So, just so I'm sure that I understand, you're finding that the inconsistencies in velocity response are in the NUMA and not in PianoTeq?

Yes Jake, in the NUMA. Actually, the template will only be for me a kind of 'stop gap measure'. I'm still going ahead with the PNOscan midi strip which will be far more sensitive/accurate anyway. I will post my best velocity inconsistency compensation .fxp as soon as I've completed it.

Best regards,

Chris

In any case, the sound you are getting in your mp3's is great.

the above is a link to an .fxp and .mp3 in FXP Corner demonstrating how it IS possible to make two individual note's vastly differing velocity curves play as if identicle using hammer hardness parameters (in this case, mezzo and forte) and volume.  The .fxp contains the adjustment information in it's relevent hammer hardness and volume note edit windows. The two notes (B and F) were the two most contrasting in terms of velocity curves so I knew that if I could sync up these so that they played identicly, it would be possible to match all the lesser contrasting notes. There IS room for improvement, but as a demonstraion of the fact that it IS achieveable, it suffices for now. I will hone it later.

It took me a whole day to accomplish this, just TWO notes! But now it's going to be a whole lot easier 1, knowing I can do it, and 2, knowing exactly which parameters need to be adjusted to get the job done. It will still be intensive, but I now know preciselt what I'm doing.

As to the 'pattern', I believe there IS an overall pattern, but am not convinced that just dicovering it and applying the same parameter adjustments reletive to this pattern, will result in a perfectly even range. I could be wrong, but I still think there might need to be individual incremental adjustments made, pattern or no pattern. I'm now not too interested in the pattern anyway, it just gives me a better orientation to know where the variations are.

I have also now determined my 'control note' i.e. the one whose velocity curve I wish to mimic throughout the range. I will work from there. I have matched the note adjacent to it already and check octaves from these shortly. I think the best way is first of all to get a global skeleton from which I can hang the rest of the work which will also serve as a check for global evenness.

When this project is finally complete, I would desire that the final result be tested on other Fatar boards to establish whether or not it may be used as a corrective template for all Studiologic keys.

Sincere regards,

Chris (back to work!)

Not happy to sit on my backside and twiddle my thumbs till I have saved the cash. bought and installed the PNOscan strip, I thought I'd ask some further advice on achieving a uniform result with what I've got now i.e. the original keyboard strip. However, before I do, I need to address incorrect information in my last post! There is NO such continual adjustment by the midi brain (boy, I should get my facts straight!). There is however a pattern that runs throughout the whole range. I don't know how fatar programmed the individual note velocity curves, they differ - not randomly but organised.

Is it possible to fully mask irregular velocity curves using pianoteq parameters (hammerhardness for example)? Given that these variations in individual note velocities are fixed (as confirmed above), and arranged in particular order, it will be easier to compensate for these inconsistencies via a template that will be good for anyone with a Fatar board (presuming that all Fatars board's individual note curves are programmed the same way).

Now the pattern sequence, after concentrating on it again, has become crystal clear. I will attempt to descibe it in detail as follows...

All octaves of any given note are consistent

all notes D-G#-D-G# etc. are the same

all notes G-A and C#-D# are the same

B's are the loudest notes (easiest to reach 127) while F's are the quietest (hardest to reach 127)

more to follow

There are TWO distinct causes of velocity inconsistancies in the NUMA Nero (as with other Fatar / Studiologic boards.)

Cause 1. has to do with the sensors. This is the mechanical cause.

Cause 2. has to do with the midi brain / keyboard cpu / firmware. This is the programming cause.

As most of you will know, I have sorted the mechanical cause by meticulously cleaning the senors etc..

However the second cause is inbuilt into the onboard electronics and this is what has been causing me problems. I have discovered that the midi brain /keyboard cpu / firmware actually continually 'listens' to what velocities you are playing at and adjusts accordingly. I did some experimentation and found this to be the case. I'm certain it is inextricably linked to the 'YouPlay' facility as this also behaves in the exact same way.
This second cause is by far the most irritating in that it's not just a simple cleaning procedure. I suppose it is possible that a firmware update may address it, but I'm not going to hold my breath.

For this reason, I am unable to finalise my work until I have installed the PNOscan midi strip. This will give me reliably consistent velocities with which to analyse. Presently playing my keyboard is like playing a Nasa foam matress! - continually reshaping itself as it wishes! Unreliable at best, downright annoying at worst!

Apologise to all who were / are hanging in there for a quick fix - there ain't one! At least I've discovered the causes of the problem(s). Now I will work on the solution I've chosen. It seems like I didn't avoid the financial outlay after all!

watch this space...

Regards,

Chris

It appears that tweaking is far from being a thing of the past! Neither the work that it involves!!!

I think I will have to take a further look at the issue.

Well, back to work eh!

Regards,

Chris

I'm puzzled by the velocity curve. It shoots up to a mf velocity response of 75 at a keyboard velocity of 36 or so. Does the NUMA have a hard action that encourages raising the response in this way?

Do you see why I'm asking? By raising the curve early, don't you just about skip the entire piano velocity layer. Don't misunderstand. I like the sound that you're getting. I'm just unclear on why the velocity curve is advantageous.

Did you try using the Dynamics slider to bring up the amplitude of the lower velocities? Moving it to the left will raise the  amplitude of the soft strikes (possibly forcing you to reduce the overall volume). Of course, you might find that you wanted to then impose your mezzo strike pattern to the soft strikes...

The sound you're getting in the mp3 is just wonderful, sigasa.  I'm not someone to gush, but the midrange in your recording brings out a sense of strings and hammers that I don't think I've heard before in PT.

On the other hand, to make a too rapid transition, this recording makes me want to know more about how you made your choices. What were you listening for as you edited each note?

Did you go entirely by ear?

Yes I went entirely by ear, but once I realised there was a scientific pattern, it became easier. In the end the entire toning / levelling was achieved using soley note edit mezzo hammer hardness. The real brakethrough came when I discovered this pattern in the mezzo hammer hardness range. I am still unsure exactly why this hammer hardness inconsistency exists within the PianoTeq model, if indeed that is where it originates? The only other possibility is that it is a velocity programming issue with my keyboard? I still don't know the answer to this question.

I had to listen to a lot of things to achieve what I have. Each note was so different to the next when compared laterally. Of course every note had to be different, but there has to be coherence in the progression in the context of the full keyboard range. When adjusting mezzo hammer hardness, among many things I also discovered that there is a sound close to either side of the model default hardnesses that is very similar, but also different in quality. This is the same as in tuning when either side of 'perfect' can sound almost identicle. However a trained ear can differenciate, and having been trained to tune I had somewhat of an advantage. This 'either side of true' aspect aided me in unveiling the pattern which existed and it's necessary compensatory mirror. The latter can be seen visually in the posted .fxp mezzo hammer hardness note edit window - a simple pattern that took such very hard work to discover!

Jake, I am pleased that you appreciate the sound. It is, as I intimate, the product of about eight weeks of intense work. This was reflected in my absenses from the forum. I intend for others to benefit from all I have achieved, and as with all my posts, there is an underlying desire to make it easier for readers to get the results that I have, without the hard work! PianoTeq is so beautiful and as such, must be heard at it's full potential. Anything less is a diservice to the Modartt team and their absolutely astounding accomplishment. All credit goes to Modartt as the creators of such a wonderful, and indeed perfect, musical instrument which I merely brought out the best in my recording (not the playing I hasten to add!).

Thank you to each and all of the Modartt team. You have made a young man very happy!

Warm regards to all,

Chris

On the other hand, to make a too rapid transition, this recording makes me want to know more about how you made your choices. What were you listening for as you edited each note?

Did you go entirely by ear?

Not only have I finished / completed / perfected this project, but in doing so I have also created a perfect velocity curve and discovered a "code" which when implemented unlocked the whole thing.

I will not comment on the sound, I will leave that for listeners to judge. But I will say that I am so very grateful for the whole experience.

What I thought was a random velocity inconsistency issue within my NUMA Nero turned out to be a simple pattern, but I had to discover this pattern first before I could perfect PianoTeq.

A lot of work, but well satisfied with the reward!

Judge for yourselves,

Regards,

Chris

Much more work done now and after coming to to the wrong conclusion that this was an impossible feat, have found myself starting again from scratch. I have done 4 notes!!! I'm not a snail, honest! If I can get this done, I will save myself at least £600 which is the price of the cheapest optical strip.

More news later,

Regards,

Chris

Could you upload the mp3 and the fxp to the PianoTeq fxp corner? (Why did you use Livid when you've used our fxp corner site so far?)

Looking forward to hearing what you've done. I like the results so far.

I am now sure that the issue of velocity inconsistencies in the studiologic/fatar boards is purely mechanical. 99.99% sure!

I was playing some stuff tonight and generally checking the keys performance and when I noticed that there was a key that sometimes would produces louder sounds than the surrounding keys and sometimes not, I jabbed it at ffff. It stopped it's louder sounds. I tried this some other keys and, sure enough, they too started to behave normally. I also remember a post by John Rule in which he says that the more he's played his NUMA Nero, the better his aftertouch velocities have functioned.

I'm convinced that these things need playing in like a real acoustic grand. After all, they have mechanical parts that need to settle just like any machine.Obviously, if the sensors are not properly snug on the circuitboard, this can cause problems, but I seated mine meticulously thoroughly and still had variations. It was late when I was playing so I couldn't jabb all the keys, but I will tomorrow and let you know the results.

(I meant of course that I was not trying to be critical of your work. Sorry. Needed sleep...)

Jake, not at all offended! No worries!

My goodness, after what must be at least 2 weeks of work, trial and error, scruggle, late nights etc. etc., I just did the whole thing in less than an hour - from scratch!

I'm not sure whether any of my previous post was accurate but I did realize I was heavlily compensating for 'inconsistent velocities'! However, the beauty of PianoTeq PRO is the note edit function(s) and this has meant the equivelent of a note for note velocity curve. I now have a brand new keyboard!!! Albeit with sticky label residue on the keys after much work on it!!! It certainly feels like I've got a new board when playing it now!!!

I'll shut up now for a little while and let the .mp3 I'm about to post do the talking.

Regards,

Chris