Playing back MIDI files through the PX560

[Rusty]

Hi and thanks in advance

 

I would like to play back MIDI scores (say from Reaper) through my PX560, and need to understand a couple of things.

 

1. the voice mappings of PX560 voices to MIDI - so that I can call the correct voice

2. do I need to assign each voice (piano, cello, flute etc) to a separate channel?

 

I've been looking but haven't found a guide that points me at what I need to set up a score to play back through the PX.

 

Would greatly appreciate the forum's advice, pointers & links to relevant information.

 

Rusty

[Jokeyman123]

Take a look at the manual if you have it-on page A-5 in the appendix-this shows how the PX560 channel assignements are arranged. This is one of the trickier aspects of the 560 to understand-as there are 3 possible 'groups" of midi channel routings and assignments. an important distinction-the internal midi/song recorder will use tracks in part "C"-so when you play or record a song from the 560-you must go to the mixer screen, and select "C" to monitor levels, pan, effect sends and tones, as the built-in recorder will record and play back form that group of midi channels-unless you record parts in the system track-you will see those tracks in the mixer screen part "A" or "B"-as you play live from the keys, and record to that track. Confused? remember, there are 48 possible midi assignment channels-very powerful since the 560 can employ up to 256 note polyphony. Now to answer your question.....finally....when you play a midi file from a computer connected-to the 560-it will route those tones and settings from your DAW to mixer part "B" and to monitor what is happening as you play-again to the mixer screen and switch to part "B". One more important detail at least from me, I'm sure others will weigh in-unless you prgram "bank change" and "program change" assignments in your midi file-with your software DAW or midi file player/recorder-the PX560 will revert to the basic "GM" soundset-not the best selection of sounds. In other words, "smf" or .mid files-unless you specifically program the bank and program changes in that file-will send the basic "GM" tones to the 560-it is only selecting what it recognizes-even if you cange the tones, pan, levels etc. from the 560 mixer screen-which you can do-the midi file will always reset the 560 back to whatever assignments for tones were in the original midi file. One other detail-unfortunately you cannot use the internal 560 mid/song recorder to record from its midi in port, but you can import a midi file to the 560, change the settings from within the 560 and save the changes as your edited file for playing specifically with the PX560 tones and other settings.

 

Alot to understand-but then I look at the 560 as almost a full "workstation" although Casio did not specify that it is-it is a stage piano, with some very nice additional 'workstation" type features.  There is one more aspect to this you might want to use.There is a file that you can import into Reaper designed to set Reaper to a default group of settings/tones designed specifically for the PX560. Once you install this file-you will now see exact tones in reaper that will match the PX-560 tones. In other words-when you load a midi file into your software-you will be able to set all the individual instrument tones in the software so these match the tones you would like to hear in the 560. remember to switch the mixer screen to part "B" to see what is playing live from your midi file. Remember-the 560 can't record this file to its internal recorder-it is acting on part "B" midi channel assignments as a separate tone module-and you can record to the song recorder live at the same time as your midi file is playing, pretty nice I think. And if you keep all your parts in sync-now you can record all this into the "audio recorder" in one shot. Hypothetically-16 DAW and 16 PX560 song recorder tracks combined can all be played back and recorded into the audio recorder. I haven't done that...yet...but it is supposed to work that way.

 

I think I got this all right-but its early-certainly open to corrections!!!

 

Here it is.....

 

https://www.casiomusicforums.com/index.php?/files/category/42-privia-px-560/ Scroll down the page look for PX-560 patch script. 

[Rusty]

Awesome!  And thanks so much.

 

It'll take me a while to digest all that, but thanks for reaching out & sharing this.

 

I'll come back with a progress report as I try to put this into action.

 

Thanks again

[Vorbis]

Question regarding polyphony.   Do GM midi tones count towards the polyphony count?   Example on the WK7500, that keyboard is a 64 note polyphony but in drawbar organ mode that drops down to just 21 note polyphony.   Plus they drop a few polyphony notes if it's doing a file playback or other functions like playing a DSP.   It is my understanding that if you exceed your maximum note polyphony, the keyboard drops any additional notes played till the demand drops.   I would think this applies to all Casio products but for example in the case of the PX560 with 256 note polyphony  getting to that number would be tough to approach but with layers and other programming it might be feasible to max out.  In listening to some songs playback and with a drawbar organ being played along side,  I cannot say I have ever heard any notes drop out.  Jokeyman your description and channel assignments  brings up some valid concerns however.

[Jokeyman123]

Technically-a "voice" in the somewhat vague definition of what a voice is-usually is defined in the classic synth sense as a distinct oscillator-a monophonic oscillator that will create one "voice" or tone at a time.  At least as far back as when synths were using "discrete" components for one oscillator circuit-I know this as I had to build simple analog oscillators when I first began studying as a tech-to create basic sine, sawtooth or square wave oscillator.s I haven't studied this voice allocation on the Casios much-but many synths have or had  "voice allocation" schemes-when synths still did not have too many discrete oscillator cicuits for early multi-track playng and recording-and each designer had slightly different versions of how voices were allocated when you ran out of oscillators-might be for example-first note played would be first note "stolen' or cut off to play an additional note. So on the WK-it must be using once "voice" or oscillator per drawbar-wbich makes sense as each voice in a drawbar organ is adding or taking away sine waves to create the various tonewheel simulations-or whatever wave Casio is using for each drawbar. As far as other Casio tones-some are using multiple wav samples to create one tone-the stereo pianos i think might be using 4 voices-2 for the lower octaves and 2 for the upper octaves, and in stereo. Each individual wav sample must use one oscillator or voice. Mike Martin would know if I am correct or not. And other Casio tones might be doing the same-I am not sure if the voice "count" is including the oscillators that are used for example for the LFO, phasing and other effects, but since effects such as phasing are also created (again I had to build these circuits form scratch early on) using oscillator circuits, it might be. And yes, of course the GM tones will use a voice, although I'm pretty sure some of the better stereo tones beyond the basic Gm tones are using more oscillators than the GM tones. So again technically, you should be able to play more complex arrangements or live parts with accompaniments if you are using mostly GM tones, which the arranger functions I believe are doing already as are almost all smf standard midi files, unless you change the bank change designated numbers.

[Vorbis]

I took some time today to play a MIDI file that has 14 tones in GM that have up to 23 notes triggering at the start of a measure.   I also switched the WK7500 into drawbar organ mode that should have reduced the polyphony down to 21.   I played all 10 fingers along with the MIDI file playing.  I still could not detect any dropped notes.   I'm pretty sure however that notes were dropping off somewhere.

 

Doing some research today on polyphony, it seems the electrical engineers have some options to impact how the instrument sounds.   They can drop a beginning note that started or is sustaining in a FIFO (first played first out).  Or they can drop a note LIFO (last played first out) and hide the dropping notes.   Some of the luxury end equipment can analyze and drop a note being played as a harmonic.   Seems that high end keyboards that have hex layers and multi polyphony to make up a single note can zoom past a 256 note polyphony with ease.    Throw in some auto accompaniment, sustain, stereo DSP and we're really taxing the system at this point.   Seems Casio has addressed this with their keyboards and minimized any impact as far as I can detect.

 

Your description of the resulting note log jam is spot on.  How are you supposed to make a purchase decision with so much happening behind the scenes?    Playing the keyboard before purchase would maybe reveal any hardware deficiency.

 

[Brad Saucier]

The virtual drawbar organ on Casio products is always using DSP to generate sound, rather than samples.  The 64 note polyphony count is referencing wave sample playback.  I'm thinking Casio is saying the drawbar organ itself has 21 note polyphony, in addition to the 64 note wave sample polyphony.  So, 85 note polyphony total?

[Rusty]

Getting there, will write it all up later today (got the PX560 patch file, enabled Reaper for CASIO MIDI & program changes, triggered  individual voices as desired, but with multiple Reaper tracks one voice will affect voices on other tracks)  -

 

1. what port should the MIDI in port be ideally set to - A, B or C?    Or does it not matter?

2. should I send each voice / part / stem / track to separate MIDI channels?

 

3. Should the Keyboard Channel parameter in the MIDI section (second screen of menu's) remain at 1 (default) or be changed to something else? 

 

Thanks to all for your input & advice, greatly appreciated

[Vorbis]

4 hours ago, Brad Saucier said:

The virtual drawbar organ on Casio products is always using DSP to generate sound, rather than samples.  The 64 note polyphony count is referencing wave sample playback.  I'm thinking Casio is saying the drawbar organ itself has 21 note polyphony, in addition to the 64 note wave sample polyphony.  So, 85 note polyphony total?

 

I've used the WK7500 as a MIDI file player in small venue gigs to play backing tracks.   The Casio WK has never been a problem using it in that way and  I will play along with the file as it's playing.  So as you suggest, maybe the engineers have something else going on inside on all the keyboards.  My first thoughts were MIDI files playing did not use any of the polyphony.   I do believe the MIDI tones indeed use polyphony notes as the file plays.

 

The drawbar organ has 9 slides and percussion settings as well as any DSP effects you program on top.   I could see the organ using 21 notes of polyphony easily.  I would see that as a reserve for the organ to be played not a limiting set up to eliminate a 64 note polyphony or more depending on which Casio Model number you have the drawbar organ feature.   Keyboard manufacturers still like to publish the number of polyphony notes available.  

[Jokeyman123]

Then Brad-how does it generate a tone with a DSP? isn't the DSP itself a form of an oscillator?

 

For example-when I design a sine wave oscillator (I've built these early on) to be used for an LFO-or a ring modulator that generates sum and difference harmonics to create its effect-aren't these oscillators? If not, I'm stumped as to how these are being created-even if it is all in the realm of a software oscillator-a program to simulate a hardware oscillator-is it using a voice in this sense, or something else? it is difficult to know, since the CPU is (I think) generating all the code to create all the waves-whether simulating the classic sine, sawtooth, or square waves-or LFO's for vibrato/tremelo or even phase distrotion or flanging.

Would be interesting to know-but then, this may be Casio code that is purposefully not being allowed to be "open source" for developers to tamper with. after all, many programmers know theory behind the Casio phase distortion voice generation-but unless i'm mistaken, i don't think Casio ever revealed the software routines used to create PD synthesis. How virtual CZ machines are duplicating PD, I have no idea unless one is playing sampled examples of the Casio CZ tones in a virtual Casio instrument, or someone has been able to replicate the actual software routines to generate the various PD waveforms. And Ihaven't even mentioned AHL used in the WK7600 (I told you not mention that!). Compounding this riddle I myself have probably created-is that many Casio tones-whether in the PX Privias, CTKs, CTX's or WK's-have DSPs as part of the tone-I wonder if these are using an oscillator for the tone, and another for the DSP built into it, or not. Might even be the same oscillator generating the tone and the DSP attached to it. Yeah, I'm fun at parties.....

[Brad Saucier]

I don't know for sure.  Things are always changing and the details are blurry.  It's something only an engineer at Casio can explain.  My take (and I'm only speaking of modern Casio products made in the past 20 years or so): Sine waves, saw waves, etc. are usually from wave samples on modern Casio products.  Aside from the drawbar organs, and the PWM oscillator in the solo synth section of XW synths, which I'm almost certain is also DSP, it's all done with samples, read from sample ROM. 

 

Another use of DSP generated sound is in Casio digital pianos, the piano resonance effects.  I've always felt this is partially why you don't see drawbar organ on Casio's digital pianos.  The DSP resources have been repurposed to piano resonance.  DSP is pretty amazing with what it can do.  

 

MZ-2000 for example.  It was the first model with drawbar organ.  It has all 9 sine waves separately adjustable, plus percussion and key click.  If this was done with waveform samples, each note played would be using over 9 samples, using 12 notes of polyphony.  The MZ published spec is 64 notes polyphony.  You can put a drawbar organ in 2 zones, a dual manual type setup, each one with it's own drawbar settings, and play big chords, run auto accompaniment as well, and it will handle all of it.  There is no way 64 notes of polyphony could handle all of that if the organ was using 9+ notes of polyphony per key.