VelezStudio

VelezStudio We help marketing and content teams enhance their brands, reduce costs, and move faster with industry

07/08/2026

This week, I was lucky enough to be invited to Figma's Maker Collective in Sydney, the local recap of everything that shipped at Config 2026. And the biggest takeaway.

Weave.

Coming to Figma soon, it lets you chain multiple AI prompts together to get one finished outcome, not just a single generation. Product shots, mockups, variant sets....the stuff that used to eat an entire afternoon.

As someone who does this daily at Woolworths, that's not just a neat feature. That's hours back to focus on bigger and better things.

A few other highlights worth knowing from the day:
→ Code layers, Figma Motion, and shader fills/effects
→ Local design teams from Atlassian and REA Group shared their vision of the future of AI in design
→ And of course, the best part of events like this isn't the slides, it's the room. Coffee, conversations, and some Figma swag to slap on the desk.

hashtag hashtag hashtag

03/08/2026

Someone built a file transfer system that needs nothing but a camera and QR codes. Using Claude Code.

One phone outputs rapidly-changing QR codes. The other reads them with its camera and reconstructs the file in real-time.

No WiFi. No Bluetooth. No connection at all.

It's called Decimen Optical Transfer. Built by a developer going by Alstroph on Reddit, bashalarmistalt on GitHub.

Here's how you can turn an old headshot photo into a new one with AI. No studio, no invoice, under an hour.Turns out I h...
23/07/2026

Here's how you can turn an old headshot photo into a new one with AI. No studio, no invoice, under an hour.

Turns out I have no good side. Sitting for a photographer just confirms it slower. So I reworked the one I already had.

Here's how to do it yourself.

STEP ONE
Pick your reference photo. Front-on, eyes visible, decent focus on your face. Ideally include more than one with multiple angles and expressions. Background and outfit don't matter, you're replacing those. Blurry or heavily filtered won't work.

STEP TWO
Get the prompt written properly. I asked Claude for a photography brief, not a request. Give it the role you want to read as, then make it specify light source and direction, depth of field, colour grade, wardrobe, camera height and distance, and background. Add a hard constraint that facial features stay identical to the reference.

STEP THREE
Run it in Google's Nano Banana 2 with your photo attached as the reference image.

STEP FOUR
Iterate, and only change one thing per round. This is the step people skip. My first attempts drifted off my actual face, then the lighting went theatrical, then the colour went glossy and synthetic

Expect five or six rounds. Change three variables at once and you'll never know which one worked.

The whole trick is that the prompt should read like something you'd hand a photographer. Vague in, vague out.

Full prompt is in the image. Swap in your reference and go.

One honest caveat. This works because there's a real photograph underneath it. You're relighting, not inventing.

21/07/2026

Pixels are having a moment right now, and it shouldn't be.

We spent twenty years chasing crisp. Retina screens, vector everything, infinite scalability. The whole industry optimised for smooth perfection.

Now, I've watched that 8-bit typeface find its way back into the limelight, and I don't think it's really about the pixels.

You don't get nostalgic for a font. You get nostalgic for being eight years old in a lounge room, rewinding a tape before school because someone forgot the night before.

When everything can be perfect, perfect stops meaning anything. A pixel font is a flex in the other direction. It says a human made a decision here, on purpose...which is rare right now.

There's also the AI of it all. Slick and generated is starting to look the same everywhere. Chunky, deliberately "low-res" design ironically reads as handmade in a way polished creative never will.

I felt all of that enough to make my own. It's called VCR Mono, built off the blocky on-screen displays burned into every camcorder and cable box from the 90s.

Check it out here, https://tinyurl.com/z5akvb2s

21/07/2026

Pixels are having a moment right now, and it shouldn't be.

We spent twenty years chasing crisp. Retina screens, vector everything, infinite scalability. The whole industry optimised for smooth perfection.

Now, I've watched that 8-bit typeface find its way back into the limelight, and I don't think it's really about the pixels.

You don't get nostalgic for a font. You get nostalgic for being eight years old in a lounge room, rewinding a tape before school because someone forgot the night before.

When everything can be perfect, perfect stops meaning anything. A pixel font is a flex in the other direction. It says a human made a decision here, on purpose...which is rare right now.

There's also the AI of it all. Slick and generated is starting to look the same everywhere. Chunky, deliberately "low-res" design ironically reads as handmade in a way polished creative never will.

I felt all of that enough to make my own. It's called VCR Mono, built off the blocky on-screen displays burned into every camcorder and cable box from the 90s.

Link in the comments if you want to check it out.

Address

Sydney, NSW
2000

Website

Alerts

Be the first to know and let us send you an email when VelezStudio posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Featured

Share

Category