Imagining Los Angeles during the holiday season - made in Midjourney

We used Midjourney to craft a vision of what Los Angeles feels like during the festive holiday season. We uploaded our own street photography to Midjourney using the describe feature to understand what qualifiers and keywords could generate works to match our original works.

Midjourney description of our holiday LA street photography

It's always fascinating to see how computer vision interprets our original works and likens it to the training material that Midjourney has been fed - we always find new artists and genres we never knew about when the descriptors show up in the /describe output - the qualifier of "found film" and "found footage" ended up yielding interesting results in future generations.

Here is an example prompt from the series, made using Midjourney v 5.2 and 6.0 alpha:

christmas time in los angeles, christmas, los angeles, angelenos, a view of the hollywood sign in the style of street photography shot on color film --v 6.0 --style raw

It was fascinating to see the same looking christmas tree copy and pasted all over different parts of los angeles, we would have to add more qualifiers to change the appearance of the tree like "tinsel covered" or "presents everywhere", etc.

Further it was interesting to see the differences between Midjourney 5.2 and 6.0 - 6.0 is markedly less surreal and more photo like without the need for additional photography qualifiers like "f/22, UHD, 8k photography...." check out our results below:

Check out the rest of our AI experiments in our Generative Gallery!

Leave a comment

Please note, comments must be approved before they are published