A new trailer for a film I am working on - "The Lost Prince - Arthur's Last Stand"
SOUND ON
I hope to make an entire feature-length film in the coming months.
The list of tools I used I used is in the comments - I mainly used
@midjourney
and
@LumaLabsAI
's Dream Machine.
@themaxburns
What has happened to the NYT?? If one wants a solid understanding of current events in a systematic, contextual and in-depth perspective, the NYT is pretty away from providing such an understanding.
@themaxburns
I look at the web front page of the NYT, I think, ok, I can skip that story, and skip that one, and so on until I almost reach the bottom of the page
Here is my latest revision of "The Lost Prince - Arthur and the Woods of Avalon" using generative AI tools,
@pika_labs
1.0 for image-to-video and
@midjourney
for text-to-image. I also used
@elevenlabsio
for voice-over. Generative AI has come a long way and will only get better
And yet another comparison for image-to-video. This time I'm looking at how the different platforms handle special effects scenes - this time car chases
. The comparison is between Runway Gen-3, Luma Labs, and Kling professional mode.
LET ME KNOW WHAT YOU THINK IN THE
Here is a quick comparison for image-to-video between (1) Luma Labs Dream Machine
@LumaLabsAI
, (2)
@runwayml
's new image-to-video Gen-3, and (3)
@Kling_ai
image-to-video.
I used pretty simple prompts for all of them.
Let me know what you think. I'll continue to do more
Here is a quick test of
@pika_labs
using Lip Sync. I generated the images using
@Midjourney
, created the original videos in
@runwayml
and then used Pika to lip sync an audio clip created using
@elevenlabsio
THere is an AI video I did for the Hackathon for my company, demonstrating what I can be done for AI video. Used Midjourney,
@runwayml
and
@pika_labs
and
@D_ID_
A new short film exploring different cinematic styles.
I am working on a feature-length film , 'The Lost Prince - Arthur and the Woods of Avalon" and trying out different cinematic styles.
Images were created in
@midjourney
, image-to-video clips were created in
@runwayml
and
I'm working on a new project and I wanted to test the combo of
@suno_ai_
music and
@runwayml
lip sync to see how well it works. This was a really quick test.
You can make a character smile in both Luma Labs and Runway Gen-3 and Kling AI (but watch the movement of the characters in the background) with image-to-video
prompt was "smiling young man in medieval tavern"
First Luma Labs
Here's another comparison between Luma Labs, Runway Gen-3, and Kling AI. This time for cinematic shots. Music was created in
@udiomusic
SOUND ON
Let me know what you think. Which tool do you prefer?
@nickfloats
Great comparison Nick! One challenge is going to be translating v5.2 prompts to v6. I’m in the middle of creating a story using MJ images and I have thousands of prompts which are tied to specific compositions. I need to “translate” them all
@mreflow
I managed to get Code Interpreter to rotate, change and brightness and interpolate between images, as well as added an audio track, by generating python code and then plugging it into Jupyter notebook (the images are from Midjourney)
Here is my latest short film, "Across the Cosmos" - one possible future of human space exploration across the Solar System and Beyond...
I was inspired by "Wanderers" by Erik Wernquist, and the "Pale Blue Dot" by Carl Sagan.
I used
@midjourney
for all of the images,
@runwayml
Here is a method to get reasonably good results for a consistent character and then placing the character in different settings. Not perfect but works
The formula is [text prompt of setting] + [text prompt of character] + [style ref of character] + [style ref of setting]
Testing AI Lip Sync tools -
@runwayml
,
@pika_labs
,
@syncdotso
(Sync Labs), and
@D_ID_
. For Runway, I also tested out different zoom levels. Runway is nice because it extends the video for you. Sync Labs apears to loop the video in order to extend it), D-iD doesn't allow video
I am experimenting with recreated the styles of well-known cinematographers that could be used a style reference in Midjourney (using --sref [reference images]).
I tried to generate several different types of shots. First, I am presenting the cinematographer and then the shots
@AlexBMorse
Great to hear. I’m really looking forward to visiting Provincetown for Carnival Week in two weeks time. Vaccines work and are highly effective in preventing symptomatic Covid. As a scientist (PhD in Biochemistry, Oxford), I think the CDC jumped the shark
Here's a test of using a cinematographic style across a lot of establishing / background shots.
My apologies for repetitive shots. The idea was to really push camera motion as much as I could without losing too much coherence and maintaining consistent lighting. I have a
@StevenGlinert
I’m in Tech and no way I’m on the Trump train. It’s just going to be 4 years of chaos and social / political unrest and economic uncertainty.
Songs from "The Lost Prince - Arthur and the Woods of Avalon"
The inages were created using
@midjourney
, image-to-video was done using both
@runwayml
Gen-2 and
@pika_labs
, the songs were created using
@suno_ai_
, the lip sync was done using D-id studios (
@D_ID_
), and some of
@zachdcarter
The Democratic leadership really needs to stand up to Trump now and do everything we can to win, or else we face 4 years of chaos, tears and anger.
@PurzBeats
@pika_labs
I've been experimenting with
@pika_labs
as well and have come up with a rough walk cycle. I upscaled the videos and images during each iteration and used the prompt "walking away -motion 1 -camera zoom in"
@ClimateHuman
I just can’t understand why anyone would support these actions. These paintings are great cultural patrimony of all humanity and to deface them in misguided protest is abhorrent. It is a totally false choice to state u have to choose between either art and climate disaster
@justin_hart
@LumaLabsAI
You can do a three-panel story using
@LumaLabsAI
. Somewhat mixed results. I simply stitched together three images from Midjourney (one has to click in the video to see all three panels)
How to change the clothes on a consistent character in
#Midjourney
. (1) Create a character, (2) Create the Clothes by itself, (3) Use the image prompts from 1 & 2, and pre-fix "wearing [the clothing]" to a combined prompt!
Here's a quick scene using
@LumaLabsAI
image-to-video. I used ending keyframe images on some of the clips and used camera movements in the prompts (SOUND ON)
@Kling_ai
I think Runway Gen-3 and Luna Labs are pretty comparable, Gen-3 appears to throw in a lot of artistic artifacts though. Hopefully this puts pressure on Luma Labs to have an unlimited plan
I'm working a new project. Here is a teaser trailer. I am hoping a have ten-minute action sequence. This trailer is mainly establishing shots while I work on the whole sequence.
Here is my latest revision for the trailer for "The Lost Prince: Arthur and the Woods of Avalon."
I used Midjourney mostly (along with DALLE-3 for a few figures) and
@Pika_Labs
for image-to-video.
@icreatelife
"Guinevere's Lament - Do not forget me in the Night" - really an experiment to see how well RunwayMl can do Lip sync for singing. I used
@suno_ai_
for the music and voice. I wrote the lyrics.
@chaseleantj
Great work Chase! I’ll try this method out. Inpainting works really well for consistent characters too. I wrote up a Medium post on how to do it.
A re-make of a short film, "The Lost Prince - Arthur's Last Stand."
The images are from
@midjourney
.
For Image-to-Video I primarily used
@LumaLabsAI
, but some clips are from
@runwayml
Gen-2 and
@pika_labs
.
I used Gen-3 for opening and ending titles.
The music is created using
Here is another quick experiment with establishing shots, trying to get movement out of characters. I am finding that you need to crank up the motion to 10 to get characters to move using Runway's Gen-2 motion brush.
Images created using
@midjourney
. I used a cinematic style
@CodeByPoonam
Another experiment with
@LumaLabsAI
, using a three-panel layout, with three clips stitched together using their new keyframe method (click on video to see all of the panels.) With a little more work, it make a really cool effect.
@nickfloats
I’ve managed to get consistent characters in interior scenes using a combination of seeds and image prompts. See
I can also get a consistent character along with another character in a setting but haven’t written this up yet. It is tricky.
@ClimateHuman
@ClimatePsych
I hate, I mean hate these actions. It is really lazy activism that accomplishes nothing and potentially damages priceless art, I’ve been doing activism for 20+ years & having effective strategy is hard. These activists need to think long and hard and come up w/ better actions.
Here is my latest short featuring scenes from the "The Lost Prince" which take place in"The Forest Primeval"
"Fáinne geal an lae" means "The bright dawn of day."
Image-to-Video clips were created in Luma Labs Dream Machine
@LumaLabsAI
I would try to hone some of the video
@Noahpinion
This is pretty astounding:
“The eurozone economy grew about 6% over the past 15 years, measured in dollars, compared with 82% for the U.S., according to International Monetary Fund data.”
@TurkMatthew
@LangChainAI
Actually this evening I got a pandas data agent working with Chat agent executor. So I can do Q&A with a CSV data file. No need to do embeddings or use a vector store. Added bonus is that you can view the thought process of the agent. Now to add memory & more tools
Here is a quick demo using Consistent Characters in
@midjourney
, Lip Sync in
@pika_labs
, Sound Effects Using
@elevenlabsio
and Video using
@runwayml
Gen-2. Lots of tools but pretty quick to put it all together. I think the weak link is Lip Sync
@icreatelife
Vary (Strong) is really good. Just use an image prompt, add background prompt and image weight of day 0.25, and change aspect ratio. Here’s a medium article about it
@alexiskold
@bhorowitz
@pmarca
Whether to support Trump or not is a basic test of character and judgement and a lot of people of failing that test (by supporting him). Disappointing and leads one to question their judgement about everything.