How to make Sora music videos with David Sheldrick
9:25

How to make Sora music videos with David Sheldrick

OpenAI 23.04.2025 52 178 просмотров 1 606 лайков

Machine-readable: Markdown · JSON API · Site index

Поделиться Telegram VK Бот
Транскрипт Скачать .md
Анализ с AI
Описание видео
Sora Fellow David Sheldrick shares an inside look at his latest project—a vibrant, multicultural twist on Victorian elegance wrapped in a K-POP dance reel. David, a British-Korean visual artist based in London, walks us through his workflow for crafting music videos in Sora, revisiting an award-winning photo shoot from his early career, and blending past inspiration with contemporary creative storytelling. David Sheldrick is a graduate of the London College of Fashion in Fashion Photography with a keen interest in image assembly, nature, and technology. He runs PS Spaces in London. We hope that you get inspired to remix your own creative visions with Sora!

Оглавление (5 сегментов)

Intro

Hi, my name is David Sheldrick and today I'm going to show you how I make videos like this. Oh my god, I'm just playing with them now. Yeah. I used to do DP and direct a lot of music videos before COVID. We followed like a kind of a bit of a format to make sure we could film the whole thing in one day. And then after we've sort of shot one location in one way, we then move to location two. We do the same thing again. Location three, same thing again. I basically decided to kind of like take this approach when making videos or music videos with Sora. So step one is

Step 1 Creative

creative. Before we even start assembling anything, we spend a good sort of at least half a day or a day just messing around with different creative directions. One of the best places to you know do research or to find new creative is on the explore pages of you know for example Sora. So if you go to the Sora website you will see there's a big feed on the homepage of what everyone else is rendering and you get to see their prompts as well. So this is one of the fastest ways not only to sort of look at a whole massive variation of creative but also a very fast way to learn how to prompt. You know there's two parts to this. There is sort of what I would say the scenes which is you know location one, location two, location three. We're going to leave that for now. But one thing we are going to think about is what does this world look like? So, you know, is it a period piece? Is it a sci-fi? Is it set in the countryside? grand palace? At this point, we are going to essentially do a bit of world building. So, I'm going to use chat GPT to help us with creating the style or the preset. Um, so for this I put expand this prompt and make it more detailed for use in Sora video rendering as a preset. I have created a little diagram here for you which basically shows you know our style as the whole. So 18th century Marian Twuinette as like uh the overall or overarching style and within that we have creative one, creative 2, three, four, five, six and you can keep going. So for creative one I've got hair and makeup close-ups. Creative two, uh, shots within sort of hallways and ballrooms in these sort of like grand palatial rooms. Uh, creative three, hunting, and then creative four are gardens. So, in Hampton Court Palace in London, if you've ever been, has a big hedge maze. It's always perfectly cut. So, uh, creative five is horses. And then creative six, kinugi models. So, kinugi is a form of Japanese art where they essentially they fix broken pottery using gold. So, you see these beautiful sort of old wares that have been recrafted uh with gold to fill the cracks. And I thought that's, you know, it's quite beautiful. So, next we are

Step 2 Rendering

moving on to rendering. We've got all of our creatives. Uh, we got our style. You know, just take that style, you go to manage presets, and you paste it in here. And done. And then you can come back and modify that. I just run this many I run each creative many times. And I do chuck in sort of like a second prompt here, which is usually a dance sequence. So, here you got close-up of a Korean K-pop model getting her hair and makeup done, wearing a huge 18th century Marian Twinette wig, white powder makeup, bolt camera shot of ethnically diverse K-pop couture fashion while dancing in unison, dancing in a Queen's bedroom, crunk dancing, street dance, dancing with attitude, dynamic dance, movement, dynamic music, video, camera work. Next is the music. I

Step 3 Music

usually use artlist. io for my stock music, but the quality is generally really good or at least for now much better than AI generated music. So the first thing that I do is I drag the track in.

Step 4 Assembly

The assembly is essentially taking all of our footage that we've rendered and we drag it all into the timeline and we create what I call a sausage. uh sausage is just all of the footage in one straight line just to show you kind of like a basic structure that we'll try to stick with which you know is sort of comes from our creative structure that we talked about earlier and I showed you earlier. So I have already dragged all mine in. So you can see here if I quickly scrub through. So, what I'll first do is with these shots that essentially have two cuts in them, chop that there. And so, we'll listen to the intro one more time. There's the bass hit. Right. But this piece of footage is like three times longer. So, go into this command R accelerate. Let's try accelerate it 300%. I'm going to trim the front a bit. I want to just make this piece of footage. I want her eyes to just open just as the base hits. I then want it to go into this macro shot of this girl with the lace gloves. So, let's see. Cool. Okay. So, I have added a few more in. Now, I'm just speeding things along now. So, I'm not going to do that entire assembly in front of you. Uh that will take one or two hours. So, I've spent sort of another 20 30 minutes just getting more of an assembly in. And yeah, so we have now uh a few more scenes dropped in. So, I'm going to press play uh just from Creative 2. We'll press play and then I'll show you sort of how the next scenes line up. and I'm just going to keep on adding in. I'm working through these creatives. I'm going to make sure I'm cutting at the drops. Cool. So, I have now jumped ahead and I have finished the assembly. It took me about 4 hours in total from the B to the A with the in the valley up six Kelly F you pay me like my name was Sally made I hopped on the planet catch a play in the don't drown trying to catch this wave Ain't no stopping me. It's a strong rubbery. It's been a patch of company design on your feet just to me. Ain't no stopping me. It's a strong robbery. They can't do is for me just to say the least. I mean your hair free and I ain't sign the least. And you know we blow big bags hauling them all. Yeah, we popping them big tags. Make a pro do whiplash. If we dripping too hard like we leaning on kickstand. Have you ever blew a couple stacks? Forget that. Lost 100 racks. Ain't tripping no you get it back. I'm the bounce back kid. I invented that. I don't like energy. So I'm burning up the sage and I'm burning bridg just to play friends trying to hate. I came in first place but still showed up later trying to couple too like someone up the lad and I ain't going to play baby girl I can't slay. Kill them all dead no ain't no party till I walked in and turned out the play. Ain't no stop me going.

Другие видео автора — OpenAI

Ctrl+V

Экстракт Знаний в Telegram

Экстракты и дистилляты из лучших YouTube-каналов — сразу после публикации.

Подписаться

Дайджест Экстрактов

Лучшие методички за неделю — каждый понедельник