How I Crafted the Jabberwocky Video
Brian Sykes
I Teach Creative Pros to UNDERSTAND / INTEGRATE AI while Retaining the Human Element | AI Consultant + AI Educator for Creative Professionals | Keynote Speaker
A.I. Explore: Illustrations in MJ
Here is a challenge of fun exploration. When prompting, if you name the subject (person, product, etc) it helps the AI maintain a slightly improved level of stabilization. If you have a unique product, an interesting character - name it. Midjourney also LOVES (so it seems) to try and guess the meaning of a word, and interpret it visually.
A great place to explore this is the work by Lewis Carroll named The Jabberwocky. It is a nonsense poem full of made up words - yet presented as actual language. The AI is very explorative in how it interprets the words and ideas presented. For reference, here is the poem... give it a try along with me following my process this time.
Jabberwocky
by Lewis Carroll
’Twas brillig, and the slithy toves
??????Did gyre and gimble in the wabe:
All mimsy were the borogoves,
??????And the mome raths outgrabe.
“Beware the Jabberwock, my son!
??????The jaws that bite, the claws that catch!
Beware the Jubjub bird, and shun
??????The frumious Bandersnatch!”
He took his vorpal sword in hand;
??????Long time the manxome foe he sought—
So rested he by the Tumtum tree
??????And stood awhile in thought.
And, as in uffish thought he stood,
??????The Jabberwock, with eyes of flame,
Came whiffling through the tulgey wood,
??????And burbled as it came!
One, two! One, two! And through and through
??????The vorpal blade went snicker-snack!
He left it dead, and with its head
??????He went galumphing back.
“And hast thou slain the Jabberwock?
??????Come to my arms, my beamish boy!
O frabjous day! Callooh! Callay!”
??????He chortled in his joy.
’Twas brillig, and the slithy toves
??????Did gyre and gimble in the wabe:
领英推荐
All mimsy were the borogoves,
??????And the mome raths outgrabe.
How I did it...
Rather than "here is my prompt - do this", I want to walk you through my creative process. This one is particularly easy to start with, because the referenced writing has already been produced by master penman - Lewis Carroll. His writing was so expressive and outlandish, it needed a look and feel that captured the wonder expressed. Reading over the poem, it is full of make-believe words used like legitimate terms, that I decided I really did not want to drive the direction Midjourney might take in exploring visual solutions to the poem. How do I let the words have weight from Carroll, but still prod it in a favorable direction?
I used this simple Prompt at the core:
Graphic Novel Illustrative Painting by [Peter Mohrbacher] from the poem "[title]" written by [author], for the lines: "[referenced line]" --ar 16:9 --v 5
There are a few terms here that sets the tone and maintains consistency for the collected set of images. That would be all the words used outside of the [brackets]. Let me explain.
Graphic Novel illustrative painting - this lets the AI know my direction is a story driven, visually rich image. The phrase is great for visuals because it calls on the AI to 'tell a story' with the renders it creates.
by Peter Mohrbacher - Peter Mohrbacher is a personal connection that I have spoken with at length about his ideas around AI, what IS art and the role of an artist, and the use of his name in prompting (which he welcomes and gave to David and the team at Midjourney, copies of his images to train the model). Not only is he an exceptional artist (check out his books and YouTube videos!), the use of his name in a prompt references his color palette and and a textural quality that I really like. Does it look like HIS style? Does it bear an uncanny resemblance to one of his paintings? As with all referenced artist, a name conjures 'elements' in the prompt renders, not a replication - so my answer to that is - No. Does it bear flavors of similarities - yeah.
I have a bracket around his name, because it IS an optional replacement. It could have just as easily been Piotr Kowalski (an amazing comic artist in Warsaw, Poland), J.C. Lyndecker, John Singer Sergeant, Norman Rockwell or a host of other potential names of reference of which a representative body of works can be referenced. I do seek to use those who are deceased (Peter is still alive and creating, but I do not have his personal permission to prompt with his name.)
The remaining elements - from the poem, written by, for the lines... these can conjure some pretty interesting results - because Midjourney and AI text2img processors are trying to interpret and lend meaning to their results. 'Lewis Carroll' for one is known for his vivid imagination and wild characters, so the results (like running a fun filter in Photoshop) lends a weight to the type of imagery that is particularly... interesting!
Graphic Novel Illustrative Painting by Peter Mohrbacher from the poem "Jabberwocky" written by Lewis Carroll, for the lines: "’Twas brillig, and the slithy toves Did gyre and gimble in the wabe: All mimsy were the borogoves, And the mome raths outgrabe." --ar 16:9 --v 5
I did this for each segment of the poem. That means a minimum of 7x I ran this prompt variation to create the illustrations to match the 'referenced line'. For the "cover" illustration, I used the following prompt - since I was not drawing from specific lines:
Graphic Novel Illustrative Painting by Peter Mohrbacher from the poem "Jabberwocky" written by Lewis Carroll, Cover Illustration for the Graphic Novel --ar 16:9 --v 5
Let there be sound...
Music and sound play a particularly powerful part of my makeup, and any time I have had opportunity to search out and discover the building blocks to develop a sound-bed for a creative project, I dove in with excitement. As a personal observation - this is on my 'list of 100 things about myself'... I love sound editing. I love music. I love the nuance of inflection, the cadence of speech, and story told with sound. For this project, I went to my trusty resource - Epidemic Sound.
I wanted a musical base that set a powerful tone for the overall feeling I wanted the slideshow of images to convey. Words are important here as in A.I.. Were I to ask you to define the SOUND you heard when watching the short video I crafted, how would you describe it? The words I typed as my search base were:
haunting, eerie
I listened to the beginnings of several dozen options. The whole was to be under 2-minutes, so too must the music tells its auditory story in that measured span. Not too short to leave you hanging, or too long to have to cut it abruptly and not complete the audible story being shared. I found the one I was after that had slower, feverish intensity that continually grew until it succumbed to a crashing fade. "The Clearing" by Golden Anchor.
I brought this over into Adobe Audition (love this program) and began to compose the sound bed. The purple (2nd line) is the music base from Epidemic Sound.
Spoken Word...
There are a lot of speech synthesizers out now, but my personal favorite for spoken word, is Eleven Labs. Not only do they have rich voice characters, but you can train it on your own voice... and it does a pretty darn good job! The voice heard in the telling of Jabberwocky, was crafted in Eleven Labs using the premade voice of Antoni. I collected the entire read - and that is the top line (in green) in Adobe Audition. The other sound slices are effects that help enhance the story - big screech, bloody body drag, wood crack 4, sword impact wood 1 and 2, villain laugh 3, zombie bite hiss - all of these were perfectly placed to emphasize the story line, and each effect was from Epidemic Sound. And no... I am not paid nor do I benefit from name dropping Epidemic or Adobe - I just like them both that much.
Putting it all Together...
I composed the whole of this video inside of Adobe Premiere. Yeah, there are ways I could have done this faster - but I was after effect (not the software - but the visual idea being conveyed in a way to leave a feeling.)
I like the structured layout and intricate details one can apply within the Premiere framework. To micro-adjust timing, to vary volume levels (if needed), the fine-tune color grading... all of this is inside of here. It was in Premiere I copy-pasted each line of text into the block of text to synchronize with the spoken word. Premiere has a means of listening to the audio and crafting its own transcription - but I knew with so many 'made up words' from Carroll's work, it would struggle with accuracy.
Once I was satisfied, I exported the whole and uploaded the results to my YouTube channel and shared the video across to my LinkedIn and Instagram pages as well.
I know... this is a lot to read. Most will not make it this far. But for those wanting to know the process used - this is how I did it... this time. It varies each time, but this is one process that may be an asset for directing your own future exploration. Enjoy!!
A.I. Explore with Brian Sykes
Generative AI Artist & Visual Storyteller | Bridging Art & Tech for Strategic Brand Impact
1 年Thanks for sharing Brian Sykes! Your process is bite-sized and easy to understand. ?? ??
Creator and Artist of Abandoned Teddies | BASc (Psychology), BSc (Info Sys)
1 年Thanks for the insight into your process, Brian… just brilliant! ????
UX/UI Product Designer | Senior Graphic Designer (Print, Digital, Web) | Maven of MidJourney, ChatGPT & Musavir | Creative Innovator | AIGA
1 年I always wondered where Jabberwocky came from. I had assumed it was Star Wars.
UX/UI Product Designer | Senior Graphic Designer (Print, Digital, Web) | Maven of MidJourney, ChatGPT & Musavir | Creative Innovator | AIGA
1 年Neat. Really cool