Slashdot Mirror
MIDI Association Announces MIDI 2.0 Prototyping (hackaday.com)
MIDI was introduced at the 1983 NAMM show as a means to connect various electronic instruments together. Since then, our favorite five-pin DIN has been stuffed into Radio Shack keyboards, MPCs, synths, eurorack modules, and DAWs. The standard basically hasn't changed. Now, ahead of the 2019 NAMM show, the MIDI Manufacturers Association (MMA) in conjunction with AMEI, Japan's MIDI Association, are announcing MIDI 2.0.
From a report: The new features include, "auto-configuration, new DAW/web integrations, extended resolution, increased expressiveness, and tighter timing." It will retain backwards-compatibility with MIDI 1.0 devices. The new initiative, like the release of the first MIDI spec, is a joint venture between manufacturers of musical instruments. The company lineup on this press release is as follows: Ableton/Cycling '74, Art+Logic, Bome Software, Google, imitone, Native Instruments, Roland, ROLI, Steinberg, TouchKeys, and Yamaha.
Sounds good to me.
And google will fuck it up for everyone as usual
Does Google just sign up for every multi-company initiative as a matter of course? Are we going to hear next about their participation on the ratings panels for small gas engines or feminine hygiene products?
#DeleteChrome
Nobody thinks that, not even the Microsoft idiots. Not even the straw man I caught.
MIDI hasn't been updated in over 20 years, and even then it was incremental type updates. The spec itself hasn't really changed since...1983? It's kind of impressive that the music industry is still entirely reliant on 35+ year old tech that hasn't changed, but at the same time, what modern tech hasn't evolved in that timeframe? We've even done away with BIOS at this point, I'm trying to think of something else that's deeply computer integrated and has remained largely unchanged in the past 20 years.
"Don't meddle in the affairs of a patent dragon, for thou art tasty and good with ketchup." ~ohcrapitssteve
I can remember reading about the possibility of adding microprocessors to electronic keyboards someday. The article mentioned the problem companies will have with compatibility. That was a LONG time ago. Anybody remember MIDI adapters for sound card joystick ports? I have one of those somewhere. Worked good.
MIDI malware. /oblig captcha : nonzero
Likely MIDI approved vendors will be the only ones who can communicate on the bus, or better yet, buy the license for Korg event reception to get "full features", and yes, Midi Content Protection ... hope your wallets are ready.
> , what modern tech hasn't evolved in that timeframe? We've even done away with BIOS at this point, I'm trying to think of something else that's deeply computer integrated and has remained largely unchanged in the past 20 years
Mice, keyboards. I have Model M (1984) on my wishlist, as an upgrade from my current keyboard.
ASCII is still king - over 95% of web pages are ASCII. In the Unicode wars, UTF8 (which is ASCII, plus more) won because it's technically superior in a strict sense - it does everything better than competing encodings do.
On a midi-related tangent, I have a box with about 20 midi cables, still unopened in original packaging. Perhaps someone who clicked on this midi article has a use for them?
https://en.wikipedia.org/wiki/Open_Sound_Control
Mice have moved from trackballs to optical sensors primarily, keyboards have adopted USB as the standard connectivity, ASCII has evolved into UTF8 as you implied, but even beyond that, ASCII was last updated to the current ANSI standards in '86 as I recall (so damned close).
I don't have much MIDI cable oriented these days, believe it or not, CV is coming back stronger than MIDI right now.
"Don't meddle in the affairs of a patent dragon, for thou art tasty and good with ketchup." ~ohcrapitssteve
Summary is incorrect when it says "MIDI was introduced at the 1983 NAMM show... The standard basically hasn't changed." The first half is true. The follow-on isn't.
MIDI received a major upgrade in 1991 in the form of General MIDI, which dealt with many of the things previously left open to interpretation by manufacturers, such as what order the instruments should go in the patch bank. That's the original MIDI 2.0, we just weren't quite as keen to use that particular notation (outside of software) yet.
I don't object at all to extending MIDI, but I think they should have called it Global MIDI or Universal MIDI or Modern MIDI or something in keeping with the General MIDI naming tradition.
How is the Riemann zeta function like Trump rallies? Both have an endless number of trivial zeros.
I know what most of them mean - but seriously, if you're going to post something, _especially_ an article/summary, define each acronym as it's used (except perhaps in the headline). It's just common sense, and a basic journalistic principal.
"National Security is the chief cause of national insecurity." - Celine's First Law
..
http://saveie6.com/
ASCII is still king - over 95% of web pages are ASCII.
Wrong. Even back in 2012, 60% of the web crawled by Google was Unicode. That figure would have only gone up since.
https://googleblog.blogspot.co...
Slashdot is in a tiny minority of ASCII-only websites.
So, the reason why MIDI has served the test of time is because it's a relatively simple protocol that does one(ish) thing and does it well - it provides data from musical instruments that is easily readable by other things which understand this very-well-defined spec, and allows things to be daisy-chained so that data can be manipulated and the data altered as a function of its sequence on the bus. It's very well understood, and that data can be manipulated however the devices wish to do so. Its simplicity is why it has stood the test of time, it is well-documented and as such is easily implemented by anyone, and I'm unaware of any royalties required to pay anyone for its use, making it possible to use in everything from open source projects to multi-thousand-dollar DJ controllers and keyboard workstations. Let's see what they're going to replace it with...
auto-configuration
There is no real configuration needed in most modern MIDI implementations, to my knowledge. The configuration is primarily on how the software interacts with it, and let's be real - that's the sort of thing the user should be doing anyway.
new DAW/web integrations
Ehm...this sounds like code for "a protocol that can allow for things like Mainstage to run in a web browser rather than on the device", to which I'd generally say, "stop making a web browser into an operating system". I really don't see the point of web integrations otherwise, and I don't see how "DAW integrations" is a problem to be solved. A DAW that doesn't support MIDI is like a word processor that doesn't support printing - fails at its core purpose and thus generally doesn't exist. If the plan is to be able to connect DAWs to each other, that's already a solved problem with Rewire...or existing MIDI files...or bouncing tracks...or running a DAW as a VST plugin. I fail to see the unsolved problem here.
extended resolution
I mean, I guess...but that's like saying we need a new standard for smaller MicroSD cards. Sure it's possible, but usability starts being counterproductive. Is there really that pressing a need for a higher-res MIDI protocol such that its human interface justifies it? Current MIDI doesn't seem to be a problem for DJs on controllers or have note limits that extend beyond what human hands can accomplish; anything much more than that I'm pretty sure is already solved in data-to-data sorts of ways that don't require MIDI.
increased expressiveness,
What even does this mean?
and tighter timing.
Again...*maybe*...but I'm open to scenarios where current MIDI timing is an actual-issue. It's like saying that RS-232 serial at 115,200 is too slow. It is if the intent is bulk data transfer, but it's plenty quick for its most common contemporary use case - typing commands into Cisco routers and similar appliances; it's far faster than I can type commands or read output and its simplicity means I am not limited to a particular solution.
Because it is such a wash.
Seriously though, they should have fucking made this MIDI update 25 years ago when Yamaha/Roland were making their General MIDI enhancements for their respective product lines.
Now, what is the point, unless they want everyone to be forced to replace hardware and have to support DRM/join the association to buy/license the standard.
Hint: The Roland/Yamaha extensions already cover more patchsets than any individual composer, or even groups of composers are able to competently use in a single composition, and probably in all their compositions. Short of actively defining a hundred or a thousand examples of every instrument on the planet, and giving them their own unicode-equivalent 'typeset range' it is entirely pointless to update the MIDI standard at this time.
You are right, that is how people learn new things...
That's memorizing, not learning.
FROM THE SOURCE (CTRL+U) OF THIS PAGE ITSELF!!!!!
"<meta http-equiv="Content-Type" content="text/html; charset=utf-8">"
Yes I'm using caps, it's because I'm shouting.
Don't know why they still kling to the old DIN-port, USB(-c) was the way to go for MIDI 2.0, backwardcompatibility would be done through USB to DIN (which already exist and work perfectly).
I code MIDI quite a bit. The spec could use some more modern updates, but at the same time it simply suffices because it is quite flexible. A much bigger problem is the lack of device specific documentation; most MIDI controller vendors fail to publish details on how to configure their products over MIDI and simply refer to OS-specific configuration software or don't even bother to do that and just rely on DAW vendors to fix it. That includes some big players like Korg, Novation and M-Audio (the latter being the company that's been selling a broken MIDI interface for years without even bothering to fix some major but easy to fix bugs).
0x or or snor perron?!
Pretty sure those were MIDI ports that we plugged joysticks into.
Back at the time of the IBM PC, these were dedicated game port, featuring 4 analog axis and for button (and thus usable for 2 players, with an analog stick and 2 button each, provided the correct Y splitter cable. Or much later more complicated 1 player peripherals, with the Gravis Gamepad being the first with such popularity).
It came on a dedicated separate expansion card.
IT way much later, once MIDI got added to the port that a few jostrick decided to use it for extra feature (I think Microsoft Sidewingers ?)
That’s why they were on your sound card... or was that a joke
The idea of packing MIDI + Game port together came from Creative.
Like I said before, the game port began as a dedicated expansion card under IBM.
This ate a whole expansion slot just for analog sticks.
So instead some manufacturer tried to make multi function expansions that packed together game ports with a few other functionnality (e.g.: my parents' 386 back hten had serial, parrellel, floppy and game all on a single expansion board, with a few extra headers to the ports that couldn't be held on the card's edge.)
Creative is the first one who had the idea of making cards for gamers, containing many functions on the same board.
The first one was Sound Blaster: an OPL2 FM synth popular with the then popular AdLib, a game port, a DAC to playback samples, and a MIDI interface (for external synths such as the then popular MT-32 from Roland).
Instead of needing 4 different expansion card eating basically every single available port, it's just 1 single card. A real space saver.
(or even more, when they also started putting CD-ROM controllers).
Given the limited amount of space on the side of the ISA slot, MIDI was routed through game port pins, and required one extra adapter cable.
That's the point at which a couple of manufacturer jumped in and decided to use MIDI as one possible way to expand the possibilities of joysticks beyond the 4 analog + 4 digital channel offered by the port. (While at the same time still retaining compatibility with games relying on the classic interface).
(The other strategies where using a completely different port : some complex joysticks used a serial connection, but this was at the cost of lost compatibility with older games.
Yet a different strategy was joystick communicating solely over the classic game port, but being able to switch into a proprietary protocole that send ditigal packets of data over the digital pins instead of straight 4 axis, 4 buttons - Logitech's ADI protocole is an example. This had the benefit of working with pre-MIDI game port, and still also be compatible with old games when the joystick isn't switched into ADI protocol mode).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
I still don't get why they didn't just add a real MIDI port to the sound card.
Limited physical space on the card. You can't even fit MIDI connectors on an ISA slot.
(Even professional midi adapter back then used a flat connector between the ISA board and an external box that had the actual MIDI DIN connector.
Using an adapter on the gameport wouldn't feel that out of place)
Also, back when the first Sound Blaster was launched there were many more gamers likely to plug a cheap $20 joystick into their cards, than gamers owning an expensive MT-32 MIDI synth.
So it made much sense for Creative to make the game port ready to use, and as that port uses already half of the physical space of the slot, then find a way to route MIDI through the unused pins.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Seriously, stop dicking around with near useless spec updates and rip up and redo *everything* properly.
Audio should now be carried digitally by the cable itself, at a decent bitrate, uncompressed. Timing signals can also easily be incorporated into this.
Thunderbolt is the way forward here. Audio, Timing, everything in one bloody quick interface, saving f*cking about with cables. It's not moog-esque, but you should easily be able to route this shit in software between instruments, with a hub.
Do a spec, get everyone to buy in, move forward (and drop the dead stuff). It's still firmly stuck in the 80's no matter how well it works. The only real difference is there are now a f*ckton more greybeards than there were before, and OH NOES WE NO LIKE CHANGE!
Deal with it.
ASCII is still king - over 95% of web pages are ASCII.
Wrong. Even back in 2012, 60% of the web crawled by Google was Unicode. That figure would have only gone up since.
https://googleblog.blogspot.co...
Slashdot is in a tiny minority of ASCII-only websites.
You've confused "ascii" with "ansi" and confused "unicode" with "utf8".
That's pretty damn impressive considering you only wrote three sentences.
I'm a minority race. Save your vitriol for white people.
Great news that 2.0 is on the way !
But why on earth do they've chosen to go to JSON for some data interchange ?
It is not compact: takes lot of bandwith for little. There are better alternatives such as ASN.1 or even EXI.
It is no straight forward outside the web. You will need heavy marshaling (a good parser and generator). It will consume a lot of memory compared to alternatives such as FlatBuffers or ole Protocol Buffers.
It is neither easy to secure nor secured by default. In a close loop Din 5 Midi config, it is fine. But as soon as you include IP gateway, this will become an issue. Think IP worm from updating the instrument firmwares to broadcast everything that is played on them for instance.
And, I don't think it is future proofed at all.
JSON is just a complete opportunity technology.
for a new Atari ST with midi 2.0 ports!
On a long enough timeline, the survival rate for everyone drops to zero.
Now if the DAW manufacturers could just come up with a similar standard for storing song data. Trying to share songs between DAWs is still a total pain in the ass involving stemming the audio, lots of time normalising MID then writing it off to new files etc.
Best of the bunch is Reaper as at least it's song file format uses plain text.
it really does sucks balls trying to collaborate on projects where you both have different DAW software.
They need to start selling sound cards with game ports so we can use our midi again.
Okay, so I have a Casio WK-210 that strictly uses USB MIDI (internal driver in the keyboard, no driver software necessary nor available for download), while the Yamaha DD-55 I also own still has the legacy 5-pin DIN connector. Did anyone besides Voyetra/Turtle Beach make a USB-to-5pin DIN adaptor kit... and if so does the latest Windows 10 and MacOS still support that? Yes, the WK-210 also has several on-board drumkits, but I would still like to use the DD-55 as a dedicated drum machine even if I don't live play the drum pads for sequencing. 15-pin Joystick/MIDI to 5-pin DIN has been deprecated, so that kind of compatible connector is no longer available. (When it was available, it was a splitter type cable that provided a Joytick connector end with all 15 pins accepted, and a second available connector for the 5-pin DIN.)
I can't stand software synth latency... and that includes the built-in Microsoft GS Wavetable Synth in later versions of Windows and that trying to run something like Re-Birth RB-303 (Propellerhead/Steinberg) in a Virtual Machine is just awful for performance and control latency.
GM, and its successor XG, are nothing more than standards that says "Program 01 should be a 'Piano 1'" sound, Program 41 should be a 'Violin' sound", etc. A device with the GM MIDI logo will have a 'Piano 1' sound predefined in the "Program 01" space. You might be able to redefine or overwrite it if the keyboard has that capability, and probably can reset to defaults.
It's not a modification to the base MIDI protocol. It's simply a standard for program numbers, but if you are programming and playing sounds on your own synth you don't have to obey it.
> any number (such as a time-delta) can be encoded in one, two, or three bytes, depending upon how many bytes are needed (the most-significant bit is used as a flag to indicate that more bytes follow). This may have made sense when MIDI was introduced...
This is exactly what utf-8 is about.
If you're still looking to get rid of those MIDI cables, the owner of pigswill.org may be willing to compensate you for them
I managed to find the sources of where does my vague remembering of MIDI joysticks comes back from....
Sidewinders never used MIDI.
They did, the force feed-back is sent as MIDI messages.
Instead, they did what everyone else did, and they used the second joystick inputs to add additional features. That gave two more button inputs plus two more axes, so even without doing anything tricky you could implement four buttons and four axes.
...which would still limit you to 4 axises and 4 buttons. Going exclusively from the joystick to the PC.
Sidewingers rely on MIDI-out for force feed-back (sending information from the PC to the Joystick).
(Meanwhile, Logitech ADI protocol relied on rhythmically querying the port in some pseudo-morse-like patterns to trigger behaviors)
But by using the four button signals to make a binary number, you could either send four-bit numbers synchronously, or three-bit values asynchronously. I believe both approaches were used, but I'm not 100% on that.
Several joysticks used "buttons" to encode the HAT position. "CH Flightstick Pro" used that. (If button 1 and button 2 signal both "pressed", that actually means the HAT is actueated, and button 3 and 4 encode a 2-bit number telling which cardinal direction is beting pointed at).
Some did try to encode digital informations on analog channels "Thrustmaster FCS" used that instead: while the analog 3rd axis is simply the throttle, the 4th axis jumps to specific position on the axis, depending on which direction the HAT is currently pressed.
These where apparently popular methods, because other stick tried to be compatible with these. Some MS-DOS simulators can recognize this kind of sticks and use them directly in the game without requiring any 3rd party driver.
Fun fact: when in analogue mode, the Logitech Wingman Digital can select whichever of the 2 above methods you'd like.
(or Logitech ADI specific drivers can send the correct "pseudo-morse-like probe" and request the joystick to switch into ADI mode, at which point it completely drops any backward compatibility and starts speaking its own digital protocol end sending packets using 2 buttons signals. But that requires specific drivers and is thus only available in Windows or Linux. Old classic MS-DOS games cannot use that)
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Okay, so I have a Casio WK-210 that strictly uses USB MIDI (internal driver in the keyboard, no driver software necessary nor available for download), while the Yamaha DD-55 I also own still has the legacy 5-pin DIN connector. Did anyone besides Voyetra/Turtle Beach make a USB-to-5pin DIN adaptor kit... and if so does the latest Windows 10 and MacOS still support that? Yes, the WK-210 also has several on-board drumkits, but I would still like to use the DD-55 as a dedicated drum machine even if I don't live play the drum pads for sequencing. 15-pin Joystick/MIDI to 5-pin DIN has been deprecated, so that kind of compatible connector is no longer available. (When it was available, it was a splitter type cable that provided a Joytick connector end with all 15 pins accepted, and a second available connector for the 5-pin DIN.)
I can't stand software synth latency... and that includes the built-in Microsoft GS Wavetable Synth in later versions of Windows and that trying to run something like Re-Birth RB-303 (Propellerhead/Steinberg) in a Virtual Machine is just awful for performance and control latency.
You mean like this:
https://www.amazon.com/Proster-Adapter-Interface-Converter-keyboard/dp/B07H2D3127/
That took exactly 5 seconds to find on Amazon. Boy am I tired!
If you are having latency problems with soft-synths, try not using Windows. Its MIDI implementation is pants as far as latency is concerned, compared with macOS. In fact, the whole Audio/MIDI world is MUCH better on Macs than on Windows. Can't speak to Linux; but its audio support is so horrible, who cares?
https://www.reddit.com/r/edmproduction/comments/31fjnm/mac_vs_windows_in_terms_of_audio_latency/
But, due to the low speed of MIDI protocol (33 kb/s), and the relatively long length of MIDI messages, latency has always been a problem with MIDI. About 5 ms. is as fast as you can expect to get a note-on OR a note-off to happen. This means that a typical 3-note chord is actually an arpeggio, taking about 15 ms. to complete. Quite frankly, that amount of delay is already noticeable to humans. Add a 4th note to that chord and it now takes a whopping 20 ms (!!!) to "construct" over MIDI.
Now, add to that the response-time of the software or hardware synth, and it is amazing that any real "music" can transpire over standard DIN-based MIDI.
You've confused "ascii" with "ansi"
No, I didn't. You're the only one that has conflated the two.
and confused "unicode" with "utf8".
UTF-8 is a Unicode encoding. How could one confuse one for the other? This is as dumb as saying that I confused a car with a Toyota Corolla. When a Corolla is a car.
You nailed it. If it didn't wrap MIDI, OSC would be absolutely useless, instead of just mostly useless like it is now. At this point, it's safe to label OSC as "abandonware:"
"OSC 1.1 Specification
The most accurate document reflecting the 1.1 vision is the NIME 2009 paper." http://opensoundcontrol.org/spec-1_1
Maybe grandpa meant the MMA should have endorsed OSC as "MIDI 2.0" and then it would have had some traction and not be abandoned. Doubtful.
http://opensoundcontrol.org/implementations?sort=desc&order=Node%3A+Updated+Time
The newest hardware on that list is 7 years old, and it's an OSC/MIDI translator. All but one of the newest software on the list is 6 years old.
Abandoned. Dead. Over.
Cue yourself.
Different AC here...no, not like that. That is a MIDI interface for devices that use standard MIDI DIN connectors to connect to a computer that has USB ports.
The issue here (I have a Casio WK-3800 with the same problem) is that USB is not peer-to-peer. The keyboard is a USB peripheral – it has a USB-B socket to connect to the USB-A port on a PC, i.e. it is a "device" connecting to a "host."
If it's class-compliant, you don't need a special driver for it to appear as a MIDI device. My Casio is not class-compliant.
That means I would need an adaptor that not only has a USB-B plug on one end and standard MIDI DIN plugs on the other, but acts as a USB host to translate the proprietary Casio data to standard MIDI commands. A simple physical adapter will not work; it requires logic.
USB Serial can introduce a serious 1 or more millisecond lag. FTDI, a very popular USB-Serial bridge chip, has default 16ms delay until transmitting its serial data when it has received less than 64 characters. MIDI characters send at around 0.32 milliseconds a character, so this would cause a delay of 20milliseconds. I work with USB based motion controls for CNC and pick and place machines. With all the USB-Serial bridge chips out there are some seriously detrimental real-time issues such as: latency timers (FTDI 16ms for example), then the is the resync from USB 1.x devices connected to USB 2 hubs, or even USB 3 hubs. Then you get Windows involved, add more. Native serial ports connected to the PC bus work well. I have seen horrendous delay/latency issues hooking our equipment to a USB port on the front panel of a PC. Put certain USB 2 hubs in series, you get a different set of latency, sometime better. Then there is the noise issue. The differential USB signalling is better than the single-ended RS232 type, but probably not arc welding friendly. The MIDI current loop is pretty good with that regards as well as getting rid of those pesky ground loops.