Is MIDI obsolete?

[HelgeG]

could MIDI via firewire be better?

Going full circle, MIDI has been implemented in OSC, so MIDI can be sent via OSC over a regular network

[lanfear]

I bought a CME UF8, which is great in every way, except for the stuck notes, randomly losing connection etc. With the MIDI cable it works, but then I can't use the transport controls on it which was one of the reasons I got it. The driver here is up to its 4th version but sill almost everyone has problems with it (if you look at the CME forum).

I'm sorry to hear this, I was a bit tempted by the CME line of controllers! I'm going to replace my old Roland PC-200, and basically I just want a similar but _with_ aftertouch.

So CME controllers aren't recommended at all?

Any tips for others?

/lf

[deaneager]

Hmmm. There's a very simple answer to this that has to be answered in 2 parts...
Firstly,will people be able to cope without MIDI? Well, there's so much old, but not defunct, hardware out there that people will still need MIDI. Therefore, it cannot become obsolete.
Secondly, is there a demand for technology that goes beyond what MIDI can do? The answer, surely, has to be "Yes."
So, MIDI should not be made obsolete, but should have an alternative defacto standard successor.
It would be like making LRP Petrol (Gas to our US buddies) obsolete, when there are many people out there with old cars that run on unleaded.

[Meffy]

Hm, I've been thinking about a CME controller too. Maybe I'll hold off a while longer. :-( Disappointing, that.

[loachm]

...broke the USB-plug of my Midi-keyboard this morning. Now I will have to use the Midi-I/O again...

[olja]

I'm sorry to hear this, I was a bit tempted by the CME line of controllers! I'm going to replace my old Roland PC-200, and basically I just want a similar but _with_ aftertouch.

So CME controllers aren't recommended at all?

Any tips for others?

/lf

Well I can use my UF8, I do it everyday, it just has these random quirks that pops up every now and then. It all seems to be issues with the MIDI USB driver and they're working on a new one as far as i know..

If they fix it the UF8 IMO is the greatest deal for the money out there. Aftertouch, sustain pedal included, hammer action, and I just got a breath controller which works great with it. You should really go through the CME forum: http://www.cme-pro.com/forum/ first though, and read about peoples issues with it before you decide..[/quote]

[ew]

So Is it really hard to write MIDI USB drivers? Could it be USB MIDI will never work for everyone?

It's usually not the drivers or the USB hardware that's at fault- it's usually the OS's problem (specifically power management schemes). 95% of the USB/MIDI issues I've seen on other people's machines are due to power management of the hubs. XP turns on power management by default- the first step is to go into device manager and turn off any power management to the hubs. 95% of the time, this will fix any issues. The same issue exists with Firewire, and the cure's the same.

Going full circle, MIDI has been implemented in OSC, so MIDI can be sent via OSC over a regular network

True- in fact, I can't think of an OSC implementation that DOESN'T include MIDI. OSC was developed as a protocol to allow timing, etc. to be transmitted over a network so remote terminals could sync together with sample-accurate timing, etc. You STILL have to get the notes into the network- hence the need for MIDI. I can't think of a keyboard or controller that's OSC compatible, after all...

ew

[Cosmic Bandito]

Asking if MIDI will be obsolete is asking the wrong question (it will always have it's place like a 303). It is time however for a protocal with more accurate timing and definitly less cabling involved. I think the technology is there and hopefully the "will" of the manufacturers. At least Yamaha is trying with their M-Lan (which I don't know much about). These companies probably are reluctant to agree with eachother for power stuggle reasons I suppose, but it will happen sooner or later.

[olja]

It's usually not the drivers or the USB hardware that's at fault- it's usually the OS's problem (specifically power management schemes). 95% of the USB/MIDI issues I've seen on other people's machines are due to power management of the hubs. XP turns on power management by default- the first step is to go into device manager and turn off any power management to the hubs. 95% of the time, this will fix any issues. The same issue exists with Firewire, and the cure's the same.

I've really searched around the net for possible solutions and that was one of the first I came across, and it did nothing for me.. But the deeper I dig into the 'USB MIDI quagmire', the more people like me I find, who have tried everything in the book and still can't get their USB devices to work 100%.

It seems like USB MIDI drivers only work perfectly under very specific conditions and are much more 'unforgiving' than the old MIDI I/O.

[Ildon]

Please read my post again. Here, I'll make it easy for you:

Or more precisely, should it be obsolete?

I enjoy using MIDI and I don't think it should be obsolete.

[NAD]

Smoke and idiots love high places.

[kritikon]

OK, here's the deal:
I WANNA DO SLIDES LIKE IN FRUITYLOOPS! fcukin' damnit! Why can't somebody just develop a new interface that will allow me to DO SLIDES LIKE IN FRUITYLOOPS?! Yeesh...

Errrr.....then use FruityLoops.

Personally, I have no problem with midi. Yep, it's timing is not absolutely instantaneous, but it's good enough to work very well. Slides...I also have no problem with portamento - midi can automate portamento time, and even portamento mode, so there's not an awful lot you can't do with it (I was doing that sort of thing with my Kenton on ancient analogue synths 15 years ago). As for having portamento going in opposite directions - it's not something I'd want to do enough to get rid of midi for. And if I do want to do that, I can use my host and simply load 2 identical instances of the same synth. And some h/w synths allow you to do things like opposite portamento as well (e.g. a Wavestation can - although it's much the same thing as using 2 instances, as you'd do it by copying the same part into a performance but with different portamento settings).


If a new standard comes along...that's OK, but I for one would not buy it unless it had midi with it. I still use alot of older midi gear, and I'm not going to pack it all away just because a new communication standard has arrived. Whatever turns up simply has to be backwards compatible. Music is not all done on a PC at home. Ask any live keyboardist. Sure, the manufacturers would love us all to throw away our old synths and FX units and sequencers and drum machines and buy new ones with the new standard - but that just is not going to happen - music is still too expensive to start all over again every time some marketing manager decides it's time for a new line.



And for all the issues with timing - I still haven't heard any professionally released song made with midi that is flawed because of obviously dodgy midi timing. For 30 years it has benn perfectly adequate, and IMO still is. Any lags etc can be skewed with one simple mouse click on the track delay, if you have latency problems that are difficult to overcome.

The complicated live stuff where timing might be crucial...can still be done live.
The midi stuff (in musical forms where the myth is that timing is absolutely crucial - such as alot of the electronic dance stuff)....has for a long time been done on midi - so it can't be a problem.


Why change it just because we can? It does pretty well everything I need of it - there are a ton of features available with midi that I just don't even get near to.

I used to have 11 midi items daisy-chained back in the day - and I really did not have problems with timing etc. Apart from my OSCar - which was solved by simply putting it at the end of the chain instead of at the front or middle. When I had too many to chain - easy, I got a midi patchbay and a parallel router - not expensive. I still use my midi patchbay even though I have less midi units now. It's rare you need to send midi messages to everything that has it - e.g. FX units - you don't often tweak reverb or delay settings on the fly. So they don't even need to be in the chain.


Especially midi is nowadays even less of a problem - DAWs can record and playback so many audio channels and FX that it's a piece of piss to render to audio. Maybe the problem some have with midi is that they just can't be bothered using DAWs for the thing they are best at - recording audio.


Midi is fine and dandy. It's old but it works...it works well, and all my machines (both old and new) respond to it without expensive modifications. I actually trust it more than anything digital - it's electrons moving down a wire - unless the wire is unplugged or a component is broken, it simply works. Even now I have the occasional PC problem that is head-scratchingly difficult to fix sometimes - and it's usually to do with conflicts, bad drivers, PC having a brainstorm, magical software crashes....it's not exactly what you'd call the most reliable system (even though I much prefer it to recording analogue, note well).

If a midi unit crashes - it's always because the cable is out or broken. I replace the cable and bingo! It works. Yes, the whole unit can break - so can a PC.

[o'malley]

gm midi sax
that alone should be enough to keep midi alive forever

and orchestra stabs

[mandolarian]

I really think MIDI is just another passing fad. A mere footnote in slow serial transfer protocols.

Midi will only really become popular when we can hookup our guitars to the refrigerator and play for beer in a more direct and satisfying way.

[Markleford]

I just have to laugh at the "serious discussion" here which lacks both historical perspective and delineation of components. kritikon is definitely on the mark, though: definitely sounds like he's been in the trenches.

Anyway, there are two matters here: the protocol (messages), and the hardware implementation (transport). Doesn't sound like many people are knowledgeable about both at once!

The timing issues with MIDI are hardly substantial, and are in fact usually caused by either 1) your PC interface card, or 2) a slow synth in your chain. Now, it *is* true that the actual voltage consumption and number of wires in a MIDI cable means that it can only go so fast, and that speed is significantly less that most other wires going to and from your computer. As the Electronic Prettyboy Messiah tells us, your mouse transmits data faster than MIDI.

But does that mean it's obsolete? Well, for certain purposes it definitely is: anybody who's done a MIDI Sample Dump can attest to that! And dumping and loading bank information via SysEx can take a lonnnnnng while: go get yourself a drink and a cookie while you're waiting for that firmware to update! As pointed out, the data is serial instead of parallel, so it's going to take that much longer to suck out the queue of messages.

On the other hand, how much longer will we need capabilities like this? The newer workstations also have USB or Firewire for data-intensive transfers, but can still use MIDI for sending actual performance data to other devices.Is it more the case that hardware synths, with their own sound generation capabilities, are reaching the end of their life and will be supplanted by generic controllers hooked up to laptops? That's probably why a lot of people see obsolescence in MIDI setups. Time will tell.

But in the meantime, there's there's still a lot of "legacy" gear out there that needs to play with others, and MIDI does just fine.

OSC, on the other hand, has no transport specification. So to say it can "replace MIDI" is a misnomer. It can wrap up MIDI, certainly. But until the industry sits down and says, "This is our new hardware interface spec that everyone should play nice with", it won't replace it. All the big names will need to agree that new keyboards have a USB, Firewire, or ethernet style port (or something new), that has a specific power requirement, and can be chained in a certain configuration (I imagine they'd do IN/OUT and make the THRU a soft config issue). But as it stands, OSC is categorically uninterested in the wire.

Protocol/Message-wise, MIDI is still quite powerful. Why are people whining about 128-value controllers? Only because hardware and software manufacturers are fecking morons and haven't decided that it would be a swell idea to actually talk to each other to use everything available to them. Each knob on your controller *could* send out 14-bit RPN data, just as the pitch wheel expresses that range. Heck, it could send out unlimited N-bit SysEx data (like the Yamaha EZ-EG MIDI guitar does) to express more than the human body could sense.

But keyboards don't do this because they shrug and say, "No software (or other hardware) supports this". Why waste the time and money developing something that will never be used? And on the other hand, software companies say, "Oh, we only support 7-bit CC values because that's all that keyboards will send us". So it's just a cycle of suck, where neither party wants to agree to make EVERYTHING that much cooler for musicians.

The same thing sadly happened to polyphonic Aftertouch. Only a few keyboard manufacturers supported this part of the spec, so other sound module makers (and, later, software devs) thought it was worth their while it implement a synthesis engine to respond to it. And since nobody responded to poly aftertouch, eventually the keyboard makers who implemented it before said, "Why bother?"

So don't blame the MIDI spec on this. Blame lazy, shortsided developers on both sides of the fence.

So what's going to replace MIDI? Certainly not OSC any time soon: not only will they have the same issue with hardware and software vendors not agreeing, but they'll have to overcome the desire to keep MIDI as "already there and easy to implement". Remember: this spec has been around for more than 20 years, so it's going to be hard to budge!

Mind, MIDI is still evolving. There is a transport spec for MIDI Over Firewire. There's a new "high definition" HD-MIDI proposal on the table. And there *was* the cross-platform/-company GMPI plug-in spec that was designed with an eye on eliminating MIDI's limitations in the software world, but Apple and Steinberg were big f**king babies and didn't want to participate because they think that AU and VST are perfect as they are, which is complete bullshit. So short-sighted: if they can't control a standard, they think it's unimportant.

So there ya go: MIDI might be showing some age, but that's primarily because nobody wants to cooperate at a level beyond the mid-80's. If there's any emergent standard that will get everyone to agree, then sign me up! Otherwise, I just don't see the status quo changing until developers and manufacturers can see past their own quarterly earnings.

- m