Hi, It's quite a corner case but I should start to look at this.

Edit: This is not about Midi keyboards but Midi in the context of microtonal, polychromatic, music theory, music/integration/software development. It's the limit of Midi itself and a work around to fix that problem. So please if you like or don't like midi, or like don't like ultrasonic sound is not the subject of this post. try to be constructive.

I am working with raw frequencies, more experimentation than music at this level, but I am planning to add a "music notation converter" so frequencies can be exported as midi, score, scientific notation with cents, commas, poly-chromatic notation, etc

But Midi reach a limit at Midi 127 that is around 12kHz, now Midi it's the first target as this allows to easily export/import to any music production software, it's there any solution or work around to go beyond this limit?

Also in a wide spectrum, new recording technologies allow to record at 512kHz that allow 256kHz freq
quite beyond the 20kHz of human limitation 4 octaves up at least 16, 32, 64, 128, 256.

A hacky way could be using pitch bend value of Midi but I don't know the limitations there, and anyway that will be my last alternative.

 

Last edited: Jul 6, 2021

Hello @GabsIT, good luck for your new Idea, a " music notation converter ".

Midi 2.0
The MIDI standard is being completely overtaken for the first time in almost 40 years.
Innovations in MIDI 2.0 could spur the development of electronic instruments. www.midi.org

 

Last edited: Jul 5, 2021

I checked the info before and today again, I think midi have the worst documentation that I ever seen.

For what I read some values are now 16 or 32 bit, but looking at any software that uses midi (maybe not midi 2.0) the max value is hardcoded to 127

top of MIDI tuning range G#9/Ab9 gis’’’’’’/ges’’’’’’ 13289.75
127 G9 g’’’’’’ 12543.85
Well it's pretty recent tho (2020), maybe is just not widely implemented

 

Maybe the "bad" MIDI documentation is based on a misconception: There is nothing like a fixed frequency tied to a note number in MIDI (except suggestions about how it can be done), and as you may know, different manufacturers have different ideas about how to map the frequency scale to MIDI notes, except that "Middle C" relates to note number 60 and can map to one of the frequencies 131, 262 or 523 Hertz for example. You're free to transpose octaves in your sound generator to achieve higher or lower frequencies.

In other words: MIDI is a language that can drive your sound generators and what these make of it is something completely independent.

 

MIDI is just data. You can transpose your oscillators (fundamental frequency) as high as you want. Not to mention the overtones that many waveshapes have.
Or maybe I'm just misunderstanding you (royally).

Also, I don't know what you mean with "512kHz" and "256kHz"(?) as it makes little sense. I'm sure you mean 96kHz and 192kHz sample rate (which has a Nyquist frequency of 48kHz and 96kHz). This has very little to do with MIDI per se.

 

Last edited: Jul 5, 2021

Spoiler: Comparison of MIDI standarts

Comparison of MIDI standarts

https://en.wikipedia.org/wiki/Comparison_of_MIDI_standards

Spoiler: 0/29/01 MIDI history

0/29/01 MIDI history

The story of MIDI is inevitably, as the format is "industry-generated", so to speak, a business-oriented story. MIDI is not a standard created by independent institutions, but a manufacturer standard. For the user, this means that the compatibility density I mentioned at the beginning again? in individual cases considerably - is restricted. Here, the manufacturers' economic and tactical considerations play a role, and some things are simply done out of vanity. But one can say: In summary, despite all the calls from critics, the MIDI standard is the one that works best because it is so simple.

The first version of the MIDI specification from 1983 is the result of a process of unification between the manufacturers Sequential Circuits and Roland, i.e. an agreement between companies and competitors. The content of a standard created in this way must also be shaped by striving for supremacy in the market. Substandards such as General MIDI (GM, Roland), General Standard (GS, Roland) and Extended General MIDI (XG, Yamaha) are attempts to induce customers to buy products from certain brands or to force other manufacturers to move in a certain direction work. If this coercion works, the customer can even benefit. In the case of General MIDI, which can be found as a compatibility feature in sound modules from many manufacturers, it is even the case that many MIDI users with products that are not GM-compatible,can't do anything (solo entertainer / keyboard player).

The creation of a universal interface with the spread of synthesizers in popular music seemed to make sense in terms of market technology.

Minimoog
This can be traced back to the development of the Minimoog (type A in 1968, prototype; type D in 1970, series model, built until 1980; type E is considered to be the first minimoog with stable tuning). This instrument was still offered in 1996 as the Minimoog F, equipped with a standard MIDI interface and RAM and ROM memory. In parts it still serves as a model for software synths such as Access Virus or Clavia Nord Lead. Artists like Klaus Doldinger, the Beatles, Wendy Carlos are Minimoog users; House music producers still use it today.

The little Minimoog excursus is useful because of the instrument's tough longevity, as its builders gave it more than one control interface, including Gate / CV and MIDI.

Forerunner of MIDI
The somewhat thin sound of earlier (affordable) synthesizers often made it necessary to play two instruments, slightly out of tune against each other (7 cents often sounds very good), in unison (at the same pitch). Since this (especially in live operation) should be done from a keyboard if possible, a kind of electronic coupling was required, which, similar to a church organ, connects two works by mechanical connection, in this case two synthesizers.

The Gate / CV method, also known as CV / Gate (Control Voltage) is based on a system in which the pitch is determined by voltage values. Such trigger / gate outputs were definitely available on many synthesizers.

The transmission of further control commands was only possible with the acceptance of a very complex cabling. The transmission of polyphonic control signals alone would have required a cable for each voice to be transmitted. Furthermore, CV / Gate only made it possible to transmit other commands to a limited extent, since with analogue control there are major compatibility problems with different structure architectures and control hierarchies. If you wanted to avoid excessive cabling in studios and especially on stages, the solution had to be a digital, serial interface. It enables (apparently simultaneous) control of two differently designed instruments in master / slave mode via one cable.

USI
The USI interface (Universal Synthesizer Interface), 1981, was developed by Dave Smith and Chet Wood of Sequential Circuits with Tom Oberheim (Oberheim) and Ikutaroo Kakehashi (Roland). It was presented at the AES (Audio Engineering Society) congress in 1981 and is the immediate forerunner of MIDI.

MIDI
As mentioned last time, the necessity of such a control as MIDI is based on the fact that popular music had to be served with electroacoustic music. Popular music must have qualities that are publicly available. (Lightness, catchiness etc. are parameters whose virtuoso application can be called "art"). For the technology that is used for execution, this means that it must function properly.

After the USI interface appeared in 1981, a MIDI prototype was presented at the 1982 NAMM Show (North American Music Merchant Convention). The year of market introduction is 1983, because this is also where the first two MIDI-compatible synthesizers are; SC Prophet 600 and Roland Jupiter 6 went into series production. In the same year the IMA (International MIDI Association) was founded. 1983 also saw the appearance of the most successful synthesizer model of all times, the Yamaha DX7, which worked on the FM synthesis principle and for a decade (and longer) shaped the sound of all music tracks in which an electric piano was used.

The FM synthesis used in the DX7 was largely developed by John Chowning, composition teacher at Stanford University. Stanford University continues to benefit from the ten-year sale of FM's licenses to Yamaha. An example of how a university can deliver to industry and at the same time take care of its own well-being, admittedly only in a roundabout way, because Chowning's efforts were not initially rewarded by Stanford University. His research on frequency modulation goes back to 1967. However, the university was of the opinion that the teaching of composition was Chowning's task and not sound research. He was fired and only hurriedly brought back from Europe after Yamaha became interested in taking out a ten-year license.

MIDI boom
With the appearance of the DX7, there was a MIDI boom, combined with a software business without equal. The programming of sounds was, due to the abstract input method with the help of numbers on a tiny display with a confusing soft button system, so complicated and laborious that even people who had "understood" the instrument were definitely lucky when they did programmed a sound for the DX7 (and its slimmed-down versions) that immediately had the characteristics that were imagined. As a result, a sub-market developed in which sounds for DX7 and other synthesizers were offered.

This is due to the "programming laziness" of the users of MIDI technology. It was quite common not to make sound designs on the instrument yourself, but to use presets and sounds offered by the industry and have no idea about the expensive synthesizer. The purchase sounds were then used to create songs with the help of (software and hardware) sequencers, which were then dubbed onto compact cassettes and sent to music production companies in the hope of landing a hit.

In Germany, the best sound programmers were found in what was then the GDR, where a sound market either existed as a black market or where sounds could only be bought at dizzying prices. The devices were smuggled in from the West (eg via people whose diplomatic baggage was not checked) and then illegally acquired; At that time a DX7 cost between 40,000 and 50,000 East Marks. When the longed-for instrument was finally in possession, it was not only "tested" as in the West, but thoroughly tried out and its operation learned.

In the West, "trade journals" that published reports and tests on newly released MIDI instruments shot up like mushrooms, and only those that published friendly reviews over the long term held up. Such magazines as "Keyboard", "Keys" etc., which still exist today, also consist to a large extent of private and business advertisements. It is advisable for today's MIDI users, despite the not always serious behavior of such sheets, to read them from time to time, as they provide information about the market offer.

Developments in the MIDI protocol
As already mentioned, the development of the MIDI protocol is not the main vehicle for the increasing professionalism of MIDI technology, but the hardware and software instruments that have turned the original synthesizer control into a system for almost any audio application. Version 1.0 of the MIDI protocol from 1985 has meanwhile arrived at version 4.x.

In key words:
1985 MIDI 1.0, published by the IMA, next to the companies Sequential Circuits and Roland join: Big Briar, Bontempi, Kawai, Korg, Lexicon, Moog, Octave Plateau, Passport Design, Siel, Yamaha.
1986 MIDI Sample Dump Standard
1987 MIDI Time Code ( MTC ) Standard MIDI-File ( SMF ).
1988 Conversion to: MIDI Version 4.1
1991 General MIDI (GM), General Standard (GS)
1992 MIDI Machine Control

MIDI and developments in the computer market
The management of MIDI has always been dependent on the type of computer on which it was operated. When computers are used in music, the proverbial poverty of musicians plays an important role. The low price of a computer is an important argument for its suitability for use by musicians.

Initially, software sequencing was made possible for Macintosh computers, but it was not very reliable.

With the development of software for the Commodore C64 ("people's computer"), MIDI sequencing had become affordable for widespread use; popular program was the Pro-16 from today's Cubase provider Steinberg.

In 1987, the Atari ST appeared on the market as a popular home computer with a built-in MIDI interface. Compared to the Commodore computer, it offered a desktop surface. As a MIDI sequencer in Germany and Europe, the Steinberg company offered the Twenty-four, later Cubase and the company C-Lab Creator / Notator and C-Lab (later Emagic) Notator Logic (later Logic).
The Atari ST has become the most popular music application computer in Europe and is still used today as a pure MIDI station in many small studios. As a computer it was already considered obsolete three to four years after its publication and was ridiculed by computer fans as a toy, but has served music users (including professional) music users - especially MIDI users - well for ten years and longer.

It fulfills the price criterion mentioned above to a high degree and is an example of a productive refusal attitude that can to a certain extent be recommended to every computer user in the field of music. In many cases it is wiser to keep your old computer. If it is operated with virtuosity, it is always faster than a new model that poses operating problems.

The composer Trevor Wishart produces his extremely elaborate compositions on two old Amiga computers and welcomes the sometimes long computing times during which one can think well.

https://www2.ak.tu-berlin.de/~fhein/Alias/Studio/Referate/Gecshichte.html

 

Last edited: Jul 5, 2021

The answer given by fixion is absolutely correct. I'm just adding a bit.

1. In practice we don't need synth' patches that play in all the MIDI range from 0 to 127. A typical bass part in most cases requires one or, may be, two octaves. You can associate this range of frequencies with any range on your keyboard. The same is with other instruments. Even if we need to simulate the piano the MIDI range is absolutely sufficient.

2. I doubt that patches with the lowest (fundamental) harmonic above 10-12 kHz would sound very musical and could be useful in musical context. There are 1) the range of frequencies that human ear can hear (this range is less than 20 kHz in many cases) and 2) the range of freqencies that are musical and therefore could serve as fundamental ones in a synth patch. The latter is significally less than the former. I don't know the numbers but you can check it up by generating a patch with fundamental frequency (not the highest one) above 10-12 kHz. You'll get mosquito's buzzing. This means, it's the the fundamental frequency of a sound that is impotant while working with synths and MIDI. The higest frequencies is imortant in audio (not MIDI) as they are required to reproduce overtones.

 

Last edited: Jul 5, 2021

MIDI Note to Frequency Conversion Table
https://musicinformationretrieval.com/midi_conversion_table.html

MIDI Note to Audio Frequency Calculator / Tuning Fork
https://www.colincrawley.com/midi-note-to-audio-frequency-calculator/

 

Last edited: Jul 5, 2021

I got a recommendation to use OSC instead
I was checking and all DAW support OSC
Much easier to implement and less restricted, float point etc

 

notes beyond 12 kHz ?
Are you joking ?

Generate a 12 kHz sine before speaking about anything in this thread ... and just after generate a 15 kHz one.
You will ruin your ear before earing anything stable.
If it is not enough, generate a 20 kHz one.
May be your monitors/headphones will not even generate them ...

https://www.szynalski.com/tone-generator/

enough time lost here
Bye !

 

Thank You !

 

low quality answers :/
I guess people is just giving their opinions without knowing the subject at all

 

1. If you know better, why you pose questions? Your question demonstrate your ignorance.

2. How this video is related to sounds beyond 12 kHz?

 

Source: https://en.wikipedia.org/wiki/Scientific_pitch_notation

Keyboard and frequencies: www.sengpielaudio.com/calculator-notenames.htm

 

The formula to calculate equal temperament it's easy by coding.
freq = base_freq*(2**(interval/tet))
** means pow() or exponent power
tet can be any equal temperament, western music uses TET12, Arabic TET24, Hindu TET22 old Thai TET7, TET31 is usually using in microtonal compositions because approximations to perfect thirds, fifths etc, Dolores the woman on the video uses TET106, she also uses a hack to use Midi actually using EDO instead, 127 Midi values is just to little, for polychromatic and even worst for dynamic scales.

I wrot this little script to calculate the frequencies myself, I prefer to use the formula so no dither.
interval 0 is the tonic
in tet 12 interval 7 is the fifth, etc

with the current setting it calculates from 440 to 880 in TET 12 with the Midi default

Code:

#python
tet = 12
base_freq = 440
base_midi = 69

for x in range(tet + 1):
    midi = int(x + base_midi)
    pn = base_freq*(2**((midi-base_midi)/tet))
    print(midi,pn)

Code:

69 440.0
70 466.1637615180899
71 493.8833012561241
72 523.2511306011972
73 554.3652619537442
74 587.3295358348151
75 622.2539674441618
76 659.2551138257398
77 698.4564628660078
78 739.9888454232688
79 783.9908719634985
80 830.6093951598903
81 880.0

But changing TET 31 for example

Code:

#python
tet = 31
base_freq = 440
base_midi = 69

for x in range(tet + 1):
    midi = int(x + base_midi)
    pn = base_freq*(2**((midi-base_midi)/tet))
    print(midi,pn)

Spoiler: 32 frequencies of TET31

69 440.0
70 449.94903166455805
71 460.12302521789434
72 470.5270673712696
73 481.1663598537887
74 492.04622201311946
75 503.17209347501654
76 514.5495368629822
77 526.1842405794221
78 538.082021649686
79 550.2488286304182
80 562.6907445836683
81 575.4139901182517
82 588.4249264998795
83 601.7300588316126
84 615.3360393062311
85 629.2496705321435
86 643.4779089345004
87 658.0278682332117
88 672.906822999607
89 688.1222122935167
90 703.6816433825938
91 719.5928955457342
92 735.8639239624969
93 752.5028636904715
94 769.5180337325787
95 786.9179411963377
96 804.7112855471813
97 822.9069629579453
98 841.5140707567043
99 860.5419119751808
100 880.0

I Just double checked with Scale Workshop

 

New recording technologies as this one, (I know it's rather rare for musicians average knowledge) "The second hydrophone was a top-quality research instrument manufactured by Oceansonics Ltd., that recorded high-resolution audio data continuously to our hard drive storage array for later analysis. It recorded 512,000 sound samples per second, deciphering audio in the 10 to 256,000-Hz frequency bandwidth." https://moo-antarctica.net/audio/
This kind of device was specially developed for this research, but there are many sources of ultrasonic sound, and many things can be done with sound besides damaging human ears lol

I am also exploring ambisonic and other stuff, I am more into engineering and science than music, music as hobbie

BTW maybe you could get inspired with these recordings, are amazing.

 

Hello GabsIT, hard-working - congratulations.

There is a topic, also here in the forum somewhere, it's about tinnitus and hearing aids.
For example, if you can no longer hear the high frequencies, there is a hearing aid from the hearing aid acoustician, which is set in such a way that you can hear the sound of leaves again and hear the sweet, sibilant sounds again.

You just can't make music with a hearing aid. Sometimes you turn the high frequencies
down a bit with the EQ. when you find it unbearable. Maybe this is a topic for you?

Maybe something where you can transfer the values from the audiogram in an EQ. EQ for hearing aid users?
https://en.wikipedia.org/wiki/Audiogram

 

Last edited: Jul 6, 2021

Looks like "kHz" is getting thrown around to refer to everything. Or is it just ironic that MIDI and 127 should cause one to think about "velocity"? And 16, 32, 64, 128, 256, 512 should cause one to think about "buffers sizes"?

And this 12kHz? I don't know what kHz might be used as a substitute for there. Semitones?

Even if referring to keys and pitch/frequency.. I haven't seen any MIDI boards with 127 keys. For what? MIDI doesn't have strict control over the pitch triggered by the keys. You can already shift the octave down so C0 triggers a fundamental pitch that is well over 12kHz if you want to. And assignment wise.. you can break every pitch/frequency standard you want to.

 

Last edited: Jul 6, 2021

Why can't a piano have more than 88 keys?

Although the human ear can discern sounds from 20 to 20,000 hz, it can only determine pitch up to a maximum level of around 4,000 hz. The piano that Bartolomeo Cristofori originally invented in Italy only had 54 keys. As piano music evolved over time, the keyboard was gradually expanded to meet the needs of composers who wanted greater potential for musical expression. Around 1890 the modern keyboard with 88 keys for 7¼ octaves prevailed (from 2C to C5; 27.5 Hz to 4,186 Hz *).

The human ear can perceive frequencies in the range of around 20 Hz to 20,000 Hz, but the upper limit that the brain can consciously differentiate is around 4,000 Hz. So even if the keyboard were expanded by adding more keys to the piano, there would be the additional tones in the bass range just a dull rumble. The additional tones in the treble range, on the other hand, would be perceived as unpleasant, jarring sounds without tonal harmony, so they would be practically meaningless from a musical point of view. Bösendorfer manufactures pianos with 97 keys, the additional nine keys are located in the bass range (2C to C5).

However, the strings of these keys in the deep bass range are actually only there to produce a fuller sound when other keys are struck by resonating. In practice, these additional keys are almost never played. * When A is tuned to 440 Hz. The pitch of each key depends on the mood.

https://de.yamaha.com/de/products/contents/musical_instrument_guide/piano/trivia/trivia007.html