Reverse (left-handed) keyboard?

Try MIDI-OX:
http://www.midiox.com/

It can do that.

(I'm left handed)

What advantage would this be if the firmware has graded hammer weights selected for right hand keyboard?

Ian

Thanks, etalmor, I can't wait to try it!

And Ian, yes, graded hammer weights would interfere with the experiment.  But it is still interesting to try -- it is like seeing the world upside down!  Practicing this way can perhaps be fruitful as well as entertaining.  I like to practice anyway with mirror-image patterns;  I find that I make somewhat different motions by habit in my left hand than my right, and often I can solve technical problems by comparison with the other hand.  Switching hands completely around may give some insight into aspects of how we usually play that we would normally ignore.

To try out the kbm file I piped a preexisting midi file through it, and the results are quite amusing:

https://soundcloud.com/smulloni/wrong-s...gal-tracks

When I save this as .KBM file, and try loading it in Pianoteq, it doesn't want to load it. Says "error while reading - invalid keymap size at line 1"... Devs?

Ah, that was the trick. Converting the file to ANSI worked!

For the record, the way to mirror the keyboard around middle D in Pianoteq now is: in the advanced tuning section, select “reverse keyboard direction”, change "string tension" to "full rebuild", and choose G2 as the diapason note (click on the invisible button immediately to the left of the C3 label in the keymap editor).  Without full rebuild, reversing keyboard direction produces useless results, and the diapason note can't be changed to G2 with the diapason note button, which won't accept values below 220.

Thanks to Niclas for explaining how to do this!

I'm glad you find it stimulating! Inversions can be fascinating, and obviously composers have found them fruitful for centuries -- the mirror fugues of Bach's Art of Fugue being a kind of limit case (maybe the most spectacular instance, while also being the least suggestive of further uses).  Cases when inversions either succeed somehow (some Bach pieces, like the fugue I posted years ago in this thread) but with different meaning than the original, or completely fail (like the Mozart Sonata K333 I've been inverting recently so as to work on my left hand technique) are also interesting in how they differ or fail.  It is really worth paying attention to what happens when an isomorphism is applied to a work of art.  I think it was Edward Cone who pointed out that a conventional analysis of Chopin's 1st Prelude transposed up an octave would be exactly the same as an analysis of the original, yet the two works would hardly be equivalent.  Any attempt to capture the significance of the difference is non-trivial, leading you into areas of human experience outside so-called music theory.   A similar but less drastic and more familiar transformation is change of tempo, which, because it is more familiar, we can become desensitized to, but to what we have been desensitized we may become resensitized — that's what music is for! Cheers, js

An additional way to get a left-handed keyboard is to use SuperCollider, the software used by jazz pianist Dan Tepfer that I wrote about some time ago. Here is a short program that inverts around Middle D:

MIDIClient.init;
MIDIIn.connectAll;

m = MIDIOut.new(0);
m.latency = 0.0;

n = 62; // Middle D

MIDIdef.noteOn(\On,
    {
    arg vel, nn, chan, src;
    if (chan == 0, {
            m.noteOn(1, n - (nn - n), vel);
    }
    );
}
);

MIDIdef.noteOff(\Off,
    {
    arg vel, nn, chan, src;
    if (chan == 0, {
    m.noteOff(1, n - (nn - n), vel);
    }
    );
}
);
)

Here is part of Bach's Prelude no 1 from WTCI played forward on channel 1 and heard backward on channel 2:
Nice chords, but not so baroque anymore...

https://forum.modartt.com/uploads.php?f...relude.mp3

Of course, using the Advanced Tuning as described is much simpler.

In fact, my short demo at twice the speed and an octave higher works better as a harp part.

https://forum.modartt.com/uploads.php?f...20harp.mp3

Well, you asked for it

Here is the complete prelude 1 reversed. I had to go back to the score since memory failed and, amateur that I am, there is a splice at the end due to a stupid mistake, but the chord progression is still interesting even in reverse!

https://forum.modartt.com/uploads.php?f...elude1.mp3

Since it goes too low forward so too high backward I left it untransposed. The interesting part in using the SuperCollider method is that I can listen to the normal version on channel 1 while playing and record the reversed one from channel 2.

This thread has been an interesting read. I've had a similar question that I've been looking for an answer to for quite a while. Would any of the tools mentioned on this thread be able to reverse half of the keyboard? If you take D4 and set it as the "middle" of the keyboard, the black and white key patterns will be mirror images on both sides. I've been trying to figure out a way to change the notes below D4 (the left hand notes) to mirror the notes above, and vice versa. This would make it possible to use some specific practice techniques, such as hands together practice, by using the same fingers (and thus the same fingerings) in both hands.

Do you mean that you want middle D to be either the lowest pitch, or the highest one?   Using Python or SuperCollider for that would be easy.  In Python, one would replace a couple of lines in invert_pitches() in the snippet above as follows (I haven't tested this, but think it should work):

def invert_pitches(port):
    for msg in port:
        if msg.channel == IN_CHANNEL and msg.type in ('note_on', 'note_off'):
            # only reverse pitches above middle D
            if  msg.note > 64:
                msg.note = 124 - msg.note
            msg.channel = OUT_CHANNEL
            yield msg

This would have the lower half of the keyboard ascend, and the upper half descend.

Perhaps this can be done with Scala files but I never got the reverse.kbm file to work with recent versions of Pianoteq so I'm not sure.