Ho RISPARMIATO 800€ euro in un anno con QUESTI TRICK su Higgsfield — Transcript
Full transcript
- 0:01Higsfield is one of the most powerful
- 0:02AI generation platforms available, and
- 0:04once you learn how to use it well, you
- 0:06can create content with incredible
- 0:07speed. Today I'll explain how to use it
- 0:11as efficiently as possible, from
- 0:12choosing the right tools for the
- 0:14situation to creating cinematic videos
- 0:16that look like they were shot with €
- 0:18200,000 cameras. This is the platform I
- 0:22use the most, and once I understood how
- 0:23it really works, I started creating
- 0:25content faster and better than with
- 0:27almost any other tool on the market. If
- 0:30you're interested in discovering tools
- 0:32and news in the AI world, I post a new
- 0:34video here on YouTube every Tuesday, so
- 0:37leave a like, subscribe, and hit the
- 0:39bell so you don't miss any. When you
- 0:41log in to Higsfield for the first time,
- 0:44the first thing you'll notice is that
- 0:46it's organized into three main sections
- 0:48: image, video, and apps. Image is
- 0:50where you generate static images, video
- 0:52is where you animate those images, and
- 0:54apps are tools that improve your
- 0:56workflow with specific features, like
- 0:58creating multiple camera angles or face
- 1:00swaps. The key to not wasting credits
- 1:03is understanding which tool to use
- 1:04based on the work to be done, without
- 1:06making unnecessary generations. Now
- 1:08I'll explain everything to you in
- 1:10detail, step by step, but before
- 1:11starting, I wanted to tell you that if
- 1:13you're taking your first steps with AI
- 1:14and want to turn this skill into a real
- 1:16job, you'll find the link to join my
- 1:18community in the description. Now we
- 1:20can get started. Click on "image" at
- 1:22the top of the page, and it will take
- 1:24you to the image generation interface.
- 1:26Most beginners go wrong here because
- 1:28they think one model is as good as any
- 1:29other. They don't realize that if there
- 1:32are so many models on Higsfield, it's
- 1:33because each one is designed for a
- 1:35specific function. If you choose the
- 1:37wrong model for what you're trying to
- 1:39create, you'll waste credits generating
- 1:42images that don't even match your
- 1:44vision. For example, for
- 1:45ultra-realistic images, you can use
- 1:47Nano Banana Pro. If you want a more
- 1:49artistic look, Sedrive 4. Gre is an
- 1:52excellent choice. Or, if you are
- 1:54generating images of a specific
- 1:55character, it makes sense to use Xfiel
- 1:57Soul with the character feature, which
- 1:59is built specifically for that. So
- 2:02before starting, think about what type
- 2:04of image you need because that choice
- 2:06will determine the model you're going
- 2:07to use. Once you've selected the model,
- 2:10you need to write the prompt. If you
- 2:12have trouble doing that, this formula
- 2:14will definitely make your job easier.
- 2:16Develop the prompt like this: subject,
- 2:18plus action, plus environment, plus
- 2:21lighting, plus camera angle. Instead of
- 2:24writing "a man playing at the casino,"
- 2:26write: "A man in a suit with a loosened
- 2:29tie, visibly tired and angry, is
- 2:31sitting at a Las Vegas casino table
- 2:33playing roulette." Behind him are other
- 2:35tables with other people playing
- 2:37various games. Light comes from above,
- 2:39illuminating the player's face well.
- 2:41The camera captures him from the side.
- 2:43The more specific you are, the better
- 2:44the result will be. Now set the aspect
- 2:46ratio and then choose the quality. Nano
- 2:49Banana Pro allows you to generate in 1K
- 2:51, 2K, or 4K. Then decide how many
- 2:53variants you want. I usually choose two
- 2:56, so I can then select the best one
- 2:57without wasting time. And then click
- 3:00generate. Once the image has been
- 3:02generated, there is a very simple way
- 3:04to transform that single image into
- 3:06nine completely different professional
- 3:08shots without having to generate
- 3:09anything from scratch. And this is
- 3:11where Xfield apps come into play, which
- 3:13, if used well, really make the
- 3:15difference. Go to the apps section at
- 3:18the top; you will see several tools,
- 3:19but the ones we will see today are
- 3:21Shots, Skin Enhancer, and Angles. Let's
- 3:24start with Shots. You just need to
- 3:25upload your image and generate. This
- 3:27tool. Will generate nine unique camera
- 3:30angles starting from that single shot.
- 3:32Choose the ones you like, upscale them,
- 3:34and now you have multiple professional
- 3:36shots without having to generate
- 3:37anything from scratch. It is super
- 3:39functional for creating variety while
- 3:41maintaining consistency. For example,
- 3:43if you are building a video, you can
- 3:44get nine different perspectives in a
- 3:46few seconds that all make sense with
- 3:48each other. Now let's see the second
- 3:50app. If there is a person in your shot,
- 3:52this is the step that makes the
- 3:54difference between an image that looks
- 3:55like AI and one that looks real. The
- 3:58app is Skincer, and it perfects the
- 4:00skin texture, making the image look
- 4:02significantly more natural. You have
- 4:04three options: soft skin, realistic
- 4:06skin, imperfect skin. For most cases,
- 4:09realistic skin is the best choice. Now
- 4:12let's see the last app, Angles, which
- 4:14allows you to create custom
- 4:16perspectives from a single shot. If you
- 4:18want a drone shot, a low angle, or a
- 4:20close-up, this tool reconstructs the
- 4:22image from the new angle you chose,
- 4:24while keeping everything else
- 4:26consistent. Images are only a small
- 4:28part of what Higgsfield can do. You see
- 4:31the real power when you start animating
- 4:33them. Click on video at the top. As you
- 4:35can see, the interface here is
- 4:36different; the controls are on the left
- 4:38and the generations appear on the right
- 4:39. You have two options: text or image
- 4:42to video. Text to video creates a video
- 4:45from scratch based on your prompt.
- 4:47Image to video takes a static image and
- 4:49animates it. In most cases,
- 4:51Image-to-video is better because you
- 4:53have full control over the starting
- 4:54frame. You know exactly how the
- 4:57character looks, the environment, and
- 4:58the lighting. You just have to tell it
- 5:00how to animate it. I'll show you both
- 5:02methods so you can see exactly how to
- 5:04use them. I'll start with
- 5:06image-to-video because it's what I use
- 5:07the most. I'll type: shallow depth of
- 5:10field and a cinematic shot. The
- 5:13character looks down at the roulette
- 5:15table, breathes deeply, and slowly runs
- 5:17a hand through his hair in frustration.
- 5:20Slight camera shake. Realistic ambient
- 5:23casino sounds. As you can see, I also
- 5:25specified the sounds because VO 3.1 has
- 5:28integrated audio. If you don't say what
- 5:30you want, it might add random dialogue
- 5:32or weird noises. Set the duration to 6
- 5:34seconds to avoid wasting too many
- 5:36credits, then click Generate. Look here
- 5:38, everything is exactly as requested
- 5:40and the final result looks like it came
- 5:42straight out of a movie. Now let's see
- 5:44what happens when you don't have a
- 5:46starting image and have to create
- 5:47everything from a single prompt. In
- 5:50this case, I have to describe
- 5:51everything since the tool has no
- 5:52reference. I'll type: a blonde woman
- 5:55touching her hair as if posing for a
- 5:57photoshoot. Behind her is a skyline.
- 6:00It's nighttime, and the light is a
- 6:01flash pointed directly at her. The
- 6:03camera vibrates slightly. I click
- 6:05Generate, and here is the result. The
- 6:07AI created the entire scene out of thin
- 6:09air. By doing this, you don't have full
- 6:11control like with image-to-video
- 6:13because it makes decisions on
- 6:14everything you don't specify. But for
- 6:17quick concepts or when you don't have a
- 6:19specific image in mind, it works very
- 6:21well. The key is always knowing when to
- 6:23use each tool. If you need consistency
- 6:26and control, use image-to-video. If you
- 6:29are testing ideas or need something
- 6:31quickly, text-to-video works just fine.
- 6:33Higsfield also has a feature that
- 6:34competes with the quality of Hollywood
- 6:36productions. Cinema Studio. This tool
- 6:39allows you to generate images and
- 6:41videos as if you were using
- 6:42professional cinematic equipment. You
- 6:45can choose tools like Modular 8K
- 6:47Digital, large format 70mm, Vintage
- 6:50Prime lenses, or Warm Cinema—gear
- 6:52that would cost hundreds of thousands
- 6:55of euros in real life. Click on Cinema
- 6:58Studio from the homepage and you will
- 6:59see two modes: image and video. Let's
- 7:02start with image. Select the
- 7:03combination you want between camera and
- 7:05lens. For example, in this case, I'll
- 7:08use Full Frame Cine Digital with
- 7:10Classic Anamorphic at 35mm because it
- 7:12gives that cinematic movie look. If I
- 7:15wanted something cleaner and sharper, I
- 7:17would have chosen the Modular 8K
- 7:19Digital with Warm Cinema Prime. This
- 7:21combination is perfect, for example,
- 7:23for commercial work. Then type the
- 7:25prompt, for example, a sports car
- 7:27speeding down the road, then click
- 7:29Generate. As you can see, the AI
- 7:31creates an image that looks like it was
- 7:33shot with a $ 200,000 camera. Now take
- 7:35that image and switch to video mode.
- 7:37The AI will animate it while
- 7:39maintaining the same cinematic quality
- 7:41and keeping everything consistent. This
- 7:43is what sets Higgsfield apart from
- 7:45platforms like Kling or Runway. You're
- 7:47not creating those typical
- 7:48plastic-looking videos; you're creating
- 7:50videos that look like movies. But, even
- 7:52though you can create visuals that look
- 7:54like high-budget films, you still need
- 7:55your characters to remain consistent.
- 7:58Before I explain how to do that, I
- 7:59wanted to remind you that in the
- 8:01description, you'll find the link to
- 8:02join my community, where you can find
- 8:04in-depth training on every aspect we're
- 8:06looking at today. For this, we'll also
- 8:08use a specific function: the character
- 8:11function. Go to the character section,
- 8:13click create character, and upload 20
- 8:16to 30 reference photos. Make sure they
- 8:18are clear, well-lit, and show your
- 8:21subject's face from multiple angles.
- 8:24I'll show you how to create a custom
- 8:25character completely from scratch,
- 8:27which you can use as a consistent model
- 8:28across the various campaigns you'll
- 8:30create. I go to the image section and
- 8:33enter this prompt. I write: "A man in
- 8:36his 20s with dark hair styled forward,
- 8:38blue eyes, square jaw, designer
- 8:41streetwear, neutral expression, white
- 8:43background." In this case, I'm using
- 8:46Nano Banana Pro for a super realistic
- 8:48result. If you've been generating with
- 8:50AI for a while, you'll realize how easy
- 8:52it is to lose track of things and be
- 8:53disorganized. That's why you need to
- 8:55use the Assets Library. Click on Assets
- 8:57Library at the top. This is your
- 8:59library. Every image and video you've
- 9:02created is here, and you can create
- 9:03folders to organize your projects
- 9:05properly. This keeps everything tidy
- 9:08and makes it easy to find what you need
- 9:10when you have to edit something or look
- 9:12back at an old campaign. Perfect, now I
- 9:15have the base, but as we saw a moment
- 9:17ago, one image isn't enough; I need
- 9:19more angles, so I take the shot to the
- 9:21Angles app, upload it, and generate
- 9:23different perspectives. The app creates
- 9:26all these images while keeping the face
- 9:28perfectly consistent. If you are
- 9:30creating a custom character, this is
- 9:32exactly the workflow you need to use.
- 9:34You generate the base and then use
- 9:36angles to create more references. After
- 9:38downloading the images, go to the
- 9:40character section, click create
- 9:42character, and upload them all. I'll
- 9:44name him Gino. It processes the images
- 9:46and creates a model based on that face.
- 9:49Once completed, Gino will be saved in
- 9:51my character library. I go back to the
- 9:54image section, select Xfield Soul as
- 9:56the model, choose Gino as the character
- 9:58, and type: Gino wearing a Japanese
- 10:00robe in front of a Japanese temple.
- 10:03Cherry blossom trees are visible in the
- 10:05background, morning light. Click on
- 10:07generate and there it is. It’s
- 10:09identical, but in a completely
- 10:10different setting and with different
- 10:12clothes. Creating consistent characters
- 10:15has always been a huge problem for
- 10:17creators, but Xfield solves it with
- 10:19this feature. Well, that was a general
- 10:21overview to learn how to make the most
- 10:23of Xfield. If you liked the video,
- 10:25leave a like. If you want to watch more
- 10:27similar videos, follow me and we’ll
- 10:29see you in the next one.
About this transcript
This page contains the full transcript of Ho RISPARMIATO 800€ euro in un anno con QUESTI TRICK su Higgsfield by Lumonboy, generated from the public captions YouTube serves with the video. The transcript has 1,975 words across 304 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.