 |
Figure 1.
Murdoch Mysteries |
Back
in 2022 I blogged about a 1966 episode of
The Wild Wild West called "The Night of the Flying Pie Plate", which featured green-skinned visitors to the American Southwest from Venus. Sadly, there are no space aliens, just a con-man trying to steal some gold. Agent James West is able to get to the bottom of the criminal scheme by seducing one of the "Venusian" girls dressed in exotic clothing and made-up with green skin paint. I don't know to what extent
Murdoch Mysteries was inspired by
The Wild Wild West, but a Season 1 (2008) episode of
Murdoch Mysteries was called "The Annoying Red Planet" and it also involved a UFO (in addition to
crop circles and an
eviscerated cow). The text prompt: "
Murdoch Mysteries" resulted in
Mr. Wombo generating an image of the world's narrowest street (see
Figure 1).
 |
| Doctor Ogden |
Equal Hat Rights. The image shown to the left was generated when I provided
Mr. Wombo with the text prompt: "
Murdoch Mysteries, Doctor Julia Ogden, Hélène Joy". It seems strange that Mr. Wombo generated an image of
Dr. Ogden wearing a hat that is so similar to the one in
Figure 1.
 |
Figure 2.
Dr. Ogden in "oak valley".
|
Using a more complex text prompt, "
photo-realism, intricate details, Murdoch Mysteries, Doctor Julia Ogden, Hélène Joy, long wavy hair, photo-realistic highlights in her hair, dressed Victoria's Secret style in a low-cut silk dress," Mr. Wombo generated the image that is shown to the right (
Figure 2). I used
WOMBO Dream's "Botany v3" style for this image because according to
Gemini, "Ogden" means "oak valley".
As can be seen in
my previous blog post, I have so far been unable to get Google's
Whisk to generate images in anything other than the default "landscape" aspect ratio. I tried selecting "portrait" from Whisk's aspect menu and then uploaded to Whisk the two "subject" images from
Figure 1 and
Figure 2 and the "scene" image from
Figure 3. The text prompt for
Figure 3 was: "
photo-realism, intricate details, a dirigible of the year 1930, Photo-realistic highlights on the surface of the dirigible, green hills of a Canadian Spring in the background".
Secret Military Project. In "The Annoying Red Planet", government agents are secretly developing an airship, but they get caught up in a contorted scheme to make people believe that aliens from Mars have arrived in Canada.
Punk Style. For the Whisk "style", I asked for, "
The overall style is reminiscent of magazine cover illustrations with a high degree of photo-realism, with a focus on rich colors and intricate details. The colors are vibrant and saturated. The image is sharp and clear, with a high level of detail visible in the subjects and the design of the scene. The overall mood is futuristic evoking a spirit of adventure and wonder. The subjects are depicted photo-realistically. There is fractal complexity in the living plants and animals of the scene. There is photo-realistic detail in the hair and clothing of the subjects. The airship technology is similar to that of the 1930s but with exaggerated mechanical details similar to a Steam Punk science fiction scenario," but Whisk claimed that it could not use that text prompt to generate a style. Instead, Whisk spontaneously invented a different style, shown below in
Figure 4.
 |
Figure 4. Whisk's preferred style. I have been unable to change the aspect ratio. 😒
|
Whisk ignored the text prompt that I provided for my requested style and changed the style to: "
The image is a digital painting, rendered in a style reminiscent of classic fantasy illustration with steampunk elements. The color palette is warm and muted, dominated by earthy browns, greens, and golds, with touches of lighter, pastel colors in the flora. The lighting is soft and diffused, creating a sense of gentle, atmospheric illumination. The overall style is highly detailed and painterly, with a focus on smooth blending and textural variation. The level of detail is consistent throughout, creating a cohesive and immersive scene. The image has a slightly vintage or antique feel, enhanced by the ornate frame and the overall composition." The image generated by Whisk to depict this style is shown above in
Figure 4.
 |
Figure 5. The second style generated by Whisk. Guns 😖 Why does it always have to be guns?
|
 |
| Generated by Mr. Wombo. |
After several attempts and altered versions of my text prompt (which were repeatedly not rendered by Whisk), I finally got Whisk to accept, "
The overall style has a high degree of photo-realism, with a focus on rich colors and intricate details. The colors are vibrant and saturated. The image is sharp and clear, with a high level of detail visible in the subjects and the scope of the scene. The overall mood is futuristic evoking a spirit of adventure and wonder. The subjects are depicted photo-realistically. The level of detail is consistent throughout, creating a cohesive and immersive scene," for which Whisk generated the image shown above in
Figure 5. I don't like Whisk's penchant for inserting guns into the images that it generates. However, I now has my "subject", "scene" and "style" images all assembled so that Whisk could generate a new "storyboard" image, which is shown below....
 |
| Figure 6. One of the two "storyboards" generated by Whisk. A cropped part of the 2nd one is in Figure 7. |
 |
| Figure 7. Second of the two "storyboards" |
My attempt to get Whisk to generate images with a tall portrait aspect ratio again failed. A work-around for this Whisk bug is discussed in my next blog post. Here is the description that Whisk generated for the "storyboard" in Figure 6. "Photorealistic image, vibrant saturated colors, sharp focus, high detail, futuristic aesthetic, spirit of adventure and wonder. A large, spherical, green airship, gondola resembling a dark brown train car with numerous windows, translucent dark green body revealing interior structure, small flag atop, floats over a verdant valley. Rolling green hills, evergreen trees, flowering plants, meandering river reflecting the sky, distant snow-capped mountains under a mostly clear sky with fluffy clouds.
A middle-aged, light-skinned man with dark hair, wearing a dark blue top hat, dark blue double-breasted coat with gold buttons, dark vest, burgundy necktie, dark pants, dark shoes, and dark gloves, stands with a serious expression. A light-skinned woman in her thirties with long, wavy blonde hair, wearing a red satin dress with short sleeves, lace neckline, fitted bodice, flowing skirt, gold necklace with pendant, and gold ring, stands with a serious expression, gaze slightly averted. Both figures are rendered with photorealistic detail, consistent with the overall scene. Daytime lighting illuminates the scene."
Here is the Whisk-generated description of the first "subject" (Murdoch) image: "The overall style has a high degree of photo-realism, with a focus on rich colors and intricate details. The colors are vibrant and saturated. The image is sharp and clear, with a high level of detail visible in the subjects and the scope of the scene. The overall mood is futuristic evoking a spirit of adventure and wonder. The subjects are depicted photo-realistically. The level of detail is consistent throughout, creating a cohesive and immersive scene."
Here is the Whisk-generated description of the second "subject" (Ogden) image: "
A light-skinned woman with long, wavy blonde hair is depicted in a photorealistic painting. She is wearing a red satin dress with short sleeves and a lace neckline. The dress has a fitted bodice and a flowing skirt. She is wearing a gold necklace with a pendant and a gold ring on her left ring finger. She is standing in front of a background of lush green plants and red flowers. The woman appears to be in her thirties. Her expression is serious and her gaze is directed slightly away from the viewer."
Here is the Whisk-generated description of the "scene" image: "
A large, spherical, green airship floats over a verdant valley. The airship's gondola is a dark brown, resembling a train car with numerous windows. The main body of the airship is a translucent, dark green, revealing an interior structure. A small flag is visible atop the airship. The valley below is characterized by rolling green hills, dotted with evergreen trees and some flowering plants near the bottom of the frame. A river meanders through the valley, reflecting the sky. In the distance, snow-capped mountains are visible under a mostly clear sky with some fluffy clouds. The lighting suggests it is daytime, with the sun illuminating the scene."
I changed the storyboard description to: "
Photorealistic image, vibrant saturated colors, sharp focus, high detail, futuristic aesthetic, spirit of adventure and wonder. A large, spherical, green airship, gondola resembling a dark brown train car with numerous windows, translucent dark green body revealing interior structure, small flag atop, floats over a verdant valley. Rolling green hills, evergreen trees, flowering plants, meandering river reflecting the sky, distant snow-capped mountains under a mostly clear sky with fluffy clouds. A middle-aged, light-skinned policeman with dark hair, wearing a dark blue top hat, dark blue double-breasted coat with gold buttons, dark vest, a police badge, burgundy necktie, dark pants, dark shoes, and dark gloves, stands with a serious expression. The policeman looks at the airship and points at the airship.
A beautiful light-skinned woman in her thirties with long, wavy blonde hair, wearing a red satin dress with short sleeves, lace neckline, fitted bodice, flowing skirt, gold necklace with pendant, and gold ring, stands with a serious expression, gazes at the man. Both figures are rendered with photorealistic detail, consistent with the overall scene. Daytime lighting illuminates the scene." My four changes to the Whisk storyboard description were: 1) The male character is specifically identified as a policeman. 2) I specified that the man "looks at the airship and points at the airship". 3) The woman's gaze is now directed towards the man. 4) The woman is described as "beautiful". See the Whisk-generated image for this updated text description that is shown in
Figure 8, below.
 |
Figure 8. Modified storyboard; Murdoch investigates a secret Canadian government cloning project.
|
 |
Figure 9. Storyboard "card" by Whisk for the image in Figure 8.
|
Whisk was able to put not one, but two badges on Murdoch, but he is seemingly looking at and pointing at the twins (
Figure 8, above). I have no idea why Whisk decided to make two copies of Dr. Ogden for this storyboard image.
 |
Figure 10. Whisk image library.
|
The image in
Figure 9 shows the "card" that got stored in my Whisk image library for this storyboard (
Figure 8). Although these cards show a "seed", I could not find nothing in the interface that would allow a user to alter the seed.
 |
Figure 11. Alternate airship.
|
The image in
Figure 11 is an alternate version of the airship with only Dr. Ogden in the image. Here is the Whisk-generated description of the "storyboard":
"
Photorealistic image, vibrant saturated colors, sharp focus, high detail, futuristic aesthetic, spirit of adventure and wonder. A young woman with long, wavy light brown hair in loose curls, fair skin, light blue eyes, subtle smile, wearing a low-cut red satin v-neck dress with gold embroidery, gold earrings with ornate detailing, and a gold necklace with a large dark green gemstone pendant, stands on a grassy hill. She gazes upward at a large dirigible airship floating above a verdant landscape. The airship has a greenish-glass exterior, divided into paneled sections reflecting the scenery, a dark brown wooden gondola with mechanical details and portholes, ropes and wires connecting the gondola to the main body, a pointed metallic nose, and a small raised structure at the top. Below, rolling green hills, a winding river, an evergreen forest, and a light-colored stone manor house on a hill overlooking the river are visible. Distant hazy mountains and a mostly clear sky with wisps of clouds complete the background. The lighting is bright daytime light."
 |
| Figure 12. Generated by Whisk with an adjustment prompt: "The subject is pointing at the airship". |
 |
| Figure 13. "Change the material of the airship from brown wood to golden reflective foil." |
 |
Figure 14. A Whisk-generated version of the UFO seen at night. The image was manually darkened by me.
|
 |
| Figure 15. SSS7 |
Here is the version of the text description for the storyboard shown in Figure 14: "Photorealistic image, night-time subdued colors, sharp focus, intricate details, futuristic aesthetic, spirit of adventure and wonder. A large, dark brown and gold dirigible airship floats in a starlit night sky, illuminated by a moon in the upper left. The airship lacks an inset image. Below, a landscape of rolling green hills, a river, and a large mansion-like building stretches out. A similar building is visible on the right near the river. A young woman, fair-skinned, with light-brown wavy hair in loose curls, is positioned in the lower third of the image, her upper body visible. She wears a low-cut red dress with gold floral embroidery, short sleeves, and a V-neckline. A gold necklace with a dark green pendant and small earrings adorn her. She smiles slightly, her gaze directed toward the moon, her right index finger extended as if pointing. The woman is dimly lit by the moonlight, enhancing the overall subdued color palette. The scene is detailed and cohesive, emphasizing the futuristic and adventurous mood". The night-time subject, scene and style images are shown to the right on figure Figure 15.
The first of the Whisk-generated night-time storyboards is shown below in Figure 16.
 |
Figure 16. Whisk-generated highly reflective dirigible and a large Moon.
|
Here is the description of the storyboard shown in
Figure 16, above: "
Photorealistic image, night-time subdued colors, sharp focus, intricate details, futuristic aesthetic, spirit of adventure and wonder. A large, dark brown and gold dirigible airship floats in a starlit night sky, dominated by a large, bright moon. Below, rolling green hills, a river, and a mansion-like building are visible, mirroring a landscape painting within the airship's transparent section. A similar building is seen on the right near the river. A young woman with fair skin and light-brown, wavy hair styled in loose curls, smiles slightly. She is depicted from the waist up, her gaze directed slightly left, her right index finger extended as if pointing. She wears a low-cut, red dress with gold floral embroidery, short sleeves, a V-neckline, a gold necklace with a dark green pendant, and small earrings. The woman is positioned against the airship and night sky background, her pose and expression conveying a sense of wonder and intrigue. The overall lighting is soft moonlight, casting subtle shadows and highlighting the intricate details of the dress and the woman's features."
The style for the night-time storyboard (Figure 16) is shown in Figure 17, below.
 |
Figure 17. Whisk generated image for the night-time style.
|
 |
Murdoch's alien abduction.
|
Once again I was dismayed by the appearance of blasters in the Whisk-generated image for the "style" that is shown in Figure 17. However, I was fairly happy with Figure 14. I liked the reflective "metal foil" balloons that Mr. Wombo generated with his "Mechanical v3" style (such as the one in Figure 3). However, that led to Whisk making some absurdly reflective balloons such as the one shown in Figure 16. For the image shown to the right, I pasted a Detective Murdoch into the "mirrored" image of the landscape on the side of the balloon where it looks like a display screen.
 |
a Whisk doll
|
Using altered text descriptions in Whisk, I was able to make simple changes such as shrinking the diameter of the Moon, but I did not find a way to get Whisk to switch from daylight to night-time illumination, so I had to manually created two new night-time "subject" and "scene" images (see Figure 15) that I provided to Whisk.
Next: making humanoid aliens with Whisk.
No comments:
Post a Comment