Dame Oh Musa - WIP

[mobius017]

I wonder if anyone might have some feedback on my current WIP?

My most specific question relates to the guitar. I'm using the "Nylon Guitar Reverb" that came with Studio One, at the moment. I actually like how it sounds in the song pretty well, but I had to turn up the sustain a lot to make it sound this way, and I'm not sure how realistic that is. Here's a short sample with just the guitar:

Dame Oh Musa - Guitar Sample

For full disclosure, I'll admit that my guitar credentials are basically, "I think I held one once." That, and a bunch of flamenco research I did for this song to try to imitate some flamenco guitar techniques with MIDI. (Though I didn't choose to incorporate flamenco's rhythmic cycles.) So I'm sure there are many here who know more about guitar than I :) .

I'll also include the full song as it currently is. There are lots of other things I'm going to do with the song as a whole--some tuning, mixing, mastering--but if anyone has thoughts on the full version or the short guitar bit above, I'd appreciate it!

Dame Oh Musa - Full WIP

For anyone wondering about the lyrics--the song is in Spanish!

Thank you for reading/listening!

 

[inactive]

I actually like how it sounds in the song

That's your answer right there.

I saw an interview with Chris Lord-Alge where he states that mixing is an opinion. And I'm now at a point where I can understand what he meant by that statement.

Not so long ago I used to worry about "what's the correct method/level/frequency/etc.", and that sort of information can be useful for beginners—at least to a point. But as you gain experience, you'll slowly learn what is meant by "use your ears."

In regard to your guitar sustain, you've used your ears and decided that you like the setting. Consider that a small step toward training your own ears.

Now the trick is to shape all the other elements to your liking.

And it will take time to learn what to listen for and how to identify/fix problems. But step-by-step you'll work your way out of the Dunning-Kruger valley of despair, and before you know it you will be ascending the slope of enlightment.

 

[mobius017]

Thanks very much @parallax_fifths! I really liked your post; it makes me feel interested and excited for this process of continuing growth! The bit about Dunning-Kruger was particularly encouraging.

It's interesting that guitar VSTs seem to want to use a low sustain value; I personally dislike how clipped it makes the sound (i.e., it ends right away when the note ends, whereas if you pluck an actual string, it should sound and then gradually fade away, at least I think moreso than what the few guitar VSTs I've tried want to provide by default). Personally, I think the higher sustain is better. I've been looking for guitar VSTs that might have better-sounding samples and more sound effects/more realistic scripted behavior than my PreSonus one, so I'll just have to keep an eye out for guitars that allow higher sustain values or else handle that aspect without the manual sustain adjustment.

Thank you again!

 

[AALLF]

Hey, I think this is the first time I've listened to your music! Can't believe I didn't notice that you already had a song on here, so I listened to "At Home With Miku" as well. Cool, I'll give some feedback on your WIP below. I hope I can be of some help.

The guitar sounds realistic to me on its own, until it gets to the part with the repeated notes. Does the "Nylon Guitar Reverb" use round robins? Or just one sample per note? There may be only so much you can do if it's just one sample per note. It sounds pretty good in the full track, since the vocals kind of mask the artificiality of the repeated notes. Unfortunately I don't use sampled guitars much, except as placeholders in my own WIPs, so I don't have any other recommendations. Guitars are notoriously hard to sample from what I understand, but you can get a pretty realistic sound out of them with some tweaking, depending on the part. It sounds good though, when the vocals are added to it.

Both vocals I believe could benefit from an EQ boost in the presence area (2k-4k) and upper treble area (10k and above). The vocals sound a little quiet because there isn't much there in those frequency areas, compared to the guitar. It's a tendency that I hear a lot in many vocaloid mixes, so you're not alone. It's probably just the default frequency response for a lot of vocaloids (I only own Gumi, though).

I liked the song! Do you speak Spanish yourself, or have someone translate the lyrics?

 

[mobius017]

Wow, thanks so much, @AALLF!

Ah, I think I know the part you mean with the repeated notes--it's probably the higher-pitched tremolo bits (or at least what I intended to be tremolo bits, though the notes they accompany are often short enough that you don't get many accompanying tremolo notes) in the second part. As best I can tell, it sounds to me like "Nylon Guitar Reverb" only uses one sample per note, but I think those notes would benefit from humanization of the velocity, and possibly the timing as well. Especially when you get a run of them, they can sound a bit machine gun-y. It's probably a bit like the rasgueado strumming from the bridge, which didn't sound right at all without humanization.

Thanks for the feedback on the EQ! I'll keep those ranges in mind when I get into mixing and see about boosting them. Most likely I'll add some compression and maybe an overall volume boost, too. I'm planning to put iZotope's Nectar on the vocals, at least, and it's amazing how much louder it usually seems to want to make things, at least by default.

I'm really glad you like the song! I took Spanish in school, so I was able to write the song myself. Do you speak the language?

 

[AALLF]

Yeah, that's what I do often with my piano parts is just slightly change the velocity and timing of every note. In the program I use (Reason), I highlight all the midi notes and it allows you to randomize the velocity and timing. I normally only do 5-10% velocity randomization and about 10-15 ms timing randomization, then I'll go in and nudge notes here or there if needed.

Another trick you can do, if you're able to automate pitch shifting/bending somehow, is to bring some of those MIDI tremolo notes up or down a semitone in the grid, but then pitch shift them back, and voila you have more than one sample per note! This may require more than one instance of the instrument to do this smoothly, and have them each one take turns playing every other tremolo note. It's a lot more fiddly work, and may not be worth it, but I've done that before.

No, I don't speak Spanish myself, though I wish I paid more attention in high school Spanish class sometimes!