paulnasca

tuningmeister · Post by **tuningmeister** » Wed Nov 16, 2011 6:18 pm

Hi, I've just installed ZynAddSubFX and look forward to using it to playing MIDI keyboard in historical temperaments. I happened to notice an error that is made when loading values from .scl files - I have no idea if this is something that has been already reported.
The error occurs when a zero follows the decimal point. For instance, C# in quarter-comma meantone file is 76.049 but it appears on the screen as 76.49. I tried several other files where I created the same situation, and the same occurs. Furthermore, the frequency rendered is altered as well. Here's a copy of the .scl file to save you looking for a test instance.

1/4-comma meantone scale. Pietro Aaron's temp. (1523). 6/5 beats twice 3/2
|
0: 1/1 0.000 unison, perfect prime
1: 76.049 cents 76.049
2: 193.157 cents 193.157
3: 310.265 cents 310.265
4: 5/4 386.314 major third
5: 503.422 cents 503.422
6: 579.471 cents 579.471
7: 696.578 cents 696.578
8: 25/16 772.627 classic augmented fifth
9: 889.735 cents 889.735
10: 1006.843 cents 1006.843
11: 1082.892 cents 1082.892
12: 2/1 1200.000 octave

Just a chance discovery; I had the above printed out in another window and I just happened to compare them.

Looks like a wonderful program, I was able to get started with my MIDI keyboard with not too much struggle. I suggest a retuneable keyboard for musicians learning to play or sing music from before 1800, since they don't easily get to hear an accurate rendition of the way the tuning systems sounded then.

Thanks very much,

deleted · Post by **deleted** » Thu Nov 17, 2011 11:50 pm

tuningmeister · Post by **tuningmeister** » Sat Nov 19, 2011 1:16 am

Thanks, Sid, for drawing attention to your posts.
I wrote some long reply but got a spam filter, have no idea why, so just a few thoughts now.
I don't know about the KBM issue; do the tunings come out wrong? [I can't think that I would have a need for keyboard mapping].
It seems that a simple kludge around this issue is to rewrite any entry of xxx.0xx as xxx.0 or xxx.1, which leaves an error at most of .05 cents, an interval barely perceptible from a fixed keyboard scale, and far far beyond the skill of any singer or instrumentalist to reproduce. So this will work great for me; and I just have to look over the few temperaments to be used to fix some entries.

Michael

fundamental · Post by **fundamental** » Sat Nov 19, 2011 3:20 am

After looking through the scl file loading, it appears that there is no issue in loading the file, but there is an issue in showing the results in the GUI.
I cannot attest to the inner workings of the Microtonal system much as I work on normal musical scales, but the file does load with correct values.

The current git head is patched to resolve this issue.

deleted · Post by **deleted** » Sat Nov 19, 2011 1:03 pm

tuningmeister · Post by **tuningmeister** » Sat Nov 19, 2011 2:05 pm

Thanks for looking at this.

I looked it over again, and must apologize, I do now find that the frequencies generated are correct to the values in the file. They are not changed by the mistake in showing the number in the GUI. It was a subtle effect, in any event.

Michael

tuningmeister · Post by **tuningmeister** » Fri Nov 25, 2011 2:32 pm

[quote="Sidulacra"]tuningmeister,

The Scala microtuning format uses two files:

1. SCL - As we know, the actual microtuning file.
2. KBM - This is used to specify the MIDI Note on which the 1/1 starting note of the loaded SCL file will be placed - and - the Reference Frequency, which most musicians place on either 69.A 440 Hz or 60.C 262 Hz, although the reference frequency can be any pitch one cares to specify.

Importantly, the SCL portion of the Scala microtuning format alone does not specify where the 1/1 will be placed or the reference frequency. This is why any implementation of the Scala standard *must* include the KBM portion, otherwise, musicians have no control over how the microtuning is being mapped to their MIDI controllers.

What usually happens, when the KBM part of the standard is not used, is that any loaded microtuning is mapped with the 1/1 on 60.C @ 262 Hz, and the reference frequency also is placed on 60.C @ 262 Hz. So in other words, the instrument will force any loaded intonation to be mapped on middle C...

But what if you want your reference frequency to be on 69.A 440 Hz - or any other arbitrary pitch required by the music - or if you wanted to bring your VSTi into tune with others that are using something other than middle C - or you want to change where the 1/1 is placed - by what method would one get the results they require for a particular composition?

The only way is by figuring out how to offset the pitches of the oscillators of a synth to bring it into tune with the target pitches; and good luck figuring that out easily. All said to explain why one must have the KBM part of the Scala format to do any serious work with alternative intonations.

fundamental,

Is this update available as a VSTi?

S[/quote]

Actually, Zyn seems to map all scala scales as if they begin on A.
The selection "Shift" on the Scales dialog page changes this mapping. So, if I want the scales to begin on C, as they seem to be designed, I set this "Shift" to 9, and it comes out correctly.
Furthermore, I can use this 'Shift' to make other possibilities available within 'regular' temperaments such as 1/4 or 1/6 comma meantones, which is what I am using. Thus, if I desire to have D# instead of Eb (that is, 4 sharps and one flat instead of 3 sharps and 2 flats), I set the 'shift' to 2. Other possibilities include 5 sharps (shift = 7), 2 sharps (& 3 flats) from shift = 4), one sharp from shift = 11, and all flats from shift = 6.
So I think this 'shift' parameter works the same way your KBM entry would.

Michael

paulnasca

KVR Audio

Error in rendering scala files