Are there "safe" inversions for a simple melody?

Question

I've noodled around on a keyboard, and created simple melodies with chords. I don't have the skill to play them well. So, I eventually port them to LilyPond and have it generate a score and a MIDI file.

Now, I'm thinking of taking some of those and "flipping them upside down and backwards".

I just tried a retrograde (a term I did not know applied to music) of an entire wee melody and it turned out just luv'ly! (I had to do it manually, as the \retrograde command made a mess of things.)

Now I'm thinking "inversion..." but I'm wondering if there are common points of a key or scale around which inversions typically work, even if they're less than brilliant.

Answer 1

Now I'm thinking "inversion..." but I'm wondering if there are common points of a key or scale around which inversions typically work, even if they're less than brilliant.

There aren't really any "unsafe" inversions. But there may be some that are unplayable, or which don't sound good. It doesn't take long to try different inversions, to see which you like best.

Sometimes an inversion is a handy way to keep the melody and accompaniment a suitable distance apart so that it sounds right. A chord played too high can crash into the melody (e.g. on a keyboard, if the left and right hands collide). Transposing the chord down an octave might sound too muddy. But an inversion allows you to create a chord between these two extremes.

Answer 2

There are quite a few melodic transformations of this type. One is just playing a melody backward (called "retrograde motion"), the length of notes need not be the same.

Another method is to invert the melody about an axis. The most common is to invert about the first note. For example, a rising melody C-D-F could be replaced by C-B-G. This method usually uses diatonic steps. It's possible to invert using the same number of half-steps: C-D-F would become C-Bb-G.

In either case, the rhythm and harmony may need adjustment.

Just for fun: the name for inverting and reversing is called cancrizans. (It seems common enough to get a name.)

Answer 3

Safe/unsafe is not the way to think about it. When you apply devices like inversion or retrograde you needed to assess the outcome, and whether it's musically useful, or how to make the outcome useful.

Depending on the details, when something is transformed by inversion or retrograde you get a kind of musical "opposite" harmonically, rhythmically, or melodically.

If you chromatically invert a plain major scale you get a phrygian scale. Similarly, if you chromatically invert a C major chord, you get an F minor chord. The inversions transform the mode, and in the later case change the chord root. None of those transformations is bad, or unsafe, per se, but you need to decide how they can be worked into the overall piece.

In terms of functional chord progression such transformations can result in retrogressions, where the normal function progression of IV V is reversed to V IV, which is normally a "problem" in functional harmony. That doesn't necessarily mean the transformation is completely bad, but you may need to reharmonize it, or reposition it within the key to make it work as good functional harmony.

Similar effects will come to play with melodic contour. If a line that follows the common "arch" contour is inverted, it of course becomes a drop down to a low point and then rises back up, an inverted arch. Not necessarily a problem, but you may need to figure out how to fit it into the whole piece. Also, if the top notes of a melodic line are more important to its good design, then inversion will take lower, subordinate pitches and put them into the high position, and the result may not be a good melodic contour.

Once I tried to invert the first 8 measures of Over The Rainbow and I heard the problem with melodic contour which can result from inversion. The lower pitches of the original line, those on the words "some" and "way", which are subordinate in the original, became too prominent in the inversion.

The top pitches of the original gently trace out a descending major scale, but when inverted, instead of getting a gradually ascending scale, the two Cs I circled in red become too prominent and ruin the melodic contour. The repeated pitches create a static line rather than a ascending one. That doesn't mean the inversion is totally unusable. I decided to move those pitches, and the E5 to chord tones low enough to not interfere with the contour of an ascending scale.

So, if you're hoping that some function in LilyPond will invert or reverse lines and automatically create great musical output, that is not likely to happen. You should expect to work further the output of those functions to figure out how to make good music from it. Also, be aware of the specific kinds of transformations LilyPond has built in and it may not automate all possible options. You may need to do some transformation manually.