Mastering Midjourney v5.1 : A Deep Dive into the New Version! (RAW, Seeds, Tips and Tricks!)

Theoretically Media
9 May 202314:41

TLDRIn this video, Tim dives into the new features of Midjourney version 5.1, highlighting its two distinct modes: a stylistically opinionated model reminiscent of version four's shorter prompts and a separate raw mode. Tim uses the same prompt and seed across different versions to compare outputs, noting that version 5.1 is more imaginative and expressive, with warmer skin tones and more volumetric lighting. He finds version 5.1 raw to be more realistic, while version 5 appears more photographic. Tim also discusses the improved sharpness in version 5.1 and explores the use of permutations, a lesser-known feature. He concludes that version 5.1 is ideal for illustrative styles, whereas raw mode is better for photographic or cinematic images. The video ends with a teaser for the upcoming version 6 and an invitation for viewers to share their experiences with the new features.

Takeaways

  • 🔍 **Midjourney v5.1 Introduction**: Midjourney v5.1 is a significant update with two distinct modes, offering more stylistic variety and creative potential.
  • 🎨 **Stylistic Differences**: Version 5.1 has a more stylistic and opinionated model, resembling the prompting style of v4, while also introducing a RAW mode that is unique and distinct.
  • 📈 **Version Flexibility**: Users can switch between different versions, including v5.1 and RAW mode, through the forward slash settings, allowing for versatile image generation.
  • 🌱 **Understanding Seeds**: Seeds are unique identifiers for each generated image, acting as an initial pattern that shapes the final output, ensuring variation in results.
  • ♻️ **Seed Number Changes**: Seed numbers are finite and may change over time, meaning the same seed number may not always produce the same image.
  • 📷 **Image Generation Process**: The script demonstrates image generation using the same prompt and seed across different versions of Midjourney to compare outputs.
  • 🌟 **V5.1 vs V5 RAW**: Version 5.1 RAW tends to produce more realistic skin tones compared to the smoother, more uniform tones of version 5.
  • 🎭 **Expressiveness in V5.1**: Version 5.1 is noted to be more imaginative and expressive, especially with basic prompts, offering illustrative and vibrant images.
  • 📉 **Photographic vs Artistic**: Version 5 offers a more photographic look, while 5.1 leans towards a more artistic and stylized output.
  • 🔧 **RAW Mode Utility**: RAW mode is suggested as a good compromise, providing creativity without heavy stylization, and is best for specific image intentions and longer prompts.
  • 📚 **Additional Training in 5.1**: There's an indication of additional training in v5.1, leading to more accurate depictions, especially in rendering fingers playing instruments.
  • 📉 **Sharpness Improvement**: Version 5.1 has improved sharpness compared to v5, which had softer images, making it more suitable for detailed and clear outputs.
  • 🔄 **Permutations Feature**: Permutations, although less used and somewhat confusing, allow for running multiple versions of Midjourney off the same prompt to explore different stylistic outcomes.

Q & A

  • What is the significance of Midjourney version 5.1 in the history of the software?

    -Midjourney version 5.1 is significant as it marks the first time a numbered update has been released between versions. It introduces two different modes within the version, offering more stylistic variety and options for users.

  • What is the purpose of the 'seed' in the context of Midjourney's image generation?

    -The 'seed' in Midjourney is a number that initiates the image generation process, acting like an initial pattern of static that forms into the final image. It ensures that the same image is never generated exactly twice, but variations will be relatively similar.

  • How can users find out their seed number in Midjourney?

    -Users can find out their seed number by adding a reaction to any image using the Envelope Filter. They will then receive a message containing the seed number.

  • What is the main difference between version 5.1 and version 5.1 raw in terms of image output?

    -Version 5.1 is more stylistic and expressive, with warmer and brighter skin tones and more volumetric lighting. Version 5.1 raw, on the other hand, tends to produce images that are slightly more realistic, with subtle differences in skin tones and lighting.

  • How does the use of permutations in Midjourney work?

    -Permutations in Midjourney allow users to run the same prompt through different versions of the software simultaneously. This is done by adding a specific set of commands at the end of the prompt within curly brackets, specifying the versions to be used (e.g., V4, V5, V5.1, V5.1 raw).

  • What is the recommended approach when using version 5.1 raw mode?

    -Version 5.1 raw mode is best suited for users who have a specific image in mind and prefer a balance between the creativity of version 5.1 and a less heavy stylization. It works well with longer prompts and is ideal for those looking for a more photographic or cinematic image output.

  • How has the sharpness of images improved in version 5.1 compared to version 5?

    -The sharpness in version 5.1 has significantly improved compared to version 5, which had softer, more blurred images. Version 5.1 brings back the clarity and detail that was present in version 4.

  • What aesthetic improvements have been noted in version 5.1?

    -Version 5.1 has shown improvements in sharpness and has become more imaginative and expressive. It also seems to have better handling of fingers and instruments, particularly guitars, appearing more accurate and realistic in the generated images.

  • What is the 'least used feature' of Midjourney that the speaker thinks is often misunderstood?

    -The 'least used feature' that is often misunderstood, as mentioned in the script, is permutations. It is considered confusing to use and can consume a significant amount of fast hours, leading to it being underutilized.

  • What did the speaker find when experimenting with emulating the style of Charles Peterson in Midjourney?

    -The speaker found that while Midjourney had not been trained on Charles Peterson's images, there were noticeable improvements in the depiction of fingers playing instruments, particularly guitars, in version 5.1. Interestingly, the closest energy to Charles Peterson's style was found in Midjourney V4, which captured a grainy, black and white, chaotic look.

  • What is the speaker's recommendation for users looking for different styles in their generated images?

    -The speaker recommends that if users are looking for illustrative or artistic styles, version 5.1 would be the best choice. For those seeking photographic or cinematic images, version 5.1 raw should be considered. The choice between versions is becoming more of an aesthetic preference as all produce solid results.

  • What is the speaker's final note regarding the future of Midjourney?

    -The speaker's final note is that it won't be long before users will be discussing version six of Midjourney, indicating that further updates and improvements to the software are on the horizon.

Outlines

00:00

🚀 Introduction to Mid Journey Version 5.1

The video begins with an introduction to Mid Journey version 5.1, highlighting it as a significant update with two distinct modes: a stylistically opinionated model resembling V4's shorter prompting style and a 'raw mode' that is often mistaken for version five but is unique. The host explains how to switch between different versions using forward slash settings. A demonstration follows, using the same prompt and seed across various versions to compare outputs. The importance of seed numbers in generating images is also discussed, with a brief tutorial on how to retrieve them using the Envelope Filter. The host then generates an image with a Bond-inspired prompt and compares the results from version 5, 5.1, and 5.1 raw, noting the differences in style, lighting, and skin tones.

05:00

🎨 Exploring the Creative Outputs of Mid Journey Versions

The host continues to explore the creative outputs of Mid Journey by running basic prompts through version 5 and 5.1, noting the differences in style and expressiveness. Version 5.1 is described as more imaginative and expressive, especially with basic prompts, while version 5 provides more standard, photographic outputs. The discussion moves on to the use of 'raw mode' as a compromise between the heavy stylization of 5.1 and the more straightforward style of version 5. The host shares their opinion that raw mode is best for those with a specific image in mind and works well with longer prompts. Additionally, the host touches on the potential for additional training in 5.1, as evidenced by improved depictions of fingers playing instruments, which were previously a challenge for the software.

10:00

📈 Sharpness Improvements and Permutations in Version 5.1

The video script discusses the improved sharpness in version 5.1, a feature that was lacking in version five. The host expresses satisfaction with the return of sharpness and explores the use of permutations, a feature that is often misunderstood and underutilized. A detailed explanation of how to use permutations is provided, followed by an example using a complex prompt involving a cinematic still by Stanley Kubrick. The host also shares their personal experience with the software, including an experiment emulating the style of Charles Peterson, a photographer known for his work in the Seattle grunge scene. The script concludes with a mention of a PDF of the video available for free on Gumroad, an opportunity for viewers to donate, and a teaser for the upcoming discussion on version six.

Mindmap

Keywords

💡Midjourney v5.1

Midjourney v5.1 refers to the latest update of a software or tool that is being discussed in the video. It is significant because it introduces two different modes within the version and is considered a turning point in the history of the software. The video explores the features and differences between these modes.

💡Seeds

Seeds are unique identifiers used in the generation of images by the Midjourney bot. Each generated image has a corresponding seed number that acts as an initial pattern or 'static' from which the image evolves. The concept is important as it ensures that even with the same prompt, the resulting images will be unique, never exactly replicating the same image twice.

💡RAW Mode

RAW mode is a feature within Midjourney v5.1 that is distinguished from the standard v5 mode. It is described as its 'own thing' and is explored in the video for its unique image generation capabilities. RAW mode is praised for its more realistic skin tones and is presented as a good compromise between the heavy stylization of v5.1 and the more standard look of v5.

💡Sharpness

Sharpness refers to the clarity and detail of the images generated by the Midjourney tool. The video discusses the improvement in sharpness in version 5.1 compared to the previous version, noting that it had been an issue with 'paper quilling' type images in v5.

💡Permutations

Permutations are a feature of Midjourney that allows for the generation of multiple variations of an image from a single prompt. This feature is noted as being potentially underused and misunderstood, but the video suggests that the new versions of the software open up opportunities for exploring permutations further.

💡Cinematic Still

A cinematic still refers to a static image that resembles a frame from a movie, capturing a moment with a strong narrative or emotional impact. The video uses the term in the context of generating images with a prompt that includes a 'cinematic still', aiming to produce images with a similar feel to those from a film directed by Stanley Kubrick.

💡Illustrative Styles

Illustrative styles are a type of image generation that leans towards artistic and playful representations, often deviating from photorealism. The video compares the outputs of different Midjourney versions in producing images with illustrative styles, noting that v5.1 is particularly adept at creating such images.

💡Photographic Styles

Photographic styles in the context of the video refer to image outputs that closely resemble real photographs, with a focus on realistic lighting and detail. The video suggests that for those seeking a more photographic or cinematic result, v5.1 RAW mode is the preferred choice.

💡Aspect Ratio

The aspect ratio is the proportional relationship between the width and the height of an image. In the video, an aspect ratio of 2-1 is specified in a prompt, which would result in an image that is twice as wide as it is tall, a common format for cinematic images.

💡Volumetric Lighting

Volumetric lighting refers to the way in which light interacts with the volume of a scene, creating a sense of depth and atmosphere. The video discusses how version 5.1 introduces more volumetric lighting in the generated images, contributing to a more expressive and dynamic visual output.

💡Grainy Black and White

This term describes a specific aesthetic characterized by images with a rough texture and a monochromatic color palette. The video mentions an experiment to emulate the style of Charles Peterson, a photographer known for his grainy, black and white images of the Seattle grunge scene. The term is used to discuss the aesthetic choices between different versions of Midjourney.

Highlights

Midjourney version 5.1 introduces two distinct modes, offering more stylistic variety and control.

Version 5.1 is a significant update, marking the first numbered update between major versions.

The RAW mode in 5.1 is unique and should not be confused with version five.

Users can switch between different versions, including 5.1 and RAW mode, through forward slash settings.

Seed numbers in Midjourney determine the initial pattern for image generation, ensuring unique outputs.

Seed numbers are finite and may change over time, affecting the consistency of image generation.

Version 5.1 produces more stylistic and expressive images compared to version five.

RAW mode offers a balance between the stylization of 5.1 and the more standard look of version five.

Version 5.1 RAW has improved sharpness and is better suited for specific imagery and longer prompts.

Experiments with Charles Peterson's style show 5.1 has better finger and instrument accuracy in images.

Version 4 captures a grainy, black and white chaotic look better than version 5.

Choosing between version 4 and 5.1 is becoming more of an aesthetic choice due to their solid results.

Version 5.1 has notably improved sharpness, particularly in paper quilling type images.

Permutations, a lesser-used feature, allows running multiple versions of Midjourney off the same prompt.

Version 5.1 is ideal for illustrative or artistic styles, while 5.1 RAW is better for photographic or cinematic images.

The speaker, Tim, will continue experimenting with version 5.1 and RAW mode and encourages sharing discoveries.

Version six of Midjourney is anticipated to be discussed soon, promising further advancements.