I recently discovered @greghmerril's program for detecting common errors in traditional four-part harmony.

The following rules of voice leading are checked:

  1. Vocal Range;
  2. Spacing Between Voices;
  3. Parallel Fifths and Octaves;
  4. Voice Crossing;
  5. Consecutive Fifths and Octaves by Contrary Motion; and
  6. Hidden(Direct) Fifths and Octaves.

I am (first of all) very impressed with the program. I am curious, however, what I should be aware of before letting a program influence my progressions. The goal is smooth voice leading, however that may be achieved. In other words, I am not concerned with the principle of it.

I am however, wondering what the human in me should be aware of or notice when I swiftly edit progressions according to the program's suggestions. Might something (don't ask me what) get lost in the process of correcting the progressions? Or, will the progression that has been "program-approved" always sound better than the original, flawed one?

If you wouldn't mind reviewing--the following are two of my progressions after making the program's suggestions. What do you see? Are there any red flags or things that jump out to you as "correct but not right"? I would just like to know what I am getting myself into (this is for my own personal composing as a hobby--not just to complete music theory homeworks).

example "program" progression 1

example "program" progression 2

Thank you!


Response to current answer:

Thank you very much for your response.

I just wanted to mention (if it wasn’t clear) that the two progressions I posted are separate ideas—they don’t go together.

Additionally, I am not specifically trying to emulate the style of Bach, with his fast harmonic rhythm and highly independent voices—I just want to focus on writing progressions that are interesting yet air-tight (so yes—focusing on chords here).

Can you please elaborate on point #1? I am not familiar with the term “contralto”—I’m thinking in terms of SATB.

As for #2, I was under the impression that “accented passing tones” were an acceptable melodic device. That being said, are there certain treatments (be it voicing, surrounding melodic motion etc.) I should be aware of?

For #3, is there something incorrect about the I69? I do not intend for it to be a iii chord. You said “it comes to another problem”?

For #5, what do you mean I shouldn’t have “that treble clef one octave down”?

I refer to the Aldwell & Schachter text. Do you know where I can find a copy of Diether de la Motte’s? I am having trouble finding an English version online.

Thank you again for the response.

  • You are going to have to program a computer to know the difference between the three types of fifths, only perfect 5ths are verboten. diminished and augmented 5ths are fine.
    – Neil Meyer
    Commented May 6, 2020 at 15:06

2 Answers 2


The goal is smooth voice leading... I am not concerned with the principal of it... I am however, wondering what the human in me should be aware of...

First, you need to get straight what you are trying to do.

If you want smooth voice leading, learn voice leading.

To say your aren't concerned with the principles of the goal, then try to do it with software, and ask how a human should react to the software's corrections without understanding the principles to make a judgement, it just going in circles.

The last four of the six rules you list are the typical prohibitions. Personally, I think a big stumbling block for understanding voice leading it to conceive it in terms of prohibitions about what to not do instead of positive statements about what you should do.

Some positive principles...

  • Voice all chord tones
  • Hold common tones and move voices the smallest distance
  • If there are no common tones, move the upper voices in the opposite direction of the bass

Walter Piston's Harmony provides some elaboration of those principles using only two pages (including notation examples) and then states that's all you really need to do for proper voice leading.

There is more nuance regarding doubling tones and dissonance, but those principles can also be started positively...

  • double the root of root position chord
  • resolve the seventh of seventh chords
  • prepare dissonances, etc. etc.

The one prohibition that seems to make things easier is: don't double tendency tones.

(You can get around the negative/prohibition wording usually with some word smithing and clever application of theory: voice tendency tones with a single voice, or move to perfect consonances only by dissimilar motion which restates the dreaded parallel/direct octave/fifth prohibition.)

In the first bar of your example you have F5 the seventh of a dominant seventh that doesn't resolve. In bar three you leap to the dissonant F3 in the bass.

You can try a revision to apply the principles at those points. Resolve the seventh. Handle the dissonance with either a preparation or as a non-chord tone. Keep in mind these are voice leading changes, so the voices will move differently by necessity. The music will change. That's part of the revision process. Don't look for voice leading revisions and then balk at changes to your original parts.

enter image description here

Resolve the seventh down, treat the F3 in the bass as a passing tone...

enter image description here

...I think that brings it closer to typical voice leading.

If I understood your post correctly, my suggested revision is on the computer program's output. It probably would be more valuable to you to post your original, and then examine the voice leading you used.

  • crystal clear--thanks @michael curtis. Just so you know, by "principle of it", I was not referring to the principles of voice leading but, rather, the principle (or lack thereof) of being assisted by a machine in writing harmonies. Additionally, my original attempt (by hand) did not look much like the above at all--that is why I was curious about the programs corrections. It appears to be great at pointing out the issues mentioned above, but it does not seem to handle the resolving of tendency tones--would that be your take-away as well?
    – 286642
    Commented May 6, 2020 at 19:25
  • Ah! the principle of the software. Right, I wouldn't spend much time think of that. Not unless I was really convinced that the programmer understood music very well and the software was a demonstrated success. Commented May 6, 2020 at 20:35
  • I'm not sure about this particular program, in part because it apparently depends on the input. So, my suggestion for a test: put in some of the simpler Bach chorales or simply textured Baroque minuets - Handel, Telemann, those kinds of unassailable composers - and see what it gives back. On the one hand it shouldn't suggest changes, on the other hand if it doesn't suggest changes, it should be clear both why the composer didn't do it and why the software suggests it. Commented May 6, 2020 at 20:46

I suppose by the rules you've selected that the desired exercise is on baroque choral in the style of Bach. If so, you should be reminded that in counterpoint you want all voices to be independent, and this is not happening here, as all voices are tied to the idea of chords. To have independence, voices should have different rhythms, and the "chords" are usually conceived in the down beats. It might help to write every voice in a different staff.

The most important step of writing music that fits into a style, for me, is to do analysis in that style. If it is Bach Chorales you want to recreate, try to analyze as much as you need to feel comfortable with the style before trying to write something. Just by taking a look at some of them you might notice that what you've written is too far away from the style.

Some errors you've made:

  1. Going for unison in the highest note of bar 2 in parallel movement of the soprano and contralto.
  2. Seconds as harmonic intervals. They would be considered really harsh, specially if located on down beats and held for a long time as you've done.
  3. Speaking of seconds, you got some modern sound in bar four. It's a I69 with that D on top. You might want that to be an E, but it leads to another problem:
  4. Not everything that is a "rule" can be applied in practice. One rule that would make my statement incorrect for the last error is that you can't double the third in a four-part chorale. But if you get Bach's oeuvre you might notice that he had more doubled thirds than he had doubled fifths, which is not considered an error by most music teachers.
  5. It's important to mention that you shouldn't have that treble clef one octave down.
  6. Finally, why did you end on the dominant (V)? It's really common to finish the first phrase in the dominant, but you always want to finish the piece in the tonic (and maybe, just in some cases, in the relative minor - iv).

Have you read any book on it? Diether de la Motte has a nice one, focused on historical perspective, and Edward Aldwell has a harmony and voice leading book more hands down on practice.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.