Since MIDI 2.0 only support USB (for now), will MIDI 2.0 need a new or updated OS class driver for the 32/64 bit universal packet interface and device descriptor for devices ? Page 16 of the USB MIDI class driver spec defines support for 1,2 or 3 bytes transfers only, not 4 or 8. How will this have to be implemented by MIDI USB device manufacturers and well as on the Mac, Windows and Linux OSes side ? Treat it like an Sysex packet ? that would be inefficient and cause processing code overhead. How will Windows Win32 API support the packet interface . AFAIK Microsoft does not plan to make changes to the Win32 MIDI API anymore but only for Windows UWP. so MIDI 2.0 is only for MIDI UWP applications , that would mean most DAWs are left out in the cold, How will this impact Linux ALSA ? Many folks may not have the funding for an annual MIDI membership like larger companies, but like me, many may have these questions and some clarification for this would help. There seems to be no information on the MIDI website to address any of such concerns. P.S Would the first byte of the message with b7=0 not confuse the driver of running status ? Why going through MIDI-CI protocol negotiation if a new descriptor or drivers is needed that could just indicate 2.0 capabilities ?.
MIDI 2.0 uses a Universal MIDI Packet (UMP) format for MIDI 1.0 Protocol messages and MIDI 2.0 Protocol messages.
There is a Version 2 of the USB Device Class Definition for MIDI Devices in development to support the UMP format. It will require new class drivers.
All of the major OS vendors are aware of the progress. I think the ALSA team is also aware.
Chair of MIDI 2.0 Working Group
Some more questions
1. Will Microsoft revise the win32 API or only going to support MIDI 2.0 UMP for the UWP API or will there a totally new MIDI API with WIn32 MIDI API backward compatibility ?
2 . How will Chrome support MIDI 2.0 for the WebMIDI ?
3. Will this mean a change of MIDI USB device descriptor for MIDI USB devices
4. Will this be a totally new class or just some added flags/field in the current header?
5. and is the descriptor definition part of the MIDI specification ?
Some excellent questions, Happy.
I look forward to receiving some answers.
Meanwhile I shan't be holding my breath for MIDI 2. It seems years away.
Hello Mike
Last week I sent a PM to you if MIDI.org could get into contact with the ALSA team to work out the issues that should enable the Linux ALSA development community to implement MIDI 2.0.. The primary ALSA contact was provided in there.
P.S It seems that low level drivers should be able to send and receive MIDI UMP packets but the higher level API may need changes.
if that is true, then why a new USB class and changes or is this Microsoft/Apple specific ?
Thanks.
1. Ask Microsoft.
2. Ask Google.
3./4. The old class specification does not define UMP, so old devices/drivers would not work. (It does not matter whether you add to the old class specification or create a new specification; old hardware/software must be updated in any case.)
5. The descriptors are defined in the USB device class specifications.
Would have expected that MIDI.ORG would be a proxy to the music world with Apple/Microsoft/Linux community for answers. Sad to read this. Wonder if being a paying MMA member you would get more insights due to the opacity of these large software companies. MS probably will never update their WIn32 API, so it will be a real question we will ever see MIDI 2.0 in the current DAWs. It is probably is going to be an Apple only thing for very long if not forever.
I don't think UMP-capable devices would need new Class IDs, but the protocols will have to be updated since current USB-MIDI Device Class Definition 1.0 only supports 32-bit MIDI 1.0 messages and none of the 32/64/128-bit UMP messages (though 32-bit messages are very similar). The work on an updated USB-MIDI specification is currently underway AFAIK.
I would expect OS vendors to update their APIs to support MIDI 2.0 UMP formats by 2021. Remember how MIDI 2.0 was heavily endorsed by Microsoft and Google at WNAMM 2020 , besides well-known MIDI Manufacturer's Association members like Yamaha/Steinberg, Roland, Ableton, Native Instruments, and ROLI.
Microsoft will most likely only update the UWP API definitions, but I'm certain they will also allow legacy Win32 programs to call these new APIs - and now that Microsoft Edge is based on Google Chrome, WebMIDI should get an update as well.
Considering Apple, they own Logic Pro since 2002 and they supported RTP-MIDI (aka AppleMIDI) since 2005 - so I'd expect them to follow suit as soon as RTP-MIDI is updated for MIDI 2.0.
Thanks, perhaps those lucky to visit NAMM were at the session but on Youtube this session Strategic Overview and Introduction to MIDI 2.0. isn't to be found AFAIK. There is only the MMA and VST session and these don't go into the detail or impact to the transport layers or OS hardware drivers and commitments but largely repeat what was already published quite a long time ago, Was Apple there as there as the frontrunner of music production hardware and software support, if not, why not ? There is nothing formal to be found on OS vendor's oron USB-IF https://www.usb.org/search/node?keys=MIDI about plans. RTP has 32 bit blocks MIDI data payload, so maybe only the higher level (client) software needs to translate to it UMP packets for further processing to/from the MIDI device. Sysex can be done using the current USB 32 bit MIDI packets, so shouldn't it be possible with just a driver change. The driver API change could tell the driver, and "now your protocol is 32 bits based data" for a packets from the USB device, for example from the client view : openMidiUMP(device). and openMidiUMPort(port). Then with some mechanism to tell which one is the 1st, 2d.. 4th 32 bit UMP packet. The callbacks may need some new event names too. Microsoft has already backtracked on WIN32 API support for UWP and the API wrapper for Bluetooth had been problematic and with bugs for year and only one DAW implemented UWP. The Win32 MIDI API may need too many changes to support this without breaking the Win32 API. I see MIDI 2.0 coming in 2020 for Apple/Google products (build in MIDI Bluetooth/ RTP), but not for Windows. 2021.. maybe even later. MIDI 2.0 ready products are for the moment more like hype-ware that may already replaced when MIDI 2.0 UMP arrives. Until May 2016 the MIDI 1.0 spec from 1996 wasn't formally available, developers had to buy for 30 years books written by people who had the insights, or reverse engineered it, from Compuserv or some other .TXT files posted elsewhere. Without search engines that wans't easy then. [iL]os Angeles, CA - May 2016... The MIDI Association, the global community of people who work, play and create with MIDI, is offering the MIDI 1.0 specifications for download at no charge to their members at www.midi.org. Until now, 300+ page document has only been available at a charge of $60 to cover the cost of printing and shipping. will developers need to wait another 30 years for MIDI 2.0 spec, and even if it's released as promised by Feb 2020, the USB , RTP and other details won't probably not be there just a copy of the pages already on the MIDI.org website ? I would not hesitate to pay $60 to get access to all the MIDI 2.0 specs, USB and any other transport layer plans , details, progress and spec. though the only option now is a $600 annual membership fee and then it still guessing what you get for it or still be kept in limbo. https://www.midi.org/forms/mmaapp.pdf
I'm not sure what are you complaining about. Even if preliminary work is undergoing on the updated USB-MIDI protocol and AVBTP/RTP transport formats, these proposals will not be submitted to USB-IF, IEEE and IETF for review and ratification until the MIDI 2.0 specification is officially published (and available free of charge on this site, as was announced).
That said, early drafts are probably available right now to MMA members, so vendors and OS makers can begin implementation and testing efforts - this is critical since MIDI 2.0 does not define its own hardware protocol anymore and relies on USB and Ethernet as transmission media.
Once review and ratification of USB, AVBTP and RTP protocols is complete, these will be available from their sites (free of charge from USB-IF and IETF, for a small fee from IEEE).
Note that companies like Microsoft, Apple, Google, Yamaha, Roland, and Native Instruments are members of the USB Implementer's Forum, but the MIDI Manufacturer's Association is not. Also IEEE and IETF are associations of engineers, so all members are individuals and serve in a personal capacity with no corporate or governmental obligations, unlike ISO and IEC.
This naturally gives priority to USB-MIDI as the easiest to get through, since IETF and IEEE processing will probably take quite a bit longer.
I'm not sure what are you complaining about.
First of all, I am not going to lower my standards by "YOU"-ing / attempted to inflict some kind of personal motivated argument. Secondly, above were laid out some questions to obtain an idea what the impact is of MIDI 2.0 related to USB devices (OS. driver etc.) and an attempt was made to also have the Linux community involved. Apparently providing insights or support for these are not on the radar or MIDI.ORG is not planning to share unless shedding out $600 annually. Not much to nothing about this has been published on MIDI.ORG website on in Youtube video's. For sure, MIDI.ORG does not have to but giving a better picture / roadmap of MIDI 2.0 would be more fair than telling less than half of the MIDI 2.0 story..
The community effort by MMA could sure could be better, but so far it's the best they have undertaken in the last 25 years. Where else you would get answers directly from the author of the USB-MIDI spec? You might not like these answers, but asking the same questions over and over is not going to end with different ones.
And just to be clear, this is not a technical support forum for some paid service or product - it's the forum for the user community, so no-one here is obliged to serve your requests. If you think otherwise, you've taken the pledge of the MIDI Association - "to nurture an inclusive global community of people" - a bit too far.
Paying the annual MMA membership fee of US$600 will get you some early access to specs and presentations, but you'd first need to start company that builds MIDI products.
So for now, you'll have to wait for the final specs to be published and see how device manufacturers and OS/API developers will respond.
Hi,
Sorry, we haven't responded here recently. We have been focused on getting the MIDI 2.0 specs signed off and ready for public download which happened TODAY!
They are available in a link at the top of the article on MIDI 2.0.
Regarding OS/API developers, we held meetings at NAMM with Apple, Google and Microsoft. All of these companies are public traded and so have to be somewhat close to the chest in terms of discussing what they are working on.
Pete Brown from Microsoft gave a presentation at NAMM and discussed many of the issues brought up here.
A3E Workshop: Saturday, January 18th • 12:00 PM - 1:00 PM
The Hilton: Level 4: Avila 6
Update for Windows Music Creation App Developers
• How is Windows doing?
• What's this I hear about Linux on Windows?
• What about MIDI?
• How is the app model evolving?
We can reach out to him to see what he would be allowed to say publicly.
Apple, Microsoft, Native Instruments, Roland, and Yamaha have all contributed to the USB-MIDI 2.0 specification that is already under development.
Also bear in mind that companies like Roland and Yamaha often develop their own drivers and don't have to rely class compliant drivers.
"Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning."
Winston Churchill
I have been too busy in the last few weeks to come back to this thread. I apologize. But the core MIDI 2.0 specifications are now publicly available as of today. Download link is on this page: https://www.midi.org/articles-old/details-about-midi-2-0-midi-ci-profiles-and-property-exchange
> Wonder if being a paying MMA member you would get more insights
> due to the opacity of these large software companies. MS probably
> will never update their WIn32 API, so it will be a real question we will
> ever see MIDI 2.0 in the current DAWs. It is probably is going to be
> an Apple only thing for very long if not forever.
I’m not so pessimistic. The 3 largest OS vendors are active members of the MMA and following or contributing directly to MIDI 2.0. MMA members have direct conversations with the influencers and developers of MIDI features in the operating systems. You can try to read between the lines and maybe gain insights, but these large companies generally do not clearly announce their plans, even to MMA members.
> Thanks, perhaps those lucky to visit NAMM were at the session
> but on Youtube this session Strategic Overview and Introduction
> to MIDI 2.0. isn't to be found AFAIK.
This video we gave at ADC in November gives a good overview, aimed at developers or manufacturers: https://www.youtube.com/watch?v=K2dAIvrI8zg&t=1s
The sessions at A3E/NAMM expanded on the presentation we gave at ADC. One was more focused on information for MIDI users. The other was more aimed at developers or manufacturers, a slight variation of the ADC presentation.
The USB MIDI 2.0 specification is still under development in the USB-IF.
Thanks,
Mike.
Chair of MIDI 2.0 Working Group
start company that builds MIDI products
For sure there are many developers that would like to bring innovative ideas to the market and utilize MIDI 2.0 and apply it in new products without having to face steep costs.
Beside USB and SYSEX IDs there are so many other costs involved with commercialization. however, bringing cost to the table that cold be avoided and unnecessary for the product to function is something any small and large enterprises looks at (I work at both sides and know how easy it is when speaking from the perspective of the latter one and how hard it is from the perspective of the former one). Practically limiting devices via cables to hosts MIDI 2.0 with USB ,for now, pushes creators of MIDI devices into a $6000 USB vendor ID application + the $600 annual membership for SYSEX that is now mandatory for MIDI 2.0. This is not a small feet for niche developers and difficult to amortize.
There might be many others that don't speak out but look at this in the same way that this will be very restrictive in the creativity for instrument and device development.. Hopefully MIDI 2.0 will not be restricted to USB 2.0 and up devices which would push up the costs for devices as well. MIDI 2.0 UMP could well be transacted over DIN (or a 3.5mm plug which is a standard) if for example encapsulated in MIDI SYSEX An alternative what MIDI.ORG/MMA could have considered was for MIDI.ORG to apply a USB vendor ID and then make the USB device ID's available for "niche" developers for a small fee. Another idea for USB is to have reserve one ore more bits in the SYSEX device ID to inform the Sysex vendor ID is the same as the USB vendor/product ID in the UMP SYSEX header , thus removing the need for an alternate/duplicate vendor ID for sysex (and the US$600 annual fee). Another way, the UMP SYSEX Vendor ID is once again limited to 16 bits - "640K is sufficient ").- If a larger set of bits would have been chosen (like ethernet MAC address or IP addresses) the SYSEX ID would have been sufficient to limit restriction and have any developer just apply one via the website. Concluding, seeing MIDI 2.0 penetrating into low volume specialized products from SMB's or even individuals seems unlikely, MIDI 2.0 seems more intended for the purpose of monetization ("make everything a service" as young students are now brainwashed with at MBA) and for corporations that have all the bandwidth for commercialization than for the purpose of innovation. and P.S:
, so no-one here is obliged to serve your requests
Regarding the "You-ing" once again, happily ignoring that.
Appreciate the fruitful discussion and replies and fully realize being just another nut (case) in a bag full of bolts and no-one here is obliged to read or reply on these comments. We live in a free world, as in -there is no free lunch. not even with MIDI 2.0