EP 048 · 60 MIN MIDJOURNEY VIDEO GOES HD FOR EVERYONE

Midjourney Video Update: HD for All & Batch Size Changes + Our Style Explorer Wishlist

August 16, 202560 min Rory Flynn & Drew Brucker

The short version

Episode 48 catches up on Midjourney's video update: batch sizes of one or two instead of four, and HD video mode now open to standard-tier subscribers. Rory and Drew also dig into Object Weight (Omni-Reference) tuning, preview the teased Style Explorer feature, and compare GPT-5 to 4.0, concluding it feels largely unchanged. They close by weighing Veo 3's native dialogue and Kling's optional sound-effects download against Midjourney's still-silent video.

Key takeaways

  1. 01

    Midjourney video now supports batch sizes of one or two (not just four), so testers can burn fewer fast hours per generation.

  2. 02

    HD video mode is now open to standard-tier subscribers, but Rory still tests in 480p first since HD burns a month of credits fast.

  3. 03

    For Omni-Reference, Rory finds Object Weight above 400 stops helping; 50-150 works best for swapping a character's clothes without losing consistency.

  4. 04

    To keep an object still in a Midjourney video extension, name its exact position (e.g., 'locked in the right side of the frame') instead of just saying 'static.'

  5. 05

    When extending a Midjourney video, describe the action, camera motion and environment interaction, not style: Midjourney already knows and keeps the style consistent.

  6. 06

    Rory and Drew say GPT-5 feels no different from 4.0 despite the hype, and expect Midjourney's own V8 hype cycle to face the same skepticism.

  7. 07

    Kling lets you download a video raw or pre-mixed with sound effects; Rory wants Midjourney to add the same either/or option once it ships audio.

Receipts · said on the record

“I honestly think like 70% of AI video is understanding how to troubleshoot it. It's like the prompting, like a lot of times the models are good enough to make up for bad prompting on certain things.”
Rory Flynn Why most AI video difficulty is troubleshooting, not raw model power Play from 57:02
“The higher you go on, on OW, like if you go OW 1000, it pulls in background too.”
Rory Flynn How Object Weight intensity changes what Midjourney pulls into a shot Play from 15:20
“Nine out of 10 of these ain't gonna work for what I'm gonna actually show, right? And so I think like the curation aspect is maybe the key of that.”
Drew Brucker Why most third-party style-code libraries fall short, Drew says Play from 24:33
“Why do we end up going down like a gruesome rabbit hole majority of time on this? Like it's like 0.2% of what we probably both generate and it somehow ends up on here.”
Rory Flynn Why the hosts' oddest Midjourney tests always go dark Play from 29:10
“There's four billion codes. So Style Explorer, what are you gonna type in blue? And then you've got 500 million style codes to go through. Blue illustration, now I've got 50 million to go through.”
Drew Brucker Drew on why simple search won't fix Style Explorer's scale Play from 21:20

Chapters

  1. 0:00 Cold Opener
  2. 8:35 Midjourney Update Rundown
  3. 9:04 Video Batch Sizes 1–2 + HD Mode
  4. 9:49 Moodboards Sidebar Update
  5. 11:27 V7.1 Notes + O-Ref Talk
  6. 12:14 Early V8 Teasers
  7. 18:53 Style Explorer Preview
  8. 20:16 tyle Explorer Wishlist + Trends
  9. 27:21 Midjourney TV
  10. 29:38 Token Talk: “Homesteader”
  11. 32:14 EXP Parameter Live Test
  12. 33:45 EXP vs Stylize Workflow
  13. 34:30 Stylize Confessions + Image Ranking
  14. 38:39 GPT-5 Hype vs Reality
  15. 40:02 Sora + Coherence Issues
  16. 45:38 HD Video in Practice
  17. 46:37 Missing “SD to HD” Button
  18. 48:21 Manual Prompting for Extensions
  19. 54:44 Text + UI in Midjourney Video
  20. 56:40 Video Troubleshooting Lab Idea
  21. 57:33 Fixing Video Coherence
  22. 58:57 Outro + Subscribe

Questions this episode answers

What changed in Midjourney's video update covered in this episode?

Midjourney added smaller batch sizes (one or two clips instead of four) and opened HD video mode to standard-tier subscribers; mood boards also got a dedicated sidebar tab instead of being buried under Personalize.

How does Object Weight (Omni-Reference) affect Midjourney results?

Rory found that pushing Object Weight much past about 400 stops adding fidelity and starts looking 'weirder'; for swapping a character's clothes while keeping them consistent, he found roughly 50-150 worked best.

What is Midjourney's Style Explorer?

Style Explorer is an unreleased Midjourney feature the hosts discuss based on Office Hours notes. Rory and Drew argue it needs trending or categorized style codes and community features, rather than plain text search across what Drew estimates are billions of codes.

Did Fast Hours think GPT-5 was a big upgrade over GPT-4?

No. Rory and Drew say GPT-5 feels largely unchanged from 4.0 for how they use it daily, and compare the hype cycle to what they expect Midjourney's own V8 launch to face.

How do Kling and Veo handle audio compared to Midjourney video?

Veo's native dialogue, which Rory says 'changed everything' when it launched, and Kling's option to download a clip raw or with baked-in sound effects both outpace Midjourney, which still ships silent video as of this episode.

Show notes

Fast Hours Podcast Ep. 48—Midjourney just unlocked HD video for everyone, added new batch size controls, and teased more of Style Explorer, V7.1, and V8. Here’s everything you need to know.

Two bleary-eyed nerds test Midjourney’s newest toys so you don’t torch your credits. In this run-and-gun lab, Drew Brucker and Rory Flynn cover the video update (batch size 1–2, HD now available), why Moodboards finally lives in the sidebar, and what V7.1 is poking at while V8 hogs the spotlight.

They preview Style Explorer, debate O-Ref's re-emergence, and trade practical tactics for coherent, extendable shots without wasting fast hours. Expect detours into style codes, tokens (hello, “Homesteader”), and EXP vs Stylize in real workflows.

Translation: a friendly chaos sandwich with updates, experiments, and a little self-roasting, while laying out clear takeaways you can steal today.

Every word Read the full transcript

If you liked this one