🤖🖼️DALL·E / DALL·E 2 Archives - Good Shepherd News - Fastest Growing Religious, Free Speech & Political Content

Midjourney V5 is Out Now – Next Steps in Photorealistic Experience with AI Art

The Truth News — Tue, 17 Oct 2023 09:31:48 +0000

Midjourney V5 is Out Now – Next Steps in Photorealistic Experience with AI Art

Arecent breakthrough in AI, you might have missed: the highly awaited Midjourney V5 is out now. The independent research lab has just released their latest version of the famous AI art generator. Some already call it “a world of photorealistic wonder” in terms of creating breathtaking images from text prompts. Wonder or not, the newly trained model promises significant improvements in language understanding, accuracy, and stylistic flexibility. Let’s try it out together and see what this update is capable of.

V5 is the second deep-learning model from Midjourney and has been in the works for the past five months. It claims to use completely different neural architecture and new aesthetic techniques compared to its predecessor. As developers put it: “You might hear it characterized as newly trained, bigger-brained, that it knows more, understands more, or listens better. All these things are true of V5.“ Of course, we had to try for ourselves. And lo-and-behold, this release does create wonders, even if it is still just an alpha test.

Midjourney V5 is out now – but how to use it?

In our previous article, we touched on Midjourney and how it could be helpful for creating artistic mood boards with AI. You will also find thorough instructions for starting out and using this generative neural network to make art. In case you haven’t tried Midjourney yet, I suggest heading over to that article first before continuing to read.

How do you combine filmmaking with traveling? How do you make sure you can come home in one piece? And how do you turn it into your job? Join Russ Malkin on Adventure Filmmaking.

Assuming you have read it, the question may arise: how can you add the new V5 model to your workflow? There are two ways of doing it: you can either type “/settings” in any #newbies Discord channel and then select the „MJ version 5“. Or – if you want to try it out first – just add the parameter “– v 5” at the end of your prompt.

Screenshot from the Midjourney Bot on Discord, highlighting the correct use of the “v 5” parameter. Image credit: Mascha Deikova / CineD

Improved coherency, especially involving faces and limbs

One of the biggest changes the Midjourney community was eagerly waiting for was no more six-fingered people. In the new model, you will notice fewer chaos blobs, as it possesses significantly more knowledge on how to handle limbs. Also, V5 should be better at generating faces and eyes down to the smallest details. We ran a few portrait tests below, and indeed, they look truly amazing.

Some experiments with realistic portraits. Image credits: created with Midjourney V5 by CineD

These are not real people! Insane, right? Yet, not flawless. For instance, in the image with a model making a selfie in a mirror, note some serious problems with reflections. The hands of the playing boys are also not perfect, but it’s like chalk and cheese compared to Midjourney’s previous deep-learning model. And the new art generator really can count to five while creating fingers.

Photorealism as the major improvement

As you see above, the V5 model can offer much more realistic imagery in general. In fact, if you don’t specify an art style, the name of your reference artist, or a media source, you will get the default system setting, and the result will look completely photorealistic. The prompts in the new release are designed to look like photos, so, if you need something more abstract or artistic, we recommend adjusting your wording.

Another possibility in fine-tuning how Midjourney reads your text description is to add the parameter “stylize N” to the end of the prompt (where N varies from 0 (min) to 1000 (max)). The lower the value, the less AI’s house style will influence the final picture. Meaning, if you need to generate a Vincent van Gogh-like image, try putting “– stylize 1000” as I did in the test below.

The result of requesting an image in the style of Vincent van Gogh. Image credit: created with Midjourney V5 by CineD

Looking at our experiments and also at the generated pictures from other members on the Midjourney community post on the showcase channel, I can’t help but notice: they are incredible. And very scary. Will there be a moment when no one can tell the difference between a real photo and one created by artificial intelligence? I guess only time will tell.

Midjourney V5 is out now: higher detailization and improved performance with prompting

Worrying thoughts aside, another claimed feature of the new Midjourney model is its enhanced integrity. The details of the generated art are more likely to turn out correctly, and also, the V5 tends to be very responsive to your inputs. Creators call this version “much more unopinionated” than V3 and V4. For a fair comparison, I took an old prompt of mine which I used for the last Midjourney-related article and fed it to the latest release. See for yourself:

On the left: the old output from V4. On the right: the result of the same prompt in the new V5. Image credits: created with Midjourney by CineD

It seems that Midjourney can now produce more controllable and predicted results. At the same time, this improvement means that it may be harder to use. Short prompts may no longer work as well as before. In the V5, developers suggest writing sentences rather than lists and using more explicit text describing what you want. Think of it like the very short essay assignment we all had to compose in school from time to time.

Higher resolution and other aspect ratios

As promised, the 5th version of Midjourney offers a two-time resolution increase and can now produce images up to 1024×1024. What you should note is that the “upscale” button works differently in this model. All images you get with V5 are at their maximum quality, so this button just separates your pre-upscaled image from the grid. Eventually, there should be the possibility of generating 2048×2048 pictures.

A further interesting update includes broader aspect ratio options. Previously, the custom ones were not supported at all. In the new model, you can create much wider or much narrower canvas sizes and try any aspect ratio you like. Oh, that is exactly what I was hoping for on the mood boards—now also available in Cinemascope!

An example of a picture generated in Cinemascope by adding “–ar 21:9” to the prompt. Image credit: created with Midjourney V5 by CineD

Image weights are back

And last but not least, I want to talk about the image weight feature which is back in the newest version of Midjourney. If you use a photo or a picture as a reference, you can now increase its leverage using the parameter “iw N”, where N ranges from 0.5 to 2.0. Let me show you how it works in the photo example of Jamie Lee Curtis receiving her Oscar for Best Supporting role last week. I uploaded her picture to the Midjourney bot and asked it to imagine her as a beautiful alien from another planet. I also put a “– iw 1.0” at the end so that we can still recognize the actress. This is one of the results the AI came up with:

A generated picture based on the reference photo of Jamie Lee Curtis. Photo reference source: A.M.P.A.S., 2023. Created with Midjourney V5 by CineD

Important reminder: if you use the created image for any purposes other than mood boards or idea development (and especially for commercial ones), please make sure you have the rights for your reference picture, as well as don’t forget about the correct attribution.

Midjourney V5 is out now: what are the main issues?

Users of previous versions of Midjourney will have to learn how to prompt again. I mentioned it before, but I think it’s an important thing to bear in mind. There is a lot of experimenting going on in Discord right now, as community members try to figure out how to phrase their requests to the V5 model as correctly as possible.
Some users note that V5 is a bit slower than V4, which shouldn’t be a surprise since there is also increased quality and detailization.
And yes, it’s still fresh, and the hype is tremendous. So beware that the servers can be overloaded. It may help to get a paid subscription that allows you to chat with the Midjourney bot directly. Or just be patient – it’s worth it.

Conclusion

Midjourney V5 is out now, and we are already blown away by some of what we’ve seen, even though, as the developers note, this is just an alpha test and things will change in the final release. They say this exact model might not be available in the future, but seeing what it’s already capable of might mean the future has come a bit earlier than expected.

What are your thoughts? Have you had a chance to try out V5? How do you like its photorealism? Do you also occasionally get this worrying feeling that AI tools are getting too good? We’re open to discussion in the comments section below. source

Midjourney Topics:

What Is Midjourney?

Text To Image with AI Art Generator

Easily turn your ideas into stunning visuals using the Imagine AI art generator. Just enter your text, pick a style, and see your imagination come alive!

Play with the AI below https://www.midjourneyai.ai/

Join their Discord Server below https://discord.com/invite/midjourney

How to get consistently great results with Midjourney

The Truth News — Mon, 16 Oct 2023 09:26:31 +0000

How to get consistently great results with Midjourney

Ever wondered how some people get much better results than your own with AI image generators? Here’s some tips for obtaining photorealistic results with Midjourney.

I’m going to say right off the bat, let’s park the debate about whether AI image generators are the death of art or photography, or whether they should exist or not. The fact is, that if I want an image that I can’t get anywhere else, an AI image generator can often give me what a stock photo site cannot. It’s a tool that you can choose to use or not, and I’m going to leave the debate there.

Instead I want to focus on how to get results from one of the most popular, and capable AI image generators around; Midjourney.

Midjourney has built a bit of a reputation for some of the stunning images, both photoreal and stylised, that it can produce. I’m going to focus here on results that look like they were taken with a camera or as a still image from a movie. So here’s a few guidelines and some prompts you can use to help improve the results you’re getting if you’re new to this.

Midjourney works through Discord, so you’ll need to create a Discord account first of all. You can then use the system via a web browser or by downloading the Discord app. When you sign up for Midjourney you’ll get a certain number of ‘processing minutes’ for free. After that you’ll need to choose a subscription tier. Note that unless you sign up for the Pro account tier, all your image generations will appear on the public message board, even if you prompt with a private message to the Midjourney bot. Only Pro accounts have the ability to use Stealth mode. This is, of course, a good thing because it prevents the dodgier uses of the system.

Tip 1. Keep your descriptions short and ordered

The first thing to know about Midjourney is that it places less importance on the later things you put into your prompt versus the beginning. What this means in practice is that if you absolutely need something in the image, it needs to be put near the beginning or middle of the prompt. It’s no guarantee, but it is more effective that way.

Furthermore, you don’t need to be too wordy. The shorter and to the point the prompt is, the better the results generally. At least this has been my own experience when using it.

Tip 2: Learn lighting and lens descriptions

One of the primary ways to get realistic results from Midjourney is by using lighting and lens descriptions. Midjourney will understand prompts that include the mention of specific cameras, such as a Sony A7S or Canon R6 etc, as well as specific lenses and focal lengths. It will also respond to film stock specifications. For example, specifying Ilford film stock will give you a black and white image.

Using phrases like “high key” or “low key” as well as “high contrast” and “low contrast” can also help you to get the type of result you want. However, some of the most powerful keywords to use in a prompt comes with the use of “global illumination”, “natural light”, “uplight”. Using the phrase “uplight” can give a softer look to the image.

Below is an example of a plate of food using some of those descriptions. The prompt I used was “Close-up of a plate of food consisting of sea bass fillets sitting on a bed of broccoli with a few capers. Natural light, global illumination, uplight f/1.8”.

You can see that it still doesn’t look very realistic, so what can we do to improve that?

Tip 3: Give Midjourney a style to use

One of the most powerful phrases to use in Midjourney is “film still” and “in the style of” or “style by”. Let’s take the prompt used above and add a style description. Both these tell Midjourney that the image you want could be a still taken from a TV programme or movie, while the style description lets it know a programme or film style reference. Let’s add “Film still in the style of a 2020s cooking programme” to the first prompt, making the full prompt “Film still in the style of a 2020s cooking programme, close up of a plate of food consisting of sea bass fillets sitting on a bed of broccoli with a few capers. Natural light, global illumination, uplight f/1.8”

Hmm, it’s slightly better, but it’s still not something you’d describe as ‘photoreal’. Now, we could specify the phase “photograph” or “photoreal”, and even “extreme detail. However there are a few other parameters we can use instead, or in addition to these.

Firstly, let’s stop Midjourney generating a square image. For this I’m going to use the “–ar” instruction. So, if I type “–ar 16:9” into my image description prompt, Midjourney will output an image in the specified aspect ratio. I’m also going to force it to use the highest quality by using the “–q 2” instruction. This uses double the amount of computational credits, but it gives a hugely improved result. Additionally I’m going to add in a seed command to add in some variation to the generations, for example “–seed 3000”. Lastly, I’m going to force Midjourney to version 5.

The result is this prompt and the image below, “Film still in the style of a 2020s cooking programme, close up of a plate of food consisting of sea bass fillets sitting on a bed of broccoli with a few capers. Natural light, global illumination, uplight f/1.8 –ar 16:9 –seed 3000 –q 2 –v 5”.

I think you’ll agree that this is a hugely improved result, and could be classed as a realistic ‘photograph’. We can use these parameters to get some startlingly realistic results, and Midjourney understands a vast number of styles. Want to create a still frame from a 1980s Hong Kong comedy action film? No problem! The images below came from the prompt “Film still, style by 1980s Hong Kong comedy action film. Starring Sammo Hung, Jackie Chan and Yuen Biao. Eighties clothing. Sharp and detailed, film grain, Kodak 200T, no dramatic lighting, global illumination, uplight, 16-35mm lens, f1.8 –ar 16:9 –seed 1500 –q 2 –v 5”.

What’s staggering about these is how it has recreated the lighting style, film look, and characterisations extremely faithfully.

Compare these to the image below, which I specified as a more up-to-date style with “Film still, style by 2020s Hong Kong martial arts thriller. Starring Bruce Lee. Sharp and detailed, fine grain, Kodak 200T, global illumination, uplight, 16-35mm lens, f1.8 –ar 16:9 –seed 2000 –q 2 –v 5”.

Tip 4: Give Midjourney a reference image

The above images are all very well, but what if I wanted to capture an image in the middle of an actual fight sequence? Midjourney has huge problems when it comes to generating certain types of image and interaction. For It doesn’t have much comprehension of how human limbs work, which can create some utterly hilarious generations such as the one below!

We can reduce (not eliminate) such ludicrous generations by giving Midjourney a reference image. To do this, find an image showing roughly the kind of position you want the generation to show. This can be one from the web, or even better, one you’ve taken yourself, even if it’s just with your phone. Double click the “+” icon next to the prompt box and select the image from your computer to upload, and then send the message.

Your image will appear in the list of messages you’ve sent. Click on it to enlarge it, then right click and select “Copy image link”. Now, when you start an image prompt, simply paste the link at the beginning, then continue your image description. It won’t give you perfect results, but you can usually get far, far closer to what you want. You can give the system multiple reference images, but I would say to use three maximum, preferably two. The below image was generated using a reference still image from a Scott Adkins movie.

Right, so those are some tips you can use all round, and it’s worth experimenting with the “style by” instruction. But, what can we do to improve portraits and more traditional imagery?

Tip 5: Improving portraits with facial feature descriptions

Another command that Midjourney understands is “clear facial features”, and this can come in very useful for portraits. I’ve had some exceptional results by including this in descriptions. The images below resulted from the prompt “Craggy old male surfer sitting next to surf board on a beach, black wetsuit, barefoot. wide-shot, stormy weather. Sand in hair. Clear facial features, 16-35mm lens, f/1.8, global illumination, natural light, uplight, Ilford film stock –ar 16:9 –q 2 –v 5”

The image below was created with the prompt “Portrait of a female surfer with surf board on a beach. Sand in hair. Clear facial features, 35mm lens, f/1.8, global illumination, natural light, uplight, Ilford film stock –ar 16:9 –q 2 –v 5”

For some reason it gave her two surfboards, but this could be rectified with multiple generations, which brings me onto…

Tip 6: Regenerate, and regenerate some more

Most of the images above took quite a few regenerations to create. So another top tip is to have patience and to not only use the regenerate function from your initial prompt, but also create variations of the image you like the most from the initial four, and then create a variation of that, and so on. If you activate the “–prefer remix” command, you could create a variation of the image you like, but specify a different seed value, or add something else to the description. There are endless possibilities. Remix mode offers some very powerful ways to manipulate existing images. For example, you could change a photograph of a pile of footballs and make them into a group of puppies.

Final thoughts

Midjourney has some incredibly powerful possibilities and I’ve included some of the better results I’ve had below with some of the phrases used in the prompts, but it’s about more than just creating realistic images. In terms of cameras it can understand everything including GoPros, but the “style by” parameter has some stunning potential. For example you could specify anything from superhero comics to the photographic style of Ansel Adams. You can also experiment with lighting styles, particularly in relation to the colour of the light and its position, such as front lit or back lit.

As I mentioned earlier, there are some things that it doesn’t do well. Another example of this was when I asked it to produce an image of a surfer on a wave or carrying a surfboard along a beach. Midjourney doesn’t understand the relationship between a person and actually standing on a board or holding it, so it took me a large number of regenerations before I finally got an image that did what I wanted it to do. The key is patience, and attention to detail in how you make your descriptions. Vary them a bit, and tweak them to see what has an effect and what doesn’t. But I hope some of the tips I have described above help you to get better results.

More examples

“Style by 1970s music documentary”

“Portrait, clear facial features”

“Hyperdetailed photography, photorealistic. Canon EF 16-35mm f/2.8L III USM lens on a Canon EOS 5D Mark IV camera”

The result of many generations of asking for a surfer on a board, followed by asking for variations on different results.

source

How to use Midjourney to generate amazing images and art

Midjourney can conjure up impressive logos, artwork, and other graphics based on your own imaginative descriptions.

Looking for a logo for your business, artwork for a project, or an image for a report? One way to get a helping hand is to turn to an online AI tool. You can choose from an array of sites.

How to use: Craiyon AI | Bing Image Creator | DALL-E 2 | Stable Diffusion

But one service that offers truly impressive results is Midjourney. With this AI image creator, you describe the type of image you want to by entering text. In return, the site delivers four high-quality renderings.

How Midjourney works

Initially, Midjourney offered a free trial through which you could test the service by requesting a limited number of images. Unfortunately, the site ended the free trial for now, with the CEO blaming the move on a surge of new users. That leaves you with no option but to sign up for one of the paid subscription plans.

Also: How to use Bing Image Creator (and why it’s better than DALL-E 2)

A basic plan will run you $10 a month or $96 a year, a standard plan is $30 a month or $288 a year, and a pro plan is $60 a month or $576 a year. Each tier ups the speed of the responses and offers other benefits. To get a taste of Midjourney, you may want to start with the basic plan to see how it well it works for you.

How to get started using Midjourney

1. Sign up

Getting started with Midjourney can be confusing as you have to jump through a couple of hoops. To kick things off, go to the Midjourney website and click the link at the bottom for Join the beta.

You’re then taken to the website for Discord, which provides the server on which the responses are generated. If necessary, click the button for Continue to Discord. At the sign in window, click the Register link. Enter your email address, type a username and password, and select your date of birth. Click Continue.

Create an account