I'm so done with my Midjourney Pro subscription for numerous reasons. I've barely even used MJ since Dall-E 3 debuted........ and now that Open AI dropped word that they are developing a Dall-E 3 dedicated API, that's even further reason for me to abandon MJ. Additionally, the MJ devs are wasting enormous amounts of our subscription $$$ on focusing on things I don't want or care very little about, especially David's heavy focus on integration of community interaction functionality into the MJ API. I don't need that because I already have many options for community interaction on Discord and numerous other social interaction platforms.
Helpful, thank you. I hope your daughter liked the character that you created. I suppose the cartoon character is going to look different because Midjourney and ChatGPT Dalle 3 pull images from different places online (source images). I'm leaning towards Midjourney - only because whenever I find art that I like, the product description always mentions that the seller used Midjourney to create the art - so I think Midjourney will be more like my style. I found your video very helpful in showing me around ChatGPT Dalle 3 - it does look a more streamlined system :) x
MidJourney is great at making pretty images and can really dial in when you want it to. I think if you have no clue how to prompt or need something in recent media then DALLE is great. But Midjourneys images do come out more precise where DALLE looks very soft and simplistic. The new tuner feature on MJ is a game changer too but DALLE definitely has the upper hand on text. I think each have their strengths. I prefer MJ because text is so easy to photoshop anyway that AI is like barely fast honestly. Plus MJ has 4X upscaling, variations, aspect ratios etc. DALLE feels more like a “can do anything but not well” in its current state. Great video
I'm just curious about celebrities though. Can midjourney do most celebrities or are they banned? I cant do every celebrity i want to do with my current setup. Dall e 3 blocks the majority of celebrity pictures.
the big problems i've had with Dalle compared to midjourney are that Dalle likes to make mockups alot and it also cuts off parts of the images when I want them centered for a tshirt, even when I specify not to do these things. I'm sure this will be amended in the future but i've already made the switch from midjourney for alot of other reason.
I think Dalle3 is fine for artistic or stylized output....the text is impressive, when it works...for complex, long prompts D3 is better and MidJourney skips over long prompts...... but for photo look D3 is really inconsistent and I prefer a photojournalistic look for most of my work. Plus MJ is much less picky about styles and forbidden words. Both are tools...both have strengths...it's not really an either/or for a full toolset.
Don’t forget about tuning styles not, aspect ratio, 4X upscaling, and variations. But I like DALLE’s image referencing. I wonder how DALLE treats photobashing
Bing removed - I’ve been hearing this from others. Next Monday 6th Nov, they have developer conference. Hopefully they release “all tools” for everyone and restore bing.
I got the the error when i tried the same prompt: I'm unable to generate images based on your request due to content policy restrictions. If you'd like to request another image or modify your prompt, please let me know!
Hi, I use Bing, 'image creator' as it is free and uses Dall-e 3. I get great results but it is very restricted when it comes to content, no fight scenes, no gore, no blood etc. I totally get why Microsoft want that as kids will be using it but I would like a more adult version. I simply wanted an image of a woodsman who had fought off a pack of wolves leaving some of them dead and himself injured.. Not allowed. So yeah, our creativity is now decided by someone else, not a good thing, what if I want to write about history and make depictions of actual events, someone else now decided what I can and can not show. A lot of middle eastern oil money is going into AI, they might have a very different ideology than westerns values.. Can you understand where it might lead?
I already quit Leonardo since their business model is so messed up. I needed a few more credits for a project after I'd run out. I upgraded my account a few days before the end of the month and they DISAPPEARED instead of cashing at the end of the month. I only got to use a few of them from the upgrade before they vanished without warning! There was no way for me to buy any credits alacarte on top of my membership tier! So they don't offer ANY flexibility. I doubled the cost of my credits for the month because I ran out before the month was up. It's not like we can plan ahead how many generations it will take to get the final result we need. So even if you have a set project, it's impossible to predict how many credits you'll need. I tried to reach them through chat several times but they never got back to me as promised. I don't want to have to learn a different AI system, but I'll go back to Dall-e (I was a beta-tester) now that I'm seeing how much better it's become
Agree Dalle 3 is better especially if you need wording in the actual pictures. For example: If you wanted a pic showing signs with actual wording, or hats with wording, Midjourney doesnt seem to know how to do that, and always comes out to look like a foreign language. However, I cannot get Dalle 3 to create anything based on a copywrited material, like Darth Vader, or comic characters. It can only make an interpretation of those things; whereas Midjourney doesnt seem to have any issues creating images of copywrited character designs.
Ended my Midjourney subscription weeks ago. It’s not consistent, copy paste generates same style even if you changed the prompt and seeds. ChatGPT/Dalle cooks with the gen_id.
If you're trying to make comic books, I think Dall E 3 is the best. The stuff I can do with Bart Simpson in Dall E 3 is magnificent. I haven't used Midjourney, but can it do DC characters like Batgirl, and can it do Bart Simpson comics? Does Midjourney struggle with adding two or three characters in a picture? Can it do Bart Simpson with Sylvester Stallone as Rambo, but also with E.T. in it? Dall E 3 can do that. Can Midjourney? I haven't used it. Can Midjourney do Will Smith slapping Chris Rock or not really. Thanks for any help.
Have you tried Dalle 3 yet?
how did you drop in your image into the dall-e cht section? i wasnt able to
@@raycon284 I have the new "all tools" version. I think they'll be rolling it out to more users next week.
@@daraghmwalsh thanks! love your content
I'm so done with my Midjourney Pro subscription for numerous reasons. I've barely even used MJ since Dall-E 3 debuted........ and now that Open AI dropped word that they are developing a Dall-E 3 dedicated API, that's even further reason for me to abandon MJ.
Additionally, the MJ devs are wasting enormous amounts of our subscription $$$ on focusing on things I don't want or care very little about, especially David's heavy focus on integration of community interaction functionality into the MJ API.
I don't need that because I already have many options for community interaction on Discord and numerous other social interaction platforms.
Helpful, thank you. I hope your daughter liked the character that you created. I suppose the cartoon character is going to look different because Midjourney and ChatGPT Dalle 3 pull images from different places online (source images).
I'm leaning towards Midjourney - only because whenever I find art that I like, the product description always mentions that the seller used Midjourney to create the art - so I think Midjourney will be more like my style.
I found your video very helpful in showing me around ChatGPT Dalle 3 - it does look a more streamlined system :) x
MidJourney is great at making pretty images and can really dial in when you want it to. I think if you have no clue how to prompt or need something in recent media then DALLE is great. But Midjourneys images do come out more precise where DALLE looks very soft and simplistic. The new tuner feature on MJ is a game changer too but DALLE definitely has the upper hand on text. I think each have their strengths. I prefer MJ because text is so easy to photoshop anyway that AI is like barely fast honestly. Plus MJ has 4X upscaling, variations, aspect ratios etc. DALLE feels more like a “can do anything but not well” in its current state. Great video
I'm just curious about celebrities though. Can midjourney do most celebrities or are they banned? I cant do every celebrity i want to do with my current setup. Dall e 3 blocks the majority of celebrity pictures.
Hi guys, is it possible to achieve consistent characters in different actions and conditions with DALL-E 3 via API? Thanks in advance :)
That is an interesting question! I don't know but please let us know if you find out.
@@daraghmwalsh Hopefully it is!
the big problems i've had with Dalle compared to midjourney are that Dalle likes to make mockups alot and it also cuts off parts of the images when I want them centered for a tshirt, even when I specify not to do these things. I'm sure this will be amended in the future but i've already made the switch from midjourney for alot of other reason.
interesting! thanks for sharing
I think Dalle3 is fine for artistic or stylized output....the text is impressive, when it works...for complex, long prompts D3 is better and MidJourney skips over long prompts...... but for photo look D3 is really inconsistent and I prefer a photojournalistic look for most of my work. Plus MJ is much less picky about styles and forbidden words. Both are tools...both have strengths...it's not really an either/or for a full toolset.
Don’t forget about tuning styles not, aspect ratio, 4X upscaling, and variations. But I like DALLE’s image referencing. I wonder how DALLE treats photobashing
I didn't get get gpt 4 all tools update. Also gpt4 browse option is just disappeared. How can i get this again?
Bing removed - I’ve been hearing this from others. Next Monday 6th Nov, they have developer conference. Hopefully they release “all tools” for everyone and restore bing.
I got the the error when i tried the same prompt: I'm unable to generate images based on your request due to content policy restrictions. If you'd like to request another image or modify your prompt, please let me know!
Yes my friend I am on the same page with you about Midjourney. Just to much going on there for me.
Hi, I use Bing, 'image creator' as it is free and uses Dall-e 3. I get great results but it is very restricted when it comes to content, no fight scenes, no gore, no blood etc. I totally get why Microsoft want that as kids will be using it but I would like a more adult version. I simply wanted an image of a woodsman who had fought off a pack of wolves leaving some of them dead and himself injured.. Not allowed.
So yeah, our creativity is now decided by someone else, not a good thing, what if I want to write about history and make depictions of actual events, someone else now decided what I can and can not show. A lot of middle eastern oil money is going into AI, they might have a very different ideology than westerns values.. Can you understand where it might lead?
I completely agree with you. I am also considering ending my Leonardo subscription.
Interesting to hear!
I already quit Leonardo since their business model is so messed up.
I needed a few more credits for a project after I'd run out. I upgraded my account a few days before the end of the month and they DISAPPEARED instead of cashing at the end of the month. I only got to use a few of them from the upgrade before they vanished without warning! There was no way for me to buy any credits alacarte on top of my membership tier! So they don't offer ANY flexibility. I doubled the cost of my credits for the month because I ran out before the month was up.
It's not like we can plan ahead how many generations it will take to get the final result we need. So even if you have a set project, it's impossible to predict how many credits you'll need.
I tried to reach them through chat several times but they never got back to me as promised. I don't want to have to learn a different AI system, but I'll go back to Dall-e (I was a beta-tester) now that I'm seeing how much better it's become
Agree Dalle 3 is better especially if you need wording in the actual pictures. For example: If you wanted a pic showing signs with actual wording, or hats with wording, Midjourney doesnt seem to know how to do that, and always comes out to look like a foreign language. However, I cannot get Dalle 3 to create anything based on a copywrited material, like Darth Vader, or comic characters. It can only make an interpretation of those things; whereas Midjourney doesnt seem to have any issues creating images of copywrited character designs.
Thanks
Welcome
The ChatGPT interface is great and I made some great images with ChatGPT + Dal-E 3. However Midjourney makes better images overall
I love this
Thanks Manef!
Ended my Midjourney subscription weeks ago. It’s not consistent, copy paste generates same style even if you changed the prompt and seeds. ChatGPT/Dalle cooks with the gen_id.
gen id is amazing!
how do you use the `gen id?
@@Oxes Create a prompt and ask for the gen_id
You didn't convince me Dalle 3 is better than Midjourney, none of your reasons did. Midjourney is still the best text to image app.
Nope
If you're trying to make comic books, I think Dall E 3 is the best. The stuff I can do with Bart Simpson in Dall E 3 is magnificent. I haven't used Midjourney, but can it do DC characters like Batgirl, and can it do Bart Simpson comics? Does Midjourney struggle with adding two or three characters in a picture? Can it do Bart Simpson with Sylvester Stallone as Rambo, but also with E.T. in it? Dall E 3 can do that. Can Midjourney? I haven't used it. Can Midjourney do Will Smith slapping Chris Rock or not really. Thanks for any help.
nice
Chatgpt: LCD = Happiness
tldr: Main reasons I can't use Midjourney: I'm a tech noob.
All of your complaints about Midjourney come down to you not understanding how to use it. And to be frank, I don't think that's the fault of the tool.
It looks like it's not the dalle3 that is better, rather then your prompting in midjourney is weak
Exactly. You shouldn't need to learn prompt engineering to make Ai images.
the cap on gpt 4 is just kinda unreasonable tbh, kinda prevents me from subscribing.
yes a bit annoying. hopefully they will raise when platform stabilises!
why do you speak so slowly? i had to watch this one in 1.25X