On March 7 and 8, 2024, Midjourney released an updated version of its /describe feature, now offering longer and more detailed prompts tailored for version 6. This enhancement is available on both Discord and the website, allowing users to generate text prompts from images with improved accuracy. Users have shared their experiences, noting the feature's enhanced descriptiveness and its ability to produce prompts that are much closer to the original images, even with random groups of objects. The update also introduces the use of the --sref parameter, enabling a new workflow where users can generate a text prompt using the original image and then run this prompt with the --sref parameter for potentially more accurate descriptions. The community has tested the feature with various images, including faceswaps of celebrities, noting the improved performance over the previous version.
Midjourney just released the new version of /describe and it's wayyyy better than the previous version It produces longer and more detailed prompts that are specifically tuned for v6 (see below, it's way less "vibey") Currently only works in Discord using the /describe command https://t.co/9doaftfVI5
Midjourney testing new algorithm for /describe https://t.co/UzhXe7hmNV
We're testing a new version of /describe today. Basically, show it an image and it'll try to describe it. This new algorithm is made to give longer, more detailed prompts, in the style of our community and is specifically tuned for our V6 models. Enjoy!