Midi message for changing bank - and some frustration

DAW's, Audio, MIDI, VST's and all other music software.

Moderators: Derek, parametric, Saul

User avatar
SeaGtGruff
Global Moderator
Global Moderator
Posts: 624
Joined: Wed Jan 29, 2014 4:24 am
United States of America

Re: Midi message for changing bank - and some frustration

Unread post by SeaGtGruff » Sat Apr 22, 2017 9:50 am

I think modern Yamaha models (PSR-S and Tyros) generally default to 1920 PPQN.

On the other hand, when I record a User Song on my lowly PSR-E models, the keyboard stores my performance using a measly 128 PPQN to help save memory by keeping the delta times small! Then, when the internally-stored User Song gets written to a USB thumb drive as a Standard MIDI File using the keyboard's "Save SMF" function, the 128 PPQN delta times get converted to (if I remember correctly) 384 PPQN delta times.
Michael Rideout
Current keyboards: Yamaha YPT-400, PSR-E433, PSR-E443, PSR-EW400, MX49 BK
Current controllers: M-Audio Axiom 61-II
Previous keyboards: Farfisa Matador 611; Casio CTK-710
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Sat Apr 22, 2017 12:07 pm

Ah, yes, Michael!

384 PPQN would be a good one, divisible by three to render good timing for triplets.
I suspect you're absolutely right about 1920 for the latest models, I'm quite sure I'm living in the past with my copy of Sonar 7 PE from many years ago.

Here's a screen shot from HexEdit showing the delta time of '05H' from the sysex file I created with 4/4 time sig., 120 bpm and 480 TPQN.
Each successive sysex mesage was spaced at 5 ticks.

Image
Last edited by SysExJohn on Tue Aug 08, 2017 4:44 pm, edited 1 time in total.
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
User avatar
SeaGtGruff
Global Moderator
Global Moderator
Posts: 624
Joined: Wed Jan 29, 2014 4:24 am
United States of America

Re: Midi message for changing bank - and some frustration

Unread post by SeaGtGruff » Sat Apr 22, 2017 8:14 pm

Regarding 384 TPQN, I thought it was a very oddball value when I first saw it used by my PSR-Es, but that was before I discovered that User Songs are recorded and stored internally on the PSR-Es using 128 TPQN! Now that I know that, the 384 TPQN value makes more sense.

As for 1920 TPQN, most modern Yamaha PSR-S and Tyros models definitely use 1920 TPQN by default, but I think that many DAWs still top out at 960 TPQN.

Two more observations about TPQN:

(1) If a manufacturer recommends waiting a certain number of ticks after a particular MIDI message, they really mean a certain number of microseconds, so you might need to adjust the specified number of ticks by some factor depending on the TPQN setting that you're using, since 5 ticks at 480 TPQN is going to be four times as many microseconds as 5 ticks at 1920 TPQN.

(2) As parametric pointed out, there's no such thing as truly simultaneous events with MIDI, since it uses serial communication rather than parallel. I've read that some musicians claim to be able to hear the slight timing differences between the parts being played on different MIDI channels, although I suspect they're most likely referring to older synths and sound modules that had lower maximum TPQN capabilities. In any case, an important benefit of using a higher TPQN, such as 1920 TPQN, is that the parts being played on different MIDI channels-- and possibly being played by separate synths-- can be more closely synchronized.
Michael Rideout
Current keyboards: Yamaha YPT-400, PSR-E433, PSR-E443, PSR-EW400, MX49 BK
Current controllers: M-Audio Axiom 61-II
Previous keyboards: Farfisa Matador 611; Casio CTK-710
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Sun Apr 23, 2017 10:03 am

SeaGtGruff wrote: Two more observations about TPQN:

(1) If a manufacturer recommends waiting a certain number of ticks after a particular MIDI message, they really mean a certain number of microseconds, so you might need to adjust the specified number of ticks by some factor depending on the TPQN setting that you're using, since 5 ticks at 480 TPQN is going to be four times as many microseconds as 5 ticks at 1920 TPQN.
Yes, quite so, and also tempo.
When authoring MIDI files I always used to use 480 TPQN and start the first measure at 120 bpm. The actual song tempo goes into measure 2.
Doing the math, a measure takes 2 seconds, one beat therefore takes half a second, working from that, one tick is approximately (N.B.) one millisecond.
Reading one of Yamaha's XG specs., way back (from memory), they wrote that a 'GM system on' message can take up to 200 milliseconds to implement.
(It sounds a lot, but think, an MU128 has 4 MIDI ports, that's 64 channels where all the MIDI initial settings have to be reset to default).
An XG system on can take 50 milliseconds.
SO...
put the GM system on at 01:01:000 (it doesn't matter about the other meta events at the same point, they are read when the file is opened), then the next event goes at 01:01:200. If it's an XG system on, then put the next event at 01:01:250.

(But if the tempo is increased to 240 bpm, then each tick is roughly (just less than) half a millisecond.
At 60 bpm each tick is 2 milliseconds
.)

Tsk! Simples. 8)
SeaGtGruff wrote: (2) As parametric pointed out, there's no such thing as truly simultaneous events with MIDI, since it uses serial communication rather than parallel. I've read that some musicians claim to be able to hear the slight timing differences between the parts being played on different MIDI channels, although I suspect they're most likely referring to older synths and sound modules that had lower maximum TPQN capabilities. In any case, an important benefit of using a higher TPQN, such as 1920 TPQN, is that the parts being played on different MIDI channels-- and possibly being played by separate synths-- can be more closely synchronized.
I assume you mean 'ports' when you say "channels".
All messages on the MIDI interface, no matter what channel they're addressed to, have to wait in line to be serially transmitted in the order in which they were generated.
Only if they're distributed across multiple ports does the possibility of parallel messages become possible.
BUT...
If you haven't got a multi-core processor the program can only generate the messages serially! :?
And if the program doesn't distribute the message generation acroos the processors? 8O
What if the multiple devices are connected to a multiport MIDI device across USB. Oh oh! USB is serial too. Aaaaargh!
Of course, these days programs run SO much faster than a MIDI interface that, potentially, thousands of messages could be generated in second.

But, as usual, WAY too much detail. :lol:

JohnG.
Last edited by SysExJohn on Sun Apr 23, 2017 10:15 am, edited 2 times in total.
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
Outof
Member
Member
Posts: 22
Joined: Mon Dec 14, 2015 10:42 am

Re: Midi message for changing bank - and some frustration

Unread post by Outof » Sun Apr 23, 2017 10:12 am

I really see I have a lot to dig into. I am still reading my way up - Johns Midi-tutor was really a great advice and a nice place to start, thank you. Always nice when people use their pedagogy skills when explaining.
However, maybe some people will look at this in the future in their search for solutions. I am now very happy to pronounce that I have found one, yuhu. Well, at least one of two common ways of doing it. The scanned "Program change chart" from Sonus and the little more recipe way of explaining it from SysExJohn did the trick. The result goes like this:
Solution.png
Whit this programmed I succeeded to automaticly let the Yamaha W7 play one note with "P.." (Preset voice bank 2), and the seccond note with "I" (Internal bank).

However, I had some problems with playing the first not correctly though. And this is also relevant for the ticks issue i guess. I thought I had inserted enough tick "space" between each event, but with this setup - my first note did not have the right attack. I thought the mentioned 5 ticks delay by SeaGtGruff should be enough, but I am not shure. Maybe I have it wrong and have to read it again. However, it was solved moving the note forward.
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Sun Apr 23, 2017 10:25 am

Great stuff OutOf,

I feel I have to say, Logic displays its M/B/T song position in a very weird way IMHO.

It takes a little time for it all to sink in, but "it's Logical Captain".
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
Outof
Member
Member
Posts: 22
Joined: Mon Dec 14, 2015 10:42 am

Re: Midi message for changing bank - and some frustration

Unread post by Outof » Sun Apr 23, 2017 10:37 am

I have not succeeded in the second way doing this. I guess this is the best and most comfortable way doing it, since it is possible to change bank and choose a specific (named) voice from the hardware synth just by sitting in the studio chair.

What have I done wrong here? I guess this do not work because I really don't know what should be happening here. When choosing Bank 1, I have succeeded (or did it happen by default) to let it choose/find the GM bank. I have also succeeded in changing to all different voices (like Celesta) in the voice menu. However, choosing Bank 1, 2, 3, 4, or more (I am playing around with the bank select button underneath the 8 = Celesta place) - just give me a "off" display on my hardware synth:
Solution_Close.png
Do I miss some programming here?
User avatar
SeaGtGruff
Global Moderator
Global Moderator
Posts: 624
Joined: Wed Jan 29, 2014 4:24 am
United States of America

Re: Midi message for changing bank - and some frustration

Unread post by SeaGtGruff » Sun Apr 23, 2017 5:31 pm

SysExJohn wrote:I assume you mean 'ports' when you say "channels".
All messages on the MIDI interface, no matter what channel they're addressed to, have to wait in line to be serially transmitted in the order in which they were generated.
Only if they're distributed across multiple ports does the possibility of parallel messages become possible.
No, I did mean channels. I was thinking of the "old" way of doing things where people send MIDI to one device and have it play a channel, then use MIDI Thru to send the MIDI on to another device so it can play one of the other channels, and so on. Of course, since the development of 16-part multitimbral sound modules and instruments, one device can play all 16 channels. Yet there are still plenty of new synths being designed and manufactured which aren't able to generate 16 different timbres at once, so the practice is still relevant today.

EDIT: When I said "more closely synchronized," I was referring to the number of microseconds between two or more consecutive "Note On" events, where the second and subsequent events have delta times of 0, such that they're intended to be played "simultaneously." At 480 TPQN an attentive ear might indeed be able to hear that the notes are being sounded like a fast arpeggiated chord, one after the other-- especially if the notes are on different channels and each channel is being handled by a different device in a daisy chain, since I assume that would add a tiny bit more latency due to the MIDI having to travel a bit further to reach the next synth down the chain, and the next, etc. But at 1920 TPQN the "simultaneous" notes won't be separated by as many microseconds-- they'll still be played like a really fast arpeggiated chord, due to the serial nature of MIDI, but they'll be closer together in time-- not truly simultaneous, but more closely synchronized.
Michael Rideout
Current keyboards: Yamaha YPT-400, PSR-E433, PSR-E443, PSR-EW400, MX49 BK
Current controllers: M-Audio Axiom 61-II
Previous keyboards: Farfisa Matador 611; Casio CTK-710
User avatar
parametric
Global Moderator
Global Moderator
Posts: 3056
Joined: Mon Jan 18, 2010 1:00 am
Contact:
Great Britain

Re: Midi message for changing bank - and some frustration

Unread post by parametric » Sun Apr 23, 2017 11:59 pm

Its probably a good time to mention here that too many iterations of midi data via MIDI thru to additional devices CAN result in latency that may

become troublesome.

Although we're talking about light-speed, nothing is for nothing, and there WILL be SOME latency with the data travelling in and out of TWO devices

before it reaches the third IYSWIM . . ?

I have heard it said that beyond the THIRD device this may become a problem.

BEST arrangement therefore is to have a midi HUB, so that the path to EACH midi device is no more than:

Source (Keyboard) - Hub - to Target device . . . .

This is probably only an issue if you are planning to have LOTS of midi-modules attached . . .

parametric
Alesis Fusion 8SSD AND 6SSD - BOTH are 384Mb/120Gb SSD/Akai ADVANCE61/Yamaha MOXF6/1024Mb Flash Ram/Yamaha SY85/8.5mb vol/1024k non-vol/DX21/Roland MT32/Bachmann double overstrung Baby Grand Piano/Win10 Pro/Ubuntu MATE 15.0.4/iBook G4/Mac OS 10.4.6/ProTools 7.4/MBox2/M-Audio MicroTrack 24/96

NI Komplete11 Ultimate

Sector101 2x SYEMB06 / 4 x EXM-E3 128MB DRAM Module

BRAND NEW DSDD (720k) FLOPPY DISKS FOR SALE - http://www.yamahaforums.co.uk/forum/vie ... =22&t=9217

Watch out now! take care, BEWARE of the greedy leaders! They'll take you where you should not go - (George Harrison)

IT'S TRUE - "MONEY TALKS" - TO ME, IT MOSTLY SAYS "GOODBYE" ;-)
http://www.chrisnmiller.co.uk/Chris
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Mon Apr 24, 2017 10:46 am

As Chris has pointed out, 'cascading' MIDI devices, out to in, thru to in and so on can introduce its own problems.

The whole MIDI stream is intercepted by device one and is then regenerated, I'm sure you've looked at the mechanism in the MMA spec.

Image

(Starts ramble...

Through the opto-isolator then off, via pin 4, to the UART for the interrupt driven input where the first device, set to channel 1, obeys just those messages pertinent.
But the program/processor still has to read the channel number off every message before ignoring many.
Meanwhile, all channels are forwarded, via pin six, and regenerated for the thru socket.
The next device processes all channels, and obeys perhaps channel 2, possibly forwarding all channels again via the thru.
There will inevitably be some small measure of delay introduced, although it should be very small, but possibly noticeable as one gets further down the chain.

The interface, being balanced, should not be susceptible to noise to any large extent, providing that manufacturers conform to the suggested earthing arrangements.
In my experience many do not, and I've found frequent examples of pin 2 of the DIN socket on the MIDI in being connected.
Only out and thru should pin 2 be connected.

With regard to MIDI messages with no delta time between them, the TPQN is irrelevant.
A well written program should output the messages to the transmission buffer as fast as it can.
As there is no inbuilt delay between the messages in the file, they should be processed in order and output probably in microseconds.
(And this is why it's important, in my view, to make sure there's a tick between the bank select messages and the ensuing Program Change.
Otherwise, if the file is edited, the messages could become disordered and one, other or both Bank messages could follow the program change.
Editing a bank number within three co-located messages will normally move that message to be the last of the three.
)
These days much faster than the MIDI transmission rate.
It will then be up to the UART to stream those messages, one after the other, as nine bytes, for three simultaneous note on's, of data.
So they should all arrive within three milliseconds. Which, I would have thought, would be unnoticeable.

The problem, I believe, that people experience today, is an unrealistic expectation of the performance capability of old hardware.
Let's take an ancient drum machine for example.
Possibly an eight bit processor. designed to work like a drummer.
So, no more than four simultaneous hits. Simultaneous???? Have you ever met a drummer that good?
Now we want the same device to process, bass, hi-hat, half a dozen toms, a couple of cymbal hits and a snare roll all at the same time.

My own belief, and it is just that, is that people's expectations are set too high.
I was always taught to work within the limitations of the equipment you are using.
So try experimenting with the order in which simultaneous notes will be processed, maybe put some a tick earlier or later.
If one checks out human playing one finds it is often wildly far away from the exact beat.

Sure we could use multiple in/out sockets, but at what cost to real estate on the back of the machine, not to mention cost.
And would we then need multiple processors to handle all the interfaces, or is it operated by interrupt, thus negating any advantage gained?
Or, we could design a parallel MIDI socket.
The problem with parallel sockets is that, as speeds get higher, so timing errors between the various channels become problematical.
And cost.
That's why all high speed links are serial these days, think USB, FireWire, SATA, etc.

I've lost what I was originally trying to say. :? :roll: :wink: :lol:
...Ends ramble.)

JohnG.
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Mon Apr 24, 2017 3:42 pm

Oh, yes!

A MIDI hub. Quite so, Chris.
It takes the output from one MIDI port (might be USB) and splits it out to multiple MIDI interfaces dependent upon channel. Clever.
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
User avatar
SeaGtGruff
Global Moderator
Global Moderator
Posts: 624
Joined: Wed Jan 29, 2014 4:24 am
United States of America

Re: Midi message for changing bank - and some frustration

Unread post by SeaGtGruff » Tue Apr 25, 2017 4:54 am

Outof wrote:I have not succeeded in the second way doing this.
Outof wrote:When choosing Bank 1, I have succeeded (or did it happen by default) to let it choose/find the GM bank. I have also succeeded in changing to all different voices (like Celesta) in the voice menu. However, choosing Bank 1, 2, 3, 4, or more (I am playing around with the bank select button underneath the 8 = Celesta place) - just give me a "off" display on my hardware synth:
Do you have the Music Synthesizer W5/W7 Technical Information document from Yamaha? I'm looking at page 40-- which is apparently page "42 of 63" in the PDF file, due to the way the printed pages are numbered-- and it gives the Bank Select MSB/LSB values as follows:

GM Normal Voice: Bank Select MSB = 0, LSB = 0; Program Change = 0 to 127
PRESET Normal Voice: Bank Select MSB = 1, LSB = 0; Program Change = 0 to 127
INTERNAL Normal Voice: Bank Select MSB = 2, LSB = 0; Program Change = 0 to 127
SONG Normal Voice: Bank Select MSB = 3, LSB = 0; Program Change = 0 to 127
EXTERNAL Normal Voice: Bank Select MSB = 4, LSB = 0; Program Change = 0 to 127

GM Drum Voice: Bank Select MSB = 127, LSB = 0; Program Change = 0, 8, 16, 24, 25, 32, 40, or 48 *
PRESET Drum Voice: Bank Select MSB = 17, LSB = 0; Program Change = 0 or 1 *
INTERNAL Drum Voice: Bank Select MSB = 18, LSB = 0; Program Change = 0 or 1 *
SONG Drum Voice: Bank Select MSB = 19, LSB = 0; Program Change = 0 or 1 *
EXTERNAL Drum Voice: Bank Select MSB = 20, LSB = 0; Program Change = 0 or 1 *

* Note that Yamaha adds 1 to the Program Change values in the document, so the numbers as they would actually appear inside of the MIDI messages should be 1 less than what Yamaha shows. The numbers listed above have already been adjusted for you, which is why they appear to disagree with the numbers listed in Yamaha's document.

In the screenshots that you posted, the Bank Select MSB and LSB values are okay for Bank 0, but not for the other banks-- for example, Bank 1 should have the MSB value set to 1, not 63.

On the other hand, I think the EX5 does use the Bank Select values shown in those screenshots. So I think the problem may be that you're trying to choose the Voice Bank on the W7 using the Bank Select values for the EX5, which isn't going to work. Likewise, it wouldn't work if you tried to choose the Voice Bank on the EX5 using the Bank Select values for the W7. Make sure you're using the correct Bank Select MSB and LSB depending on which synthesizer you're trying to change voices on. :)
Michael Rideout
Current keyboards: Yamaha YPT-400, PSR-E433, PSR-E443, PSR-EW400, MX49 BK
Current controllers: M-Audio Axiom 61-II
Previous keyboards: Farfisa Matador 611; Casio CTK-710
User avatar
jima
Global Moderator
Global Moderator
Posts: 328
Joined: Tue Oct 18, 2011 11:54 pm
Cayman Islands

Re: Midi message for changing bank - and some frustration

Unread post by jima » Wed Apr 26, 2017 12:15 pm

SysExJohn wrote:Oh, yes!
A MIDI hub. Quite so, Chris.
It takes the output from one MIDI port (might be USB) and splits it out to multiple MIDI interfaces dependent upon channel. Clever.
With apologies for hijacking the OP's post. I have one running under El Capitan and it is wonderful. Would have to be USB as splitting one 31.25Kbits/s MIDI channel over 8 would likely result in horrible latency.
Attachments
MIDISPORT.png
Jim
Yamaha EX5, AN1X, TG500, Moog Werkstatt, Korg Wavestation SR, Sigma, Legacy Digital, Nord G2 Engine, Focusrite PRO 40, MIDISPORT 8x8/s, Uno
Loves keyboards but so little time :(
User avatar
SysExJohn
Global Moderator
Global Moderator
Posts: 624
Joined: Tue Dec 02, 2008 1:00 am
Contact:
Uganda

Re: Midi message for changing bank - and some frustration

Unread post by SysExJohn » Wed Apr 26, 2017 2:25 pm

jima wrote: With apologies for hijacking the OP's post. I have one running under El Capitan and it is wonderful. Would have to be USB as splitting one 31.25Kbits/s MIDI channel over 8 would likely result in horrible latency.
Hmm!
For sure there's gonna be some latency, but "horrendous"? Not sure.
The processing would just have to look for bytes with the status bit set, then the last four bits of that byte ...

Nope, a little more complex than that!

Ah! If it's a channel message, i.e. 8nH to EnH, then examine the channel nybble (why do I always think of BlackAdder?) and forward the message to the appropriate port.
And I don't think being over USB makes any difference.
What I'd heard is that the MIDI data only gets streamed at around 31k anyway.
If it's sent more quickly, and through a USB to MIDI converter, how does the receiving MIDI port deal with it?
It just flows faster, but more erratically, through the USB parts.

And what does it do with SysEx messages that have no embedded channel number?
Forward to all, I suppose?

Further thought needed. :idea: :|

Hmmmmmmmm!
MIDI Tutor forum at http://midi-tutor.proboards.com/index.cgi.
UW500; 2xMU1000 + PLGs AN, DX, VL, VH & PF; SC8850; EWI4000s + VL70m; Samson Graphite 49.
AN1x, Novation X-Station 25. E-MU 02 cardbus + 1616m x 2. Sonar 7 PE. XGworks ST. & SQ01.
Garritan GPO4, COMB2, JABB3, IO, Organs, Steinway, World, Harps + GPO5 ... and now CFX lite.
User avatar
jima
Global Moderator
Global Moderator
Posts: 328
Joined: Tue Oct 18, 2011 11:54 pm
Cayman Islands

Re: Midi message for changing bank - and some frustration

Unread post by jima » Wed Apr 26, 2017 3:57 pm

Well, it occurs to me that if it is a MIDI-based hub/splitter then all 8 (e.g.) channels worth of data has to traverse the first leg from the keyboard or computer to the hub. No matter the intelligence of the hub, all 8 channels worth of data is serialised and transmitted across that first leg with a max bandwidth of 31250 bps. That is why I say potentially horrible. In the USB incarnation, the first leg is high bandwidth, as in the MidiSport, then routed and serialised over 8 x 31250 bps channels keeping all the attached devices happy.

Perhaps best to split this off to another topic if it should continue?
Jim
Yamaha EX5, AN1X, TG500, Moog Werkstatt, Korg Wavestation SR, Sigma, Legacy Digital, Nord G2 Engine, Focusrite PRO 40, MIDISPORT 8x8/s, Uno
Loves keyboards but so little time :(
Post Reply

Return to “Music Software Forum”