Mixing my own music with AI

Suno AI has an “upload audio” feature, allowing users to upload up to 60 seconds of their own content to be extended by the AI. So earlier this month I had some fun feeding it 45-60 clips of my own music and having the AI write lyrics and turn the clips into choir songs. It’s interesting to hear how the AI uses the melodies, chord progressions, and orchestrations provided in its own creations. The lyrics are a bit amateurish, but serviceable; I was too lazy to write my own. I’m calling the project Hannifin x AI. Here’s the first installment, based on my classic piece “Hour by Hour”; the first 60 seconds are from the original piece, while the rest is AI-generated.

I did the same with 18 other of my pieces. Some things I noticed:

  • The AI works best with simple 8-bar melodies, or 4-bar phrases. It doesn’t seem to “parse” weirder phrase structures very well.
  • It’s not very good at extended the input instrumentally, in my opinion; it quickly starts to sound too repetitive. Having it produce lyrics and turning the music into a song seems to work better. (Melodic repetition seems easier to bear with alternating lyrics.)
  • If you want the AI to generate the voice of a choir, feeding it music from the louder, more energetic and melodic parts of a piece seem to work better, especially if it features a prominent string section. Otherwise you’re more likely to get a soloist, and the music it generates is less likely to sound like a natural continuation of the music you provide.
  • For whatever reason, some tracks just seem to work better than others; maybe it depends on how “familiar” the AI is with the melodic and harmonic patterns? For some tracks, it gave me some pleasant results right away. Other times I had to roll the dice over and over to get something acceptable.

There were some pieces I tried for which I could not get any output that I was happy with, including The King’s AssassinThe Moon Dreamed By, and On the Edge of a Dream. And there was one track, Silver Moon Waltz, for which I got a couple songs I was pleased with. Anyway, I’m done trying for now.

As for the video above, I made it with Blender 4.2, which took a little time figuring out, mostly with various tutorials on YouTube. I’m not completely satisfied with the results. What’s supposed to be “dust” looks perhaps too much like snow and moves a bit too fast, and the text looks a bit weird. Turns out trying to create a sort of “drop shadow” effect on text in Blender is pretty much impossible; I had to sort of fake it with compositing cheats, and I’m not sure I did the best job. (I could’ve just put the text on the background picture and used an image editor to create the drop shadow, but I wanted the animated frequency bars to have it too.) Also, the text might be a bit too bright, but I blame that on the VR display I get with Immersed in the Meta Quest 3.

I’ll upload the other 19 songs I created soon!

 

My first AI music album: “The Shadow Age”

I’ve been enjoying writing songs with AI songwriting tool Suno for the past few months, and recently put together a full-length album of some of my favorite tracks so far. While AI wrote the music and provides the performance, I wrote the lyrics, which are very deep and profound. (Though two of the tracks are from old famous poems.) The symphonic metal album is free to download here (ZIP file, MP3 V0, 111.2MB) or on Bandcamp.

Don’t like AI music? Well, I’m sorry, but I’m going to create even more AI albums, bwahaha!

Prayer to St. Michael with Suno AI

I turned the Prayer to Saint Michael into some epic choir music with Suno AI:

It would have been a lot easier for me to learn my prayers as a kid if it had been so easy to turn them into music.

I actually wanted the whole prayer to be sung by the entire choir, but Suno AI seemed to insist on featuring a solo vocalist for the second part (“May God rebuke him…”), as you can hear above. I also had to try quite a few times to get it to pronounce “wickedness” clearly and correctly; it kept wanting to sing “winess” or “wicks”. But I like how it ended up.

Here are some other versions it come up with, though I didn’t quite like any of them as much as the above.

V3 with the little “….amen!” at the end sounds almost comical.

Anyway, I’ve been thinking about posting some lyric videos of my Suno creations to YouTube. I made the St. Michael video above with Shotcut, but that seems impractical for a video with changing lyrics. Perhaps if I can make a template in Blender, I can use that. But I haven’t played around with Blender in a long time, and I don’t want to spend too much time on it… something to play around with later this month.

For now, it’s almost time for the 2024 eclipse! Though the weather might not be so good… we’ll see…

Fun with Suno: AI Song Generator

Wow, this is my first blog post of the year. That’s pretty sad.

This week I’ve been playing around with Suno, an AI song generator. As far as music-generating AI goes, it’s definitely the best I’ve seen so far, as it actually generates melodies, which is what most musical AIs stink at.

Of course, it’s got its weaknesses, but this is new tech, so that’s to be expected. And I haven’t seen competition that really does anything similar yet, though I’m sure that will come.

Anyway, here are some of the songs I’ve generated with the app. You can have it generate its own generic lyrics, but I find it more interesting to provide my own.

The first three are symphonic metal, one of my favorite genres. Maximus is an epic choir singing in another language. A Song Unsung and The Road Inside are some relaxing indie folk. The Owl and the Dragon is a folk-ish lullaby. A boys’ choir sings The Crystal KnifeAbout the Cats is in the style of a generic 90s pop song. Finally, Boop! is an Irish folk song with nonsense lyrics. Links to the lyrics for each song can be found at the bottom of this post.

Weaknesses

Perhaps the biggest weakness is lack of control. Other than providing the lyrics and style, you don’t really have much control over the details, which you’d likely want if you were a serious composer or songwriter.

Styles are also limited; I asked it for the style of a Russian folk song (“The Owl and the Dragon”), and it just gave the singer a Russian accent.

The format is limited. For best results, it seems good to stick to four-line verses and chorus, from which generates standard generic 8-bar melodies.

It’s text-to-song isn’t perfect. Sometimes it repeats a syllable, ignores a syllable, or puts emphasis on a weird syllable. Sometimes it will sing a line from a verse as though it’s part of the chorus; its “parsing” makes mistakes.

Sound quality is another weakness. You can probably tell from the examples that it outputs some pretty low-quality sounds, especially with the bombastic symphonic metal, which can sometimes make the lyrics hard to understand. But musical sound data has even more information than images, and image AI generators themselves still output a lot of noise. With images, however, it’s easy to discount the noise as texture or something. With musical sound, noise gets in the way; with professional recordings (especially if you’re an audiophile), we’re used to hearing nice clean sounds; even the hissing high frequencies of cymbals matter to a degree.

In some output (not the ones I’ve showcased here), I could swear I could hear overtone artifacts of other words or singers faintly in the background; I’m guessing the AI is doing diffusion with frequencies / Fourier transforms, and generating little fragments of training data it should be ignoring. Or it could just be weird auditory illusions.

Is it useful?

Given all these weaknesses, is Suno a useful tool? Honestly, it’s probably not super useful for professional musicians yet, perhaps other than a quick and easy way to get some ideas. Otherwise, it’s perhaps still more of a toy at its current stage.

Granted, such a musical toy can still be a lot of fun, and I’m excited to see the app develop further. I’m not sure who’s behind it or even what country it’s from, but I do hope they don’t get bought out too easily.

TuneSage

What about my own music AI, the development of which I’ve been procrastinating on? Has Suno beat me to the punch?

My approach is a lot different as I’m not really dealing with the sound of music. My focus with TuneSage is more about the actual notes and musical structures of a piece.

Lyrics

Here are links to each song on Suno, where you can see my profoundly beautiful lyrics:

Close Your Eyes
A True Heart
The Shadow Age
Maximus
A Song Unsung
The Road Inside
The Owl and the Dragon
The Crystal Knife
About the Cats
Boop!

Coming Soon: The Archives

I’ve got quite a few compositions that I’ve never uploaded to streaming services, and I’ve been meaning to for a while now. In fact, I’ve got over 4 hours of tracks unavailable on Spotify and other streaming sites, which over the years has probably cost me some 12 cents or so in lost profits. OK, maybe not that much, but still. So before the end of the year I hope to release “The Archives”, four volumes of my early music. As there are about 4 months left of the year (as of this writing), I’ll probably aim to release one a month.

As most of these tracks have been available to freely download on my outdated MP3s page for a long while now, digital copies of the albums will also be free.

Here’s the current plan in terms of track listings:

The Archives: Volume 1

  1. The Workshop
  2. Hatching of a Dragon
  3. Grandeur
  4. Flight of the Dragon
  5. The Silver Knight
  6. The Aeneid
  7. Canon (Not Really) No 1 in C major
  8. Canon (Not Really) No 2 in C major
  9. March of the Canterbury Tales
  10. Short Piece for Orchestra
  11. Largo for String Quartet in A major
  12. Serenade for Strings No 1
  13. Octet in A minor

The Archives: Volume 2

  1. Journey of a Steed
  2. End of the Road
  3. Knights of the Round Table
  4. Lullaby for Cello and Piano
  5. Short Piece for Strings
  6. Trio for Harp, Flute, and Oboe No 1
  7. Woodwind Quartet in G minor
  8. The Gafradalasha String Quartet: I
  9. The Gafradalasha String Quartet: II
  10. Waltz of the Penguins

The Archives: Volume 3

  1. Waltz of Pegasus
  2. Twilight Fantasia
  3. Dragon of the Mist
  4. The Neuschwanstein Suite: 1. Dance of the Ice Sprites
  5. The Neuschwanstein Suite: 2. Dance of the Woodland Sprites
  6. The Neuschwanstein Suite: 3. Battle and Defeat of the Trolls
  7. The Forgotten Wish
  8. Serenade for Strings No 2
  9. Broken Swords
  10. The Banquet
  11. Trio for Harp, Flute, and Oboe No 2
  12. Hour by Hour

The Archives: Volume 4

  1. Guinevere’s Lullaby
  2. The King’s Assassin
  3. Lullaby for Harp and Whistle
  4. Mozart’s Dream
  5. Trio for Harp, Flute, and Oboe No 3
  6. This is the Pizz
  7. Inside the Android’s Dream
  8. Dance of the Fools
  9. Fairy Music
  10. Lullaby for Harp and Vibraphone
  11. Melody for Harp and Piano
  12. Piano Concerto No 0
  13. It’s a Rondoful Life
  14. Short Piece for 2 Violins
  15. Island of the Dragons

So be on the lookout for these amazing albums collecting your favorite golden oldies on your favorite streaming platforms! Stream them and allow me to make 0.0002 cents each time you listen, which will help me pay for a single M&M every five years.

P.S. There are still a few tracks missing, mainly more experimental and less-melodic works that would likely sound jarring and out of place mixed in with the rest. I may upload these as part of a separate album later.

New album “Moonwish” released!

I recently released my third album, Moonwish!

The album is now available on Bandcamp, AmazonSpotify, and YouTube!1

9 tracks about 50 minutes long. The cover features an image of the 2017 solar eclipse, though I tried spicing it up with some color and texture. I think it came out OK, if still a bit amateurish.

I’m distributing it digitally with DistroKid, which charges about $20 a year, but does not charge fees for royalties earned. Still, in the five years I’ve paid them, I’ve only made about $35 in return, so still a net loss (and that’s before taxes). But that’s with only two albums (A Dream Half Lost and Storybook Overture). As DistroKid does not charge more for each new release, I really need to release more stuff so that it will at least pay for itself.

I still hope to produce some music animation videos for YouTube for some of the album’s tracks, particularly Second Star to the RightSomewhere Far Away, and Star of Wonder. I’d also like to create a video that features the entire album.

Patreon

On a side note, I also haven’t updated my stagnant Patreon for a long while. All of my patrons dumped me for not delivering, save one single patron who must be a saint. But I was composing about two new tracks a month, so now I owe around 80-something tracks for all the months I collected tips and did not deliver anything. So I still plan to deliver all those tracks! But it will likely take a while. Other than delivering those tracks, I doubt I’ll keep using Patreon; I just don’t seem to be able to stick to a monthly schedule. (Actually, I think there’s a newer option now where you can just deliver stuff when you can, which I guess I could try, but it’s probably still not really worth the effort for my niche audience. Better to focus on TuneSage.)

A bit of new Mozart!

A short but previously lost piece by Wolfgang Amadeus Mozart was recently performed just last week on January 27th, Mozart’s 265th birthday. The brief allegro can be heard below:

A bit more about the piece from the Mozarteum Foundation:

The present Allegro in D major for piano was hitherto known only from 20th-century sales and auction catalogues, where it had been described as a “sketch for a composition for orchestra or chamber ensemble”. On the basis of this vague information, musicologist Alfred Einstein assigned the piece the number App. 109g/16 in the third edition of the so-called Koechel catalogue, which finally became K. 626b/16 in the sixth edition. A music-loving engineer bought the autograph manuscript in an antiquarian bookstore in Paris in the late 1920s; his descendants kept it in the  Netherlands for 90 years. When the manuscript was offered to the Salzburg Mozarteum Foundation it became immediately evident that it was not merely a sketch, but a complete work for piano in Mozart’s own handwriting, a piece unrelated to all known compositions. Finds of this kind have become extremely rare; the last comparable case was the rediscovery of the Allegro in F major for piano, K. 33B in 1937.

Although short, I think the work is ingenius. It’s quality work. And there are simply too many notes, that’s all. Just cut a few and it’ll be perfect.

Radical Face and The Land Before Time

Musical artist Radical Face, one of my favorites, recently released a new EP: Therapy. It’s great stuff, catchy melodies, memorable lyrics. While listening to the third track, “Personal Giants”, a simple four-note phrase that appears at the end of the main melody caught my ear. You can hear it first appear at about 12 seconds in:

Just those four notes there. “And kept the light…” And again at 30 seconds in. “You told me time…” Sounds like a simple ascending major triad, with a minor chord on the second beat. Something like this:

This simple phrase stuck out to me because it reminded me of one of my favorite film scores, James Horner’s score for The Land Before Time. The “Great Valley” theme begins with a similar phrase, an ascending major triad with a minor chord (iii?) on the second beat. You can hear it enter at 3:07 in this track:

Other than those four notes, the melodies are quite different. But to me they’re memorable enough that hearing them in Radical Face’s song immediately conjured up images of Little Foot and rocks and a great valley and dead cloud dino Mama beckoning… And the lyrics in “Personal Giants” perhaps could apply to Little Foot. “To me you’re a giant, some distant lighthouse” … maybe a stretch, but it could work, yep yep yep.

So then just the other day Radical Face does a livestream Q&A, and what does he say at 37:07? Behold…

“Ooh, I love movie soundtracks. Some top ones would be, I really specifically adore The Land Before Time soundtrack by James Horner. I think it’s so good.”

Aha!! You see?! Clear and undeniable evidence of musical influence here! And only I understood, only I could see the secret of those four notes, only I made the connection, haha!

By the way, one of my pieces also features some clear and undeniable influence from The Land Before Time soundtrack, if you can find it…

More font fun and other random stuff

Font rendering in OpenGL

Haven’t done so much programming in the past week, but I did try rendering fonts with NanoVG (in lwjgl). Unfortunately it’s really not much better than just using OpenGL’s NV path rendering extension. Small fonts look slightly better, but not really good enough for me to want to use them. See the example below, a zoom-in of an 8-pixel high rendering of the font “Verdana”, NanoVG rendering it on top, NV path rendering below. NanoVG is better, but it’s still way too fuzzy to look any good.

So I might just use bitmap fonts for small text; I can’t see any alternative. (Bitmap fonts basically means loading in each letter as a pre-rendered picture and plopping it on the screen. The disadvantage is that they don’t look very good when resized or positioned between pixels, but they’ll at least be guaranteed to render small fonts clearly and crisply.) I’ll continue to use NV path rendering for larger fonts or fonts that need to be animated more dynamically or rendered with 3D perspective. I do want to try using oversampling with stb for small fonts before I move on from font rendering and further develop a GUI system; the demo doesn’t look too bad.


Kanopy film streaming service

I just realized our local library offers free access to the film streaming service Kanopy, and they’ve actually got a decent selection. Not the latest blockbusters, but some good foreign and classic cinema. (They’ve even got The Red Pill, the controversial anti-radical-feminist documentary which Netflix refuses to stream. (Though they did just recently finally add the DVD to their catalog.)) Since it’s free (for library card holders of participating library systems), we’re limited to 10 streams per month, but the streaming quality is decent. It’s not full HD (at least not on my PC), but it’s better than DVD quality. Interestingly they also allow you to embed videos. Here’s “Kumiko the Treasure Hunter”, which I really enjoyed (which I guess you won’t be able to see without an account, haha):

Interesting indeed!


Some metal music

Finally, I recently discovered the band Elvenking. I couldn’t quite get in to some of their older work, but their last album, released last year, is quite catchy. Disregarding the bizarre sense of fashion metal band members tend to share, this song is some power metal awesomeness:

I also came across the latest album from the symphonic metal band Leaves’ Eyes, and found it to be quite catchy as well. Love the use of choirs, the female lead’s operatic voice, and the cheesy fantasy lyrics. (I’m not a big fan of the growl singing, it sounds so gross and ugly and demonic, why is it so popular? It sounds so awful, so unmusical. Bah!)

OK, that’s all for now.

Second full-length album coming soon!

Storybook Overture

I hope to release my second full-length album next Friday, December 2nd, 2016. A digital version will be available through my bandcamp page, and I also hope to try releasing an audio CD through Amazon’s CreateSpace. I’ve never tried their CD service before, so I’ll be interested in how it turns out. (That might take a little longer as I don’t know how long it will take for them to send a “proof” copy… we’ll see.)

STORYBOOK OVERTURE, the title of one of the tracks, seemed like an appropriate album title, as many of the tracks were inspired by my book series, Insane Fantasy. You’ll also see I’m reusing the artwork I did for the book on the album cover.

Here’s the track list:

  1. The Storm Cometh
  2. Storybook Overture
  3. A Stargazer’s Lullaby
  4. The Stormbringer’s Apprentice
  5. Broken Wings and Distant Things
  6. Journey by Moonlight
  7. Lullaby of the Westwind Woods
  8. Secrets of the Ancient Seas
  9. September Ends
  10. Lullaby for a Quiet Village
  11. Fairytale’s End

As you can see, I’ve already released all but tracks 9 and 10 to YouTube (and track 10 will probably be my next YouTube upload). The total playtime of these tracks comes out to about 70 minutes, so it’s a pretty full album.

It’ll probably be a while before I get the album on iTunes, Spotify, etc., as it’s not really worth the cost with the amount of listens I get (less than 1,000), but I’d like it to be on iTunes and Spotify eventually. It’ll just have to wait for a bit.

Also, I’ll try to email free bandcamp download codes to any Patreon supporters I have at that time. (They can already download most of the tracks on Patreon anyway.)