(Image credit: ChatGPT)
The best part of Halloween, besides the candy, is the costumes. I’ve enjoyed the rare times I’ve been able to put together particularly elaborate looks, and I always like seeing how others embrace the spooky season. I lacked the time to do anything on that scale this year, but for fun, I decided to see how well ChatGPT could digitally transform me for Halloween, perhaps inspire a look for next year based on the classic Universal Horror monsters.
ChatGPT’s image tools, when paired with a good selfie, can go surprisingly far in crafting visual alter egos. I wouldn’t call the results my undead doppelgangers exactly, but at least in some cases I could believe I’d walked out of the D-movie ripoff of a Universal classic. You can see some of the best results and t…
(Image credit: ChatGPT)
The best part of Halloween, besides the candy, is the costumes. I’ve enjoyed the rare times I’ve been able to put together particularly elaborate looks, and I always like seeing how others embrace the spooky season. I lacked the time to do anything on that scale this year, but for fun, I decided to see how well ChatGPT could digitally transform me for Halloween, perhaps inspire a look for next year based on the classic Universal Horror monsters.
ChatGPT’s image tools, when paired with a good selfie, can go surprisingly far in crafting visual alter egos. I wouldn’t call the results my undead doppelgangers exactly, but at least in some cases I could believe I’d walked out of the D-movie ripoff of a Universal classic. You can see some of the best results and the prompts behind them below.
Vampire weekend
(Image credit: ChatGPT)
I started with a vampire, naturally. I wanted to look like a 1930s vampire lord, with sharp cheekbones, a widow’s peak, and an elegant black cloak. I very quickly found that ChatGPT could only do so much before I stopped looking like me. Still, certain key phrases worked very well for all my portraits, as I discovered here. It also helped give the image the expression and the mood I wanted.
The prompt ended up as, “A hyper photorealistic portrait of the man from the reference photo as a classic Vampire, retaining his real facial features but with pale skin, red lips, and clearly visible fangs bared in a predatory expression. He wears a black cloak with a high collar and formal shirt. The lighting is moody candlelight with a golden glow from one side, evoking an aristocratic gothic castle interior.”
Hair of the dog
(Image credit: ChatGPT)
The Wolfman was next. Again, it took some experimenting to get a look that wasn’t either just me with dog ears and a nose, but also wasn’t just a wolf’s head with nothing connected to my actual appearance. The result is similar to the movie character, though I can’t help but feel like I’m upset because my hair is out of control.
The final prompt used is, “A hyper photorealistic portrait of the man from the reference photo transformed into a feral Wolfman. His face remains clearly recognizable but covered in realistic fur, with sharp teeth, yellow eyes, and a snarl mid-growl. The background is a moonlit forest interior or an old cabin, with atmospheric shadows and cool tones.”
Mummy man
(Image credit: ChatGPT)
I had similar reservations with the mummy costume, much as I love all (but one) of the many movies starring them. I didn’t want this one to be too covered to see it’s me or to lack enough bandages to make it clear it’s a mummy. After a couple of trials, referencing how you could almost see my skin made the image work out great.
Sign up for breaking news, reviews, opinion, top tech deals, and more.
The prompt is, “A hyper photorealistic portrait of the man in the photo wrapped entirely in aged, cracking bandages, his eyes and mouth just visible beneath. His skin shows faintly beneath the wrappings, preserving his facial likeness. The setting is an ancient tomb lit by torchlight, with dusty air and cracked stone walls. The tone is eerie and sepia-warm.”
(Image credit: ChatGPT)
I didn’t intend to give myself such sad eyes with Frankenstein’s Monster, but the vagaries of AI image creation meant that if I didn’t mention a serious or somber look, the final result went haywire. In addition, the coloring had to be specific lest it either look bright green like a cartoon or just my normal skin tone, arguably an even more unsettling option.
The prompt is, “A hyper photorealistic portrait of the man in the photo transformed into Frankenstein’s monster. His skin is green-gray with realistic texture and metal bolts protruding from his neck. His expression is somber, and he wears a dark coat over a sweater. The background is a dim industrial stone wall lit with moody, cinematic lighting reminiscent of a 1930s laboratory setting.
Truly, the most difficult picture to make is the one at the top with them all together. Corralling the digital characters so that they were still me, and the monster, and in the right location and looking like a portrait took roughly 20 attempts, and I still see more flaws there than in the individual portraits. Still, if you want to have a Monster Mash starring you, you, and you, try this long, complex prompt built out of a mix like Frankenstein’s Monster himself:
“Make a hyper photorealistic group portrait featuring four classic movie monsters, each transformed from this man’s face in the attached reference photo through expert prosthetic-style makeup and cinematic lighting. The monsters are a savage Wolfman with sharp teeth, and glowing eyes, a Vampire in a black cloak with pale skin, red lips, and clearly visible fangs bared;, Frankenstein’s Monster with green-gray skin, neck bolts, and detailed facial scars, and a Mummy wrapped in aged, tattered bandages that still reveal parts of the man’s face beneath. They are posed together in a well-lit Gothic library. The lighting is warm and evokes vintage Hollywood studio portraiture. Each monster should remain distinct but visibly share the same face, as though all were played by the same actor.
Follow TechRadar on Google News and* add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!*
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
Eric Hal Schwartz is a freelance writer for TechRadar with more than 15 years of experience covering the intersection of the world and technology. For the last five years, he served as head writer for Voicebot.ai and was on the leading edge of reporting on generative AI and large language models. He’s since become an expert on the products of generative AI models, such as OpenAI’s ChatGPT, Anthropic’s Claude, Google Gemini, and every other synthetic media tool. His experience runs the gamut of media, including print, digital, broadcast, and live events. Now, he’s continuing to tell the stories people want and need to hear about the rapidly evolving AI space and its impact on their lives. Eric is based in New York City.