This is a method I was told about on a certain internet forum. I am writing it here so I don't have to repeat myself.
Note: The webUI I use is ReForge (fork of Forge/Auto1111)
I've updated the method using the information from Volnovik. The results are exactly the same but the method is faster to execute.
NOTE: I've attached a custom extension that is a modified copy of the built in extension that makes the process even more brain-dead easy. Just download the Custom Multidiffusion file, unzip it, place the "custom_multidiffusion" folder into the "extensions" folder. Start up the UI and confirm that the extension is enabled.
Improved Method
0. Enable "Upscaler for IMG2IMG". To do this, first goto Settings tab, scroll to "User Interface", click on "Settings in UI", navigate to the drop down for "Settings for img2img", add "Upscaler_for_img2img" to the list, click on "Reload UI".
1.Gen your image in txt2img
2. Send image to img2img (keep all the information it sends)
3. Set Resize By -> Scale 2
4. Find "Upscaler for img2img", set it to your choice of upscaler (I use 4x_NMKD-Siax_200k)
5. Enable MultiDiffusion, Set it to Mixture of Diffusers
6a. If you have the Custom Multidiffusion extension, just toggle Use Generation Image Dimensions and don't worry about it anymore.
6b. If you are using the built in extension, set the Tile Width/Height to the image's dimensions (ex. if you gen at 1024x1024, that is what you put here)
7.Scroll to Denoising Strength and set it to 0.25 (anywhere between .15 to .45 is good, higher numbers results in more drastic changes)
8. Click generate until you get an improved image you like
9. If you like a gen but only want to slightly change the results then save the seed and alter the Denosing Strength
Old Method
1.Gen your image in txt2img
2.Send image to Extra tab and upscale it by 2x using your choice of upscaler (I use 4x_NMKD-Siax_200k)
3. Go back to the txt2img tab while its upscaling and send the original image to img2img (we want the prompt and settings info)
4.Go back to Extra tab and send the upscaled image to img2img now, replacing the old image
5.Enable MultiDiffusion, Set it to Mixture of Diffusers
6a. If you have the Custom Multidiffusion extension, just toggle Use Generation Image Dimensions and don't worry about it anymore.
6b.Set the Tile Width/Height to the original image dimensions (ex. if you gen at 1024x1024, that is what you put here, do not use the upscaled dimension from Extra tab)
7.Scroll to Denoising Strength and set it to 0.25 (anywhere between .15 to .45 is good, higher numbers results in more drastic changes)
8. Click generate until you get an improved image you like
9. If you like a gen but only want to slightly change the results then save the seed and alter the Denosing Strength
Why am I using this slightly more complicated even more easier method instead of Hires Fix?
Because it is completely better than Hires fix. It improves the image when you increase the denosing strength without ruining it like hires does. It's faster than hires. Once you understand the steps it's very easy to reuse. The only draw back is (in the default extension) you have to reset the dimensions in MultiDiffusion each time you change them on the original gen image. (This is no longer true thanks to the custom extension, praise ChatGPT!) Also, you might occasionally get unwanted object added to the new gen. (for ex, if you gen breast you might get a second pair of nipples, this is a very easy fix, just remove breast from the prompt in your img2img tab or lower the denoising strength) It's worth it, believe me.
Update Mar. 22, 2025: Updated the instructions and added a modified Multidiffusion extension to streamline the process even further beyond! (ahhhhhhhhhhhhhhhh)