Is Signal Studio coaching or consulting?

Neither, exactly. There’s no curriculum, no modules, no homework. Signal Studio is creative direction — we identify where your presence is misaligned, design the system that corrects it, and direct its execution. The relationship is closer to working with a creative partner than a coach or a consultant.

Who is Signal Studio not right for?

If your offer is still being validated, if you need deliverables in under 30 days, or if you’re looking for content volume rather than brand alignment, this isn’t the right engagement. We discuss fit on the call before any commitment.

What happens after I apply to Signal Studio?

Every application is reviewed personally within 48 hours. If there’s potential alignment, you’ll receive an invitation to a private call. No sales process — just a conversation about whether this is the right fit for where you’re headed.

Free guide

Recreate any videostarring your own AI avatar.

Build a consistent digital character in Higgsfield once — three images — then drop it into any video you want to repeat with Genjutsu. Same movement, same framing, your character.

Avatar3 images · once

ThenAny video, on repeat

Before you start

You need a Higgsfield account.

Higgsfield is where both the avatar and the videos are made. Everything in this guide happens inside it — no other tools needed.

Create your Higgsfield account

1Sign up using the link above.
2Pick a plan. Image and video generation both run on credits, and video uses them fastest — make sure you have enough for Part 2.
3Part 1 happens in Image → Create Image, Part 2 in Genjutsu — both in the top bar.

Part 1 · Avatar

Build your avatar once. Use it forever.

Three images give Higgsfield everything it needs to keep your character consistent from every angle: a clean full-body photo, a 360° body turnaround, and a 360° face sheet. Make them in order — the first image becomes the reference for the other two.

Each prompt comes as a Template — replace everything in [BRACKETS] with your character. Switch to Example to see the exact prompt I used. Keep the details identical across all three prompts; repeating them is what holds the character together.

Main photo

The master reference. One clean, full-body image of your character, head to toe.

360° body

Eight views, front to back. Holds the body, outfit and proportions when the character turns.

360° face

Six angles and a close-up. Holds the face steady from every side.

Want the avatar to be you? Upload a few clear photos of yourself as references in step 01 — the prompt does the rest.

In Higgsfield · Image

Generate the main full-body photo

Go to Image → Create Image. I use GPT Image 2.5 Sunburst.
Making an avatar of yourself? Click Upload media and add 2–3 clear photos of you — face and full body. Inventing a character? Skip this.
Paste the prompt below with every [BRACKET] filled in and hit Generate.
Pick the result you'd be happy to see in every video — this is your character's master reference.

01 · Main full-body photo

Photorealistic full-body fashion character reference photo of an original [WOMAN / MAN] in [HER / HIS] [AGE], [OVERALL VIBE]. [LOOK & SKIN], [HAIR], [EYES], [BROWS, MAKEUP, LIPS]. [BUILD], relaxed confident posture, standing straight facing the camera, arms relaxed at [HER / HIS] sides, calm self-assured expression with a hint of a smile. Wearing [OUTFIT], [ACCESSORIES], [FOOTWEAR]. Seamless neutral light-grey studio backdrop with soft even lighting, full body head to toe in frame with space above and below. Sharp realistic skin texture, true-to-life colors, editorial model casting photo, no text, no logos.

Click to pause

In Higgsfield · Image

Generate the 360° body turnaround

Still in Create Image, add the main photo from step 01 as the reference — it's under Generations.
Paste the prompt — it asks for eight full-body views, front to back, with the same face, outfit and scale.
Check every view: both feet visible, nothing cropped, outfit unchanged. Regenerate if any view drifts.

02 · 360° body turnaround

Character turnaround 360-degree model sheet of the exact same original [WOMAN / MAN] from the reference image, identical face, hair, body and outfit in every view. Eight evenly spaced full-body standing views in one row, left to right: front, 45° left, left profile, left back, back, right back, right profile, 45° right. [SHE / HE] stands relaxed and upright, arms loose at [HER / HIS] sides, full head-to-toe framing with both feet visible in every view, not cropped. [AGE], [SKIN], [HAIR], [EYES], [MAKEUP & FACIAL STRUCTURE]. Wearing [OUTFIT — SAME AS PROMPT 01], [ACCESSORIES], [FOOTWEAR]. Seamless neutral light-grey studio backdrop and floor, soft even diffused studio lighting, consistent scale and lighting across all views. Small thin uppercase grey sans-serif captions under each figure: FRONT, 45° LEFT, LEFT, LEFT BACK, BACK, RIGHT BACK, RIGHT, 45° RIGHT, and a centered caption '360° VIEW' with thin lines on either side at the bottom. Photoreal, natural skin texture, no airbrushing, single character only, no other people, no props, no logos, no watermark.

Click to pause

In Higgsfield · Image

Generate the 360° face sheet

Same flow: add the main photo from step 01 as the reference. Making an avatar of yourself? Add your clearest straight-on selfie too.
Paste the prompt — seven head-and-shoulders views plus a large close-up.
Zoom in on the close-up. If the face doesn't match step 01, regenerate before moving on.

03 · 360° face sheet

Face 360-degree reference sheet of the exact same original [WOMAN / MAN] from the reference image, identical face and hair in every view. Grid of seven head-and-shoulders portraits on a seamless neutral light-grey studio backdrop: top row — front straight-on neutral expression, 3/4 left, left profile, back of head showing hair; bottom row — right profile, 3/4 right, and a larger tight straight-on face close-up. [AGE], [SKIN, FRECKLES / MARKS], [HAIR], [EYES], [BROWS, NOSE, LIPS], [JAW & CHEEKBONES], [MAKEUP], [EARRINGS / JEWELLERY], [COLLAR / NECKLINE] visible. Calm neutral expression, same lighting in every panel: soft diffused studio light. Visible fine skin texture with natural pores, subtle asymmetry, no digital smoothing, no beauty filter, no plastic or glossy skin. Small thin uppercase grey captions under each panel: FRONT, 3/4 LEFT, LEFT, BACK, RIGHT, 3/4 RIGHT, CLOSE-UP. Photoreal, single character only, no other people, no logos, no watermark.

That's your avatar — three images you'll reuse for every video. Keep them together in one folder.

Part 2 · Genjutsu

Now put your avatar into any video.

Genjutsu takes a reference video and recreates it with your character — the movement, the framing, the timing. Pick something you've seen work, and make it yours.

In Higgsfield · Genjutsu

Recreate a video with your avatar

Click Genjutsu in the top bar.
Browse the Trending feed, find a video you want to repeat and hit Recreate — it loads into Motion transfer.
Click Add your characters, products, or scenes, open Generations and select your three avatar images. Extra generated images — a location, an outfit — go here too.
Turn on Prompt and paste the prompt below.
Check the model is Higgsfield Genjutsu, pick a quality and hit Generate. My run was 168 credits at 720p.

Genjutsu prompt

Keep everything the same in the video, just replace the [FACE / FACES] with the [PERSON / TWO PEOPLE] in the images provided.

Click to pause

Done — same avatar, new video. Pick another reference and run step 04 again; your avatar images stay in Generations.

Artur Rioni

Creative Director, Signal Studio

Work with me →