*https://s.mj.run/3SgnnfB-9BE a blonde woman sat in a kitchen, she is holding a big cereal box in her right hand, she is holding the cereal up showcasing a product, the camera is close up focusing on the product, the background is a blue backdrop, 50mm lens, dof background, immaculate, exciting, hyper realistic product imagery, canon 1dx camera, --ar 9:16 --v 6.0*

  NOTE: Always end your prompt with —v 6.0 this will utilise the most recent version of the MidJourney bot.

This is the rough formula you should follow for your prompts:

1. Is there a model holding the product or in the image?

   1. Ethnicity?
   2. Hair colour?
   3. Poses?

2. What the product is?

   1. Dimensions
   2. Colours
   3. Size
   4. Shape
   5. Material
   6. Finish

3. Where is the product located?

   1. Is it in the air?
   2. Is it sitting on an object?
   3. What is the location or environment?

4. What is surrounding the product?

   1. What objects or props are surrounding
   2. Are these objects behind, in-front, to the side?

5. What is the lighting and feel of the image?

   1. Studio lighting?
   2. Natural lighting?
   3. Is it moody?
   4. Is it dramatic?

6. What type of camera would this image be taken on?

   1. Film camera?
   2. DSLR camera?

7. What type of lens would it be taken on?

   1. 50mm?
   2. 35mm?
   3. Wide angle?
   4. Macro lens?
   5. Fish eye?
   6. 180mm?

8. What is the aperture of the camera? (what is the Depth of field)

   1. Is the product the main centre of attention and is the background blurred?
   2. Is the product image disconnected from the background?

9. What is the style of the image?

   1. Is it hyper-realistic
   2. Is it a cartoon?
   3. Is it semi-realistic?
   4. Is it artistic?

10. What is the desired aspect ratio of the image?

    1. 16:9 is widescreen
    2. 9:16 ****is a vertical IG story format
    3. 4:5 is a standard IG feed post
    4. 1:1 is a square image

How do you actually start the prompting process?

1. Type into the chat bar /imagine and click the ‘space’ key afterwards.

2. Insert the prompt that you have just crafted and click enter.

An example of a prompt we used for this is image is here:

“a woman sat in a modern bathroom, she is wearing a white and yellow bathrobe, she is holding a cylincar container in her right hand, she is holding the container up showcasing a product, the camera is close up focusing on the product, the background is a blue backdrop, 50mm lens, dof background, immaculate, exciting, hyper realistic product imagery, canon 1dx camera, --ar 9:16 --v 6.0”

 

What if Midjourney is still struggling to generate my image?

💡 TIP: You can provide Midjourney a reference image and it will learn from the image given and generate outputs that includes many of the features the reference image has.

 

These are the steps:

1. Send Midjourney a photo of the product image that you want it to generate.

To do this, simply drag and drop the image into the chat bar on midjourney and click enter.

This will send the image to the server.

💡 NOTE: You can also send Midjourney a reference photo that is not a product photo. For example, you may have found a picture you like from another brand and you want Midjourney to re-create a similar picture for your brand. This is also possible by following the same steps shown below.

2. Click on the image you sent to the Midjourney server, right click on the image and copy the link.

3. Paste the link of the image at the beginning of your /imagine prompt.

Type into the chat bar of the server and start your prompt with “/imagine”.

Then paste the link of the image after /imagine.

1. Paste the prompt of the image you want Midjourney to create after the reference image link.

Paste the prompt you created above into Midjourney, directly after your reference photo image link.

For example, the image below is what this would look like.

Midjourney will now get a visual representation of what you want to create. This creates are clearer message for the AI and helps with the quality of the output.

6. Refine and adjust prompt for final image.

Based on what Midjourney gives you, you may need to adjust your prompt slightly or create variations.

This will be broken down in the next step of the process - Please head to step 8.

8. Refine Prompts in Midjourney 🤔

Once we have crafted our ideal prompt and sent Midjourney a reference image we are 85% of the way there!

However, as Midjourney is an AI platform, it sometimes takes a few revisions and refinements of our prompts in order to create the desired output.

<aside> ⚠️ ALSO NOTE: Midjourney may struggle to generate your product image 1:1 even with highly specific prompting, reference images and refinements. If this is the case, then the aim is to get it as close as possible to what you are wanting to create and then add the final touches with Photoshop. Step 10 of these instructions will break down how to do this.

</aside>

In this step we will be breaking down how to refine our images within Midjourney.

Once we have submitted Midjourney our reference image and entered our prompt, Midjourney will generate four variations of images.

You are able to open up all four images at the same time and inspect which ones you like.

If you don’t like any of the images that have been generated you are able to click the ‘re-generate’ button and it will generate four new images.

What are the buttons underneath the four generated images?

Underneath the images are nine buttons.

U1, U2, U3, U4 - V1, V2, V3, V4

The ‘U’ is short for upscale. Upscale means increasing the resolution of an image.

The ‘V’ is short for variation. This means creating a slight variation of an image.

The associated number is indicative of what image it is referring to.

For example, if you like image 1 you can click ‘U1’. This will upscale image 1.

This image can then be downloaded from Midjourney.

This is the result of clicking ‘U1’.

If you like image one but I want it to be slightly different, you can click ‘V1’ this will generate four new variations of that image.

Keep continuing this process until you have achieved the desired output image.

You can also change the text within the prompt to also increase the variations and change the image to your liking.

1. Download and upscale with Nightmare Upscale (Refer to Step 9 on how to do this)

Once you have generated the image you desire, download the image and go to Nightmare Upscale and upscale the image to the highest quality resolution.

9. Upscale Image Using NightmareAI Upscale 📈

Now because we have downloaded the image from Midjourney. The image resolution is not the highest quality.

We can use Nightmare AI Image Upscaler to upscale the image and improve the resolution.

(Note: If you have not set up your Replicate account to use Nightmare AI, please head to Step 3 and the instructions on how to set up your account and add funds are there).

We recommend choosing the upscale as high as possible. Nightmare Upscale lets you 10x the resolution using their platform.

Note: we always want to have the highest quality images to use for when we go into Photoshop.

Here are the instructions on how to upscale an image using this platform:

Here’s the process to upscale your image:

1. Go to https://replicate.com/nightmareai/real-esrgan
2. Upload your image file

Drag and drop your image from Midjourney into Nightmare AI file upload.

Set your upscale level to what you desire. We recommend above 3.

Select the ‘Run’ option

Once your image has been upscaled it will appear on the right hand side

Download the image

Perfect!

We now have an upscaled the image and it is now the highest resolution possible.

10. Final Editing In Photoshop✍️ 

Now, if Midjourney was able to generate your product image exactly how you wanted it to look without any manipulation needed in Photoshop you can skip this step.

But if you’re wanting to get your product to be 1:1 you will need to insert it using Photoshop.

In this step we will be breaking down what tools and steps are used in Photoshop that we use in our examples.

Here is a Photoshop Crash Course to get you up to speed on the platform:

 

Now time to dive into the steps they are used in Photoshop:

1. Import your image to Adobe Photoshop

2. Use tools within Photoshop to blend and insert your product or branding onto your AI generated image.

This step sounds simple in theory, however there isn't a one size fits all approach.

However, here are some of the common tools that we use in order to achieve our results:

• Blending Tools
  1. Multiply
  2. Screen
  3. Add
  4. Lighten
• Other Tools
  1. Clone stamp
  2. Mesh warp
  3. Liquify
  4. Path
  5. Camera raw filter
  6. Gaussian blur

💡 NOTE: Because all edits require a different process, below are some tutorials that will help guide you through our process so you can learn.

💍 Jewellery example for simple shaped necklace:

FROM THIS