short staccato notes that prevent the hammer from being caught by the catcher. In an upright action aprox half of the reset force comes from the weight of the action lever, that is attached to the hammer by a bridle tape in order to tame this free bounce of the hammer leading to a more regular touch with repetitions.
I think one more thing should be mentioned: an upright action can be reset to be able to repeat a note without having to let the key go up the full way. In fact it can repeat as fast as a grand action, but only if the preceding note is held long enough for the hammer to get caught. If it is not caught you have to let the key go up the full way. You will notice the bad repetition behaviour of an upright when playin trills.
My text has become much longer than intended, but I hope I have made my point
regards,
Sebastian
Thanks Joe/Sebastian for the input.
I think we can solve this problem instaling a horizontal gravity sheet near the hammers.
Háa háaa háaa...
Ok, seriously speaking... When someone get the chance to test Yamaha NU1 Avantgrand Digital Grand Hybrid Piano, please get to this topic and report about the key sensibility, if it's better or not than most digital piano compact hammer action system.
Yamaha vertical action:
Hey, wait a minute...
The gravity sheet (Sci-Fi joke) got me a idea...
If we replace the springs for gravity (transfered from vertical to horizontal by some sort of a pulley with some lead weight) maybe it work better.
Hi again.
I've tested the NU1. The action feels nice for an upright but it the internal sounds especially through the internal speakers suck. I didn't test it with pianoteq though.
@gravity sheets: I had this idea, too some time ago, but I don't think it'd work. Like I said you can compensate for the non constant reset force by a spring, but this spring has to have exactly the right stiffness and it will raise the friction at the jack-hammer butt link and might lead to unwanted behaviour.
regards,
Sebastian
Yamaha Grand Touch use real piano actions... or close....
See how the space from the keys edge to the center of balance (with pins), are shorter in Yamaha Grand touch?
Yamaha Grand touch:
Traditional grand piano action:
Upright action:
Yamaha Grand Touch also have the hammer's arm support (golden in the picture) closer to the center of balance (pins).
Welll.... my question: Is this Yamaha Grand Touch "shorter" action similar to a real piano action of some piano brand, or no grand piano action is shorter like this?
If it's really shorter than all grand piano actions, does it make difference for play?
Look the 200% fingerforce point, nd how is different from a 9 foot or longer grand piano to Avant Grand or Grand Touch Yamaha models:
9 foot:
6 foot baby grand:
Upright:
Kwai MH3:
Roland PHAIII:
Yamaha GH : 
Grand Touch:
Avantgrand:
Is the 200% point really so important , or just a small extra confort ???
If it's really important, I supose Bosendorfer Ceus digital was said to have the longest action, so I supose it have the better 200% point of all digital pianos.
Interesting that you bring this up. I didn't know manufacturers give information about this.
It makes the difference between small and huge instruments. And it makes it easier to get a uniform touch, but I suppose you can get used to quite any key size.
regards,
Sebastian
I got the images from a topic on Pianoworld.com
A person there said the Grand Touch short action do not make much diference compared to good digital hammer action piano, just have the wood and scapment.
I understand that 100 years ago piano manufactures dreamed to eliminate scapment "clic" and get a uniforme weight along the entire keyboard from bass to trebble.
Some people use this as argument to say that we don't nedd scapment or graded keysboard...
But people play in real pianos today, especially in important events, and a digital piano with real piano sensation helps to have same ability, or be prepared to the imperfections of real pianos.
The escapement 'click' is actually a good thing, as it gives you a tactile response if you play pianissimo notes wrong. I once played on a very special action that does not have a normal escapement click: it was rather a zone in which the gear ratio between key and hammer slowly decreases.
the uniform weight is a thing that nearly all DP manufacturers mess up with: they all sell graded hammer actions, but that actually means the bass keys having a stronger reset force, which is not right, as a an accoustic grand action has indeed heavier hammers in the bass range, but this additional weight is compensated with lead pins in the keys. So on an accoustic instrument you have the bass keys being more inert than the treble ones, but all need the same weight to get depressed very slowly.
EDIT: When the sustain pedal is not pressed, you have to lift dampers as well and that makes bass keys indeed a little heavier to depress.
Take a look in this Bechstein D concert piano action:
http://sheaf1.wpengine.netdna-cdn.com/w...piano4.jpg
mh nice action, wanna have it
Take a look in this magnetic balanced action:
http://home.kpn.nl/velo68/MBA_GB.pdf
And also the magnetic acellerated action (page7):
