Was ist die Webseite?

Digital artworks by IzbacivacX

Startseite
Schweiz
Wabern
Digital artworks by IzbacivacX

AI Art Specialist / "Fiverr Pro" and "Top Rated" in "AI Artist" section of Fiverr
PM me for consulting.

AI Digital Illustrator working with Midjourney, Stable Diffusion and Leonardo AI

24/09/2026

Your reference image already describes the scene. Your video prompt should direct what changes. (Video in 1st comment)

When turning a still into video, repeating every visible detail can bury the actual motion instructions. A more useful structure is:

1. Subject action
2. Camera movement
3. Environmental movement
4. Audio in a separate sentence

Example:

“Eye-level medium shot. The bartender slowly turns toward the camera and places the glass on the counter. The camera performs a gentle dolly-in. The curtain moves slightly and reflections travel across the polished bar. Audio: quiet room tone and a soft glass-on-wood tap.”

The reference image carries the character, setting and composition. The prompt gives the shot direction.

Start with one clear action and one camera movement. Add secondary motion only where it supports the scene.

This doesn’t guarantee perfect consistency: the model may still alter faces, hands or object geometry. Google also notes that natural, consistent spoken audio—especially short speech—remains an area of active development.

23/09/2026

You can write an excellent prompt and still choose the wrong image model.

“Nano Banana” currently describes several Gemini image models—not one identical tool:

• Nano Banana 2 Lite prioritises speed and cost. Google says it is not optimised for multiple references or repeated conversational editing.

• Nano Banana 2 is the general-purpose choice for generation, editing and combining references. Its documented limits include up to 10 high-fidelity object images plus four character references.

• Nano Banana Pro targets complex professional assets, stronger world knowledge and precise control. It supports up to five character references and three dedicated style references—but requires a Google AI plan in the Gemini app.

Practical rule:

Use Lite for fast single-pass concepts, Nano Banana 2 for everyday production, and Pro when brand accuracy, several characters, typography or style references materially affect the result.

More references aren’t automatically better. Give every image one clear job: character, product, environment or style. If two references contradict each other, the model must decide which one to sacrifice.

22/09/2026

More frames won’t necessarily fix your AI video.

Catch-up: on 17 September, Runway released Enhance Frame Rate. It converts existing footage—including videos created outside Runway—to frame rates such as 30, 60 or 120 fps. Access is listed for Max+ plans, with support up to 4K and five minutes.

Why care? You can convert an existing clip to suit your delivery requirements without generating the scene again.

But frame interpolation fills the gaps between frames. Don’t assume it will repair a character changing shape or an object disappearing.

Here’s a practical check before processing your whole film:

1. Take five seconds containing the fastest movement.
2. Open Enhance Frame Rate through Runway’s Enhance menu.
3. Select your required delivery rate—60 fps, for example.
4. Keep resolution, duration and crop unchanged.
5. Compare the original and converted clip at normal speed, then inspect hands, faces, thin objects and cuts frame by frame.

Look for both smoother movement AND intact shapes.

Choose the frame rate your project needs—not simply the largest number available.

21/09/2026

Your rough sketch can be part of the prompt—not just inspiration.

If you’ve invented a shape that’s difficult to describe, upload a drawing instead of asking AI to guess it. Google documents this sketch-to-image workflow for Nano Banana 2.

Try it with a lamp idea:

1. Draw the outline clearly and photograph it straight-on.
2. Upload the sketch in Gemini.
3. Separate what must stay from what AI should add.

Copy this prompt:

“Turn this lamp sketch into a realistic studio product photograph. Preserve the bent stem, angled shade, flat oval base and their relative proportions. Use brushed brass for the stem and matte ivory for the shade and base. Keep the same viewing angle. Add realistic material texture against a plain light-grey background. Remove the sketch lines. No extra components or text.”

Then compare the silhouette with your drawing—not just whether the result looks impressive.

This creates a concept visual, not an engineering-accurate render. Check geometry before presenting it as a finished design.

Gemini image editing requires an eligible adult account; availability and usage limits apply.

20/09/2026

Generating AI clips is one thing. Turning them into a finished video is another.

Krea brought those steps closer together on 17 September: its Agent now has a timeline video editor.

You can arrange generated clips, trim them, add audio and export an MP4 inside the same workspace.

One useful detail: holding Ctrl on Windows or ⌘ on Mac while dragging a clip’s end starts a generative extension. An ordinary drag does NOT create new footage.

Why it matters: for a short promo, this could reduce the back-and-forth between generation and assembly.

The catch? The agent doesn’t operate the timeline for you. You still make the editing decisions.

Agent access is included with Max, Business, Creator Max and Enterprise; other plans need a Krea access code. Generations and extensions use compute credits.

Useful integration—not “one click and your film is finished.”

19/09/2026

Your AI edit looks pasted on? Check what’s happening around its feet.

A convincing chair needs more than the right shape and colour. It also needs to belong to the floor beneath it.

Look for three details:
① Contact shadow: the tight shadow where each foot touches the floor.
② Cast shadow: the larger shadow, following the room’s light direction.
③ Reflection: only when the floor material calls for one. Carpet doesn’t need a mirror image!

Here’s the useful part: when selecting an area for AI retouching, include the affected floor—not just the chair. The surrounding surface may need changing too.

In Magnific’s Image Editor, use Retouch and try:

“Ground this chair naturally on the floor. Add tight contact shadows beneath its feet, match the cast shadow to the window light from the left, and add a subtle reflection appropriate to the polished floor. Preserve the chair’s design and the room.”

Inspect the feet and floor afterwards: AI can alter details you wanted to keep.

Save this three-point check for your next object replacement.

17/09/2026

“Create a luxury perfume ad” leaves a lot of decisions open.

Before generating, I turn the idea into a visual brief:

① SUBJECT — what exactly should be in the image?
② PLACEMENT — where does it go, and what space stays empty?
③ CAMERA — from which height and angle do we see it?
④ LIGHTING — where does the light come from?
⑤ ENVIRONMENT — what surrounds the subject?
⑥ FINISH — what should the final image look like?

Here, the bottle belongs toward the lower right. The upper left stays dark and open for a headline. Warm light passes through the amber glass.

Those choices give the image a purpose before generation starts.

Swipe for the full prompt and the result with callouts. The first attempt added a label, so the revised brief specifies plain, unbranded glass.

A clear brief still needs checking. Compare the result with the decisions you made, then refine any instruction it missed.

Save this structure and adapt it to your next portrait, cover or product visual.

06/09/2026

Five months struggling alone. Two hours that made the difference. 🎯

“I learned more in 2 hours of instruction than I was able to muster in 5 months of struggling with teaching myself.”

That’s what my latest AI art student wrote after our lesson. He also left a ⭐⭐⭐⭐⭐ review and a $100 tip. Thank you for that trust! 🙏

Want to learn AI art—or already creating but struggling to get the results you need for your visual business?

I teach practical, personal lessons built around what YOU want to create: choosing the right tools, controlling your results, keeping visuals consistent, and turning ideas into finished work you’re proud to show clients.

Over 120,000 images created and 520 completed projects taught me a lot. In my lessons, I share that experience so you can put it to work yourself.

Message me “LEARN” and tell me what you want to create. Let’s take your AI skills further. 🚀

14/08/2026

Just me being me.

09/08/2026

Well, yeah, we in [BOTS] are just 3, random 4th playing with our nerves every evening, I swear Activision is trolling us with matchmaking. We get Potatoe in team and Lobby full of demons 😅 so i though I will try test seedance 2.5 to depict our evenings 😅🫡 .5

Adresse

Wabern

Webseite

https://pro.fiverr.com/freelancers/izbacivacx

Benachrichtigungen

Lassen Sie sich von uns eine E-Mail senden und seien Sie der erste der Neuigkeiten und Aktionen von Digital artworks by IzbacivacX erfährt. Ihre E-Mail-Adresse wird nicht für andere Zwecke verwendet und Sie können sich jederzeit abmelden.