The topic of MIDI 2.0 can be a hot minefield when discussing what people want. I know that there have been a number of owners of past Arturia KeyLAB controllers who were disappointed that the latest MK3 doesn’t support MIDI 2.0 and some of the features that were not included.
A post that I saw was about MIDI 2.0 1. High resolution data instead of only 128 values - for example very fine Velocity response and smooth no stepping parameter tweaking.
This got me wondering what people are excited about or were wanting to see in upcoming hardware. For me the idea of High resolution data is near the bottom of the list of something that I see wanting for new Firmware for the MK3 or future hardware. So I figure it would make sense to explain my reasons or rationality.
Currently we have 127 levels inside midi. How they came up with that number I don’t know but for the most part it’s been fine to live with for me. I don’t believe that the average or even the above average person can differentiate between adjacent step values that currently are used. This is something easy to test with a DAW and the current hardware/software you own. If you create a midi region and put into is a series of notes with a +/- 1 or 2 on the velocity entered to the prior note, on playback most if not all would have a hard time distinguishing any change. Especially if done blind, or better yet double blind. It would become even harder if you tried the same with instead of the same note, you changed each tone of each note. I have tried this with different trained musicians and they found it more a chance of luck rather than definitely knowing. I found it wasn’t until you get to +/- 6 or 7 that you perceive there is a change happening but its not until the +/- 10 to 12 that you definitely tell that the notes more or less emphasized.
So to me that says that yes, on paper that having a higher resolution looks better, but in practical terms its rather redundant if most people can’t tell the difference with a 1/127 step increment from quietest to loudest, then what benefit does having a 1/1024 step?
But we are not just talking about keys, this will also go to the encoders and faders too. Again, the challenge is pull up MCC and (on the Mac) hit command+M. Now turn your encoder wheel and look at the value output. How much of a turn did it take to get the value to change by 1. Same with the fader. I have seen posts on other sites that complain that the faders are just too small (strike distance) to be able to get the accuracy of the current midi 127 step accuracy. they are buying larger 100mm faders so that they can get the 1/127 accuracy on the fader. Now I will be in full agreement that some of the advanced fader control surface midi devices would get lots from midi 2.0 as some of those have a 150-200mm throw. but not on a KeyLab as there just isn’t the room for a long throw and we don’t need it.
Am I missing detals or reached a totally wrong conclusion. Please, I would love to hear a reason or rationalization for why higher resolution data is needed in the KeyLAB or likely even any of the arturia instruments.
I don’t know enough about MIDI 2.0 to have any opinion on features, but as far as the original 0-127 range goes, here’s a quick explanation I found on reddit:
MIDI is based on a series of bytes, grouped together to form MIDI messages. A note is a 3-byte message where the first byte denotes the MIDI channel and status (e.g. ”Note On”) and the next two bytes are for note and velocity. The left-most bit of each of these bytes is required to be 0, leaving 7 bits for the parameter value, hence a 0-127 range.
I don’t understand. The topic ask a question. Your first post does not reflect this question. What do you want? No midi 2.0?
Are you interested in what people want from midi 2.0 or interested in technical arguments?
It seem to me, that you say many people have spend decades of work in midi 2.0 for nothing or for scam. Time will tell.
A main thing about midi 2.0 high resolution is to reduce or perhaps eliminate stepping. Stepping is something many complain about.
Both the sending and the receiving unit need the resolution to respond properly.
High Resolution is also about getting a analog feel.
I have actually also thought about starting a topic with the question you ask in the header. But to me the technical side is another matter.
I don’t know for sure, but i assume you can find lots of technical informations and tool downloads on midi.org , if you are interested.
I have mentioned this 5 things about midi 2.0 in a none prioritised list:
High resolution data instead of only 128 values - for example very fine Velocity response and smooth no stepping parameter tweaking.
Number of midi CC’s
Very low latency midi - Also tighter midi sync/ better timing
Two way communication
No need for MPE that’s based on Midi 1.0 - MPE was probably none excisting if midi 2.0 excisted then.
There are more that some people might look forward to be implemented.
Time will tell how it all will work, and how much will be implemented widely and used.
It WILL benefit people such as e-drummers, classical and Jazz pianists, EWI players etc.
As an ex drummer myself, and someone who uses the services of an e-drummer on occasion, capturing the ‘groove’, particularly with styles like Funk and Reggae which derive their groove in large part from ‘imperfect’ timing; it’s HIGHLY attractive.
Then there’s the issue of dynamics, 127 is pretty limiting in many types of music, Jazz and Classical being the obvious examples.
‘Ghost notes’ on a snare drum never sound quite right, as do ‘drags’, there’s always something just a bit mechanical if you listen critically.
It would do away with the need for ‘articulations’ such as ‘drags’ then.
I am looking forward to higher resolution controller data > using 14 bit data rather than 7, making them equal in resolution as pitch bends, NRPNs and RPNs…
and this is the cool bit about sharing experience. Getting to expand your knowledge, challenge your beliefs and become a better musician in our case out of it.
The Mackie controller uses midi. (And all varieties and protocols like HUI (both Mackie developments) and drum computers like the Roland TD series. once you use a midi-ized studio with a drumcomputer, three midi keyboards and a mixer controller based on midi messages— that’s a lot of data. Being bounced around by ‘stupid’ machines. E.g. they don’t configure themselves.
It would make sense to speed up the connections. To open the bandwidth. To make sure usb drivers work together and if the processors in the boxes allow, to speed up configuration.
High-Resolution, Naturalistic Control Input/Output
MIDI-CI for Profile Configuration, Property Exchange, Process Inquiry
Haptic interface for real-time control of audio/visual/mechanical expression
Current Project:Rust-BOA-Servo/TAURI Architecture for Live-Performance MIDI Orchestration of Integrated Motion Capture 2.5D SVG Animation synchronized with Music.
So for me, MIDI 2.0 is no longer bound to music alone, but can now be a truly responsive, Multimodal Instrument Digital Interface.
Note: Recent Rust developments should allow me to now work around the past complexity of using C++ Crate for ROS 2 mechatronics/motion control.
They didn’t just ‘come up’ with that number: it’s the highest value number that can be represented with just 7 bits. It’s the same reason why you get panning controls that go from something like -63 to +64 (sometimes the other way round). It’s a 7-bit system so there are only 128 possible different combinations from 0000000 to 1111111.
(I know i’m quoting an ancient post, but someone might find it beneficial.)
That is very true but where the biggest pondering is… ever since moved in computing from an 8bit computer that memory and addressing was a focus so you needed to optimize and keep registry sizes small. But we long since gone through 16bit and 32bit and and now are in 64bit world. Yet we seem fine to live with an 8bit midi 1.0 protocol and think that is just fine
I don’t understand the point you are trying to make here. Also, who are the “we” you seem to speak for?
Taking that at face value; if someone is accustomed to creating music shaped by low-resolution and limited capability tools, then of course that would feel “just fine”.
However, if like myself, you have invested decades in learning to play and perform music with the subtlety and the nuance that acoustic and analog electric instruments afford, then like me, you may have long wished MIDI controllers and complementrary synthesizer with the same fidelity of expression as those acoustic and analog instrument.
And that is before addressing how much simpler it will be to connect hardware and software instruments that are network-aware of each other, can share their capabilities, and active state at any time.
The netter resolution is nice, as is more CCs considering the scope of software instruments. One thing I hope they do is make it very modular. I like to call it MHTML … Or, Midi HTML. It would be great to have a list of parameters, give them a class name, assign a CC to it, and then the instrument gets it. Then, we can share the ‘code’ and the next guy/gal does not have to do much more than edit an easy to see/edit code…
But while we are on the subject of things that need a universal upgrade… VST software control. It’s nice that most software “learns”… But you know what I’d love to see??
Two controls on every HW Workstation and VST synth. Just @#@@$@ two! Here they are:
MK On/off
MK
How does it work? You turn the Magic Knob feature on, and assign a CC value to a knob.
Now every flipping thing you click on with the mouse is controlled by that one knob.
Then we can all forget about endless CCs.
1 Like
_Need help ?
When you can't find the answer online or in the product manual, the Technical Support Team is here!.
_Stay tuned
Follow us for the hottest sounds, fresh content, exclusive offers and Arturia news as it happens.