AI videos made with Midjourney 6.1 and Runway Gen-3

Подписаться 1,7 тыс.

Просмотров 1,6 тыс.

50% 1

I used Runway Gen-3 on Midjourney 6.1 images and got some amazing results. This tech just keeps getting better and better! Here are some of my best results collected in this video.
If you like this content please like, comment and subscribe, it helps small channels like this to get noticed! Also, please comment if you are interested in any prompting details or if there's any images /videos you'd like to see more of!
#scifi #fantasy #generativeai #aiart #midjourney #midjourneyV61 #runway #runwaygen3

Опубликовано:

10 сен 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 25

@AlternateRealities4732 Месяц назад

Please leave a like and a comment to help me out with the youtube algorithm! Even just a simple "hello!" helps out massively! Subscribe for ongoing Alternate Realities content! Other videos... Midjourney V6.1 has dropped and it is INSANE : ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-YXcMgHvZiFo.htmlsi=cQ8LVB1MbloCpvFX AI FILM - Primal - AI generated short video : ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-CL2HuttPSYA.htmlsi=zrd156qDkpPy_hu6 AI Video - DYSTOPIAN SciFi Visions | Runway Gen-3/Gen-2, Luma : ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-jwm1A8ikWBI.htmlsi=f5s4PNgBvB9kHWmZ AI Video - Cyberpunk Visions | Runway Gen-3 : ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-7diwNGA9xpA.htmlsi=SOW1B90T1UFx3Hjc AI FILM - The Magi - AI generated short | Runway Gen-3 - ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-RWzbsXL_05M.htmlsi=Zdezgn7qLuhInMat AI Art and Music - Darkwave SciFi : ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-OfLl78p0thk.htmlsi=mMN8l0vgbS8DBe-C

@05minprod 20 дней назад

These are incredible. Not sure if you've done this but it would be cool to show the Midjourney image, your prompt and then the final video. Well done!

@AlternateRealities4732 20 дней назад

I could do this in future videos like this, as Midjourney releases new versions or features, etc.

@CpourVous 19 дней назад

Very good, the monkeys and war scenes are amazing.👍❤️

@AlternateRealities4732 18 дней назад

glad you liked them, I thought they came out cool to :)

@nuuttikataja666 11 дней назад

Good work!

@AlternateRealities4732 10 дней назад

Thanks! 😊

@Ai-Art-42 Месяц назад

Fantastic results, i think I still need to practise a bit

@AlternateRealities4732 Месяц назад

yeah, there are so many ways to prompt in midjourney, there's a whole universe hiding in that latent space. And Gen-3 image-to-video is fickle, sometimes it takes a lot of generations to get a few good results

@Ai-Art-42 Месяц назад

@@AlternateRealities4732 expensive hobby, or

@AlternateRealities4732 Месяц назад

@@Ai-Art-42 well it does pretty much require the unlimited plan to get useful results from Gen-3, which ain't cheap ;)

@cinerobot01 Месяц назад

Great work!

@AlternateRealities4732 Месяц назад

Thanks!

@JamesELarkin Месяц назад

Wow

@AlternateRealities4732 Месяц назад

glad you liked it :)

@pasdedanse1386 18 дней назад

You should put this music on Spotify

@AlternateRealities4732 17 дней назад

hmm, I hadn't even considered that. Will have to keep it in mind!

@akmalsultanov638 Месяц назад

Very impressive. Could you make some tutorial for us, buddy?

@AlternateRealities4732 Месяц назад

hmm, I've never tried to make that kind of video but I'll consider it!

@user-ki3je4oe6d Месяц назад

Notice how there were no videos with realistic human hands.

@AlternateRealities4732 Месяц назад

Midjourney 6.1 does hands fine, most of the time, and for a single subject. If there's multiple people in a scene it still struggles sometimes. The video generators are hit or miss with hands still, they often mangle them badly. Its still early days for this tech. I just work around the current limitations.

@plinker439 15 дней назад

character motions are very roughed and/or cumbersome and i don't think it's getting better. That would need ridiculous amount of datasets for every possible body overlapping with every pose and possible motions and etc etc, i think that can only happen by an AI with consciousness.

@AlternateRealities4732 15 дней назад

its hit or miss, it struggles with motion in image-to-video at times but usually motions are smoother with txt-to-video or with more photorealistic images, etc. Also its not just clipping stuff together, its generating original video based on its model of how things move in videos which was trained by looking at billions of seconds of video.....which is an important distinction. Its going to get A LOT better, we're nowhere near the peak.