Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The 30th Anniversary Of MIDI: A Protocol Three Decades On

thequietus.com

Post reply on HN
Print view View original

21–30 of 54 posts

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#21

Post by ominous_prime »

npsimons wrote: post #20
ominous_prime wrote: post #15

Earlier quoted context omitted.

The opening of "Rhapsody in Blue" contains a glissando, which is comprised if a fast run of the individual notes, not a continuous sweep. Since midi messages are only 8bits, you have only 127 steps of resolution. Augmenting a note beyond a certain point leads to noticeable quantization of pitch.

Many performances I've heard of "Rhapsody" were essentially a sweep (portamento? the terminology is a bit muddled). But even if it's not possible on the real instrument, or the real instrument doesn't have the range, doesn't mean that limit should be there; I agree with you that limits suck, and it would be nice if someone invented something better than MIDI (and made it popular/standardized).

True, I was thinking of the piano arrangement, but orchestral version is opened by a clarinet, and the player usually combines a glissando and a sizable portamento (and not having much woodwind experience, I have no idea how one would create that slide, like in opening of this classic recording http://www.youtube.com/watch?v=1U40xBSz6Dc)

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#22

Post by bhauer »

Funny, I remember writing custom communication protocols over the MIDI ports of my Atari ST because, at least on that hardware, MIDI ran at a higher bit-rate than the serial port.

Oh, and playing MIDI-Maze [1].

Never actually used it to drive an instrument.

[1] http://en.wikipedia.org/wiki/MIDI_Maze

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#23

Post by ChuckMcM »

I think MIDI represents a good example of what you can do if you compromise sensibly. I argued with folks over the 31.25K baud vs 38.4K (9600 x 4).

I also remember the assertion that it was going to "ruin" music because any 10 yr old could program a computer to play tempo perfect renditions of complex compositions, and yet to this day the accuracy of MIDI was always off putting.

One of my personal milestones was having a stack of instruments that were capable of 200 simultaneous voices without tape layering. That allowed the reproduction in real time of a symphony, but the biggest challenge was getting 200 separate MIDI streams running in parallel was a huge task. At one point I had 20 386 class PC's each capable of producing 10 streams. It was a silly goal since everyone else layered tracks but for me it was a weird quest to be a conductor of my own personal symphony orchestra :-)

I learned a lot about how time synchronization protocols can (and do) get screwed up in real time situations. That was fun.

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#24

Post by chops »

ominous_prime wrote: post #21
npsimons wrote: post #20

Earlier quoted context omitted.

Many performances I've heard of "Rhapsody" were essentially a sweep (portamento? the terminology is a bit muddled). But even if it's not possible on the real instrument, or the real instrument doesn't have the range, doesn't mean that limit should be there; I agree with you that limits suck, and it would be nice if someone invented something better than MIDI (and made it popular/standardized).

True, I was thinking of the piano arrangement, but orchestral version is opened by a clarinet, and the player usually combines a glissando and a sizable portamento (and not having much woodwind experience, I have no idea how one would create that slide, like in opening of this classic recording http://www.youtube.com/watch?v=1U40xBSz6Dc )

That kind of slide involves subtly easing off the holes of the clarinet, leaving partial openings. It's a skill I practiced, but never really mastered. I could slide a few notes, but it was never make it that smooth, nor could I slide that many without easing off too quickly and making the slide "jump".

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#25

Post by afandian »

ominous_prime wrote: post #15
npsimons wrote: post #12

Earlier quoted context omitted.

Additionally how could you express a solid sweep from C1 to C8? Not that you could or want to The opening in "Rhapsody in Blue" is two and a half octaves, and that's jut the first thing that comes to mind. Also, I'm sure there have been some extensions that enable this, but I don't know for sure; the Yamaha DB50XG I had came with sitar, and while I don't recall exactly the demo song included, it seemed like a reasona…

The opening of "Rhapsody in Blue" contains a glissando, which is comprised if a fast run of the individual notes, not a continuous sweep. Since midi messages are only 8bits, you have only 127 steps of resolution. Augmenting a note beyond a certain point leads to noticeable quantization of pitch.

I can't quite remember and can't be bothered to look in the spec, but MIDI has multi-byte integer values for use in some cases, and control messages such as pitch bend use this. Could be as little as 2 bytes, but more than 1.

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#26

Post by dividebyzero »

An interesting use of midi data can be seen with the Fretlight Guitar: http://www.fretlight.com (disclaimer, I work for them). Midi data is tagged so that it can light up the guitar fretboard to teach you how to play. Any Guitar Pro 6 tab can light up the Fretlight.

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#27

Post by gtani »

ominous_prime wrote: post #21
npsimons wrote: post #20

Earlier quoted context omitted.

Many performances I've heard of "Rhapsody" were essentially a sweep (portamento? the terminology is a bit muddled). But even if it's not possible on the real instrument, or the real instrument doesn't have the range, doesn't mean that limit should be there; I agree with you that limits suck, and it would be nice if someone invented something better than MIDI (and made it popular/standardized).

True, I was thinking of the piano arrangement, but orchestral version is opened by a clarinet, and the player usually combines a glissando and a sizable portamento (and not having much woodwind experience, I have no idea how one would create that slide, like in opening of this classic recording http://www.youtube.com/watch?v=1U40xBSz6Dc )

http://www.acoustics.org/press/155th/chen.htm

If you go to a 6th grade band class, to where the clarinet nd sax players are not adept at embouchure, choosing reeds and assembling the instruments, you hear all knds of interesting inadvertent things. Single reeds are fantastically flexible instruments. A few months ago i noticed for the first time how similar the tones of Dolphy's bass clarinet and Coltrane's soprano sax are on the 1961 recordings.

If anybody's interested, how to play quarter tone music on sax (pg 25 in the pdf:

http://davidvaldez.blogspot.com/2012/09/quarter-tones-by-don...

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#28

Post by TylerE »

ChuckMcM wrote: post #23

I think MIDI represents a good example of what you can do if you compromise sensibly. I argued with folks over the 31.25K baud vs 38.4K (9600 x 4). I also remember the assertion that it was going to "ruin" music because any 10 yr old could program a computer to play tempo perfect renditions of complex compositions, and yet to this day the accuracy of MIDI was always off putting. One of my personal milestones was havi…

It's funny how things have progressed, isn't it?

For relatively little $$$, one can have hundreds of gigabytes of professional samples in VSTi plugins, and be able to render playback so good it would fool almost all listeners into thinking it was an actual, real, room miced recording.

All in the end still essentially driven by Midi.

Re: The 30th Anniversary Of MIDI: A Protocol Three Decades On

#29

Post by jerf »

I got a Korg X5 around '95 and clocked a lot of time using it for composing. (Poorly, but nonetheless.)

But a fun bonus was hooking it up to video games that supported putting out music in General MIDI.

In an era where everybody was listening to Adlib FM on their Tie Fighter: https://www.youtube.com/watch?v=puvD_FDS_jE

I was listening to something more like this: https://www.youtube.com/watch?v=Ajz-m0qIDCk

Although I think my X5 actually sounded even better. (I always thought the violin and viola were a bit tinny, but the cello was fantastic, and the bass not too bad. And the choir sounded better in the lower register than that recording, though you'd still never mistake it for a real choir.)

I also recall the Day of the Tentacle theme desperately needed a good clarinet sound to sound really good. I often though the theme must have been composed on the 05/W that the X5 had the same sound set as, because it sounded that good. Alas, no Youtube love here. (Closest is the MT-32 recording, http://www.youtube.com/watch?v=YQeQzeVmN5w , but that is way worse than the X5 rendering.)

Post reply on HN