AI-Generated | The Tea House & Street Series

Lin Dai. (1934-1964)

I chose to name this portrait as the screen goddess Lin Dai. Born in Guangxi, China in 1934, she became the foremost star of Mandarin films as she won the hearts of audiences with more than 40 films. She died by suicide in 1964. I used MidJourney V4 and refined it by using ON1 Photo Raw 2023. 

The prompts are as follows: 

[A young Chinese Opera woman wearing blue cheongsam, red lipstick, full body shot, a cup of tea, teahouse background, photography by Annie Leibovitz. -- no hands.]. 

The hands are still an issue with V4, so -aside from editing the image- I also had to improve/alter the hand that rests on the table. I left the other one untouched.

Zhou Xuan, Actress and Singer. (1920-1957)

Zhou Xuan was an iconic Chinese singer and film actress. By the 1940s, she had become one of China's Seven Great Singing Stars. She was the best known of the seven, nicknamed the "Golden Voice", and had a concurrent movie career until 1954. She recorded more than 200 songs and appeared in over 40 films in her career. I've added prompts to get a full body and a 1940 street in Shanghai.

I used MidJourney V4 and refined the image by using ON1 Photo Raw 2023. 

The prompts are:

[Young beauty attractive lady wearing cheongsam, style by film Rouge, full - body shot, walking on the street in 1940s Shanghai, style by Yasutomo Oka, 8k, HD, cinematography, photorealistic, epic composition Unreal Engine, Cinematic, Color Grading, landscape Photography, Ultra - Wide Angle,Depth of Field, etc. -- no hands]

Betty Loh Ti (Hsi Chung-i). Chinese Actress. (1937-1968)

Betty Loh Ti (July 24, 1937 – December 27, 1968), also known as Le Di or Loh Tih, was a Hong Kong actress originally from Shanghai. Known as the "Classic Beauty", she was one of the most celebrated actresses of Hong Kong cinema. She -as some of her contemporaries- died by suicide.

I used MidJourney V4 and refined the image by using ON1 Photo Raw 2023. 

The prompts (very basic) are:

[A young Chinese Opera woman wearing blue cheongsam, red lipstick, full body shot, a cup of tea, teahouse background, photography by Annie Leibovitz. no hands.]

Yoshiko Kawashima (1907-1948)

Yoshiko Kawashima (1907–1948) was a Qing dynasty princess who was raised in Japan, and who served as a spy for the Japanese Imperial Army during the Second Sino-Japanese War.

The prompts are:

[Young beautiful attractive Chinese Opera woman wearing pale blue long cheongsam, looking back at camera, style by film In The Mood For Love, full - body shot, walking in street in 1940s Shanghai, photography by Annie Leibovitz, dynamic composition, 8k, HD, cinematography, photorealistic, epic composition, etc. -- no hands]

Bak Yin, Queen of Cantonese Films. (1920-1987)

Bak Yin known as the "Queen of Cantonese films" was born in Guangzhou in 1920. She started her movie career in 1936. Before the Japanese occupation forced the industry to shut down in 1941, she had starred in over 30 films. After the end of the war in 1945, she became one of the few pre-War stars to reclaim "leading lady" status.

The prompts are:

[A young beautiful Chinese Opera woman wearing yellow cheongsam, red lipstick, full body shot, a cup of tea, teahouse background, photography by Annie Leibovitz. -- no hands]

Soong Ch'ing-ling (Madame Sun Yat-sen). 1893-1981

Soong Ch'ing-ling was the third wife of Sun Yat-sen (President of the Republic of China 1919-1925). She was a politician and following Sun Yat-sen's death, she strongly supported the Communist government, and was appointed by Mao Tse-tung to various important posts.  

The prompts are:

["Young beautiful attractive Chinese Opera woman wearing black long qipao, style by film In The Mood For Love, full - body shot, walking in street in 1940s Shanghai, photography by Annie Leibovitz, dynamic composition, 8k, HD, cinematography, photorealistic, epic composition Unreal Engine, Cinematic, Color Grading, landscape Photography, Ultra - Wide Angle, etc."]

AI Generated | The Shogunate's Female Samurais

I chose to name this Ai-generated image as "Onamihime" (1541-1602), who was an Onnamusha (a female warrior) during Japan's Sengo-ku period. The MidJourney prompts I used for it is "young Japanese woman in samurai armor, 4k, hyper-detailed, studio photography, soft cinematic perfect light, ultra-realist, full body."

I chose to name this Ai-generated image as "Nakano Takeko" (1847-1868), a Japanese female warrior of the Aizu Domain, who fought and died during the Boshin War. The MidJourney prompts I used for it is "young Japanese woman in samurai armor, 4k, hyper-detailed, studio photography, soft cinematic perfect light, ultra-realist, full body."

 

I chose to name this Ai-generated image as "Komatsuhime" (1573 -1620), a female warrior during late- Sengoku period and early Edo period. The MidJourney prompts I used for it is "a beautiful Japanese samurai young woman in armor, 4k, hyper-detailed, studio photography, soft cinematic perfect light, ultra-realist, full body in the style of Annie Leibovitz ar 3:2"

Random Thoughts On AI-Generated Images

I've been exploring MidJourney over the past few days, and generating AI images that I conjure up as I go along. I've also read a few explanatory articles and viewed images made by well established AI "artists" on its chat room-like interface. I've learned to generate variations of the same image, and how to upscale them from 1024 to 2048 pixels...which are too low of a resolution to make a reasonable print.

Opinions from photographers about Ai generated images vary widely...ranging from those who fear it and describe it as a threat to their livelihood/occupation to those who fully embrace it/or try to by saying it's a new artistic path. There's also a camp that argue against it because it plagiarizes the work of other photographers or image-makers, although AI bases its output on millions of algorithms and variations.

I'm more inclined to embrace it, not as a replacement to my own photography, but as a separate artistic endeavor, and not necessarily on parallel tracks to photography. So far, the images I've generated by inputting unremarkable text prompt are certainly striking but are not as realistic as genuine photographs.

Nonetheless, the costumes worn by my imaginary characters are truly remarkable in their details and color shades. In contrast though, the facial renderings seem somewhat artificial or "plasticky", although I've seen photographs so over-processed by editing software that they closely resemble these AI-generated images.

There's no chance that AI will replace photography, but I believe it will certainly make a dent in its commercial applications. Photojournalism, event photography, wedding photography, and even travel photography will remain "real"...but commercial photography (and potentially, fashion photography) will certainly be affected.

In the meantime, I intend to dabble in Ai-Generated imagery, have fun with it but continue with my photography.

My First AI-Generated Images

AI-Generated Illustration/Image By Tewfic El-Sawy

As the title of this blog post announces, these are my very first AI generated image produced as a result on inputting a short string of "prompt" words on Midjourney. 

Midjourney is an independent research lab that produces an artificial intelligence program under the same name that creates images from textual descriptions...also known as "prompts". I created these images by just adding words such as "Chinese Opera artists realistic portraits" or something like that, and in a matter of seconds of rendering, these appeared as a result of these "prompts". Users create artwork with Midjourney using Discord bot commands.

I would quibble with them being "realistic", and I think they are more like overworked portraits with a very heavy-hand using a bunch of tools from a photo-editing software. However, they are beautiful for what they are...the details of the make-up and silken robes are truly impressive.

While I admired some AI images I had seen on social media platforms before, my  interest was piqued by this recent blog post by The Phoblographer. The article correctly (in my view) says that AI generated visuals don't deserve to be called "photographs", but that there's definitely immense potential for AI usage in art and CGI. 

Will I ever hang up my cameras because of AI and its potential? Of course not...but I will occasionally experiment and have fun with it. 

I've photographed many Chinese Opera artists in the production of my monograph "Chinese Opera of the Diaspora", from New York City, Penang, Hong Kong, Taipei...and while most of these AI-generated images are eye-catching, they're "lifeless" when compared side-to-side with the real photographs I made during my two years project. 

For this last entry, I worked the AI-Generated image by processing it in ON1 Raw software (adding a darker background -also AI preset- and some slight adjustments to the face).

The Neon Lights Are Calling

Kenmare Street/Cleveland Place

Broome Street/Mulberry Street

I've been recently influenced by a large number of Japanese street photographers (on Twitter) who are partial to night scenes of neon-lit streetscapes; some in Tokyo and Kyoto. There are other photographers from various geographical locations, especially from New York City and London, who espouse the same aesthetic.  From experience, the urban scenes with red work the best with this color grading technique; which some call "cinematic grading".

Thompson Street

Spring Street

Tashi Delek (བཀྲ་ཤིས་བདེ་ལེགས) With Kunchok

Kunchok Palmo's Tibetan heritage is front and center with a resplendent purple and fuchsia "chuba" (can also be known as chupa) outfit. It's a unisex garment, and worn all over the Tibetan Plateau with roots deep in Central Asia. The traditional "chuba" is an ankle-length robe tied around the waist by a long sash. For the photo shoot I provided a Mahākāla wooden mask from Bhutan, along with bone prayer beads and an ancient seal. Mahākāla is a deity common to Hinduism and Tantric Buddhism, and is considered as a protector deity.

Kunchok's name is Tibetan/Nepali for "goddess". Tashi Delek (བཀྲ་ཤིས་བདེ་ལེགས) is the Tibetan expression used in greeting, congratulation, and also used to bid good-luck wishes.

We had arranged to meet in SoHo for a 2-hours photo session, but as I had scouted the area a day earlier, we went first to Cortland Alley whose graffiti-laden walls I had seen on Sunday were just "begging" to be powerful backdrops. Unfortunately, a moving truck was parked in a way that we didn't have an accessible area for Kunchok to pose and for me to photograph...however, we managed a few poses using some angles.

Following the 15 minutes in Cortland Alley, we continued the shoot on the nearby SoHo streets. The black gates of the Wooster public parking garage offering the best "studio-like" backdrop.

For larger higher resolution photographs:

Kunchok Palmo by Tewfic El-Sawy on Exposure And for an audio-slideshow:

My POV On Portrait Photography

I recently met with "K" for an outdoor photo session to feature her wearing a traditional Tibetan purple & pink "chuba". I also brought with me an authentic wooden mask carved in the likeness of Mahakala (a Himalayan protector deity) as a prop.

Although I describe myself as more intrinsically interested in the "photojournalism meets travel photography" ethos, I also realize that portrait photography (whether candid or posed) was and will always a part of it...and with it comes a set of "rules" that I try to always follow.

The overriding covenant that I always keep at the forefront of my photography mindset is Sebastiao Salgado's who said "If you take a picture of a human that does not make him noble, there is no reason to take this picture".

I start off all my photo sessions by being complimentary of their appearance, by being punctual (even if the models are late due to traffic, trains, etc), by expressing thanks before and after the sessions, by having cleaning wipes and hand sprays on the ready...little things which I believe make a difference. It doesn't mean that I "click" with every person who poses for me, but that's my personal modus operandi. And doing so, allows the nice communication flow between the person being photographed and I...which is very important.

Before the photo sessions, I make sure to scout the areas for previous and new spots where I will photograph the persons. This allows me to anticipate the available light at the time we schedule the session, and to check for any impediments that might disrupt the upcoming session. I'm not a studio photographer, so I never use artificial light modifiers and rely solely on ambient sunlight.

Depending on the purpose (and the models' personal "vibe") of the session, wardrobe choice is obviously key. In most of the photo sessions I set up, the models chose what they wear with minor guidance from me. In such cases, I usually nudge them towards dark solids as I like using graffiti'ed walls as backdrops. 

As my photo sessions are all non-commercial, there are no wardrobe stylists and/or makeup artists. But most of my subjects are adept at applying their own make-up which I don't interfere with...except sometimes asking for the application of a vibrant lipstick color.

However, my most recent three photo sessions were focused on ethnic wear... South Korean hanbok, Chinese qi pao/cheongsam and the Himalayan chuba. So I was merging "ethnic fashion portraiture with travel photography"...not exactly full circle, but something akin to it.