Posted
I am suddenly seeing that chords with an open voicing and in the lower playing range can include non-chord tones. I assume this is a bug.
See example attached (screenshot and project).
Note that there is no harmony defined here, but I would assume the chords then should simply all fall into the default C major triad.
Thanks!
Tue, 2023-02-07 - 16:15 Permalink
When this workflow is possible in Synfire , then more orchestral music is possible
(25) How to write with sample libraries ensemble patches - YouTube
Mon, 2023-03-06 - 17:05 Permalink
Apologiest that I keep on harping about this, but I continue to struggle with these bugs. Attached is another example demonstrating a similar problem.
Here, chord symbols result in non-chord tones, if a certain playing range is selected (bars 1-4 in the example). With a different playing range of the otherwise same figures and chords (bars 5-8), the problem does not occur.
Remember that in the previous case, chord symbols could result in non-chord tones for chords in open voicing.
I was thinking that turning figures & chords into fitting pitches is THE core feature of Synfire. I therefore hope that such issues can be fixed relatively soon.
In case this is a feature, and it is intended that chord symbols can result in non-chord tones, then I would be grateful to learn of a way to enforce chord tones in one way or another.
Thanks!
Tue, 2023-03-07 - 10:30 Permalink
Thanks for sharing that!
So, do you think that the behaviour in this case might be a feature, and non-chord tones are intentionally added to figures/phrases consisting solely of chord symbols? But then, why would this only occur with certain playing ranges and not with others? Of course, we can only guess here without input from Cognitone themselves.
I ran into this issue when using custom scales with more than 7 tones, and there the inserted non-chord tones just did not work musically for me. Anyway, at least bypassing the interpretation can serve as a workaround here.
I am still looking foreward to a workaround for the open voicing case, though.
Tue, 2023-03-07 - 12:14 Permalink
Yes, that is a feature. With active Interpretation, Synfire does things like that all the time, and the result is also context sensitive. That is, the same figures and chords can produce completely different results at different parts of the song. Simply because the context is different in each case, i.e. different figures and chords are in the neighborhood. Also what the other instruments are playing matters. If you don't like the result you can try to tweak the settings of the Interpretation parameter. Check out also the settings in the Global Preferences >> Scale Selection. If you have set it to "Traditional", your chords and scales might just be not traditional enough :)
If tweaking the Interpretation settings does not lead to the desired result, bypassing it is the only option. But the thing with the open chords is something else. There, Interpretation bypass did nothing either. That actually appears to be a bug.
Sat, 2023-03-11 - 16:14 Permalink
This is (was) indeed a bug. Thanks for pointing out. The next update will have it fixed.
There was a fundamental mishandling of the pitch mapping that makes me wonder why this didn't come up earlier. So, expect some of your music to sound slightly different with 2.1 ;-)
Sat, 2023-03-11 - 16:25 Permalink
FYI:
Green chord symbols are supposed to be always in the chord (incl. bass, if that's different). They should not respond to voice leading rules at all.
It's different with the purple Relative symbols. There only the anchor is in the chord and additional symbols are subject to voice leading relative to that.
Mon, 2023-03-13 - 11:53 Permalink
makes me wonder why this didn't come up earlier
Because this behavior has been in the software for so long that you could think it was a feature. :) But right, the application of Voice Leading features to chord symbols (if it was about that at all) didn't make much sense.
By the way, didn't the functionality of chord symbols also change with Synfire 2? If I remember correctly, in early versions of Synfire, chord symbols worked just like vertical symbols, differing only in color.
Mon, 2023-03-13 - 17:42 Permalink
With 2.0, chord symbols are mapped to chord inversions and voicings. These depend on the instrument's middle range and the progression's settings per harmonic context, which can be overridden per individual segment settings.
And the Interpretation also has a veto whether all segment's individual settings should be ignored.
Mon, 2023-03-13 - 17:47 Permalink
With the bug now fixed, I notice how some arrangements sound more clear than before.
Of course, if you want more chord extensions, you now need to select them for the chord. Previously Synfire just sprayed a few extensions in here and there, which was fine, but not intended. The vertical scale should instead be used to "spray" chord extensions around.