Need a MIDI splitter recommendation.

[RABid]

I'm in a dire need of a portable MIDI splitter for my DAWless jam setup. It needs to be at least 1 in and 6 out with the ability to assign specific channels to each out. Example use: With the Digitone as my master it would send clock and start/stop to the Digitakt and the Roland MC-707. Those devices would use their internal sequencer but need to stay locked in. The Digitone would also send MIDI data to other devices such as Korg Volcas with each device being assigned to its own channel.

 

If MOTU still made the MIdi Timepiece AV I would get one even though it is big. I loved that device and the way you could program routings on the computer screen and save them in 128 presets. I've looked and cannot find anything like that in current devices so I guess I will simplify my goals to what was described above.

 

Question: how many devices can you chain before lag becomes an issue?

[Al Coda]

A MIDI "splitter" (1 In into several Outs) is a MIDI Thru box.

Normally, the controller keyboard does the MIDI channel mapping and then all is distributed to the connected devices.

It´s way better than daisy chaining MIDI.

 

NOW, when your conroller keyboard isn´t able to perform multi MIDI channel transmissions, you need a MIDI processor doing that task,- WHEN it´s able to shift MIDI info incl. MIDI CC controller data from an incoming MIDI data stream on a given channel to multiple other MIDI channels.

Not many can do !

The MIDI processor might lack the 6 MIDI Outs you want,- but can do the job when you add a MIDI Thru box w/ 6 -8 MIDI Outs.

Midisolutions offers one.

 

Now it´s up to you finding such MIDI processor today.

 

My experience is, only a keybpoard controller like the discontinued Viscount Physis Piano K4 / K5 can do,- or some "vintage" MIDI processor matrix routers from the 80s/ 90s.

Miditemp PMM88E is still the best and all in one box !

 

 

A.C.

[Reezekeys]

Can't help with any specific recommendations since I'm all in-the-box, but Al is right, you need a box that does filtering, if not outright processing of midi data. I used a Yamaha MEP4 many years ago. That's a single rack space unit that was a one-of-a-kind piece for its day. It has one midi in going to four outputs. Now Mac & PC apps can do everything that box did, but for its time it was unique.

 

Can you not use a simple thru box, then set the receiving devices to simply ignore the data not meant for them?

 

Question: how many devices can you chain before lag becomes an issue?

 

This "lag" question persists and from what I know it's not an issue â there is no lag introduced by daisy-chaining 5-pin midi. These are purely electrical connections that go through no processing (yes I know, some units do a "software thru" and in that case there can be lag). What can be an issue is degradation of the signal due to poor quality or too-long cables, or maybe too many cables (if you're chaining a bunch of devices). This could result in "lost bits" causing hiccups, stuck notes or just no sound. For me this is a hypothetical because I'm old enough to remember the early days of midi and one particular session where we daisy-chained something like ten Korg Poly-800 synths together. It worked flawlessly - and sounded horrible!

[KuruPrionz]

Ebay has quite a few MOTU MIDI Timepieces for sale. This one is $90, from a seller with lots of feedback and a high rating. Ebay is very Buyer-focused, they will make sure you are taken care of if there is a problem.

 

https://www.ebay.com/itm/303713790370?hash=item46b6c0a9a2:g:hgEAAOSwq39feLFp

 

You get plenty of time to test it and make sure it's working correctly. Just save the packaging until you are happy with it.

[EscapeRocks]

......... I used a Yamaha MEP4 many years ago. That's a single rack space unit that was a one-of-a-kind piece for its day.

 

I loved my MEP4! It's how I learned the dead languages like Sanskrit I do not miss programming in SYSEX on the tiny screen ......

[synthizen2]

Another single rack space unit that worked for me was the Digital Music Corp. MX-8.

 

 

I used to have 2 of them. I no longer needed them so I sold them both a few years ago, but they were great MIDI routing and processing units.

[Kurt W]

The MIDI processor might lack the 6 MIDI Outs you want,- but can do the job when you add a MIDI Thru box w/ 6 -8 MIDI Outs.

 

You can"t do the job this way, OP wants to «assign specific channels to each out».

[Delaware Dave]

this one will do 5 out:

 

https://www.genuinesoundware.com/?a=showproduct&b=51

 

also see this, maybe two?

 

https://www.facebook.com/GenuineSoundware/photos/a.537502179674328/4392394594185048/

[Al Coda]

this one will do 5 out:

 

https://www.genuinesoundware.com/?a=showproduct&b=51

 

also see this, maybe two?

 

https://www.facebook.com/GenuineSoundware/photos/a.537502179674328/4392394594185048/

 

That´s a USB2 (DAW) MIDI interface.

OP wants something standalone MIDI router w/ re-channelising from MIDI Input #1 and across up to 6 MIDI Outs.

If I understood correctly, it should be user adjustable from it´s front panel and w/o need of a computer.

 

 

A.C.

[Delaware Dave]

this one will do 5 out:

 

https://www.genuinesoundware.com/?a=showproduct&b=51

 

also see this, maybe two?

 

https://www.facebook.com/GenuineSoundware/photos/a.537502179674328/4392394594185048/

 

That´s a USB2 (DAW) MIDI interface.

OP wants something standalone MIDI router w/ re-channelising from MIDI Input #1 and across up to 6 MIDI Outs.

If I understood correctly, it should be user adjustable from it´s front panel and w/o need of a computer.

 

 

A.C.

I think it does both USB and MIDI DIN 5 pin, I think the MIDI DIN 5 pin is on the right side by itself. I could be wrong.

[Steve Nathan]

I used to use one of these. Still have it laying around here somewhere I believe it would do what you want.

Yamaha MEP4

[Al Coda]

... but Al is right, you need a box that does filtering, if not outright processing of midi data. I used a Yamaha MEP4 many years ago.

 

Can you not use a simple thru box, then set the receiving devices to simply ignore the data not meant for them?

 

A MIDI thru box woiuld be o.k. WHEN the device at it´s input is able to perform re-channelizing incl. controller re-mapping.

 

The only units I know which can do it all are the Miditemp rackmount 1HU PMM-44 / 88 and the best one,- PMM88E.

Some Miditemp MIDIplayers include a PMM88E.

 

I´m using the 88 and 88E since the 90s,- so I know.

I had almost everything beginning w/ Sycologic M16 / 16x / 16R which were simple 19" rackmount, 2U, matrix routers w/ a remote control unit.

 

Miditemp and Viscount Oberheim MC3000 were king of the hill.

Latest controller offering same functionality was Viscount Physis Piano K4 / K5.

 

Miditemp MIDI re-channelize is called "manifold" in Miditemp world.

It allows copying the on a given channel incoming MIDI data to any other MIDI channel on demand,- multiple incoming to multiple other outgoing MIDI channels on multiple ports and for 128 MIDI channels in a total !!!

 

For the full implementation of MIDI continuous controllers, that´s not that simple.

So they separate the controllers incl. PitchBend from the incoming data stream and map ´em to some exclusive channel, bypassing the "manifold" (re-channelize) process of MIDI notes and velocity data.

Then, the user decides at which MIDI Outs and on which MIDI channels he wants all, or a subset of, the controllers back (combo process of re-route and filtering the single controller data stream separately).

The big advantage is the MIDI busses inside the processor box won´t never be saturated (MIDI buffer overflow) by multiple streams of multiple MIDI CC controllers.

It works perfect, no lag !

 

Tony Banks used 2 of these, stacked by optional Fornet optical lightpipe bus.

 

There was also the very rare to find MT-16X which housed 2 PMM88Es in one case.

 

The DMC MX8 did (limited) processing on only 2 MIDI inputs,- JL Cooper Synapse did too.

Miditemp does it on all the 8 Midi Ins AND Outs individually,- means some processing appears only at the outs and on user selectable channels on demand.

 

There´s actually a PMM88E available HERE but unfortunately "pickup only" and no shipping.

 

LAB4Music SiparioX can possible do similar tasks !

 

 

This "lag" question persists and from what I know it's not an issue â there is no lag introduced by daisy-chaining 5-pin midi.

 

There is when you daisy chain more than 3 devices, the transmitter included.

It´s in fact kind of "copy-lag" caused by the opto couplers being present in 5-Pin Din MIDI systems.

And that lag adds to the usual MIDI latency (approx. 1 ms per event) !

 

So, thinning out data stream is a good idea always.

 

 

A.C.

[Al Coda]

I think it does both USB and MIDI DIN 5 pin, I think the MIDI DIN 5 pin is on the right side by itself. I could be wrong.

 

It´s not matter of USB and/or 5-pin DIN MIDI.

It´s all about if it works standalone or not and if it does, if there´s the option to program routing/processing from it´s front panel.

 

I don´t recognize the GSi is a standalone unit as a MIDI matrix router/processor.

It´s more a MIDI patchbay w/ a USB MIDI interface.

 

A.C.

[Reezekeys]

This "lag" question persists and from what I know it's not an issue â there is no lag introduced by daisy-chaining 5-pin midi.

 

There is when you daisy chain more than 3 devices, the transmitter included.

It´s in fact kind of "copy-lag" caused by the opto couplers being present in 5-Pin Din MIDI systems.

And that lag adds to the usual MIDI latency (approx. 1 ms per event) !

 

So, thinning out data stream is a good idea always.

 

 

A.C.

This didn't make sense to me so I did a little googling. I came upon a web page where this person went full nerd on this subject. The upshot - yea, there's some lag but we're talking microseconds. tl;dr - six stages of midi in-->thru and the total lag time was 34μs:

 

http://www.dubrovenko.ru/ddssite/dstmihue.htm

 

So, again, for all practical purposes, using a hardware midi thru does not add any delay to generated notes. Of course the slow speed of midi 1.0 over a DIN connection, where each event takes 1 ms, can smear things when you want more than a few events to happen simultaneously - but that's a completely different subject.

[Al Coda]

This didn't make sense to me so I did a little googling. I came upon a web page where this person went full nerd on this subject. The upshot - yea, there's some lag but we're talking microseconds. tl;dr - six stages of midi in-->thru and the total lag time was 34μs:

 

http://www.dubrovenko.ru/ddssite/dstmihue.htm

 

So, again, for all practical purposes, using a hardware midi thru does not add any delay to generated notes. Of course the slow speed of midi 1.0 over a DIN connection, where each event takes 1 ms, can smear things when you want more than a few events to happen simultaneously - but that's a completely different subject.

 

Yeah, the interwebs ...

Everything you google is true.

 

But why is a dedicated MIDI-Thru box/device better than daisy chaining via the internal MIDI-Thru of a 5-pin DIN MIDI connector trio then ???

When that guy is right,- we never needed a dedicated MIDI thru box!

 

It´s not only "μs", "ms" and baud rate,- there´s also the jitter (far more recognizable than constant latency), latency of transmitter, receiver, time the opto couplers need to copy data from In to Thru, processing time in MIDI processors and matrix switchers,- and it´s all per EVENT (single note).

Play 8 notes and already the theoretical delay it´s 8x as much,- plus the jitter and all s##t mentioned above.

But continuous controller data puts the most stress on the system and I´ve never seen any keyboardplayer just only using a single controller at a time.

Your foot might operate the expression pedal while doing pitchbends and add vibrato or open filter by aftertouch or modwheel.

That stream is much more than the notes and velocity data and the "delay" between ´em.

The jitter increases massively w/ many notes and many CCs on the bus.

And we not only play one keyboard at a time.

Most of the times 2 keyboards transmit MIDI messages and sometimes a MIDI (bass-) pedal too.

 

Also today´s MIDI isn´t any better because the baud rate never changed and USB is much more worse than any ancient ATARI´s onboard MIDI interface ever was.

USB2 just only adds more ports,- but for each port, the baud rate and all the other issues are still the same as in the 80s or more worse.

This rules especially for outgoing MIDI from computer/DAW to ext. hardware.

Ingoing MIDI from p.ex. USB controller to DAW is usually much more constant and tight.

That´s why we had the AMT and LTB technologies in Emagic and Steinberg MIDI interfaces and DAW applications,- until they decided for not supporting ´em anymore.

Up to now it´s hard to believe they really thought no one will need any 8x8 MIDI interfaces anymore in future,- just because there are now software instruments in the ballpark.

 

That said,- no one of the pro-developers I know ever recommended daisy chaining more than 2 or at max. 3 MIDI devices,- especially not when these are MIDI multitimbral and when lots of controller data is in the ballpark.

And even the "standard" MIDI channel for drums is #10,- for MIDI sequencing, they also recommended putting p.ex. MIDI drums to the lowest numbered MIDI channel possible and to the upper visible tracks in a MIDI-sequencer´s arrangement window as well,- just only because the playback algorithim processes these tracks first and that is what you want for tight percussive audio signals.

I´d say that are very lucid recommendations since we all know MIDI is a serial interface.

 

You know,- above you mention "delay" and in former post you said "lag".

At least for me, that´s a difference.

While "lag" to me is the overall "real world" issue including jitter and everything mentioned above,- "delay" to me is some constant increasing/decreasing measurable quantity,- a lab result.

 

MIDI 2 might be better in future, but isn´t available for the most MIDI gear out there up to now.

 

 

A.C.

[Kurt W]

You are not correct when implying that Reezekeys are referring to unreliable facts. The delay when hardcopying data from a midi in to midi out port are in the microsecond range and totally neglectable when adding to midi transmission rate or midi latency as you call it.

 

The reason for avoiding daisychaining is that the optocouplers slightly disturb the rise and fall times of the pulses, and this might introduce reception errors where the receptor might end up not detecting the correct bytes on the asynchronous data stream. It is not delay that is the problem but transmission errors resulting after several stages of optocouplers. The midi specifications does not restrict the daisychaining to three steps but state that if more you should use higher quality optocouplers in the hardware design.

 

A splitter adds only one stage of optocouplers and minimize the probability of receiving errors. You make this theme more confusing than neccessary.

[Al Coda]

You make this theme more confusing than neccessary.

 

maybe, but at the end of the day and when considering the OP´s demand,- it´s what IYO,-

 

a)

use daisy chaining and don´t wrap your head around anything

 

b)

buy a MIDI thru box (= "splitter"), live with what it can do and leave the other tasks like MIDI re-channelize to whatever gear already existing in his arsenal.

 

Reminder on disadvantage for daily work and according to Thru-Box/"Splitter":

 

Multiple copies of the same incoming MIDI data stream at all the MIDI Thru ports: same MIDI PrgCh.-No received from all to the MIDI Thru ports connected gear requireing special organization of patches in each unit connected; all gear has to be on the same basic MIDI channel the masterkeyboard uses for MIDI data transmission (when it´s not able to transmit on multiple user definable MIDI channels).

 

Did OP say his masterkeyboard can transmit on multiple MIDI channels ?

No, he didn´t.

 

c)

buy a used or new quality MIDI matrix switcher/ processor which can do it all.

 

Now what ???

 

I understood the OP wanted to know what´s better than daisy chaining,- no ?

 

I expect he already double checked if it works for him and before posting questions,- no ?

 

And,- Reezekeys doesn´t have any MIDI problem in his rig like the OP seems to have.

It´s because he´s using a single Roland A800Pro MIDI controller running via USB into his laptop running Mainstage and triggering software instruments.

 

I´m out now.

 

Reality check:

Use daisy chaining and be happy.

At least it´s the easiest and cheapest.

Next time try yourself before asking.

 

 

 

A.C.

[Reezekeys]

I like how we're debating this issue of multiple daisy-chaining through midi thrus, something I would bet almost nobody does anymore! And for the ones that do - I'm sure they've had enough time to confirm whether or not their setup works or not; something like 35 years if they've been in this game since the beginning (I'm close to that)!

 

As Kurt says, I was only referring to the electrical properties of optocouplers when I said they don't add perceptible delay. I also said that multiple units daisy-chained can lead to errors but Kurt explained it better and schooled me on that â I thought it was cable capacitance of cheap cables that might cause bits to be lost or misinterpreted. Now I know that optocouplers themselves can vary in quality and cause errors. Maybe I was wrong about the cables and cable capacitance is not an issue, even for cheap ones.

 

An aside on the issue of cabling: if you read that web page I linked to, the person states "to make the experiment closer to the reality we add protection diodes, and connect every two cascades via a regular 5m MIDI cable", so there were two 5m cables in this test â not an insignificant length. I use a 35-foot (~11 meters) midi cable on all AWB gigs to connect my A800 to my laptop and have been for years, and have never experienced any errors. I also read a comment from a poster on Quora where he said (cut & paste here): "I spliced a MIDI connector to XLR, plugged that into the stage snake then the house wiring. The MIDI signal traveled some 200â300 feet through the ancient in-building wiring, surrounded by audio signals, until it at last reached the MIDI interface into the sound playback computer. Never had a problem!"

 

Anyway, Al is 100% correct in describing the limitations of a heavy midi 1.0 data stream. Lots of controllers, pitch bend, aftertouch, then a bunch of notes that have to all sound at the same time - that's a problem. I was not aware that midi over USB travels at the same 32Kbaud speed as 5-pin though; is this really true? Sounds like that's what you were saying. But, as to daisy chaining opto-coupled midi by itself causing perceptible lag, delay, or whatever you want to call it, I don't think so.

 

My long-winded opinion is that since midi thru boxes are not that expensive, it makes sense to use them where you need to send a single midi stream to multiple devices - but daisy chaining with the thru ports can work just as well, depending on the factors mentioned - total # of devices, optocoupler (and possibly cable) quality. In both cases, one needs to think ahead and consider each device's ability to reject the data it doesn't need since every device gets the same data. It looks like RABid's use case won't allow for that and he needs to filter or process the midi stream so different devices get altered data. The only thing I might add to that discussion (finally ) is that the MOTU "Micro Express" might do the job - a 4x6 programmable interface that can filter specific categories of midi data and rechannelize too. You need a computer to program it, but according to what I read (an account from someone claiming to have spoken to a MOTU CSR), it works standalone with a USB power cube so you can bring it to a gig and leave the computer home. https://www.sweetwater.com/store/detail/MicroExpUSB--motu-micro-express-usb

[Al Coda]

I like how we're debating this issue of multiple daisy-chaining through midi thrus, something I would bet almost nobody does anymore!

 

Yeah,- I like too

 

But "MIDI microtiming" seems not to get too miuch interest these days anymore.

 

You know, today, it´s all about drag and drop audio clips and post pro.

 

I prefer it´s being tight before being printed to tape or HDD and I want it as tight as possible in a performance.

That´s all !.

 

I hoped it might help people, but it seems it doesn´t.

 

Anyway,- I myself benefit from my over decades growing knowledge as a "MIDIot",- performing w/ VERY large rigs in the 80s/90s and I can use this experience ´til today,- also w/ software.

 

 

A.C.

[Reezekeys]

I remember the day I discovered how to "fix" the midi timing I was experiencing in my little recording studio - striping time code on track 8 of my Tascam 388, and making multiple passes with different tracks enabled, instead of one pass with everything playing! What a revelation. :-)

 

Then came MOTU "midi time stamping", but by then I don't think I was doing much recording at home, and they said it only worked with their interfaces. I had a medium-sized rig for a home studio in the 1980s - Akai S900, Yam TX802, Korg M3R, Ensoniq Mirage, OBXa, and a DX7. All wired through a Yamaha 8x8 midi patchbay. So in actuality I never daisy chained anything!

[Zalman Stern]

Haven't had time to read the entire thread and probably don't want to, but iConnectivity's mio products are great and seem like a good solution here. The XM and XL: https://www.iconnectivity.com/midi-interfaces-1 . Routing is generally configured from a computer or iOS device, but it is not necessary for them to do the job once configured. They're fairly versatile in the routing and remapping type things that can be done. The two mentioned support RTP MIDI onto Ethernet. I doubt the plain mio and XC are useful here as they're just MIDI to computer interfaces.

 

-Z-

[TommyS]

MIDI Solutions ,,, might have what you need,

[Adam Burgess]

http://midisolutions.com/prodt8.htm

 

And, drop them a mail to ask if they can reflash it to output on different channels.

and possibly keep a couple of outs set to the money channel just in case!

 

Only thing I ever had was a 1 into 2 MIDI Thru box, and I forgot it was there - which is exactly what you need.

 

And, another vote for the iConnectivity stuff. Hardware is fantastic.

The terrible software is slowly getting better, but, you only have to configure it once!

 

My Audio2+ and MIDI4+ have amazing routing capabilities, and got me out of a few holes.