yamaha psr443 won't receive midi data when playing back recorded performances


Joined
Apr 22, 2016
Messages
2
Reaction score
0
Also, when it does play back I can't change the voice to the one I used to record, it always plays grand piano.
any suggestions?

Using macbookpro
 
Ad

Advertisements

SeaGtGruff

I meant to play that note!
Moderator
Joined
Jun 6, 2014
Messages
2,555
Reaction score
1,049
First I'm going to give you "the bad news," then I'm going to tell you how to work around it-- so try not to freak out or get discouraged as you're reading "the bad news"!

The PSR-E models and their sibling models (YPT, DGX, and YPG) are designed such that-- with certain exceptions-- MIDI messages generally don't affect the keyboard. On page 3 of the PSR-E443 MIDI Reference it says the following:

NOTE:
1 By default (factory settings) the instrument ordinarily functions as a
16-channel multi-timbral tone generator, and incoming data does not
affect the panel voices or panel settings. However, the MIDI messages
listed below do affect the panel voices, auto accompaniment, and
songs.
• MIDI Master Tuning
• System exclusive messages for changing the Reverb Type and Cho-
rus Type.
In other words, you can send MIDI messages to the keyboard, or play back MIDI files on it, and the keyboard will in fact play the MIDI messages for all 16 channels of MIDI data, since it's a 16-channel multi-timbral tone generator.

However, the sounds that the keyboard generates in response to the MIDI messages will be separate from-- and (other than the exceptions listed above) will have no effect upon-- anything you're playing "live" on the keyboard.

What this means is that "By default (factory settings)" you cannot be playing the keyboard, send it a patch change (Bank Select MSB, LSB, and Program Change), and have the voice you're already playing with change to the voice specified by the patch change. (I know-- "Aaaaaarrgghhhh!", right?) That also goes for other cool MIDI events you might want to send to the keyboard, like modifying the EG Attack Time, EG Release Time, LPF Cutoff Frequency, and LPF Resonance.

Now, the interesting thing is that Yamaha makes some apps for the iPad which can, in fact, do exactly what the MIDI Reference says you can't-- e.g., the "Sound Controller" app lets you change the Modulation, Attack Time, Release Time, Cutoff Frequency, Resonance, etc., and those changes do affect the panel voices that you're playing with! So apparently there's a way to change the "default (factory settings)" such that the panel voices and panel settings will be affected by MIDI messages. Unfortunately, I haven't yet figured out how that's done.

Someone on another board pointed me to the various "Receive" messages described in the XG documentation, which he thought were supposed to let you control whether or not the panel voices would respond to MIDI channel messages, as well as controlling which types of channel messages were to be received or ignored. However, I tried using those messages and they didn't work as we'd hoped-- which I was afraid of, since the XG documentation states that the default settings for those messages are such that they will allow the MIDI data to affect the XG "multi-part parameters." Consequently, I think the real purpose of those messages might be to let you ignore certain channels if you wish, such as allowing channels 1 through 4 to be played by the keyboard but not the other channels.

Fortunately, I know of at least two ways that you can work around this behavior. Each method has its own advantages and disadvantages.

One method is to connect the keyboard to a computer or tablet-- which you're presumably doing, anyway, if you're trying to change patches by sending MIDI events to the keyboard-- then turn off the Local Control setting and route the keyboard's outgoing MIDI events back to itself (i.e., "bounce" the MIDI off of a DAW or other MIDI utility, so to speak). What this does is silence the keyboard's panel voices (i.e., the Main Voice, Dual Voice, and Split Voice) so that playing on the keyboard doesn't produce any sounds directly, but the MIDI Note events generated by your playing will be sent to the DAW, then sent back to the keyboard-- along with any other MIDI events you're having the DAW send (e.g., patch changes)-- such that the keyboard will play those notes using the desired voice and other settings.

The obvious disadvantage of this method is that the keyboard must be connected to a computer or tablet. (Actually, if the PSR-E443 had MIDI In and MIDI Out ports then you could just use a 1-foot MIDI cable to connect those ports together without needing to be connected to a computer or tablet.)

But a big advantage of this method-- aside from the obvious one of allowing you to send patch changes to the keyboard-- is that (depending on your DAW) you can probably create more splits and layers than what the PSR-E443's panel controls allow. For instance, you might be able to set up the MIDI tracks in the DAW such that a certain range of notes on channel 1 are sent back on channels 1, 2, and 3, letting you layer three voices together. Then you could have a different range of notes be sent back on channel 4, and another range of notes be sent back on channel 5, letting you play using three split areas. Much depends on the what the DAW will let you do. However, if you do something like this then you have to be careful not to exceed the keyboard's 32-element polyphony. (I say "32-element" rather than "32-note" because some voices use 2 sound elements, thereby reducing the effective polyphony to 16 notes.)

The other method is to use the keyboard's "One Touch Setting" feature. This feature works differently than on the PSR-S models, because you don't have four OTS buttons for selecting between four possible OTS voices in a style or song. Instead, you must select "Voice 000" when playing a style, pattern, or song, which sets up the panel voices according to the OTS settings in the style, pattern, or song-- assuming it contains any OTS settings. Furthermore, the OTS settings are stored differently in the styles, because style files for the PSR-S models have a special "OTS" section which holds the settings for the four OTS voices. The PSR-E443 can't use the OTS data that's in a style's OTS section. Instead, the settings for the OTS voice must be placed inside the first measure of the song or style. When you're playing a song or style and you select voice 000, the keyboard will apply the OTS settings to the panel voices-- the channel 1 settings will be applied to the Main Voice, the channel 2 settings will be applied to the Dual Voice, and the channel 3 settings will be applied to the Split Voice. I haven't actually tried this with the Dual Voice and Split Voice yet, but it definitely works with the Main Voice. I suspect that perhaps only those settings which can be selected for the three panel voices with the panel controls will actually be applied-- e.g., there's no panel function for setting the "Main Voice Modulation," or the "Split Voice Attack Time," so I'm thinking that perhaps those settings might not get applied, since there are no corresponding panel settings for them to be applied to. But it might work. I've only recently discovered this "One Touch Setting" trick, and I still need to do some experimentation to see what is possible or not possible with it.

The disadvantage of this method-- aside from the possibility that it might not work on "non-panel" settings-- is that you can't switch voices quickly, because to apply the OTS settings you must press the "Voice" button to switch to Voice mode (if you weren't already in Voice mode), then enter "000" on the number pad to select the OTS voice.

On the other hand, a big advantage of this method is that you could create hundreds of "song" files which are actually just voice setups-- i.e., the song files wouldn't contain any Note events (unless you want to include some), just the initial "setup" measure that selects the voices, effects, and voice setups to be used, preferably with only one setup per "song" file. Then you could select the desired "song," switch to Voice mode, and enter "000" to activate that setup-- a bit like registrations, but with the ability to store hundreds of setups on a USB thumb drive and choose between them, as opposed to having only 32 registrations.

As I said, I haven't figured out all of the possibilities yet, so for now the better method might be to turn off Local Control and bounce the MIDI Note events off of a DAW, because with the right DAW and MIDI controller (e.g., a "launch pad" type of controller) you should be able to trigger patch changes and loops with a press of a button.
 
Joined
Apr 22, 2016
Messages
2
Reaction score
0
When I choose a voice from the psr443 (ex. strings) and record the midi information in my DAW and try playing it back, it only plays grand piano, not the string voice I chose to record. why?

I'm using a Macbookpro with Protools
 
Ad

Advertisements

SeaGtGruff

I meant to play that note!
Moderator
Joined
Jun 6, 2014
Messages
2,555
Reaction score
1,049
It sounds like you recorded the MIDI without capturing any patch information-- i.e., the Bank Select, Program Change, etc. When you record a MIDI track that doesn't have any patch information, it defaults to Program 0, which is the Acoustic Grand Piano.

What you could do is manually insert the desired Bank Select and Program Change events at the beginning of the MIDI track.

However, your keyboard has an "InitSend" function that will send all of the pertinent settings to your DAW so you can record them. If you've already recorded a track, you don't have to record it over again from scratch if you don't want to-- just add another MIDI track and record the "InitSend" information to it. Then you should be able to copy and paste the MIDI events from the "InitSend" information and paste them at the beginning of the track you've already recorded.

In the future, when you're getting ready to record a new MIDI track, it would be a good idea to start by using the "InitSend" function to send all of the setup information to the DAW for recording, then starting playing after the "InitSend" function has finished executing.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top