.

.

Very nice!  This sounds like a real performance or at least like you played it on a midi keyboard and that way retained some very nice humanization to the rendition.  And the 2nd time you repeat the main theme it is a little bit quieter which is a very nice contrast!  This has very nice voice leading except for the F natural in measure 10.  I think since your destination note in bar 11 is B, an F# in bar 10 would really work much better giving a strong melodic motion of a 4th up instead of a tritone which is an ear-sore to me.  I think you were probably trying to borrow from the minor mode but in that case I would have the whole beginning of the B section in the minor mode for that to work but I don't think that that is a particularly good idea for this piece and the character you establish in the A section.  If you had a Bb in bars 9, and 11 (and an Eb also in bar 9 to avoid yet another tritone) then the F natural might fit in better, but still an F# might be preferable if you want a strong harmonic resolution from D7 to G minor using the melodic minor scale.  Now that I look at the score again it seems like you might have been avoiding the tritone between the C and F# in bar 10, but that is a tritone that happens naturally in the G major scale and is an expected feature of the D7 chord which you are outlining in that measure.  I think that tritone would be much less objectionable than the one you have now between the F natural and B.  Those are my thoughts.  Thanks for sharing!

It sounds very balanced and beautiful.
The truth is that in the second part there is a bit of ambiguity in the harmony.
The F natural of bar 10 to which @PeterthePapercomPoser refers does not sound bad, although it is true that it follows a line that ends in its tritone and that is an issue to avoid in the melodic design from the cantus firmus (of course all this would be relaxed with time and time). 
The passage from measure 8 to 9 is strange if it is understood as a modulation from D major to C major. If that modulation were as it is, the F natural would make more sense. But as far as I can see there is no modulation, we are still in G major measures 10 - 11 - 12: Am - D7 - G.

Therefore, logic and hearing demand that this F be F#.
 

.

Hey Aaron,

I really find you improve much in the voice leading. Even though this one is a short piece, the melody moves very smoothly without any large leaps, probably except the F natural which @PeterthePapercomPoser and @Luis Hernández mention which I agree with, both with the tritone reason Peter suggests or the harmonic reason Luis suggests.  Now I really think you can start expand your pieces and develop your ideas, when your voice leading and harmony are start staying firm. You can also add more details to further polish your piece, for example the dynamic markings and expressions. You may even try styles other than the short classical pieces!

Thx for sharing! It's nice to see constant progress from you. Cheers for that 🙌🙌🙌

Henry

@44W74l4

As I said this is one of the principles of the basic melody design (cantus firmus). If you draw a melodic line and form a tritone (augmented 4th), a split occurs. The upper note B wants to resolve to C (ascending), and the lower note F wants to resolve to E (descending), as happens with any tritone. And this is not what happens.
In other contexts with more complex textures, perhaps all this is much more relaxed, although it is still an essential principle. But in this kind of works where the voices are clear, the effect is more noticeable.