MidJourney.com, Create Any Image you want with AI!!!

Joined
May 1, 2012
Messages
1,282
Reputation
80
Daps
2,934
Reppin
NULL
Sora is def top tier.
20250331-1836-Yuffie-s-Dual-Identity-simple-compose-01jqq9191tfezr0jtvv1sz2n7f-1.png
 

GoldenGlove

πŸ˜πŸ˜‘πŸ˜ΆπŸ˜‘πŸ˜
Staff member
Supporter
Joined
May 1, 2012
Messages
60,752
Reputation
5,890
Daps
142,963
They really dropped Gen-4 for Runway but the character reference feature isn't even out in this shyt yet
:unimpressed:

That OpenAI 4o got these platforms rushing shyt out

What the fukk is taking Midjourney so long to crush the buildings with their image and text to video model?

:feedme:
 
Joined
May 1, 2012
Messages
1,282
Reputation
80
Daps
2,934
Reppin
NULL
What was the prompt for this?
I copied it from the Arcane pic that appears from time to time. But here is what I used:

----------------- YUFFIE ------------------------------------------------------------------------------------------------------------------------------------------------------
A front-facing close-up portrait of Yuffie From Final Fantasy VII. The composition is perfectly symmetrical, with the face divided vertically into two distinct artistic styles. The left half is rendered in the FFVII visual style β€” with stylized shapes, signature linework, and color palette unique to this universe. The right half is hyper-realistic, showcasing intricate skin textures, lifelike lighting, and fine details that bring out every pore, hair, and nuance of expression. The division between the two halves is jagged and organic, resembling a cracked mirror or torn paper, highlighting a symbolic rift between fantasy and realism. The character's expression is neutral yet introspective, allowing both styles to equally convey their emotional depth. The background is soft and subdued, focusing all attention on the face and the contrast between the two visual worlds. The faces on both sides line up perfectly
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

------SEPHIROTH --------------------------------------------------------------------------------------------------------------------------------------------------------------
A front-facing close-up portrait of Sephiroth From Final Fantasy VII. The composition is perfectly symmetrical, with the face divided vertically into two distinct artistic styles. The left half is rendered in the FFVII visual style β€” with stylized shapes, signature linework, and color palette unique to this universe. The right half is hyper-realistic, showcasing intricate skin textures, lifelike lighting, and fine details that bring out every pore, hair, and nuance of expression. The division between the two halves is jagged and organic, resembling a cracked mirror or torn paper, highlighting a symbolic rift between fantasy and realism. The character's expression is neutral yet introspective, allowing both styles to equally convey their emotional depth. The background is a fiery fire of flames and subdued, but focusing all attention on the face and the contrast between the two visual worlds. The faces on both sides line up perfectly
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


It worked well with all the other FFVII characters as well.
 

Majestyx

Duck Season
Joined
May 2, 2012
Messages
17,355
Reputation
2,536
Daps
40,409
Reppin
Los Scandalous
They really dropped Gen-4 for Runway but the character reference feature isn't even out in this shyt yet
:unimpressed:

That OpenAI 4o got these platforms rushing shyt out

What the fukk is taking Midjourney so long to crush the buildings with their image and text to video model?

:feedme:
we will be talking soon bro...
 

GoldenGlove

πŸ˜πŸ˜‘πŸ˜ΆπŸ˜‘πŸ˜
Staff member
Supporter
Joined
May 1, 2012
Messages
60,752
Reputation
5,890
Daps
142,963
While Gen-4 doesn't have the character refenced for consistency enabled yet, I ran it through some of the images I'm working and with just text and it's impressive. It takes much longer to generate a clip compared to turbo tho.

It also doesn't allow you to set a beginning or ending frame like Turbo does.
 

GoldenGlove

πŸ˜πŸ˜‘πŸ˜ΆπŸ˜‘πŸ˜
Staff member
Supporter
Joined
May 1, 2012
Messages
60,752
Reputation
5,890
Daps
142,963
While Gen-4 doesn't have the character refenced for consistency enabled yet, I ran it through some of the images I'm working and with just text and it's impressive. It takes much longer to generate a clip compared to turbo tho.

It also doesn't allow you to set a beginning or ending frame like Turbo does.
I want to try this more, but it takes quite a bit of time so the trial and error aspect of it is not really feasible if you are working on a legit project compared to messing around.

Even if you're just trying it out, you'll probably be turned off cause I feel like it legit takes up to like 10 minutes for 1 Gen-4 generation to complete... then you can see what it produced and it looks like shyt or has some weird shyt happening in the video.
 
Top