New Roland Controller - MIDI 2.0

Will Pianoteq exploit this new MIDI 2.0 protocol, just as soon as it is released?  It’s like an answer to my prayers (aside those for possible world peace)!

The second version of the Roland controller seems a godsend  —under one thousand (1,000) dollars (U.S.).  For the price, you get variable note-off velocities in addition to high resolution note-on ones!  No other board manufacturer might offer so much for so little!

Yes, it does!  That is the meaning I get.

Yes, I have a thought. Have you tried re-quantising the high resolution MIDI file, to reduce the number of velocity levels down to 127? The reason I ask, is because that would prove your point - that it's ONLY the extra velocity resolution that you can detect.  There may be other differences between the two files, not necessarily related to velocity resolution.  Quantising the velocity steps to 127 would leave everything else intact. If you could still notice a difference, you could then dig deeper, to try and determine why you hear the difference.

I'll be absolutely amazed if anyone could detect a difference between standard MIDI velocity resolution, and high-res velocity.

Greg

If you're saying that "high resolution audio" is just marketing, meaning any resolution higher than CD quality, until recently I would have agreed with you. However, I have just learnt of a big study from Queen Mary University, where they re-analysed lots of other earlier studies, and determined that high resolution audio can sometimes be discriminated, when comparing to CD quality. This study is being disputed (what study isn't?), but I've decided to accept the results for the time being. I personally have zero interest in anything above CD quality for listening to music (I actually think even Apple lossy sounds excellent), but I understand the need to have high resolution formats for intermediate use - whenever the signal needs to be processed in various ways. And I accept that some people may be able to appreciate high resolution even for the final output.

Greg

Just regarding my suggestion to quantise the high-res Yamaha MIDI file - I said that as if it would be something easy to do, but I don't think it is easy to do at all. I found an app which allows the files to be edited, but what we need is a simple way to process the whole file, and I don't think this app would be suitable. FWIW though, the app is here: https://zenph-reperform.software.informer.com/
Maybe Modarrt could answer the question though: is there any difference, OTHER than resolution, between the standard MIDI file, and the extended MIDI file, that could alter the result of Pianoteq's renderings?

Greg

Thanks for sharing, Robert.
Had my eye on this but not anymore.

I wonder if Robert had a faulty unit.

This was discussed a years ago. Casio had Hi res implemented for ages, using an extra controller input. It is supported by Pianoteq for ages also. The point is that it has zero influence compared to 127 velocities. It was tested just "by experience" - not noticeable, and also with some technical tests (which potentially have some chance to have some flaws). The point for that tests is that Pianoteq simulation introduce a significant degree of randomness, or should I better say unpredictability, that completely fills the gap, say, between 78 and 79 velocity input. So even pure technically speaking the higher resolution than 128 has no point.

Where MIDI 2 will be invaluable is for the soft synths controls, where you could finally make a filter sweeps via a knob on your controller without hearing the 'steps' or the need of artificially smoothing them via the algorithms.

@Amen Ptah Ra
I suggest you may do your own tests and post the results here. One of the methods I suggest: are you familiar with basic DAW MIDI editing? Take one of your choice, open Pianoteq VST and:

1. Create a series of notes, say, every fourths, C60, ten or more in a row with a velocities of, say: 50, 50, 51, 50, 51, 51, 50, 51, 50, 51. You may loop them, or it is even possible to make a randomization solution if one is interested enough. (btw, are you able to clearly hear the difference between even 50 / 51 100% of times?)
2. For the second test make the same or another series of the same note, same range, 50 - 51 but also use the MIDI Controller 88 with random or if you want a gradual adjustment.

I mean you can make this test to check if it is worth it even if you don't have a hi res keyboard, you can still make a hi res midi files in any DAW on a market. Please check it yourself and share your results. Anyone is welcome to make it as complete as possible, checking it at higher or lower velocities, etc., or inventing any other scenarios where the hi-res implementation is obvious and clearly audible. I'm genuinely interested to know other people results. I've spent quite a lot of time testing and checking it in the past and so I made my well weighted conclusions. If one can actually prove me wrong that would be great. I'm very aware that the tests I've made may not cover all possible scenarios, and may have whatever other flaws, so if one can clearly show the advantages of the hi-res that would be just great.

One more thing, not only did I listen to the results of such tests, I also evaluated the resulted wav files in Sound Forge (you may use the Audacity or any other), checking the actual levels and what else, so that is why I'm saying that at least for the time of that tests, the randomness in the Pianoteq implementation was too small to be noticed by hearing but at the same time it is too big to make the hi-res sufficient.

I'm not saying that hi-res is bad, I would always prefer to use the higher tolerances in any area especially if they are available at a zero cost and even if I'm not able to directly appreciate it. Saying that I still highly doubt anyone would be able to appreciate it in case of Pianoteq - controller combo. I'm not saying it in a negative way, just stating what I think. I may also add that from experience Casio PX-160 + Pianoteq feels miles away purely in terms of velocity and dynamics control than any real grand I played ever, including the very top ones big names I was happy to have a chance to play on.

In case it is not clear yet, I actually own and play for many years on a hi-res keyboard, the PX-160.

Nowhere they say that it is the only keyboard that is capable, or the only officially approved one for the hi-res.

-added-
BTW just noticed, the default presets I use have the "Hammer noise" set to humanize in the settings, seems that well was the case of showing randomness. Too lazy to recheck, nor really care about the subject, it is perfectly fine for me whatever it is. Look, Pianoteq implementation is not really a precise sinewave synth emulation, it is all random all over the place, you may see it in the Pro version, ALL the parameters deliberately and carefully very so slightly messed up. And even a lot of them are still "under the hood" and not reachable even in the Pro version. The resolution for those settings is also not some gazillions of precision, it is within the reasonable boundaries, well enough to be unnoticeable. Even if you take the common dynamic resolution for many (most?) factory presets of 40 dB then divided by 128 equals 0.3125 dB difference. Quotes from the Internet:

Again, nobody says hi res is bad, it is very good, but unfortunately not that much of a deal in our case, there is absolutely definitely no 'dramatic' difference, I would dare to suggest that if any at all.