GORDON: > I can't read either of those messages, because the Yahoo! Groups site appears to be pretty badly broken right now. STEVE: As for not being able to read messages, perhaps there is a problem with your computer. I recommend getting messages as a digest. That way you can re-read posts. For your convenience, I have copied the referenced STEVE post in its entirety below. (The part which you deleted in your reply was only a part of that.) GORDON: > What exactly did you do? I find that if I send everything in one go, the CZ-1000 crashes every time. STEVE: Once you figure out the problem you are having with the website take a look in the files section for the folder named "Using Midi-Ox with your Casio CZ." Those are the screenshots that go with http://launch.groups.yahoo.com/group/CZsynth/message/5876. Out of curiosity, which interface(s) are you using (and what computer/OS/SP)? If you are using one of those really bad 1x1 usb-midi cable interfaces, that might be your problem. Also, if you are simply capturing a patch dump and trying to send it to the CZ unchanged, that will not work. If you want me to test some short sysex strings on my CZ-5000, I can give it a try. If you have some patch dumps, load them to a temporary folder and I can give those a try. Steve http://launch.groups.yahoo.com/group/CZsynth/message/5876 follows: --- In CZsynth@yahoogroups.com, "steve_the_composer" <smw-mail@...> wrote: > > Even if its a moot point, I decided to test a usb-midi interface cable. I opted for the iConnectivity mio. [It probably stands for midi in/out.] > > The test was on an HP 64-bit Win7 SP1 laptop with an AMD Phenom II P650 Dual-Core Processor 2.60 GHz. (Not as fast as this year's models, but light years ahead of what was common in the 1980s.) I used midi-ox with default settings, with no delays added between bits or bytes. > > I did three tests: > (1) Receive and then send Glide sysex > (2) Request Sequencer Dump > (3) Request Sound Area Dump and then sent it back to the CZ. > > I have screen shots, which I will post when I am not so tired. I need to draw a few things on at least two of the shots and write up some explanations. > > Two important points for the moment: > (1) With the dump requests, I changed the keyboard to MIDI = 8. However, I set the sysex request using 70 [channel 1]. It worked. > (2) The patch dump needs to tweaked before it is sent back to the CZ. > > It will probably be easier to see when I post the screen shots, but basically here's a description. Consider this send request where the computer requests the CZ to send the patch data from the Sound Area: > > F0 44 00 00 70 10 60 70 31 F7 > > The 6th byte [10] is the code to request the CZ send a patch. The 7th byte [60] tells the CZ to send the data from the Sound Area. The 8th and 9th bytes [70 31] are an acknowledge ment from the PC to the CZ to continue after after the CZ sends a dump header: > > F0 44 00 00 70 30 > > Once the CZ receives the 70 31 from the PC the patch (aka tone data) is sent with a final F7 byte. > > Notice the 6th byte is now 30. > > In order to send that data back to the CZ, you need to change that 30 to a 20 and then >>INSERT<< a byte to tell the CZ the destination. > > In my case, I changed the header in the dump I received as follows: > > F0 44 00 00 70 20 60 . > > The 20 in the 6th place of the sysex command tells the CZ that data is coming. the 60 in the 7th place tells the CZ to put the data in the Sound Area. > > You cannot just send the data that was received from the CZ back to the CZ. Without the 20 and the destination, the CZ will not know what to do with the data. > > I am not sure if that provides the missing link to solve the original problem. The resending of the Glide mode change does not need to be changed like this. Not sure why it didn't work with the Hosa usb-midi interface cable. > > In sum, on my modern laptop, the iConnectivity mio successfully received and transferred a short sysex command and a single patch dump using midi-ox without inserting delays. Needless to say if I were writing a midi patch librarian program to receive and send 32 or 64 patches I would have a delay between each dump send. > > Also, if I were making a midi file of 16 or 32 or 64 sysex dumps to send to a CZ, I'd space them a bit in the file so there is some separation between them. > > By the way, when I requested the sequencer dump, I inserted 7 sets of 70 31 acknowledgement byte pairs even though for the sequence I had in the CZ two would have been sufficient. There was no ill-effect from sending the CZ too many "go ahead" commands. Once the CZ sent the F7 to say the sequencer dump was done, it seemed to ignore those bytes. > > However, I was not doing the dump and performing at the same time. If someone else wants to see what happens when they do data transfers at the same time as sending copious amounts of performance data, be my guest. I think Casio's infamous F7 screw-up might produced unpredictable results. I don't do bulk data dumps while performing and its not a cutting edge technique I want to explore. > > (On the other hand, I have explored putting in data that is outside the range of acceptable parameters. But that's a different can of worms.) > > Hope these tests and explanations helps someone, sometime. > > Steve > --- In CZsynth@yahoogroups.com, "steve_the_composer" <smw-mail@...> wrote: > > Even if its a moot point, I decided to test a usb-midi interface cable. I opted for the iConnectivity mio. [It probably stands for midi in/out.] > > The test was on an HP 64-bit Win7 SP1 laptop with an AMD Phenom II P650 Dual-Core Processor 2.60 GHz. (Not as fast as this year's models, but light years ahead of what was common in the 1980s.) I used midi-ox with default settings, with no delays added between bits or bytes. > > I did three tests: > (1) Receive and then send Glide sysex > (2) Request Sequencer Dump > (3) Request Sound Area Dump and then sent it back to the CZ. > > I have screen shots, which I will post when I am not so tired. I need to draw a few things on at least two of the shots and write up some explanations. > > Two important points for the moment: > (1) With the dump requests, I changed the keyboard to MIDI = 8. However, I set the sysex request using 70 [channel 1]. It worked. > (2) The patch dump needs to tweaked before it is sent back to the CZ. > > It will probably be easier to see when I post the screen shots, but basically here's a description. Consider this send request where the computer requests the CZ to send the patch data from the Sound Area: > > F0 44 00 00 70 10 60 70 31 F7 > > The 6th byte [10] is the code to request the CZ send a patch. The 7th byte [60] tells the CZ to send the data from the Sound Area. The 8th and 9th bytes [70 31] are an acknowledge ment from the PC to the CZ to continue after after the CZ sends a dump header: > > F0 44 00 00 70 30 > > Once the CZ receives the 70 31 from the PC the patch (aka tone data) is sent with a final F7 byte. > > Notice the 6th byte is now 30. > > In order to send that data back to the CZ, you need to change that 30 to a 20 and then >>INSERT<< a byte to tell the CZ the destination. > > In my case, I changed the header in the dump I received as follows: > > F0 44 00 00 70 20 60 . > > The 20 in the 6th place of the sysex command tells the CZ that data is coming. the 60 in the 7th place tells the CZ to put the data in the Sound Area. > > You cannot just send the data that was received from the CZ back to the CZ. Without the 20 and the destination, the CZ will not know what to do with the data. > > I am not sure if that provides the missing link to solve the original problem. The resending of the Glide mode change does not need to be changed like this. Not sure why it didn't work with the Hosa usb-midi interface cable. > > In sum, on my modern laptop, the iConnectivity mio successfully received and transferred a short sysex command and a single patch dump using midi-ox without inserting delays. Needless to say if I were writing a midi patch librarian program to receive and send 32 or 64 patches I would have a delay between each dump send. > > Also, if I were making a midi file of 16 or 32 or 64 sysex dumps to send to a CZ, I'd space them a bit in the file so there is some separation between them. > > By the way, when I requested the sequencer dump, I inserted 7 sets of 70 31 acknowledgement byte pairs even though for the sequence I had in the CZ two would have been sufficient. There was no ill-effect from sending the CZ too many "go ahead" commands. Once the CZ sent the F7 to say the sequencer dump was done, it seemed to ignore those bytes. > > However, I was not doing the dump and performing at the same time. If someone else wants to see what happens when they do data transfers at the same time as sending copious amounts of performance data, be my guest. I think Casio's infamous F7 screw-up might produced unpredictable results. I don't do bulk data dumps while performing and its not a cutting edge technique I want to explore. > > (On the other hand, I have explored putting in data that is outside the range of acceptable parameters. But that's a different can of worms.) > > Hope these tests and explanations helps someone, sometime. > > Steve > > --- In CZsynth@yahoogroups.com, "steve_the_composer" <smw-mail@> wrote: > > > > With all of the side issues, I am wondering if you are still stuck with the CZ not responding to sysex. > > > > I called my local Guitar Center and found I can buy a usb-midi "cable" and return it for full refund withing 30 days. They have one that is open, so I won't feel bad about returning it. > > > > If you are still trying to get the bulk data transfers to work, I can stop by Guitar Center and run the same tests I described. Let me know. > --- In CZsynth@yahoogroups.com, Gordon JC Pearce <gordon@...> wrote: > > On Sat, Aug 31, 2013 at 09:58:29PM -0000, steve_the_composer wrote: > > OK. I did the comparison. I used these posts: > > > > GORDON = http://launch.groups.yahoo.com/group/CZsynth/message/5928 > > STEVE = http://launch.groups.yahoo.com/group/CZsynth/message/5876 > > > > I have written some observations about the comparison. > > > > Executive Summary: > > 1. I did exactly what you said needs to be done. > > 2. I successfully achieved what you said couldn't be done. > > I can't read either of those messages, because the Yahoo! Groups site appears to be pretty badly broken right now. > > What exactly did you do? I find that if I send everything in one go, the CZ-1000 crashes every time. > > -- > Gordonjcp MM0YEQ >
Message
Re: Sysex and USB MIDI
2013-09-01 by steve_the_composer