avatarPromptDervish

Free AI web copilot to create summaries, insights and extended knowledge, download it at here

5091

Abstract

kFp6wSMMX-3TjHIFntw.png"><figcaption>Original image after being panned right</figcaption></figure><figure id="3fb9"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*X-lvSYLahHgnzygL6ND0UQ.png"><figcaption>Upscale #4 above</figcaption></figure><p id="c1c3">For some variety, I decided to take the upscale and use “zoom out x2” on it. Notice how the original upscale part of the image is much smaller now and loses a bit of definition, and at the same time so much more has been added to the picture in all directions.</p><figure id="4c4d"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*yGw664HE1pK_TVykcGta-w.png"><figcaption>The upscale of the #4 image after clicking the “zoom out x2” button on the upscale</figcaption></figure><figure id="347e"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*jSOcV3J7ytVRQe6d3WUFow.png"><figcaption>Upscale of the squared image previously</figcaption></figure><p id="44e2">I loved the image right here but being an explorer I just had to pan right to look at those houses and what else might be out there.</p><figure id="955d"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*37MwUPGt6jlPZLBk3LVrrA.png"><figcaption>Panning right from the image above this one</figcaption></figure><figure id="45d2"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*x2JEY8sftA0MSI83n40h7w.png"><figcaption>Upscale of #4 from above</figcaption></figure><p id="c221">It just keeps getting more interesting so I panned right again.</p><figure id="47e2"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*O6A8pQJs-ppzvTGBDW6uyw.png"><figcaption>Panning right again</figcaption></figure><figure id="be46"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*6tTCeAs47j6roEjbxVHboQ.png"><figcaption>Upscale of #4</figcaption></figure><p id="0e74">And finally, one last “Zoom out” to bring in more sky and foreground.</p><figure id="4aaf"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*xhgyX_GQnU7P4tbwI1zPqA.png"><figcaption>Using Zoom on the previous image</figcaption></figure><p id="576b">And one two gratuitous lefts pans for balance and here we have</p><figure id="3b58"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*oTSgKUXUIGEX6q1HHVbWTQ.png"><figcaption>Another two left pans just for balance. note that the original train is quite small just right of center at the back.</figcaption></figure><figure id="2e7a"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*v-Ff9fQwan5mlcIKmpIG2w.png"><figcaption>The first train image that started this journey.</figcaption></figure><h1 id="fc63">More practical purposes for Pan and Zoom Out</h1><p id="f7d0">The “Pan” and “Zoom Out” features are not just for making dramatic alterations; they are also great for fine-tuning your image. For instance, if Midjourney tends to crop out the top of a subject’s head or hat, the pan function can help you include those missing elements. Having a little extra room in your image offers the flexibility to crop it down later to exactly what you want.</p><p id="c6fa">This is particularly handy given Midjourney’s tendency to center subjects squarely in the middle of the frame. Utilizing the “Pan” feature allows you to adjust the composition to better align with principles like the rule of thirds, thereby creating a more visually appealing result. Using “zoom out” gives you more area all the way around the picture for additional cropping space.</p><figure id="0ed3"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*dH1lMOpywYqfLcjwmSlimA.png"><figcaption>Original image, ears too close to the edge and more leg would be nice so we will play with pan to get more of her into the picture.</figcaption></figure><figure id="0bd4"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*2etYwfUJAVSBkBQ1LUncHQ.png"><figcaption>Pan down gives us some nice legs</figcaption></figure><p id="0fae">After one more “Pan Down,” the resulting image featured some rather odd-looking feet and additional cats. This is fairly common with panning and zooming, one must take it in stride and look first at the pan prompt to see what can be better. Occasionally I have just used a period in the field and nothing else, it is not possible to leave the field blank.</p><figure id="be64"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*vldVxZmVcK97mzgJ_Gw6VA.png"><figcaption>If you keep panning without changing the prompt, such as adding something here about her feet and shoes.</figcaption></figure><p id="d338">To address the issue of the oddly-rendered feet, I took a different approach. I completely cleared the existing text in the pan prompt window and replaced it with “deep-purple-shoes.” While this method improved the situation, the results weren’t perfect, showcasing the trial-and-error nature of working with AI-generated imagery.</p><p id="6467">About the hyphens — I often use hyphens to link adjectives and nouns or other qualifiers, ensuring they’re applied specifically to

Options

the intended object rather than the entire scene.</p><figure id="55bd"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*OF20b8jg-ua1rIYDKLmOAA.png"><figcaption>Removed all of the previous prompt and changed it to “deep-purple-shoes”</figcaption></figure><p id="cc7b">Option #1 on the top left appears to be the best fit for my needs. While the other choices could potentially be fine-tuned using the “vary(region)” function, activating that option would first require either “Zoom Out” or “Make Square” to be enabled. For this project, I’ve decided to proceed with option #1. To give myself more flexibility in the post-editing stage, I’ll use a 1.5x “Zoom Out,” which will expand the background and offer more room for precise cropping later on.</p><figure id="961b"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*y-i_EkWdH-9r57tkIKMR5A.png"><figcaption>Zoomed out 1.5x from the previous upscaled image</figcaption></figure><p id="ab1c">Having zoomed out 1.5 it would now be possible to vary region and make any corrections to the scene. I am satisfied with it as it is so I will leave it alone. I added the white frame to this final upscaled image to show how much we were able to add with a few clicks and generations. I noticed that one of her heels is super long, which could be fixed in vary(region) but it is a much quicker fix in Photoshop so I would fix that in post-work on the image.</p><figure id="be65"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*uTyIgD32j0BFgID2PM0UsA.jpeg"><figcaption>Our final zoomed out version of the panned down image. Frame shows how much we added here.</figcaption></figure><div id="e765" class="link-block"> <a href="https://pub.aimind.so/consistent-characters-and-corrections-using-vary-region-in-midjourney-54da396c39a9"> <div> <div> <h2>Consistent Characters and Corrections Using Vary(Region) in Midjourney</h2> <div><h3>Midjourney has introduced Vary(Region) a type of inpainting for changing a region of your image with masking and a…</h3></div> <div><p>pub.aimind.so</p></div> </div> <div> <div style="background-image: url(https://miro.readmedium.com/v2/resize:fit:320/1*yn3z696yOTVdwhhGiLjguQ.gif)"></div> </div> </div> </a> </div><div id="6824" class="link-block"> <a href="https://pub.aimind.so/midjourney-inspiration-exploring-describe-vary-and-pan-651fdc1c6860"> <div> <div> <h2>Midjourney Inspiration — Exploring Describe, Vary, and Pan</h2> <div><h3>The “/describe” feature in Midjourney is a fascinating tool that uses artificial intelligence to interpret and describe…</h3></div> <div><p>pub.aimind.so</p></div> </div> <div> <div style="background-image: url(https://miro.readmedium.com/v2/resize:fit:320/1*AWKszmlNa4bh7hwUDN24_A.png)"></div> </div> </div> </a> </div><div id="1232" class="link-block"> <a href="https://pub.aimind.so/midjourney-explorations-1-its-all-in-the-words-c642e36e412b"> <div> <div> <h2>Midjourney Explorations: #1 — It’s all in the words!</h2> <div><h3>Finding the right word to describe your image goes a long way. In this series of Midjourney Explorations we will dive…</h3></div> <div><p>pub.aimind.so</p></div> </div> <div> <div style="background-image: url(https://miro.readmedium.com/v2/resize:fit:320/1*PNetTSddzw62FLOc8zVxeQ.jpeg)"></div> </div> </div> </a> </div><h1 id="688e">In Conclusion</h1><p id="5d97">Thank you so much for taking the time to read through this guide. I genuinely hope you found it insightful and beneficial as you navigate the world of Midjourney. If you’d like to stay updated on future articles and dive deeper into this fascinating set of generative art tools, please do follow me here on Medium.com. Your support means the world to me, happy creating!</p><figure id="f369"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*3uoWIFkJPBNxGvt5JCCosA.jpeg"><figcaption></figcaption></figure><h2 id="417c">A Message from AI Mind</h2><figure id="ecd4"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/0*5Wm7sOfTpe5DEbhg.gif"><figcaption></figcaption></figure><p id="0438">Thanks for being a part of our community! Before you go:</p><ul><li>👏 Clap for the story and follow the author 👉</li><li>📰 View more content in the <a href="https://pub.aimind.so/">AI Mind Publication</a></li><li>🧠 Improve your <a href="https://www.aimind.so/prompt-generator?utm_source=pub&amp;utm_medium=message">AI prompts effortlessly and FREE</a></li><li><b>🧰 Discover <a href="https://www.aimind.so/?utm_source=pub&amp;utm_medium=message">Intuitive AI Tools</a></b></li></ul></article></body>

Midjourney Fun With “Pan” and “Zoom Out”

The “Pan” and “Zoom Out” features of Midjourney allow you to expand an upscaled image in a specific direction or in all directions.

Below we begin by panning left on the upscaled image shown.

Original Prompt: A tree is growing in a forest, vivid, crisp-focus, highly-detailed, optimistic pop-art-style acrylic painting on canvas — c 44 — s 444 — w 444 — v 5.2

When I did the left pan, I did not change the prompt, so there are second trees on quad numbers 1, 3, and 4. Number 2 is an example of how Midjourney just likes to randomly throw a woman into your image for no apparent reason. In this case, it is very cool, other times one just has to ask WHY?!

Since #2 is interesting I decided to upscale it and panned right, again without changing the prompt, and again one woman, an extra tree and a couple of sunsets, notice we now have two sun.

Some base images lend themselves more to panning, in most cases, you do want to consider how your prompt may need to change to get the pannings to work properly without undesirable repeating elements.

Original prompt: A sunset captures the essence of a sunset over a landscape, using shades of oranges, yellows, and reds, with irregular and natural shapes, with rough and textured brushstrokes, viewed from a low angle, and with warm and soft lighting to evoke a peaceful and serene atmosphere. Inspired by J.M.W Turner and Claude Monet. — ar 3:2 — c 70 — s 800

Starting with my base image, I used the “Pan” feature to extend the view to both the left and right. I changed the initial prompt to just “ocean” to avoid multiple suns appearing in the extended scenes. Intriguingly, Midjourney took some creative liberties, adding elements like a ship and a lighthouse to the landscape. These were not things I explicitly requested; they were simply the AI’s interpretation of what might naturally belong in such a setting.

Changed prompt and panned both left and right: ocean, using shades of oranges, yellows, and reds, with irregular and natural shapes, with rough and textured brushstrokes, viewed from a low angle, and with warm and soft lighting to evoke a peaceful and serene atmosphere. Inspired by J.M.W Turner and Claude Monet. — c 70 — s 800

When you use the “Pan” feature, the resulting image often exceeds the standard size constraints. For example, the final image size in one of my tests was 4032x896 pixels. This can be very handy if you wish to have a larger image size. Keep in mind that if you subsequently use the “Zoom” or “Make Square” functions, the image will revert back to a 1024x1024 resolution, potentially sacrificing some finer details. If you’re planning to use any of the “vary” options on a panned image, you’ll first need to select either “Zoom” or “Make Square” for those options to become available under the “Upscale” menu.

Panning is generally more forgiving with simpler landscapes, like oceans, but can be trickier with complex scenes, like countryside. One challenge I’ve encountered is that the area where the image extends can sometimes be glaringly obvious. This usually requires some prompt adjustment and rerolling to achieve a more seamless extension.

For example, my initial image was prompted by the phrase “A train is traveling through the countryside.” To avoid duplicating trains in the panned versions, I modified the prompt to just “countryside,” letting Midjourney fill in the details. Interestingly, the AI seemed to place the scene in what appears to be Europe, adding an unexpected twist to the landscape.

This image below is panned right one time. You can see the original unpanned image at the very end of this article below the final image in this set of explorations.

Upscaled image: A train is traveling through the countryside, vivid, crisp-focus, highly-detailed, optimistic pop-art-style acrylic painting on canvas — ar 16:9 — c 44 — s 444 — w 444 — v 5.2 (with one right pan from original)
Original image after being panned right
Upscale #4 above

For some variety, I decided to take the upscale and use “zoom out x2” on it. Notice how the original upscale part of the image is much smaller now and loses a bit of definition, and at the same time so much more has been added to the picture in all directions.

The upscale of the #4 image after clicking the “zoom out x2” button on the upscale
Upscale of the squared image previously

I loved the image right here but being an explorer I just had to pan right to look at those houses and what else might be out there.

Panning right from the image above this one
Upscale of #4 from above

It just keeps getting more interesting so I panned right again.

Panning right again
Upscale of #4

And finally, one last “Zoom out” to bring in more sky and foreground.

Using Zoom on the previous image

And one two gratuitous lefts pans for balance and here we have

Another two left pans just for balance. note that the original train is quite small just right of center at the back.
The first train image that started this journey.

More practical purposes for Pan and Zoom Out

The “Pan” and “Zoom Out” features are not just for making dramatic alterations; they are also great for fine-tuning your image. For instance, if Midjourney tends to crop out the top of a subject’s head or hat, the pan function can help you include those missing elements. Having a little extra room in your image offers the flexibility to crop it down later to exactly what you want.

This is particularly handy given Midjourney’s tendency to center subjects squarely in the middle of the frame. Utilizing the “Pan” feature allows you to adjust the composition to better align with principles like the rule of thirds, thereby creating a more visually appealing result. Using “zoom out” gives you more area all the way around the picture for additional cropping space.

Original image, ears too close to the edge and more leg would be nice so we will play with pan to get more of her into the picture.
Pan down gives us some nice legs

After one more “Pan Down,” the resulting image featured some rather odd-looking feet and additional cats. This is fairly common with panning and zooming, one must take it in stride and look first at the pan prompt to see what can be better. Occasionally I have just used a period in the field and nothing else, it is not possible to leave the field blank.

If you keep panning without changing the prompt, such as adding something here about her feet and shoes.

To address the issue of the oddly-rendered feet, I took a different approach. I completely cleared the existing text in the pan prompt window and replaced it with “deep-purple-shoes.” While this method improved the situation, the results weren’t perfect, showcasing the trial-and-error nature of working with AI-generated imagery.

About the hyphens — I often use hyphens to link adjectives and nouns or other qualifiers, ensuring they’re applied specifically to the intended object rather than the entire scene.

Removed all of the previous prompt and changed it to “deep-purple-shoes”

Option #1 on the top left appears to be the best fit for my needs. While the other choices could potentially be fine-tuned using the “vary(region)” function, activating that option would first require either “Zoom Out” or “Make Square” to be enabled. For this project, I’ve decided to proceed with option #1. To give myself more flexibility in the post-editing stage, I’ll use a 1.5x “Zoom Out,” which will expand the background and offer more room for precise cropping later on.

Zoomed out 1.5x from the previous upscaled image

Having zoomed out 1.5 it would now be possible to vary region and make any corrections to the scene. I am satisfied with it as it is so I will leave it alone. I added the white frame to this final upscaled image to show how much we were able to add with a few clicks and generations. I noticed that one of her heels is super long, which could be fixed in vary(region) but it is a much quicker fix in Photoshop so I would fix that in post-work on the image.

Our final zoomed out version of the panned down image. Frame shows how much we added here.

In Conclusion

Thank you so much for taking the time to read through this guide. I genuinely hope you found it insightful and beneficial as you navigate the world of Midjourney. If you’d like to stay updated on future articles and dive deeper into this fascinating set of generative art tools, please do follow me here on Medium.com. Your support means the world to me, happy creating!

A Message from AI Mind

Thanks for being a part of our community! Before you go:

Midjourney
Generative Art
Ai Art Generator
Ai Art
Recommended from ReadMedium