Translate sysex for use with pyramid

Emu did this (Mfr ID, Model ID, Unit designation).

But thats the issue with SysEx: it is by design (in my understanding) specifically NOT standardised (except leading 0xF0 and trailing 0xF7) to facilitate freedom for each manufacturer.

I still think it’s amazing that MIDI ever worked - getting manufacturers to agree on such a sweeping standard is a mind numbingly crazy concept. Im excited for 2.0.

exactly, basically it was designed for things like patch dumps… so that the data was only ever parsed by the manufactures own code.

so if you try to go beyond that you have to start providing the user with some kind o parsing tools - which frankly is often beyond what most musicians are interested in doing :wink:

yeah, it worked because it was simple… look beyond the basics (simple note on/off, simple cc’s, channel at, program control) and the ‘support’ gets thinner/ more sporadic
e.g. even simple stuff like 14 bit cc and nrpn are a bit of a minefield.

yeah 2.0 is going to clear up so much of this over time.
midi-ci in particular is hugely exciting .

1 Like

I remember in the 80s my needs were quite less demanding: Grey Matter E! Board on a DX7IIFD and an RX21 drum machine. Plus modulation needs didnt seem so strident with the music of the day.

But (and bear with me because that was a looong time ago), i vaguely recall the E! Board (which made the DX7 multitimbral to 8 voices and added an onboard sequencer) allowed me to SysEx dump because id use the entire RX21 memory for one song and dump in between sets.

Maybe it was something with the floppy drive on the DX7.

I remember this because my “trailer park industrial band” was opening for a douchemetal band and i f***ked up one of the dumps and the drums didnt come in when they were supposed to and the douchemetal band made jokes about it through their whole set.

So…maybe OP isnt so far out to lunch? LOL

But yeah: patch dumps.
And editor/library software on your 286.

Wow - like the 80s were 40 years ago or something

Oh hell.
Now i want some Rx21 samples.
Oh, the memories…!

Edit to add: and someone selling an RX21 on Reverb near me and guess who made an offer?

Anyone want samples for their Rample? LOL

Ah - nostalgia

there’s a problem: sometimes only SysEx are exposed and available.
for instance, Zoom MS pedals.
or Roland HandSonic HPD-15 effect parameters.

btw, NRPNs are not much better, because 4 CC messages is 12 bytes, which is about typical SysEx message length.
SysEx might be even better than NRPN because they are always channel-independent.

when you do automation, you usually want it to be channel dependent, since you often have different instruments on different midi channels.
I see few areas where channel independance is advantageous.

yeah, I dont think NRPN are much better, as I mentioned previously midi support, once you go beyond the basic midi messages, drops off alarmingly quickly,.

but I didnt say, sysex was less effecient, it can be very effecieint, as basically its a ‘propriatary’ protocol embedded into the midi protocol…

(data size are not quite so important as they were… but given its a serial protocol, large messages do give sync issues)

I think more important is

a) its not channelled
b) the format is completely uknown… so you cannot edit it

for sure, some pedals instruments use sysex, but that not really the point… (*)
my point was simply, your unlikely to see support of sysex in hardware sequencers (and we dont get much even in full blown commercial DAWs)

so… you have to look for other solutions, which is translating them to something that IS supported :slight_smile:


(*) just because pedals support via sysex doesn’t mean they expect tools to support automation thru these means…
to do so would be a bit odd, given Im sure Zoom know what the ‘average daw’ supports and not.
rather they support sysex primarily as a way to dump presets… or perhaps to be used via their own tools as a propriatary protocol.

i suspect Zoom did not expect external automation on these pedals, but made that for some kind of external editor.
there’s no officially published MIDI implementation, only reverse engineered — but who cares once it’s known to work …

but older digital machines (late 80s – early 2000s) are another story. they often have relatively small subset of parameters available via CC, and complete set of parameters available via SysEx — examples that i own are E-Mu P2000 / P2500 series, early models of Access Virus, and Roland ROMpler machines (e.g.HPD-15).

I only know of the Virus…
although they documented the sysex for the early virus (not the TI!), I think sysex was mainly focused on patch dumps etc (though single parameter change kind of falls out of how they did it, but is still 11 bytes)

for automation they supported two other mechanism :
a) Mod matrix
synths could only provide access to ~110 parameters via CC (remember some cc are ‘reservered’) , many like virus had many more… so they put the 100 most used on CC, and then others could be accessed via the mod matrix which could be modulate via a CC.
(… I guess close to what daws would call a midi learn approach)

b) polypressure
this was quite ‘clever’ , though a ‘grave missue’ of the standard :wink:
polypressure as a concept was not supported by the Virus, and few used it anyway. so they ‘re-used’ this message. providing another 127 parameters.
the advantage of this is, its channelled, its only 3 bytes long, its a standard midi message.

(though in the end not that useful… since because polypressure was not used widely, few daws supported it )

anyway my point previously was not that sysex was not put to use for this… it clearly was.
but that it wasnt really intended to be used like this, it was kind of an abuse (like Virus using polypressure) of the standard - which would never see widespread adoption.
(sysex is consider a proprietary message for that manufacture - which are ‘ignored’ by other midi instruments/manufactures)

the easiest/most flexible way to deal with them, is to use some kind of ‘translator’ on your midi network to turn them standard CC (or similar) midi messages.

the ‘good news’ is that midi 2.0 will also potentially help in this area…
we will be able to build midi 2.0 ‘translators’ that could turn these sysex into the new NRPN 32 bit universal messages, and use MIDI-CI to describe them.
this would enable new midi 2.0 ‘hosts’ (sequencers/daws) to automate these parameters effortlessly :slight_smile:

of couse… the point is ‘midi 2.0’ hosts, whilst software daws will just natually evolve into midi 2.0 over time.
I think we can expect that most hardware sequencers will require new hardware…
possibly because they need more powerful hardware (more memory in particular!?), but also because the firmware will likely need radical re-writes… and that development needs to be ‘paid for’ some how.

But Im definitingly looking forward to midi 2.0, and in particular midi-ci - midi parameters were a bit of a mess in 1.0… and have long needed sorting out :slight_smile:

there’s a problem with MIDI 2.0 — it will not appear on the older gear, especially older digital gear, which is in no way considered „iconic“, so no one is going to care about.

that’s why MIDI 1.0 will still be relevant for decades.

I disagree…I think we are going to see a plethora of ‘intelligent’ midi hubs that wll act as ‘midi 2.0 gateways’ … Bomebox already announced they would do this.

so what you will do is plug your midi 1.0 device into these, and output midi 2.0 on the other side.
the midi hub will have ‘instrument profiles’ (provided by communities I expect).
this approach will mean you will get MIDI-CI for your old 1.0 gear.

building these profiles is not going to be that hard… not much different from what we already do on the pyramids ‘instrument’ definitions - it will be something the muscian can do, then share with the community.
(ideally we’d develop a standard format for this, so it could be used on different hubs, but I fear that wont happen!)

(some of these hubs, like bomebox already, could allow more user code, to allow for translation of sysex to the new NRPN combined with MIDI-CI)

as you can tell from above I think MIDI-CI is going to end up being THE main selling point of midi 2.0,
I think musicians are tried of trying to remember/and mapping CC ‘numbers’ into synth parameters.

probably.
on the other hand, manufacturers would love to abandon supporting old gear in any forms to force people to buy more new gear.

time will tell…

oh definitely agree with that…

but often it creates a new market…
in this case, your right, the synth manufactures have new gear to sell.

however, the creators of the intelligent hubs (which we have seen a few of recently) have been looking for a ‘way in’ to the mainstream, and I think midi 1.0 ↔ midi 2.0 might prove to be it for them.
explaining midi processing to the ‘masses’ is difficult, but selling a magic box that does midi 1.0 to 2.0 is easy :wink:

1 Like

good point.
i’m probably a bit too paranoid ))

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.