PNOscan in TP40WOOD (Page 1 of 2)

I'd love to see what shell could be used other than the Empty Baby Grand Shells from the early 90's I've seen for 88 note controllers.
Do you plan on housing the keybed...?

So you buy a Nero then add the Story & Clark PnoScan...
I played many controllers at NAMM everyyear, and the Kawaii MP series was always my favorite, but 3 years ago I really liked the Nero.
MY problem has been I use 88r's for synths and as a Master MIDI controller so the crappy M Audio KS88's have been fine with me, but every 2 years it seems the jeys from 6 nights a week start triggering maxVal and lose their softer contact.........
But I now have a LividCode and Deopfer Fader Box, so I think the Nero will be my choice next.
Thanks for sharing...

Saturday (yesterday) and today (24-25th September[2011]), I took all the caps off of the tops of the white keys (they'd basically started to fall off!). I Adjusted the height of the PNOscan strip accordingly and found it to be the best position for secure attachment to the TP40WOOD frame. Then (this morning) it took me literally half an hour to chisel off the wooden blocks on the end of the black key triggers. Reassembled, calibrated and this is the result...

http://www.filehosting.org/file/details..._25.09.mp3

I intend to make detailed instructions with pics etc. which I can then send to anyone who wishes to do the same. These instructions would not replace QRS's PNOscan inatallation instructions, which i would recommend you read thoroughly first, but would be an aid to the installation into the TP40WOOD (NUMA Nero) action frame.

Gadgety has expressed an interest and if anyone else would like more info, contact me via the forum e-mail service.

Regards,

Chris

after a long no show on PianoTeq, I thought I'd better pop in!

http://www.forum-pianoteq.com/uploads.p...lation.mp3

the link, a rough .mp3 demo of the keybed prepared for final calibration/regulation process. I had a real tough time trying to level perfectly black and whites, but with the help of Lanny and Wim, both of QRS/Midi9, I have been able to work out a perfect solution.

This does mean that the project is a little more complicated than I thought, but with persistence it can be done. For me, the hardest part of this project has been the constant need to problem solve, but it's paid off now and if I ever needed to repeat the excercise (minus many, many hours) I am now able with much greater ease (and experience!). The calibration perfecting is time consuming and intensive, for me anyway, but I've learnt a technique which will have the keyboard 'perfect'. It will perfectly faithfully reproduce what I play like a pen perfectly translates from the hand every movement to paper. That means that any inperfections in the musical performances/recordings will be my mistakes and cannot be at all blamed on anything else!

more soon,

Regards,

Chris

I have found that subtracting an increment of 96 from the 'at rest' position of the black keys is giving good results so far. I have sync'd a group of three successive blacks and their adjacent whites (F-B inclusive). I will do further testing shortly and when carried throughout the whole range, I'll post a demo (if I'm not tempted to demo a sneak-peak mean time!). This is a development on the process mentioned in the previous post. I'm about to go back to work on this with fresh ears, but as I said before somewhere, there is a certain sound that a key/note makes at any of it's velocities (especially noticeable in the low to mid+ range) when it's set 'spot on'. I think it may sound a little weird to most ears when completely 'in' as ther will be absolute faithful (+/-)reproduction of the PianoTeq models. We're so used to idiosynracies in musical instruments that when purer sound approaching 'perfection' is heard, it sounds synthetic. Of course, most realise the importance of the human element (ever tried to listen to regimented midi files for any length of time???), but do we realise that diversity in this created realm is an intended norm! I know personally the 'pursuit of perfection', as is evidenced in my many often lengthy posts. But of late, I also desire to enjoy the journey, 'cut myself a little slack' if you will!

I digressed!

I'll report back soon,

Regards,

Chris :,

Hi Jake,

Essentially, in making a compenstory subtraction of '96' from the black key's 'at rest' measurement figure, I'm telling the PNOscan computer to 'make more shallow in ascent' the black's velocity curves. As I imagine it in my head, I 'fake' the black key heights to give them a shallower curve.

I tried using hammer hardness to level the original strip, but found that to be unworkable. The beauty of PNOscan inconjuction with the software (WinNessie) is that yu don't loose range. In otherwords, when making adjustments at the 'keyboard/sensor strip computer' end rather than by the parameters in PianoTeq, you are on the 'right side' of the setup. If your keyboard/sensor strip is not calibrated correctly, then no amount of adjustments in PianoTeq can fully and perfectly compensate.

I'm using a Windows Vista laptop.

Kind Regards,

Chris

Thank you Jake

I did download both (from your previous links I believe) when I was still trying to resolve the inconsistencies of the original velocity strip in the NUMA Nero. PianoTuner is a great 'fix' for an imperfectly calibrated velocity strip, but the downside of any 'secondary' compension is that it cannot perfectly make up for the original inconsistencies. What I mean is that it is not going to change the original strips computations. With PNOscan, I am starting at the 'primary' source. I am able to calibrate perfectly this, or infact any, keybed. The difficulty I have had in choosing the global increment by which I compensate for the black/white velocity difference(s) has been in actually finding it. It has up till now been complete trial and error. I have now changed to '88' which on testing is much more level than the previous '96', but it may be slightly out +/-   It is the 'human' element that makes it difficult to align note-for-note curves. I have found that the bass is the 'litmus test' in the fact that it, compared to the rest of the range, most clearly exposes  any remaining inconsistencies. This is the best guide which I will use today to fine tune this increment, if need be. However, this is merely the 'global' increment after which will be micro tweaks! FUN!!!

Here's the latest demo .mp3 (last night [Sunday Oct. 9th)

http://www.forum-pianoteq.com/8c68ca5bd...d341e47a7b

Thanks again Jake,

Kind Regards,

Chris

Note-On, Note-Off, Note-On, Note-Off!!! Wiil check again now to see if I got it wrong!

Chris,
Thanks for doing the test. Did you use Pianoteq to do the monitoring/recording? If not - what did you use? This is important, because some sequencers will translate the tri-sensor sequence (Note-On, Note-On, Note-Off, Note-Off) into "standard" Note-On/Note-Off pairs. My sequencer does this. (Ableton Live Lite) 

Anyway, assuming you used Pianoteq (which does faithfully record the raw MIDI from the controller), or some other raw recording utility, that means that the PNOScan system appears to emulate the double escapement in a different way to Yamaha and Casio.
The advantage of having the back to back Note-Ons is that the piano engine can in theory recognise that this is occurring, and impart a subtle change of timbre between successive notes, like occurs on a real piano, due to the hammer striking strings that are already vibrating.  PNOScan does not allow this to occur, because the piano engine will damp each note normally.   The advantage of the PNOScan method, however, is that the MIDI remains fully standard. IMHO the back to back Note-Ons is a fudge. ;^)

If PNOScan has a mode where it can generate some kind of extended MIDI format, it would be interesting to see how it behaved in that mode.  (quote from their web site: "Recordings can be output as standard MIDI files or on the on the QRS Pianomation® system for more dynamic playback." )

I've logged a feature request with Ableton to properly support tri-sensor keyboards. ;^) (it's a bit of a tricky problem, isn't it? It seems that we need a new note-specific sostenuto controller message for this feature!)

Greg.

By putting a note release velocity with a value of zero in the first note-off message, PNOScan could indicate that the note decays gradually after the first note-off message, rather than being cut off sharply. The midi rendering software could then recognise that the strings are still vibrating when the second note-on occurs. Pianoteq does recognise and take advantage of note release velocity information if it is supplied by the controller. I've no idea if PNOScan generates note release velocity but I think this would be the correct way to do it as it preserves  standard midi format (note-on, note-off, note-on, note-off).

John

Greg,

Mode 2 ('less dynamic range/loud' mode [my description]) Note-on, Note-off, Note-on, Note-off - just as Mode 1

Kind Regards,

Chris

Chris, did you also check release velocity?

Can the double-escapement release point and full release points be adjusted?

EDIT: I guess the key retrigger adjustment would be equivalent to the double-escapement - yes?

Greg.

Chris,
Thanks again, and thanks for being patient. It looks like a very good system indeed.
Just by the way, I suspect there would be a lot of room in my Kawai MP9000 for the PNOScan sensor strip, because the existing (standard) sensors are right at the ends of the keys, on the hammers, near the capstans. (nowhere near where the PNOScan strip is placed) I think the same would apply to the current Kawai RM3 Grand action (and possibly some others?).  So, because both the sensors and the hammers are well out of the way, it might be an easy install.

Regarding the release velocity, what I really wanted to know was the actual release velocities that occur when you do that specific test sequence. (the test you have already performed, earlier).   It would also be good to know whether the release velocity for the first note-off is consistent, or whether it changes with the actual speed of repetition. If, by chance, it works like the way JohnG suggested, the release velocity for the first note-off should of course ALWAYS be zero, regardless of how fast the partial repeat is played.

Greg.

Chris,

I think you will see one of the following three scenarios:

If PNOScan is transmitting release velocity it will show up in the Pianoteq midi dialogue window as a non-zero (and variable) velocity in the note-off messages.

If the velocity value in the note-off messages is always zero then it means that PNOScan is signalling a note-off event by sending a note-on midi message with a zero velocity instead of an explicit note-off message (this is a valid midi alternative to a note-off message).

If the velocity in the note-off messages is always the same (e.g. 64), then it indicates that PNOScan is sending true note-off messages but is not actually reporting release velocity.

regards,
John

So that confirms PNOScan does not send key release velocity.

Thanks,
John

I have achieved my best calibration by far today. Results here
http://www.forum-pianoteq.com/7a6866f4c...6134700e2f

This post from Joseph Felice about release velocity makes a lot of sense:
http://www.forum-pianoteq.com/viewtopic...7428#p7428

However, remember that the main reason we were interested in release velocity initially in our discussion here was an alternative means of conveying damperless double-escapement repeats.  It appears that PNOScan does implement double-escapement repeats (repeats for relatively shallow key returns), but just not damperless. I accept that this is too is probably an extremely subtle thing to be concerned about.

Greg.
p.s I think it would be interestisting to see what a piano would sound like with metallic dampers. If Pianoteq ever models this, then release velocity may be more important. Hows that for a tangential thought. ;^)

Greg, if you mean by damperless repeats, that you can play rapid repeating notes and during this control via your repeat technique whether the notes sustain during a trill (for example) or not, then YES, PNOscan does do this. I can play a trill above release point and the notes will not sustain, or below release point and they will - during the same trill. It all depends on how high I lift my fingers to trill. It can be controlled.

Regards,

Chris

I've just tested Pianoteq. There is no discernible difference in the sound between the two types of repetition sequences. ;^)  I can see that more concurrent voices are consumed for the back-to-back Note-On method, however, if there is any difference in the sound it is inconsequential.  (the polyphony count does increase a lot for a rapid single note repeat - about 16 for successive Note-Ons, and about 8 for standard Note-On/Note-Off pairs, but this will depend on the repetition rate too)

For rapid repeats, NEITHER method produces truly satisfying repeats - there are sequences of notes that sound 100% identical, interspersed with the odd note that sounds different - it does not sound very natural.  Will upload files a bit later.

Haven't tried any other virtual piano yet - I'll have to "play" the MIDI file to these, because I can't use my DAW due to the fact that it converts the tri-sensor MIDI sequence into standard Note-On/Note-Off pairs.

So, Chris - I agree with you. At the moment, Pianoteq is, to all intents and purposes, behaving the same.  Nevertheless, I would still feel better about PNOScan if it conveyed the damperless double-escapement repeats explicitly.

Greg.

The file in the above post is only 96 encoded. Pretty poor quality. I will split the file and post a 160 encoded mp3 demo shortly. May I encourage anyone with Focusrite's VRM monitoring hardware i.e. VRM Box or Saffire Pro 24 VRM, to listen to the file I will post next through the virtual AURATONE 5G monitors on the 'PROFESSIONAL STUDIO' setting.

kind Regards,

Chris

http://www.forum-pianoteq.com/uploads.p...0.2011.mp3

Best yet, close to absolute uniformity!

https://skydrive.live.com/redir.aspx?ci...5AAB24!669

escapement inherent in TP40WOOD as demonstrated here.

Regards,

Chris

Yes, on key depress there is a slight build up to escapement point at which there is a sudden give (or release) and the key finishes it's travel.

No, there is no escapement kit in this keybed. It is just designed this way. As soon as I took out the silicone plungers, I could feel this escapement point so clearly. The rubbery spongey silicone of the original midi strip disguises this feeling.

It is actually a much better 'grand touch action' than people realise. And I only found this out becasue I had to take out the original strip to install PNOscan, which of course is a 'non-contact' optical strip. This means that you only feel the action itself, and nothing else i.e. no spongey rubber/silicone.

Hope this explains Greg,

Kind Regards,

Chris

Here is a better example. The weight is perfect. Sometimes it goes past this point, sometimes not. Therefore you can clearly see the point at which this 'escapement or release or let-off takes place.

https://skydrive.live.com/redir.aspx?ci...5AAB24!671

Regards,

Chris