top of page

Midjourney V7 Arrives: Testing Quality, The New Draft Mode, and Japanese Prompts

11 minutes ago

4 min read

0

0

0

midjourney blog post image
A Midjourney generated image using Midjourney Automation Suite

Midjourney has just released version 7, and it brings some big changes. According to the official notes and initial tests, the image quality, detail, and accuracy are much better than in version 6.1. Plus, there's a completely new feature called Draft Mode. We took a close look at what's new and how it compares to the previous version. Get ready to see if V7 lives up to the hype!

Getting Started with Midjourney V7

Switching to version 7 is simple. When you're in Midjourney, look for the settings option, usually in the top right. Click it, and you should see '7' listed as an available version. Select it to start using V7.

Enabling Personalization for Best Results

To get the most out of Midjourney V7, you need to enable a feature called Personalization. This helps the AI learn what kind of images you like. It might take about five minutes to set up, but once done, Midjourney can create results that are more aligned with your taste.

How to Enable Personalization

To start, click the option to unlock personalization. Midjourney will show you two images. You simply click on the one you prefer. Repeat this process 200 times. Yes, 200! Doing this helps the AI understand your visual preferences. It's best to choose based on your gut feeling rather than picking randomly.

Comparing V7 Image Quality: Realism and Detail

One of the main claims for V7 is improved image quality, especially in realism, detail, and the structure of faces, hands, and objects. We ran some tests to see the difference.

Testing Complex Scenes and Characters

We tested prompts describing specific scenes and characters. For example, an astronaut on Mars or a medieval scholar in a library. V7 did a good job reproducing details like clothing, glasses, and the overall atmosphere mentioned in the prompts. While V6.1 wasn't bad, V7 often felt more polished and detailed, avoiding the slightly "too perfect" or "AI-generated" look seen sometimes in earlier versions.

Improved Structure: Faces and Hands

Midjourney sometimes struggles with hands and detailed facial expressions. We tested prompts with multiple people, like friends in a cafe or kids raising hands in a classroom. V7 showed clear improvement here. Hands often looked more natural, and facial expressions were more realistic. While not every single image was perfect, V7 was much more consistent in getting these tricky elements right compared to V6.1.

Managing and organizing batch generations like these comparisons can get complex. Streamline your workflow with the Midjourney Automation Suite from TitanXT. It helps handle multiple prompts and variations easily.

Better Texture and Fine Details

V7 also claims better texture rendering. We tested this with a prompt for a close-up portrait of an elderly man and images of silk cloth. The difference was noticeable. V7 captured skin texture with much more realism and depth. The way light interacted with the silk fabric also showed finer details and a more convincing look compared to V6.1, where textures could sometimes look a bit flat or less realistic.

Image Prompt Reproducibility in V7

We also looked at how well V7 uses image prompts to influence new generations. Using an existing image alongside a text prompt should help guide the new output's style or content.

Experimenting with Style and Image References

Tests showed that V7 seems capable of using style references quite well, adopting the overall look and feel from the input image. However, consistently reproducing specific character details using image references seemed hit or miss during testing. It appears there might still be development happening in this area, but the ability to adopt a style from an image is a useful feature.

Discovering the New Draft Mode

Version 7 introduces Draft Mode. This feature is designed for speed and efficiency, letting you quickly brainstorm ideas.

How Draft Mode Works

You can turn Draft Mode on with a button in the settings. When active (it turns red), image generation is significantly faster – up to 10 times quicker – and costs half the usual GPU time. It's great for quickly trying out different concepts, camera angles, or colors. While the initial draft images might be slightly lower quality than a standard generate, you can use the 'Upscale Subtle' option afterward to improve the final output quality once you have the composition you like. Think of it as a sketching tool.

Voice Input and a Remarkable Japanese Prompt Feature

Draft Mode, particularly through the web version, offers voice input. This lets you speak your prompts instead of typing.

Using Voice Input for Prompts and Parameters

There's a microphone button that starts conversation mode. You can naturally describe the image you want, and the system converts it to a prompt. What's even more impressive is that you can also speak parameters, like requesting a specific aspect ratio (e.g., 9:16) or telling Midjourney to repeat the prompt multiple times.

Native Japanese Prompting in V7

During testing, a significant discovery was made: Midjourney V7 appears to understand and process prompts written natively in Japanese very well. Previous versions like 6.1 struggled dramatically with Japanese, often only understanding a single word or producing irrelevant images. V7, however, accurately interpreted complex Japanese sentences describing specific scenes and objects. This is a major breakthrough for Japanese speakers or anyone who prefers prompting in Japanese.

If you're working with many prompts or trying different languages and styles, managing everything manually can take time. The Midjourney Automation Suite from TitanXT can help automate these tasks, making your workflow smoother and more efficient.

Final Thoughts on Midjourney V7

Midjourney V7 brings solid improvements to image quality, making outputs more realistic and detailed, particularly in challenging areas like hands and textures. The new Draft Mode is a fantastic addition for rapid prototyping and brainstorming, offering speed and lower cost. The voice input feature is a convenient way to prompt. Perhaps the most surprising improvement is the strong support for Japanese prompts, which worked significantly better in V7 than before. While some features, like character referencing in Draft Mode, seemed still under development during initial tests, the overall update makes Midjourney an even more impressive tool for image creation.

Want to explore these new Midjourney V7 features faster and more efficiently? Check out the Midjourney Automation Suite from TitanXT to automate your prompting, scaling, and organization tasks.

11 minutes ago

4 min read

0

0

0

Related Posts

Comments

Share Your ThoughtsBe the first to write a comment.
bottom of page
Midjourney Automation Suite - Automate your image generation workflows on Midjourney | Product Hunt