At this point its still easier n faster to sketch than to write and modify all those descriptions till u find something that is only KIND OF what u imagined
I agree, you can sketch faster to get your point across and would work closer to your script, but there are some benefits this way, too. As well as for those who can't draw, it's a useful tool. And that's what it is, just another tool. Thanks for watching and commenting.
Our team is just 2 people, none of us can draw. We are using AI to develop an MVP and hopefully get funding to hire real artists. As of today, we can't afford them.
@@museitorloseit there is no benefit to work with midjourney. the sketch looks "great, but there is no cinematographic language. it sux as hell for storyboard
@@blork4211I agree, but AI is changing. Midjourney just got into version 6 and a new Niji 6 mode, too, that looks fantastic. It all depends on how you use the ending prompts or ending tools such as --weird, --stylize, etc. But, you're right, there is no cinematographic language to it. Drawing it by hand would be your best option, even if it's just stick figures. Incidentally, you can upload your sketches to Midjourney and use the "image weights" tool to get your sketch to look "better." I tried it with one of my old storyboards from years ago and was awestruck at how better it appeared. I may do a tutorial on that soon. Remember, you can manipulate the render with a cinematic look, too. Thank you for watching and commenting.
This is interesting BUT storyboards should use certain rules of editing - directions on screen, wide after wide or narrow after narrow... "blocking of a scene... this is the most creative and important part of filming - this is just pretty drawing. (excuse my humble opinion)
I definitely agree. AI struggles a lot and I prefer the old method. You're more precise and like you said, blocking, scene directions, etc. Thanks for watching and commenting.
You're welcome! Exactly, these tools are just that, to help the creator with a faster process and to help them over an obstacle they're in. You can definitely create a comic strip with it. I've seen some people already creating a comic book or graphic novel. :) Check out my previous videos on creating character concepts for comic books. Thank you for watching and commenting. Have fun creating!
I started out with $10 a month. That's the base plan, I believe, and found that I needed to upgrade to the next plan over for unlimited relax hours. I believe that is either $20 or $30 a month. I forgot the pricing. I don't use up the fast hours that much, but I get 15 hours' worth if I need it. So, you get unlimited text prompting with relax mode and 15 hours with fast mode, which you can switch on or off at any time.
I have been debating if this is possible for weeks. Thank you! I will likely be using this as a highbyrid approach, getting close to my idea as possible and give the most crucial panels to my story board artist to make something based on what I have. Many people think AI is taking away jobs but it's actually creating more as in the past I might be too scared to risk budget on an artist that MIGHT capture my vision now I can come to that artist with a little more for them to go off eliminating the need for content redrafts which, let's be honest, everyone hates.
Tools like Midjourney are definitely helping out in some aspects and I think it's just going to create more jobs all around. There are going to be a lot of people who gave up on being creative finding new light with tools like these and bringing fantastic things out to the public. Midjourney will help you out in your projects and get most of your vision across, that's for sure. I've used it to get some ideas out for a comic book, but I have an amazing comic book artist that I've hired who's working on it as I type this. I hope to publish it in the next few months. Look out for it :)
For Filmmakers young and old! In this episode of Muse It or Lose It, I'll walk you through a simple and easy way to create storyboards to show your cast and crew or potential producers interested in seeing your movie concept. It's a great way to visualize your film.
@@joeribot7810 You can sign up for an account with Midjourney on Discord and check out their subscriptions. There is a free trial: www.midjourney.com/home
@@museitorloseit thank you so very much I created a story board with Procreate but using this tool can help save me a ton of time ! I’m gonna try now thank you so much for your help joe
That's the issue with AI tools, but with Midjourney you do have the ability to use different methods. A few are: Using the --seed tool when you end your prompt and add any number after it. Example: --seed 1234, or whatever number you'd like to get the same results in your rolls. Another would be to incorporate someone's real face into the image using the image upload and using that image URL. Another way is to specify the character's features. Let's say, a doctor in his 30s, sleep brown hair, blue eyes, long beard and wearing a eyeglasses. Something like that, but using real images may definitely help. You can also use the remix tool and keep remixing it until the features come out. Check out my other videos in my channel where I talk about image uploads, the Blend tool, Remix, and other cool features. Thanks for watching and commenting!
Go for it! Nothing is stopping you, but you :) This is definitely a great tool to help writers come up with stories. Check out one of my first videos where I help writers out of their writer's block and how to start writing. Here's the link: th-cam.com/video/CTmKP09xw5Q/w-d-xo.html Thank you for watching and commenting.
Hi Brian, that's a great question. I would just break it down in parts. It's a hit or miss, so you'll probably end up doing it a few times until the prompts you choose work. Let's say there is a character sitting behind a desk and you want a profile shot of it. You type, "Movie storyboard panel. A side profile of a male character, messy dark hair, disheveled appearance, business suit. Sitting behind a messy desk in a high-rise." Sometimes you won't get it on the first try, so you'll either need to re-roll or use the Remix tool, if you have it on. Also, try playing around with the Niji tool. There is currently the Niji 4 mode and the Niji 5 mode. The Niji 5 mode is VERY cool now. When you use --niji 5, you can then type this after it "--style expressive," "--style scenic," or "--style cute." Each one will give you different renders. Try them out. You can click niji mode on by type /settings or just ending your prompts with --niji 4 or --niji 5. Remember, Niji 5 is the only one that uses the three different styles I've described. Thank you for watching and commenting.
It's a hit or miss with AI tools, but with Midjourney you do have the ability to use different methods. A few are: Using the --seed tool when you end your prompt and add any number after it. Example: --seed 1234, or whatever number you'd like to get the same results in your rolls. Another would be to incorporate someone's real face into the image using the image upload and using that image URL. Another way is to specify the character's features. Let's say, a doctor in his 30s, sleep brown hair, blue eyes, long beard and wearing a eyeglasses. Something like that, but using real images may definitely help. You can also use the remix tool and keep remixing it until the features come out. Check out my other videos in my channel where I talk about image uploads, the Blend tool, Remix, and other cool features. Thanks for watching and commenting!
I used storyboard sketches in my films for shot type, composition, framing, and camera angle. My sketches never had detailed faces. My storyboards communicate with my DP and team. So for me (and my films), no character consistency is needed.
Hi, thanks for watching and commenting! Well, it's kind of tricky with AI, but it is possible to try to get the same look of the character using some tricks. One method is describing your character. Does he/she have curly hair, what color. Do they wear a leather jacket or a dress? Chubby face, squinting eyes? Etc... But, like I've mentioned, it's tricky with AI since it follows the text prompts. You can end your text prompts with a "seed" and a number, that way, your panels will have the same "style." Say for instance you write this: "A storyboard panel of a man with a leather jacket, and greasy slicked back hair, running away from the cops down a sidewalk, --seed 12345." You can use whatever seed number you want... it's a hit or miss. Another method to try to get the same face or head is by using a real image and uploading it to Midjourney and then right-clicking on the image to copy the image URL address. Then you can start using that Image URL with your character description and put him or her in certain situations. It's always best to get a good amount of your image, like left profile, right profile, forward-facing, down angle, up angle, tilting, etc, to get a better generation that will help Midjourney recreate that character. You basically, upload all of those images and copy every URL from those images. So, let's say you have 5 image URLs, you can add all of them in the beginning of your prompt and then write your character description, etc. And you can combine it with --seed 12345, too, if you wish. Try them it, one with and one without the seed. It doesn't always work, but you can always try. Now, if you want to change your characters looks or scenery, it's always best to leave the "Remix" tool on. You can turn it on by going to /settings and clicking on Remix. Midjourney is a hit or miss type of deal, and it's always a good bet to just keep messing around with it and re-rolling until you like a certain panel sequence. Check out my other videos to see how I use --chaos or --seed tools, as well as image URLs to help you. I hope that this helped.
Hi Joe, you need to sign up to Midjourney to be able to use these AI tools. The link to Midjourney is www.midjourney.com/home It is subscription-based, so you'll need to subscribe to one of their plans.
Ok but.. what if I want full control over what my coverage is going to be? Surely the prompts can understand 'Medium Shot of...' for example? If it can't do that, then it's a very anti-filmmaking tool.
While Midjourney devs are continuously updating and listening to the users about their AI program, you can actually write that in. You can write "Medium Shot," "Wide Shot," "Bird's Eye View," etc. Remember, if it doesn't come out the first try, try rerolling until it comes out. Also, I made a new Storyboard video a couple of weeks ago where you can actually sketch something up fast and let Midjourney take control to modify it. There is also a new button called "Vary (Region)" which can also help you modify things within. There are so many buttons and tools with Midjourney, that there are ways to help the creator. Check out my recent uploads. Thank you for watching and commenting!
I would sketch them out too, even if they were stick figures, to be more precise, but you can get some great ideas that you didn't have by prompting. Like creating the scenery and backgrounds and then sketching freehand. Thanks for watching and commenting.
Hi Shiffi66, you can try to use the exact characterization by using "Seed" followed by a number from 0-60000. Seed uses the same algorithm seed number used to create the image Midjourney generates. You can also use Sameseed, but that only works on older versions like V 3. I made a video on that and also the "Chaos" tool. I'll leave a link at the end. Example: "the same character in different poses who has blue eyes and blonde hair with a slightly evil smirk --seed 25474. Another better method would be to use the generated image you like and click on it to enlarge it, then right-click on it and save the image address. Then you use that image address before you write a text prompt and Midjourney will use that image to create your characters. I'm actually working on a video on that. Stay tuned. Then there is the third option you can use with a combination of the seed and/or the image prompts using the "Remix" tool. I'll also leave a link with more information about it. Let me know if these were helpful. Here are the links: Seed or Sameseed th-cam.com/video/GFlgeeCLSkw/w-d-xo.html Image Prompts: th-cam.com/video/2tyY61aVOpI/w-d-xo.html Remix Tool: th-cam.com/video/oPbqWWhtltk/w-d-xo.html
You know that you don't have to limit yourself with a pencil sketch when you can have a scene with the final look in exactly the same amount of time isn't it? :)) The purpose of the pencil for a storyboard is to be flexible for drawing and maximizing the important informations that need to be understood by the team. With midjourney, there is no reason to create a pencil render when you can actually get even more informations in the image. I suggest to try the same technique with photography or movie still. It works too.
Definitely, I've tried all of those and more. I just kept it simple. Check out my other videos :) The new blending tool is crazy cool. Thank you for watching and commenting! I appreciate it.
I don't agree tbh. Storyboarding for production, maybe, but not for a pitch or conceptualising stage. You do want to be able to 'withold' information when pitching concept visuals to a funder/client, though. Sometimes because you havent quite worked out what is needed there and by just filling it in with something you end up creating an understanding or expectation of something that you didn't intend. The sketch approach allows to to convey the essence of something, like a gesture or pose of a character, without having to lock in specifics at that stage like character race, hair length, their height or weight etc. Things like these should be left out unless you have a specific reason so that further creative decisions can be made. So indistinct faces or forms can be really useful.
Hi, Matthew... there is a free trial when you sign up and I believe you get a few couple of tries, but then you'll need to get a subscription. There are three subscriptions. Basic, Plan of $10 per month, Standard Plan of $30 per month, and Pro Plan of $60 per month. If you type "/subscribe" in the Midjourney Bot, you'll get a link to see the subscription options. Basic Plan has limited generations (~200/month), Standard Plan has 15h Fast generations and unlimited relaxed generations in which you can generate as many images as you want but will take a little bit longer to generate, and the Pro Plan has Pro Plan 30h Fast generations and unlimited relaxed generations and stealth image generations so that no one can see your text prompts or images. There are free AI tools out there, too, like OpenAI Dall-E, but I haven't tried them yet. Thank you for watching and commenting.
is there a way to create storyboard in first person so the AI recognizes the same cinematic scene so you move through the space at eye level or do i need to create each independent scene and try to stitch them. The issue is the prompt even with same seed creates alterations. I am trying to look, left and look right in city then move towards a car which is in the distance but it is throwing in too many variables even with this sequential narrative. The storyboard comes close but then when input first person i get a sketch. Have you tried to achieve this yourself?
Great question! Of course, we're dealing with AI here, so, it only understands algorithms. You'll need to "trick" it into what you want. It's still in its infancy, but there are various ways to get 90 percent of your results. Also, the developers are working on releasing version 5 soon, so stay on the look out. You can also try using the new Niji mode to get some amazing results. To get first person point of view, try using that exact phrase, but use it right in the beginning. Example: first person point of view, a movie storyboard panel, running through the streets of a rundown city. Try to be specific but detailed. When describing your characters, make them stand out. Like, a brunette woman in her 30s, wearing a floral dress, runs for her life down the street, movie storyboard panel. Something like that should help. You can also try to incorporate real images to keep a consistent character and manipulate it with each text prompt or through remix mode. Check out my previous or recent videos on using real images in your prompts. I think you'll find them helpful. For the car in the distance situation, try remixing the panels you're working on and also, don't forget to reroll, in case your first generated images didn't come out right. It sometimes needs a few rerolls to get good. I haven't messed around with the storyboards in a while, but I may have to revisit it. Remember, it's a hit or miss and the best solutions are to find the panels you like and retouch them with a program like Photoshop or other free programs like Photoshop, until the AI gets better. Thank you for watching and commenting. I hope these were helpful to you. Check out my most recent uploads, maybe you'll find something helpful.
@@museitorloseit thank you for the detailed reply, I will reply to this later with some specific detail of my findings because I am in deadline trying to problem solve it at the moment for a deadline
@@chriskilbane You're welcome! I hope these work. Remember, Midjourney and tools like these are just that, they're tools and a way to help you out of jams. If you're in a deadline, get the renders that you can and then draw or sketch over them yourself to get your vision. Like, maybe, just create the background you want, and then add your own touchups. You can always find free images online that you can use, as well. It's like photo-bashing. Or render out single individual props and then photo-bash them later. But, I do hope that you find the solutions with the techniques mentioned above :)
Hi, to use Midjourney, you need to subscribe. When you subscribe, just read their agreements per subscription plans and ownership of your AI-generated artwork. They range from $10, $20, and $60 per month. You can create images and use for your videos on social media, but read up on their requirements. Lots of TH-camrs use Midjourney to make their videos and Instagram, etc. There is also an open.ai such as Dall-e to make AI artworks.
Hi, Midjourney is a platform on discord. It generates AI Art with simple text prompts. For more information about Midjourney and to sign up, you need to go to the Midjourney Discord server to be able to use it. Unfortunately, it's not an app or downloadable, however, you can use Discord on your phone and use Midjourney while on the go. This is the website for Midjourney. www.midjourney.com/home
Hi, thank you for watching and commenting. It's an AI tool called Midjourney and available through Discord. Check out my first video on how to sign in and begin using it.
Hi M C. Thank you for watching and commenting. In the video, I do explain what text prompt I used to make the storyboards. And now, with the new version of Midjourney 5.1, you get can generate a much better rendering. Here's a great way to start off with a storyboard, "A sketched storyboard panel." Then you can add whatever text you're planning for your storyboard from your script. It's always good to keep the prompts short with the importance in the beginning. Example: "A sketched storyboard panel. A police detective walking down an alleyway." Then press enter. Then wait until it generates and see if it came out the way you want. You can describe the surroundings if you wish or just focus on the character. You can also play around with the camera positions, like "A sketched storyboard panel. A close-up of a police detective walking down an alleyway. Worried look on his face." I hope that this helps. I plan on updating the Storyboards video again to go along with the new version of Midjourney.
@@MC-ou4is For keeping the same character, there are different ways such as describing their looks, or saying, "resembles [insert person here]" you can also end your prompts with --seed 12578 or whatever number you wish. That way, your generations will have that same algorithm. I made a video on --seed. Also, you can incorporate an image of your actor and use that as reference. I've also made videos on using image prompts. Here are the videos: th-cam.com/video/BBFGuqjaKoU/w-d-xo.html, th-cam.com/video/GFlgeeCLSkw/w-d-xo.html, th-cam.com/video/9UCs8WRVGdI/w-d-xo.html, and th-cam.com/video/2tyY61aVOpI/w-d-xo.html are a few ideas.
At this point its still easier n faster to sketch than to write and modify all those descriptions till u find something that is only KIND OF what u imagined
I agree, you can sketch faster to get your point across and would work closer to your script, but there are some benefits this way, too. As well as for those who can't draw, it's a useful tool. And that's what it is, just another tool. Thanks for watching and commenting.
Our team is just 2 people, none of us can draw. We are using AI to develop an MVP and hopefully get funding to hire real artists. As of today, we can't afford them.
True but people who like me who can only do math and program but no art skills has a tool to help me out. I don’t have money to hire artists.
@@museitorloseit there is no benefit to work with midjourney. the sketch looks "great, but there is no cinematographic language. it sux as hell for storyboard
@@blork4211I agree, but AI is changing. Midjourney just got into version 6 and a new Niji 6 mode, too, that looks fantastic. It all depends on how you use the ending prompts or ending tools such as --weird, --stylize, etc. But, you're right, there is no cinematographic language to it. Drawing it by hand would be your best option, even if it's just stick figures. Incidentally, you can upload your sketches to Midjourney and use the "image weights" tool to get your sketch to look "better." I tried it with one of my old storyboards from years ago and was awestruck at how better it appeared. I may do a tutorial on that soon. Remember, you can manipulate the render with a cinematic look, too. Thank you for watching and commenting.
Thank you! I've been looking for a way to create storyboards with AI. Oh my god, thank you!😅
Glad to hear this video was helpful to you. Thank you for watching and commenting.
This is interesting BUT storyboards should use certain rules of editing - directions on screen, wide after wide or narrow after narrow... "blocking of a scene... this is the most creative and important part of filming - this is just pretty drawing. (excuse my humble opinion)
I definitely agree. AI struggles a lot and I prefer the old method. You're more precise and like you said, blocking, scene directions, etc. Thanks for watching and commenting.
Your tutorial was epic in helping make all the difference in my learning this process! Thank you!
You're very welcome! Thank you for watching and commenting!
This was very very helpful!
I'm glad to hear that this video was helpful. Thank you for watching and commenting!
Very useful video; many thanks! With some slight reworking, these storyboarding prompts could probably also be used to create comic strips.
You're welcome! Exactly, these tools are just that, to help the creator with a faster process and to help them over an obstacle they're in. You can definitely create a comic strip with it. I've seen some people already creating a comic book or graphic novel. :) Check out my previous videos on creating character concepts for comic books. Thank you for watching and commenting. Have fun creating!
As a storyboarder, I want to partner up with good directors. And I can use AI.
What subscription level to mid journey would you recommend?
I started out with $10 a month. That's the base plan, I believe, and found that I needed to upgrade to the next plan over for unlimited relax hours. I believe that is either $20 or $30 a month. I forgot the pricing. I don't use up the fast hours that much, but I get 15 hours' worth if I need it. So, you get unlimited text prompting with relax mode and 15 hours with fast mode, which you can switch on or off at any time.
I have been debating if this is possible for weeks. Thank you! I will likely be using this as a highbyrid approach, getting close to my idea as possible and give the most crucial panels to my story board artist to make something based on what I have. Many people think AI is taking away jobs but it's actually creating more as in the past I might be too scared to risk budget on an artist that MIGHT capture my vision now I can come to that artist with a little more for them to go off eliminating the need for content redrafts which, let's be honest, everyone hates.
Tools like Midjourney are definitely helping out in some aspects and I think it's just going to create more jobs all around. There are going to be a lot of people who gave up on being creative finding new light with tools like these and bringing fantastic things out to the public. Midjourney will help you out in your projects and get most of your vision across, that's for sure. I've used it to get some ideas out for a comic book, but I have an amazing comic book artist that I've hired who's working on it as I type this. I hope to publish it in the next few months. Look out for it :)
For Filmmakers young and old! In this episode of Muse It or Lose It, I'll walk you through a simple and easy way to create storyboards to show your cast and crew or potential producers interested in seeing your movie concept. It's a great way to visualize your film.
Could you please tell me how to get this Apt ASAP.
It’s fantastic
@@joeribot7810 You can sign up for an account with Midjourney on Discord and check out their subscriptions. There is a free trial: www.midjourney.com/home
@@museitorloseit thank you so very much I created a story board with Procreate but using this tool can help save me a ton of time ! I’m gonna try now thank you so much for your help joe
@@joeribot7810 You're welcome, Joe. It definitely can help in certain areas, for sure. :) Happy creating!
only thing thats missing is the ability to maintain the same face throughout prompts -- thats when it will be extremely useful for storyboarding
That's the issue with AI tools, but with Midjourney you do have the ability to use different methods. A few are: Using the --seed tool when you end your prompt and add any number after it. Example: --seed 1234, or whatever number you'd like to get the same results in your rolls. Another would be to incorporate someone's real face into the image using the image upload and using that image URL. Another way is to specify the character's features. Let's say, a doctor in his 30s, sleep brown hair, blue eyes, long beard and wearing a eyeglasses. Something like that, but using real images may definitely help. You can also use the remix tool and keep remixing it until the features come out. Check out my other videos in my channel where I talk about image uploads, the Blend tool, Remix, and other cool features. Thanks for watching and commenting!
Awesome tips, makes me want to write my first script!
Go for it! Nothing is stopping you, but you :) This is definitely a great tool to help writers come up with stories. Check out one of my first videos where I help writers out of their writer's block and how to start writing. Here's the link: th-cam.com/video/CTmKP09xw5Q/w-d-xo.html Thank you for watching and commenting.
How do you do it when you have your own script that you have already written?
Hi Brian, that's a great question. I would just break it down in parts. It's a hit or miss, so you'll probably end up doing it a few times until the prompts you choose work. Let's say there is a character sitting behind a desk and you want a profile shot of it. You type, "Movie storyboard panel. A side profile of a male character, messy dark hair, disheveled appearance, business suit. Sitting behind a messy desk in a high-rise." Sometimes you won't get it on the first try, so you'll either need to re-roll or use the Remix tool, if you have it on. Also, try playing around with the Niji tool. There is currently the Niji 4 mode and the Niji 5 mode. The Niji 5 mode is VERY cool now. When you use --niji 5, you can then type this after it "--style expressive," "--style scenic," or "--style cute." Each one will give you different renders. Try them out. You can click niji mode on by type /settings or just ending your prompts with --niji 4 or --niji 5. Remember, Niji 5 is the only one that uses the three different styles I've described. Thank you for watching and commenting.
Great, but what about character consistency through sequences?
It's a hit or miss with AI tools, but with Midjourney you do have the ability to use different methods. A few are: Using the --seed tool when you end your prompt and add any number after it. Example: --seed 1234, or whatever number you'd like to get the same results in your rolls. Another would be to incorporate someone's real face into the image using the image upload and using that image URL. Another way is to specify the character's features. Let's say, a doctor in his 30s, sleep brown hair, blue eyes, long beard and wearing a eyeglasses. Something like that, but using real images may definitely help. You can also use the remix tool and keep remixing it until the features come out. Check out my other videos in my channel where I talk about image uploads, the Blend tool, Remix, and other cool features. Thanks for watching and commenting!
I used storyboard sketches in my films for shot type, composition, framing, and camera angle. My sketches never had detailed faces. My storyboards communicate with my DP and team. So for me (and my films), no character consistency is needed.
How can I make sure the characters look like the same people from panel to panel?
Hi, thanks for watching and commenting! Well, it's kind of tricky with AI, but it is possible to try to get the same look of the character using some tricks. One method is describing your character. Does he/she have curly hair, what color. Do they wear a leather jacket or a dress? Chubby face, squinting eyes? Etc... But, like I've mentioned, it's tricky with AI since it follows the text prompts. You can end your text prompts with a "seed" and a number, that way, your panels will have the same "style." Say for instance you write this: "A storyboard panel of a man with a leather jacket, and greasy slicked back hair, running away from the cops down a sidewalk, --seed 12345." You can use whatever seed number you want... it's a hit or miss. Another method to try to get the same face or head is by using a real image and uploading it to Midjourney and then right-clicking on the image to copy the image URL address. Then you can start using that Image URL with your character description and put him or her in certain situations. It's always best to get a good amount of your image, like left profile, right profile, forward-facing, down angle, up angle, tilting, etc, to get a better generation that will help Midjourney recreate that character. You basically, upload all of those images and copy every URL from those images. So, let's say you have 5 image URLs, you can add all of them in the beginning of your prompt and then write your character description, etc. And you can combine it with --seed 12345, too, if you wish. Try them it, one with and one without the seed. It doesn't always work, but you can always try. Now, if you want to change your characters looks or scenery, it's always best to leave the "Remix" tool on. You can turn it on by going to /settings and clicking on Remix. Midjourney is a hit or miss type of deal, and it's always a good bet to just keep messing around with it and re-rolling until you like a certain panel sequence. Check out my other videos to see how I use --chaos or --seed tools, as well as image URLs to help you. I hope that this helped.
@@museitorloseit It does, thanks!
@@MusicalMarble Awesome! Have fun creating!
Could you please tell me Where I can get this
Apt ? mid journey
script to story Board ?
Hi Joe, you need to sign up to Midjourney to be able to use these AI tools. The link to Midjourney is www.midjourney.com/home It is subscription-based, so you'll need to subscribe to one of their plans.
@@museitorloseit thank you so very much. Also, can I use my iPhone to subscribe?
@@joeribot7810 Yes, you can sign in through the phone. Just follow the steps. They made it easy.
Ok but.. what if I want full control over what my coverage is going to be? Surely the prompts can understand 'Medium Shot of...' for example? If it can't do that, then it's a very anti-filmmaking tool.
While Midjourney devs are continuously updating and listening to the users about their AI program, you can actually write that in. You can write "Medium Shot," "Wide Shot," "Bird's Eye View," etc. Remember, if it doesn't come out the first try, try rerolling until it comes out. Also, I made a new Storyboard video a couple of weeks ago where you can actually sketch something up fast and let Midjourney take control to modify it. There is also a new button called "Vary (Region)" which can also help you modify things within. There are so many buttons and tools with Midjourney, that there are ways to help the creator. Check out my recent uploads. Thank you for watching and commenting!
I would rather just draw it because the idea would get out if my head while im typing the promp , but the renders are nicee
I would sketch them out too, even if they were stick figures, to be more precise, but you can get some great ideas that you didn't have by prompting. Like creating the scenery and backgrounds and then sketching freehand. Thanks for watching and commenting.
can renderings of specific characters be used repetitiously when creating storyboards using Midjourney?
Hi Shiffi66, you can try to use the exact characterization by using "Seed" followed by a number from 0-60000. Seed uses the same algorithm seed number used to create the image Midjourney generates. You can also use Sameseed, but that only works on older versions like V 3. I made a video on that and also the "Chaos" tool. I'll leave a link at the end. Example: "the same character in different poses who has blue eyes and blonde hair with a slightly evil smirk --seed 25474. Another better method would be to use the generated image you like and click on it to enlarge it, then right-click on it and save the image address. Then you use that image address before you write a text prompt and Midjourney will use that image to create your characters. I'm actually working on a video on that. Stay tuned. Then there is the third option you can use with a combination of the seed and/or the image prompts using the "Remix" tool. I'll also leave a link with more information about it. Let me know if these were helpful.
Here are the links:
Seed or Sameseed th-cam.com/video/GFlgeeCLSkw/w-d-xo.html
Image Prompts: th-cam.com/video/2tyY61aVOpI/w-d-xo.html
Remix Tool: th-cam.com/video/oPbqWWhtltk/w-d-xo.html
You know that you don't have to limit yourself with a pencil sketch when you can have a scene with the final look in exactly the same amount of time isn't it? :))
The purpose of the pencil for a storyboard is to be flexible for drawing and maximizing the important informations that need to be understood by the team. With midjourney, there is no reason to create a pencil render when you can actually get even more informations in the image. I suggest to try the same technique with photography or movie still. It works too.
Definitely, I've tried all of those and more. I just kept it simple. Check out my other videos :) The new blending tool is crazy cool. Thank you for watching and commenting! I appreciate it.
I don't agree tbh. Storyboarding for production, maybe, but not for a pitch or conceptualising stage. You do want to be able to 'withold' information when pitching concept visuals to a funder/client, though. Sometimes because you havent quite worked out what is needed there and by just filling it in with something you end up creating an understanding or expectation of something that you didn't intend. The sketch approach allows to to convey the essence of something, like a gesture or pose of a character, without having to lock in specifics at that stage like character race, hair length, their height or weight etc. Things like these should be left out unless you have a specific reason so that further creative decisions can be made. So indistinct faces or forms can be really useful.
is this free or subscription base? if both, what are limitations of free verison.
Hi, Matthew... there is a free trial when you sign up and I believe you get a few couple of tries, but then you'll need to get a subscription. There are three subscriptions. Basic, Plan of $10 per month, Standard Plan of $30 per month, and Pro Plan of $60 per month. If you type "/subscribe" in the Midjourney Bot, you'll get a link to see the subscription options. Basic Plan has limited generations (~200/month), Standard Plan has 15h Fast generations and unlimited relaxed generations in which you can generate as many images as you want but will take a little bit longer to generate, and the Pro Plan has Pro Plan 30h Fast generations and unlimited relaxed generations and stealth image generations so that no one can see your text prompts or images. There are free AI tools out there, too, like OpenAI Dall-E, but I haven't tried them yet. Thank you for watching and commenting.
is there a way to create storyboard in first person so the AI recognizes the same cinematic scene so you move through the space at eye level or do i need to create each independent scene and try to stitch them. The issue is the prompt even with same seed creates alterations. I am trying to look, left and look right in city then move towards a car which is in the distance but it is throwing in too many variables even with this sequential narrative. The storyboard comes close but then when input first person i get a sketch. Have you tried to achieve this yourself?
Great question! Of course, we're dealing with AI here, so, it only understands algorithms. You'll need to "trick" it into what you want. It's still in its infancy, but there are various ways to get 90 percent of your results. Also, the developers are working on releasing version 5 soon, so stay on the look out. You can also try using the new Niji mode to get some amazing results. To get first person point of view, try using that exact phrase, but use it right in the beginning. Example: first person point of view, a movie storyboard panel, running through the streets of a rundown city. Try to be specific but detailed.
When describing your characters, make them stand out. Like, a brunette woman in her 30s, wearing a floral dress, runs for her life down the street, movie storyboard panel. Something like that should help. You can also try to incorporate real images to keep a consistent character and manipulate it with each text prompt or through remix mode. Check out my previous or recent videos on using real images in your prompts. I think you'll find them helpful. For the car in the distance situation, try remixing the panels you're working on and also, don't forget to reroll, in case your first generated images didn't come out right. It sometimes needs a few rerolls to get good.
I haven't messed around with the storyboards in a while, but I may have to revisit it. Remember, it's a hit or miss and the best solutions are to find the panels you like and retouch them with a program like Photoshop or other free programs like Photoshop, until the AI gets better. Thank you for watching and commenting. I hope these were helpful to you. Check out my most recent uploads, maybe you'll find something helpful.
@@museitorloseit thank you for the detailed reply, I will reply to this later with some specific detail of my findings because I am in deadline trying to problem solve it at the moment for a deadline
@@chriskilbane You're welcome! I hope these work. Remember, Midjourney and tools like these are just that, they're tools and a way to help you out of jams. If you're in a deadline, get the renders that you can and then draw or sketch over them yourself to get your vision. Like, maybe, just create the background you want, and then add your own touchups. You can always find free images online that you can use, as well. It's like photo-bashing. Or render out single individual props and then photo-bash them later. But, I do hope that you find the solutions with the techniques mentioned above :)
are they free to use in an a TH-cam video?
Hi, to use Midjourney, you need to subscribe. When you subscribe, just read their agreements per subscription plans and ownership of your AI-generated artwork. They range from $10, $20, and $60 per month. You can create images and use for your videos on social media, but read up on their requirements. Lots of TH-camrs use Midjourney to make their videos and Instagram, etc. There is also an open.ai such as Dall-e to make AI artworks.
Can someone help me to download Midjourney story board
Hi, Midjourney is a platform on discord. It generates AI Art with simple text prompts. For more information about Midjourney and to sign up, you need to go to the Midjourney Discord server to be able to use it. Unfortunately, it's not an app or downloadable, however, you can use Discord on your phone and use Midjourney while on the go. This is the website for Midjourney. www.midjourney.com/home
Which software is this
Hi, thank you for watching and commenting. It's an AI tool called Midjourney and available through Discord. Check out my first video on how to sign in and begin using it.
You don't explain the intro of how you create it,it feels like you upload it,you don't specify it in detail.
Hi M C. Thank you for watching and commenting. In the video, I do explain what text prompt I used to make the storyboards. And now, with the new version of Midjourney 5.1, you get can generate a much better rendering. Here's a great way to start off with a storyboard, "A sketched storyboard panel." Then you can add whatever text you're planning for your storyboard from your script. It's always good to keep the prompts short with the importance in the beginning. Example: "A sketched storyboard panel. A police detective walking down an alleyway." Then press enter. Then wait until it generates and see if it came out the way you want. You can describe the surroundings if you wish or just focus on the character. You can also play around with the camera positions, like "A sketched storyboard panel. A close-up of a police detective walking down an alleyway. Worried look on his face." I hope that this helps. I plan on updating the Storyboards video again to go along with the new version of Midjourney.
Ok...but how can you keep the same character
@@MC-ou4is For keeping the same character, there are different ways such as describing their looks, or saying, "resembles [insert person here]" you can also end your prompts with --seed 12578 or whatever number you wish. That way, your generations will have that same algorithm. I made a video on --seed. Also, you can incorporate an image of your actor and use that as reference. I've also made videos on using image prompts. Here are the videos: th-cam.com/video/BBFGuqjaKoU/w-d-xo.html, th-cam.com/video/GFlgeeCLSkw/w-d-xo.html, th-cam.com/video/9UCs8WRVGdI/w-d-xo.html, and th-cam.com/video/2tyY61aVOpI/w-d-xo.html are a few ideas.
Thanks for you help mate...
@@MC-ou4is You're welcome :)