How I Did It

[jcfelice88keys]

Hello to Wiser Guy, Nikolas, Allan, Jerome, Nick and all other SoundsOnLine friends,

In another thread having to do with the dynamics of EWQL pianos, the first member in the above list asked, publicly, how much processing I applied to the EWQLP demos to make them sound somewhat unlike all of the others combined. I believe this discussion is important enough to warrant its own thread. As always, everyone is encouraged to read and offer commentary on what seems to work for my own case of producing what other people characterize as good sounding EWQLP demos for Nick and Doug.

My answer may appear to be slightly disappointing to you, but I must make the following formal reply: I applied precious little post production to my demos. Perhaps this is because I am intimately familiar with the feel and response of my Roland A80 keyboard.

As proof of my applying precious little post processing, I have gone back and captured a screenshot of the opening measures of Gershwin's Rhapsody in Blue to describe some of the finer points of my demo performances. Hopefully this may give you some insight you might have not thought about with respect to your own use of the EWQLP library.

B A C K G R O U N D :

Before plunging into the discussion, I wish to state that I have been using the same (viscous damped action) Roland A80 mother keyboard that I acquired back in 1990.
In effect, I have twenty years' worth of experience playing on the same keyboard in my home studio. I am keenly aware as to how it responds to my touch. The velocity curve that works for my own fingers is slightly convex instead of concave. This setting happens to match my own playing style and touch; therefore, I make no claims that it will help you. I would have been glad to share a photo of the A80's velocity curve -- but after close to 20 years of playing on it, the keyboard's green LCD display is nearly completely 100% dark. I use one main setting that I know works for me, but for the life of me, I cannot see what the parameters are!!

I would like to discuss what I remember of the way my A80 was set up when the LCD window still worked. As stated earlier, I use a slightly convex (as opposed to concave) response curve. It is delightfully ironic that Nick Phoenix may have unwittingly sensed this, because he recently suggested that I set EWQLP's on-screen sensitivity curve slightly concave!!!

To digress for a moment, for the past fifteen years, I have had constant access to a rather nice Steinway grand at the church where I am Senior Organist. Personally, I believe what may set my touch apart from many others that use electronic keyboards exclusively, is that having the real thing as a practice instrument makes a great difference in what I am able to create on the virtual piano library. This particular Steinway has quite a heavy touch, as it was built in the early 1980s; no further explanation is required here -- if you know the history of Steinways in the early '80s, you will relate to what what I am saying.

Back to the Roland A80 settings that I recall from memory:
The low end of the slightly convex velocity setting allows me to achieve reproducible velocities when using a soft/sensitive touch. I do recall that I have fixed a lower velocity limit of about 10 or 12 on my A80 keyboard, so individual notes do not completely disappear. The upper end of the slightly convex velocity setting allows for me to "dig in" to the keyboard -- without a hairtrigger problem of too easily maxing out the note-on velocity level at 127. In fact, I must really have to work at achieving 127 velocity, and this is done by design.

* * * * * * *

D I S C U S S I O N :

With the above as background, let us proceed to viewing and discussing the opening bars of Rhapsody in Blue:
Please refer to the first attachment, and take a look at the piano-roll presentation of the opening trill -- what do you notice? The two most prominent features of the trill show that these notes were performed live -- there is no quantization nor step entry present. This is the result of decades of practice!!! The first notes are relatively long (as the trill starts out slowly). Then notice how the notes get shorter, and more closely spaced (as the trill accelerates), then as the trill decelerates, the notes become longer again.

I was unaware of the next feature when actually playing this trill, but there is a slight detachment between notes of the trill. The sustain pedal (MIDI CC#64) was applied and took care of making the notes sound legato.

Now notice how the note G is played slightly louder than the F, as the trill continues to be played. It is here that I did go back and make slight adjustments to the velocity to "smooth" out the transition of notes within the trill. I must confess that, before sending a copy of the standard midi file to Nick Phoenix, I thought that some of the notes' velocities in the trill "looked" slightly uneven (even though they didn't sound terribly uneven) -- but this was my first submittal to Nick and I wanted the midi file to "look good" so as to impress him!!!! (I can LOL now, but this was an absolutely true statement!!)

Please take a look at the overall range of Note-On velocity in this excerpt -- do you notice that the highest value does not exceed something around 100? Obviously, the note-on velocities do approach 127 later in the piece, and especially at the very end. Personally, I believe that too many people's performances "hit the wall" velocity-wise, and get too loud in too soon a time. What you are seeing with my velocities is the result of the shaping of my mother keyboard -- that, plus tons of ongoing practice(!) for weeks and months prior to attempting to record the final demo performance.



Playing the Left Hand Octave jumps:
First, please note that I was not using a metronome to record this live performance. In fact, I just hit the record button and let the notes fall on the grid wherever they fell, without concern about measure number. By the way, I am playing this performance from memory, so this helps me control the "continuity" of the piece; in my opinion, listeners should not have to endure slight pauses while the performer turns pages of sheetmusic. In effect, I used Digital Performer (then Version 5.12, currently using DP Version 6.01) as a pseudo tape recorder to capture this performance in midi format.

As a point of reference, however, please have a look at the B-flat rolled octaves at pseudo-measures 6-1/2, just after 8 and at 9-1/2. What do you see? Well, you can see that I hit the lower Bb's with my pinky finger, and jumped to my left thumb rather quickly. The duration of the three lowest Bb's is very short, but you will also notice that the pink field shows that I "caught" the notes with the sustain pedal.

After hitting the short B flat octaves, and pinning them with the sustain pedal, my left hand jumped up to the three held chords. As you can tell, there is obviously zero quantization performed here! Although I was rather unaware of how my left hand hit the keys for these three chords, I was surprised how seemingly random was the order that they were hit. In the third chord (D-Ab-C), my pinky was slightly late! No matter, because this was how I hit the chords. In retrospect, some of the "life" my performances seem to take on ... is that the notes' start times are NOT quantized! (For that matter, neither are the release times quantized.)

You will also notice some variation in the velocities of individual notes played in these left hand chords. There was no intention to "assign" each note its individual velocity; rather, the velocities fell where they fell.

At pseudo-measure 7-1/2, please take a look at how the Ab-Bb-Ab mordent came out. Personally, it looks "sloppy" in midi piano-roll playback form -- but that's how it was played, and actually, it sounded quite nice.

* * * * *

Hopefully this type information sharing will give many of you a chance and some insight to analyze your own piano playing. In order to "improve", I heartily recommend that everyone makes a midi file of his or her live playing on whatever keyboard you have.

Next I would recommend that you go over your file and look for the presence or absence of some of the features cited above. By combining what you hear (or don't hear) with your ears, and what you see as a visual representation of your playing, you will be in a better position to modify your playing technique. Then I would encourage you to make subsequent recordings of your playing -- and keeping those files as records of your progress.

Thank you for reading this far. As always, critiques and comments are welcomed. After all, this is OUR forum and we each deserve to make the best of whatever it offers.

Cheers and best regards to all,

Joe -- jcfelice88keys

[peter5992]

Ha ha Joe - love this post!!

You are a super keyboard player, and I bow down deeply to you Sir, when it comes to keyboard skills. And I trust your honesty and integrity anytime - even though I don't understand the technical details.

[OneThrow]

All I can say Joe is you must be a wonderful teacher. Such care and dedication. And willingness to share with us your insight. Thanks!

[wiser_guy]

Since I feel responsible that my question in another thread got you into all this trouble, I have to deeply thank you for this detailed and honest description of your workflow. I suggest anyone who messes with software pianos should carefully read this.
I personally thank you for sharing these thoughts, Joe.

[persentio]

Very intriguing and insightful! Thanks for sharing Joe

[jcfelice88keys]

Hello All,

Thank you for your many kind commentaries to my posts in this forum. It is your collection of positive remarks that gives me the desire to continue being of help wherever and whenever possible.

At first I was thinking this thread probably should have been placed in the Midi Techniques section, but thought it might have gotten lost there. I wanted to post it in the General Discussion section because I felt this would reach the largest number of members, and to give everyone the encouragement to persevere with this marvelous piano library until the software works for them.

The original thread took me a few hours to write, and I posted it at about 1AM Saturday morning. As such, when I re-read it today at Noon, I found a wrong choice of words in the context of Nick, of all people! Regarding the convex shaped velocity sensitivity setting on the A80 keyboard controller, I wrote (in blue colored text) that Nick unwittingly sensed it and recently suggested that I set EWQLP's sensitivity curve slightly concave. Apologies to Nick; I should have written that Nick instinctively sensed it and recently suggested the curve of the software should be adjusted slightly concave.


Writing some of my thoughts to you as a way of encouragement to stick with it and do the best job you are capable of producing:

Yes, this sample library requires a good keyboard controller and a good computer with lots of RAM and a fast hard drive or two -- but the effort is worth it in the SOUND you are able to derive from it.

Yes, EWQL Pianos -- whichever of the four choices most closely suits your personal performing taste and requirements -- requires that you personally practice long and hard to wrest out your best performances ... but your efforts are worth it in the MUSICALITY you are able to derive from it.

As of this writing, IMHO there is no better virtual piano library than EWQLP -- hands down.

Cheers,

Joe

[Doug Rogers]

When I first saw the topic, I thought OJ had been released (that's the infamous OJ Simpson for our European friends).

Nice work Joe!!

- DR

[jcfelice88keys]

Hello Doug,

Until you wrote your reply about "how I did it", it hadn't occurred to me I was misquoting Mr. Simpson.

In order to be more inline with OJ's book title, perhaps I should rename the thread topic: "If I Did It, Here's How It Happened."

You made my day, Doug.

Cheers,

Joe

[A.Leung]

Great read- thanks Joe.

<>

One should buy a foot pedal to quickly change pages with programs like Sibelius for uninterrupted playing.

[A.Leung]

Originally Posted by Doug Rogers

When I first saw the topic, I thought OJ had been released

- DR

And the preQuel: "Why I did it."