Composing

Anna Henry (Producer/Dev Exec with Netflix, Amazon, CBS) is leading a free Stage 32 webinar on How to Write a 2-Page Pitch for Your Television Series.

Even if you’re not writing the show, understanding what makes a killer pitch can help you spot (and align with) strong projects—and build better creative partnerships.

Registration Link: https://www.stage32.com/education/products/free-stage-32-webinar-how-to-...

Worth a look if TV is your space.

A few weeks ago, I decided to take a look (one free week) at software that claims to be able to create a vocal track based on musicxml and lyrics.  Most of this class of software 'extends' the piano roll paradigm used by the AEM (Audio Engineering Model).  Two things are added to a vocal track by the (normally separate) software - a text syllable is placed into the rectangle representing a note and a 'curve' is added around the rectangle to designate how the note is sung/articulated.  I decided to try one of my earlier live instrumental recordings where I recorded and played the melody and arrangement on a Yamaha P80 keyboard back in 2002.

https://soundcloud.com/joelirwin/my-friend-forevermore

I chose AceStudio to try as it heavily advertised on my Facebook feed. Here is what I tried:

1. I first created a full score of the track as a lead sheet containing the melody and lyrics.  I scored the whole song longhand without any repeat sections.  The idea of this score was to allow me to export a musicxml file that would allow AceStudio to create a singing track.  I imported a musicxml file into AceStudio and chose a singer and i was indeed able to generate a vocal track that played well against the original mp3 file - 'to a point'.  Now in real life the musicians may not hold a beat exactly - sometimes they could play slightly early or sometimes slightly late.  But the singer always seems to manage to stay with the musicians and never seems to sound out of sync (i.e., early or late with the musicians).  Unfortunately, about halfway through the playing, the singer started to sing before the musicians constantly and it was noticeable.  In concept, it could be fixed by the AE (Audio Engineer) but I didn't have time or the feature in the software other than manually dragging all the out of sync rectangles.

2. I thought perhaps it was my exported music XML so I tried the AceStudio utility to create musicxml from a PDF file - the utility kept aborting.  Support said it worked for them.  I never had time to troubleshoot why they worked and mine didn't.

3. I played the music track and recorded my voice singing it and then substituted another singer.  The issue here is the same as playing a keyboard and recording it into a piano roll track except a bit more inexact - thing is my voice when recorded was slightly high or low (i.e., I sang sometimes slightly out of key) and sometimes early or sometimes late.  So they were all the syllabic rectangles there but some were to high/low or slightly to left/right.  With more time and the right software I could have recorded myself separately and put myself in key and maybe even sync on beat ("quantize") and then import.  Instead I manually moved all the rectangles in place. Here is how it sounded

https://soundcloud.com/joelirwin/my-friend-forever-more-acestudio-valentina

So if you are well versed in the AEM and like to work with piano roll and enjoy tweaking and customizing, then you will find software like AceStudio will function for you much like any virtual instrument you have worked with keeping in mind that singing is its own idiom and that what gets manipulated and chosen is vocal specific.

Is it AI?  IMHO that depends on what you call AI - my idea of effective AI is where the software does most everything (like you would expect from a real singer).  The current software (so far as I have seen) while providing 'intelligence' on the singing paradigm, still requires a significant design and implementation effort by the AE.

Take for the example the first step of using a vocalist - auditioning for one.  In the software model you get to filter the singer on certain criteria and then you need to choose the resulting ones one at a time as your vocalist until you find one that sounds 'right'.  You may end up try out literally dozens of singers. In real life there are contests the function this way but as the songwriter/music creator' I would prefer the software to analyze my song/music and knowing (i.e., intelligence) a lot about the candidate singers including some of their recordings, be able to create a preferred ranked list of 'finalists' much like the auditioning process could be for a theater or an opera.

I believe AI Vocalist software has a way to go in offloading intelligence from the AE.  For example, each singer has a different sweet spot (AceStudio does not let you know what the sweet spot is but software like "SoundID Voice AI" does).  The music key may need to be modified to maximize the singers performance.  Also, some singers have limited range and so the key needs adjusting to minimize 'octaving' which may still be necessary sometimes (e.g., singing too high may sound weird and so the vocalist may need to sing an octave lower).  And here is one more: word articulation.  Singers sing in dialects and words like "hand" or "potato" which are pronounced differently and so must be sung differently.  You can ask a singer to sing every occurrence of the word "hand" in a different pronunciation.  Right now it could be difficult for an AE to tell the software to change all occurrences of a word or syllable to something else or even something more complex - re-sing a song with a different accent.

So they may be software out there that are more sophisticated than AceStudio in implementing a virtual singer - I have only tried one so far.

And by the way, I have looked and discussed only software that create a virtual singer based on melody and lyrics.  This does not pertain to software that creates its own instrumentation and singers such as Suno.  I may look at it just to become acquainted though I much prefer composing and arranging my own material.  I have heard some country demo tracks coming out of suno and they sound awfully commercial - that's very scary!

Hey everyone, my name is Shane. I recently joined the community.

I’m a composer and producer for film, TV, and media. As a dancer as well, I bring that perspective into my compositions. Film and TV often inspired me to explore new movement, feeding off the cinematography and music.

When I approach scoring a scene, I start by reading the script, but then I really tap into what I see onscreen letting my body guide me through the rhythm and emotion of the moment, whether it's slow or fast, joyful or poignant. In a way, I’m moving in real time with what’s happening, and from there, I translate that into music, creating a sound and rhythm that’s both unique and expressive.

I’m excited to connect with filmmakers, screenwriters, directors, and others who are looking to bring a fresh and distinctive sound to their projects. Feel free to message me anytime. I’d love to meet you!

For those looking to enhance their music composition skills, I came across this insightful video: https://www.youtube.com/watch?v=Zd_sbC5bnCA

The video shares five transformative principles for arranging and organizing musical ideas into complete pieces. It’s particularly helpful for those who have loops or motifs but struggle with developing them into full compositions.

After watching, I’d love to hear your thoughts: Which of these ideas resonate with you the most? Do you have any personal techniques that have significantly impacted your composing process? Let’s discuss and learn from each other’s experiences!

This Wednesday, April 9th, we’re hosting an exciting 24-hour Ask Me Anything in the Screenwriting Lounge with

Eric Nazarian—a writer, director, producer, photojournalist, and Nicholl Fellow.

Here’s the link.

Remember to ask relevant questions.

Have a great week!

K. Kennard

/-/

https://www.stage32.com/lounge/screenwriting/Ask-me-anything-wednesdays-...

In Person Meetup in the East Valley next week

RSVP here

This is pretty cool news !

Even though a year ago, this indie Artist wins an Emmy !

https://variety.com/2024/music/news/bmi-film-tv-awards-honors-game-thron...

In your career/journey, what would make you feel like you "made it"?

Hi everyone! I'd love to hear what everyone is working on (or recently finished.)

Good article from BMI -

I like the way he shares how’s he composes with instruments he has, sometimes, before writing / doing a mockup.

Check it out!

https://www.bmi.com/news/entry/a-conversation-with-composer-mark-orton

Hi! My name is Rob Lowry and I am a music supervisor for film and television. Recent projects include Companion (New Line), Heart Eyes (Sony), and Novocaine (Paramount). Upcoming Projects include Drop (Blumhouse) and Friendship (a24). I've also worked on past projects such as Gossip Girl (HBO Max), Ramy (Hulu), The Lost City (Paramount), Do Revenge (Netflix), and Cha Cha Real Smooth (Apple).

I'll be answering your questions all day Wednesday through Thursday morning! Feel free to ask any questions about music in film & tv. Some samples that may be useful:

- What’s the best way that a writer can communicate a musical vision in a screenplay or pitch, to help the music supervisor?

- How do you decide when to use score vs. licensed music in a scene?

- What’s the biggest mistake filmmakers make when planning their music budget?

- What makes a film or series stand out to a music supervisor?

- How can emerging directors or composers get on a music supervisor’s radar?

Excited to hear your questions and start a fun dialogue about the power of music in film & tv. Chat soon!

https://www.hollywoodinsider.com/best-film-composers/