Iranian Classical Music, MIDI files ect cont'd
Mohammad R. Gharib
Here I have attached our discussions on the midi project.
The idea is to create a lot of midi files from the music
sheets available in the market. The files would be very small
, editable, and ready for play back using affordable music softwares.
The original message is at the bottom, and the last one is at the top.
We still need more volunteers with new ideas.
----------------------------------------------------------------------
Mohammad Reza Gharib telephone: (818) 395 - 4759
Graduate Aeronautical Laboratories m...@galcit.caltech.edu
California Institute of Technology
http://www.galcit.caltech.edu/~moh
----------------------------------------------------------------------
Kees van den Doel
In article <3251C7...@galcit.caltech.edu>,
Mohammad R. Gharib <m...@galcit.caltech.edu> wrote:
>Here I have attached our discussions on the midi project.
>The idea is to create a lot of midi files from the music
>sheets available in the market. The files would be very small
>, editable, and ready for play back using affordable music softwares.
There's tons of FREE midi players on the net. Quite affordable.
I have some doubts about how useful this would be. Especially in Persian
music dynamics, rubato etc. are very important and not written down. A
scanned midi file would not have any of this information.
If you're interested check out http://www.flexfx.com which has a library
of thousands of midi files, classical, popular, ethnic, including a
section with Persian music. Listening to this stuff will give an idea
of what you'll get from midi files.
Generally I find the classical music files amusing, but awful sounding,
like a computer. Dead and robotic. The best ones are presumably played
in on a midi grand piano (a normal piano which detects every keystroke
and saves the entire performance including all rubato and all dynamics)
rather than punched in from notation.
This is the way to go I think.
Anyways, I'll be happy to share my printed music with anyone who wants
to try scanning the stuff in. I have the following radif books, each
with a complete radif:
Violin:
1) Saba (2 volumes, each with a compete radif, rast and chap kuk). Have
audio recording.
2) Lashkari. No audio recording.
Voice:
3) Karimi. The notation is superb, even the most subte rubato effects
are notated. Small print notes indicate soft notes. This is probably
the best candidate for midification but will need to be typed in by hand
because of the special notation for rubato, dynamics, vibrato,
glissandi. Have audio recording.
Tar and setar:
4) Tahmasbi. Just a copy of Karimi, but he simplified the
notation. Waste of time. Have audio recording.
5) During. Borumand radif, very complete. Includes an extensive intro on
theory including lots of measurements of tunings used by various
masters. They differ quite a bit! Have 3 audio recordings, Alizadeh on
tar, on setar, and a recording on kamancheh.
6) Maruffi. Monumental edition from before the revolution. Has the most
gushe's of any radif. No recording.
7) Santur. I forget by who, can't find it immediately... Is basically a
copy of Karimi, with extra gushe's added, most notably bayat-e-kord,
which is missing altogether in Karimi. Have audio recording on CD. I
only have the first CD in the series, didn't find the rest anywhere.
8) Farhat. This book on persian music contains an appendix with a
"radif" selection. These pieces are just copied mainly from
Maruffi. Waste of time for midification.
Besides this I have a lot of rythmic pieces for all sorts of
instruments, too much to catalogue here right now...
Anyone who wants copies of the printed music or recording for this
project can send me email and I'll send you the stuff.
Anyone who has a radif version not in this list (except Talai which I
have ordered already) please contact me so I can get it from you
somehow.
Kees
PJ
Hi,
I think it is a very good Idea. I am ready to help (If i can)
some points :
if you want, you can use my FTP and Web-Server to present some Info and demos.
( http://www.pejman.com )
did you thought about Karaoke ?
how about the Text (you know the music-notes are from left to right) ,
so we (5 engineers) developed 2 years age a phonetic language called PERPHON
in germany (which is still used by few people for email and so on)
If you want i can send you the spizification to use for the text SYNCHRON to the Music
thanks in advance
PJ
Mohammad R. Gharib
Check out the tuning wrench by Artic Software at
http://earth.execpc.com/~artic/mmu5.html
download the demo and play around with it.
The tuning possibilities are great but you can
not hear them on the demo. The software is listed for
70 dollars.
also down load the midiscan or the pianoscan at:
http://www.musitec.com/musitec
Kees van den Doel
I punched in the first piece in the vocal radif by Karimi, daramad shur,
as a midi file. This may serve as an example of what you can expect
from midi files. You can get it from
http://www.cs.ubc.ca/kvdoel/radif.html.
Let me know what you think.
Kees
Kees van den Doel
I punched in the first piece in the vocal radif by Karimi, daramad shur,
as a midi file. This may serve as an example of what you can expect
from midi files. You can get it from
http://www.cs.ubc.ca/spider/kvdoel/radif.html.
Mohammad R. Gharib
Kees,
>
> There's tons of FREE midi players on the net. Quite affordable.
We all know this. We also know about Ali Asady who made the midi files
in that gpg.com page. The idea is to create easily transferable Iranian
music scores.
Midi files and .NIFF files are the smallest format for music scores.
Editting and republishing scores with a better quality at home is
another
advantage of this. Iranian music has little to offer in terms of
written
music and that can be easily be put on the web.
Mohammad Reza Gharib
Kees van den Doel
I hacked the MIDI file format and made a C program (extremely ugly) to
convert a MIDI file to any tuning you specify. Feel free to do with it
what you want, except make money. :-) The source code and some examples
can be found on
http://www.cs.ubc.ca/spider/kvdoel/radif.html.
Besides daramad shur I also put MIDI files of daramad esfahan (both from
the Karimi radif) there. There are 5 different tunings of esfahan: equal
temperament (mainly to show the difference a wrong tuning makes),
"modern", "authentic" according to Ella Zonis, "authentic" according to
Farhat, and something I cooked up myself, which I think sounds most like
performances of esfahan that I've heard by reputable Persian musicians.
I'd like to hear your comments!
Kees (kvd...@cs.ubc.ca)
Mohammad R. Gharib
I found your C program very helpful.
I used it to tune some of my work, you can find them at:
http://www.cco.caltech.edu/~moh/midifiles
Some of them sound not so good because I had written them
originally for a Tar player to read and interpret not to
sound good as a midi. The other problem is that I had more than one
melodies or multiple notes which the program does not resolve.
( I thought you said multiple tracks would not be a problem but
I have problems.)
One can download them, listen to them, import them into one's favorite
music software and transpose them for any instrument. The files are
less than 10k. The untuned files plus an approximate tuning chart
is at http://www.cco.caltech.edu/~moh/untuned .
By the way when you click on midifiles, they get automatically loaded
into
your midiplayer. To save a file in a certain directory press shift +
click.
I also scanned a few black and white scores for the midiscanner.
They are at: http://www.cco.caltech.edu/~moh/imgfiles
The midiscan.exe is also there. The recognition was not %100 accurate
and I think its better to have the midiscanner on the same machine with
the scanner. That way one can experiment with different resolutions.
I think my 300 dpi scans were not the best and something between 300
and 400 is better.
Another great page on tuning information is
http://www.dnai.com/~jinetwk/other.html
It has leads to a lot of pages regarding alternate tunings software,
etc.
Kees van den Doel wrote:
>
> I punched in the first piece in the vocal radif by Karimi, daramad >shur,
> from midi files. You can get it from
>
> http://www.cs.ubc.ca/spider/kvdoel/radif.html.
>
> Let me know what you think.
>
> Kees
--
Kees van den Doel
In article <32627D58...@galcit.caltech.edu>,
Mohammad R. Gharib <m...@galcit.caltech.edu> wrote:
>I found your C program very helpful.
>I used it to tune some of my work, you can find them at:
>http://www.cco.caltech.edu/~moh/midifiles
>Some of them sound not so good because I had written them
>originally for a Tar player to read and interpret not to
>sound good as a midi. The other problem is that I had more than one
>melodies or multiple notes which the program does not resolve.
>( I thought you said multiple tracks would not be a problem but
>I have problems.)
Actually, your file negar.mid has chords on track 2, which is not
allowed by my program. As I wrote in the description of the program,
only monophonic tracks are supported. The reason is that a pitchbend is
inserted for every note on event. If there are two notes at the same
time you should have two different pitchbends at the same time which
can't be done in MIDI. There is a way around it, namely to scatter
chords over different MIDI channels, but that falls outside the scope of
a sunday afternoon hack.
Furthermore, if you have multiple tracks, they must be on different MIDI
channels for the microtuning to work. Your 2 tracks are both on channel
one. This is a limitation of MIDI, not of the tuning program, because
pitchbend is a channelwide effect. I for got to write that in the
documentation.
I hope that fixes the problems you have.
Kees
Mohammad R. Gharib
Kees,
The program is woderful and I'm sure a lot of people around the net
can use it, especially the people mentioned at that just/alternate
tuning page. I'd be shocked if no one has done it before.
As far as the tuning problems, I was aware of the channel issue and I
thought I had put the tracks on different channels. I was also aware
of the chord problem, but somehow things did not sound right event when
I had only two pure single note tracks on different channels in one
file. I'll try to give an example later.
The great thing about the code is that the pitch-bend is invarient
of transposition, i.e. you dont have to erase pitch bend, transpose and
pitch bend again. I think thats the beauty of midi.
Now, is it too hard to modify the code to detect erase all pitch bends
as well? i.e. 0 0 0 0 ... option would make things chromatic again etc.
Once again, congratulations!!
Mohammad
P.S. As far as my Tasneef: Negar, the real concert audio is at
http://www.cco.caltech.edu/~moh/sound_files/negar.aiff 284 kilobytes for
the first melody. compare with
http://www.cco.caltech.edu/~moh/midifiles/negar_m.mid
9 kilobytes for the score/midi. Not much "Hal" as you say.