Levelling irregular velocities using piano/mezzo/forte hammer hardness

OK

This evening I had a Eureka moment. I have achieved the best results so far levelling the velocities on my Acuna88 using piano and mezzo hammer hardness adjustments. Because of this, I believe I am now better equipped to describe how to level irregular velocities.

Before I begin, I will say that the results obtained by this approach will ultimately depend on the state of the keybed/action. If your keyboard has stuck keys for example, you cannot expect to achieve the velocity consistency of a Steinway! That said, most keyboards will benefit from this technique to a greater or lesser degree. Also note that note-for-note hammer hardness editing ability is only available in PianoTEQ PRO as far as I know.

STAGE 1: 'Piano'

The method I have discovered to be most effective is to begin with 'piano' hammer hardness adjustments.
Your best friend in this method is the chromatic run! Begin to play a chromatic run (a scale of semitones) anywhere on the keyboard very quietly and as evenly as you are able. You need to really listen closely as you do this and as you do, you will hear any inconsistencies in the 'piano' (or 'pianissimo') range.

Now, those notes that stood out as sounding 'louder' or 'brighter', make a mental note of. Presuming you have note-for-note editing abilities (PRO Version), double-click on the 'piano' slider in the hammer hardness area of the PianoTEQ GUI. You will now find yourself on the 'pianissimo' note-for-note editing page.

Now you can begin to adjust the notes that 'stood out'. Note here that you shouldn't need to make big adjustments in the pianissimo range. Now pick a note that stood out and play it. You will see a marker indicating the note on the screen. First, right-click and select 'remove all control notes/nodes'. Now you are ready to make your first adjustment. At the note you have chosen, drag down the yellow line 1 notch (the smallest it will allow). This will probably be all that is needed to make the difference to the note (EDIT: IN THE PIANISSIMO RANGE). Now play the note within a chromatic run and listen to the difference. Remember to play very quietly and evenly.

Do this for all the notes that seem louder/stand out.

I will continue with PART 2 tomorrow.

Stay Tuned!

Kindest Regards,

Chris

Hi mooks

The simplest answer to you question "where do velocity errors come from" is that
they most commonly come from physical irregularities in keybed regulation. For example, if I apply exactly the same down force when playing two consecutive notes, and the second key has more inherent friction in it's downward travel, this second note is going to produce a 'quieter' sound i.e. a difference in velocity response.

I must admit as some of you already know, I've been messing with this velocity stuff for a few years now and an still learning! I guess I AM a little more fussy now as I can hear irregularities very clearly after so much practice. I've been working this morning on perfecting the Acuna88 response and still have work to do to achieve this, but I will say this, that there will probably be very little adjustment to the 'piano/pianissimo' range. Most of the adjustments will commonly be in the 'mezzo' range. I will work some more and, as indicated last night, attempt PART 2 of the 'tutorial' tonight.

As you request mooks, I will send some illustrations showing screen shots of corrected velocities and yes, there do tend to be patterns. If the keybed is twisted, there will be undulations and peaks and troughs etc. in the corrections.

More soon,

Kindest Regards,

Chris

Hi Rhombic

Yes, I've tried using volume note-for-note. What would be a brilliant addition to PianoTEQ PRO would be a note-for-note velocity curve as many, including yourself, have suggested. But what would be even better than this would be a note-for-note velocity curve WITH SLIDERS!

I'm going to have to work a bit more on the levelling before I can contribute further. I had hoped to be able to put some screen shots together etc., but I'm still not happy with my present results. It IS my current pastime so I will keep you guys posted regularly.

If anybody who has experience with 'toning' would like to contribute, feel free.
I tuned, but didn't tone, back in the day!

Kindest Regards,

Chris

It surprises me that you need little adjustment to the pp-p range. If there was a friction variation, I would expect this to affect low velocity more than high. Or maybe there's just a fixed percentage error? Do you find that for a single key, the adjustment is roughly proportional to the velocity? I guess that will be clear in your numbers.

I don't have much faith in my ability to accurately control velocity in my playing and I've been wondering if I can do something more reliable. What do the sampling people use? I thought about playing pairs of notes at the same time with two fingers held tightly together. I don't think that would really reveal the subtleties you're seeing such as the friction variation, as it's comparing the results of "same velocity" rather than "same force".

To be honest, Fatar/Studiologic have actually done a very good job in improving their velocity response so I am not dealing with a sluggish action in the Acuna88. You are right about friction and it's effect on the ppp-p range. Whilst working with the NUMA Nero, I DID have to work on this range more, simply because it was more irregular.

I have been working on the Acuna again today and have begun to realise that there is an art to toning. The thing is, the aligning of the velocities requires one to listen very, very intently. The levelling usually requires the adjustment of all thee zones (piano/mezzo/forte) to produce the desired conformity. I'm going to do some more work tonight and will update tomorrow.

Kindest Regards,

Chris

http://www.forum-pianoteq.com/uploads.p...ing_1_.mp3

In the above .mp3, two notes are repeated. They are the matching pair C# and D (next to middle C) where D was adjusted to match the C#.

I'm getting the hang of it I think. I sure hope to speed up - this one took quite a while!

Kindest Regards,

Chris

If the Pianoteq internal algorithms do not mimic the effects achieved by those who voice real pianos by hammer adjustments then you very well might. What do you think so far?

I now have two notes in sync http://www.forum-pianoteq.com/uploads.p...ling_2.mp3

Turns out it didn't need such a big adjustment after all. I will post an image shortly

Kindest Regards,

Chris