Mid journey feels clunky and messy compared to what Leonardo has become…it is a pleasure to use and has improved steadily.

I've been watching lots of videos about how to generate consistent characters and you are the ONLY ONE that explain it with high level of detail, in the other videos they skip really important details that really makes the difference. Thanks to this video, now, I can say that I'm capable to create consistent images. Thank you very much!

Wonderful informative and well scripted KZbin content

This is really something. I want to learn more about this with my writing to add videos.

the only way I see this becoming a real tool is if they come up with a better interface and standalone app with checkboxes rather than the prompting chaos you just went through.

These do NOT qualify as "videos". They are fancy 2-4 second .gifs. Nice, but utterly useless as far as needing actual videos.

It's beginning let's see competitors

Alpha is nice, but there are important features on Discord that are not available in Alpha. I find the /shorten, /tuner, /describe features in the Discord version more efficient.

Haiper make more like gift instead of video lol, 2 seconds cant be called video, it will be good when the time has no that crap time limit, actually is completely useless

Very useful video and well presented. I've used several of these platforms. I think Playground presents the best images and is fun to use. (They've recently cut the daily use to 100 per day BTW). Ideogram is best at following prompt instructions. Way ahead of others. I've been with Ideogram since it started and I am more impressed with its images and fidelity to the prompts is ojutstanding. I also use Leonardo but don't like it too well. It's not bad but the images seem rather bland and not very striking. I'm 83 years old and interested in art, not photorealistic photography, so ease of use is important to me. Some of the platforms simply won't do what you want. Earlier today, I saw a video on Pulp magazine covers which are usually racy. I decided to make one of my own. My prompt was : "Pretty terrified girl with torn clothing cowering before a knife wielding man." Playground and Leonardo came up repeatedy with images of the Girl dressed in tight fitting clothes and wielding the knife. Only Ideogram gave me images of the man holding the knife. It was very irritating. Apparently some platforms don't want to portray women as victims but rather as Mixed Martial Arts Super Ninja's. Some platforms won't allow Hitler or Stalin as characters, but will allow Mao Tse Tung or Kim Jong-Un. Weird ! Also, platforms that do allow Hitler or Stalin change their faces a little bit. Why ? Same thing with Na;poleon. All three men have easily recognizable faces, but accurate portrayals are nearly impossible to create using AI. Also, swith Napoleon, he is usually given Red (British) or Green (Austrian) Uniforms and not A Blue French Uniform. I think you should include Night Cafe, Starry AI and Yodayo . Yodayo is an anime site and unlike the others pretty much uncensored. While I complain about the censorship on some sites, there is a lot of material I don't care for. on Yodayo. I don't like the young anime female characters that look to be about 12 years old engaging in pretty raunchy acts. That said, it follows prompt instructions pretty well, is good at making Collages and can produce some very impressive images that no one would object to. All three have a good free plan. Again, I liked your review and subscribed.

Waste of time. How many of us want to make a 2 second video?

2 seconds are not enough

interresting...

Wait patiently for the next few months. I think it won't be too exploitative.

Very expensive 😢

The 4 seconds video option is available now

Nice.

2 second look like animate image

the length of the videos is truly ridiculous. No make sense 2 seconds, come on! Need at least 10

thank-you❤❤

I am excited about Sora but will definately try Haiper. Thanks for the video.

Microsoft designer does not have a daily cap in the way that you are trying to portray it or create the narrative. No, it has a cap on the amount of fast generation one can make. And that resets daily, yes. But you can generate as much as you would like all as much as you would like. It just is going to take a little bit more time after your tokens are used for the day.

The problem with your way that you did the test when it comes to gpt is that you didn't recognize that your prompt was being changed by GPT and then reprompted in a way that gave you the images that you've shown on your screen. It's not a true representation of what Dolly 3 is doing, but through GPT and you didn't even show what it changed your prompt too.

You can actually bring the posters into Canva and it will find all text layers and recreate w live text in fonts as close as can and you can correct all spelling. Amazing combination of the two platform for insane logos and text promos

Great!!!

Game changer + SREF

Great work, keep posting this kind of content, you can get many followers.

Thank you for this, you explained it simply. What about character reference of two or more characters in one imagery? And is product photography something you also do in midjourney?

Just another, greedy company that steal what bilions of people created with their energy and time. We can pack it in any nice words we can find and think of, we can change the meanings of words even but that won't change the fact.

Love your videos! Found you yesterday and I have subscribed. Thank you and keep them coming!

After a century of directors, actors, actresses, cinematographers, etc, dominating the limelight, now the dominant role of storytelling is going back to the writers / scriptwriters. A welcomed change if I may say.

close to 3,000 but not 5,000 is its required now

Wished they opened it up for everyone

is it still working very well, it needs to be updated ?

I just made over a thousand images, far away.

Thanks dude 👍

How did you get access to it?

Great video, mate! 💪🎨🖌

keep the good work going

Great explaination!

If you are almost but not quite getting the result you want because the art style is changing your image a bit too much, or the if the art style is not coming through strongly enough, you can weight both the image and the text prompt with double colons. For example: adding ::1.5 after your image link would add a bit more emphasis to it. Just adding :: is the same as adding ::1

Thank you, I love your video.

Your tutorial was fantastic! The use of side-by-side visuals and enlarged prompts made the presentation highly effective and helpful. Thank you!

I’m in! I like your content!

How do I access the image URL? You copied it so fast in your video that I could not follow even though I replayed over and over. Love you and what you're doing here. Thanks in advance

Good job!

Great tutorial and primer in lighting (midjourney or not). Thanks.

One good tip is to fix seed as you test chaos or any other changes

You sir have the best description and to the point tips I have seen thus far!

I noticed when you switched from street style photo to street portrait, the distance and angle changed a lot. Portraits were all head and shoulders looking at camera while street style photo showed more of their torso and they looked in different directions. Would be nice if you could change a single parameter at a time so we could see the immediate effects. Thanks for the vid!