Let's talk about vintage lenses.
Here is your cool samurai show with modern lenses.
Here is your cool samurai show with vintage lenses.
Hollywood is no stranger to fads.
We are currently in the middle of a "make everything too dark" fad. But that fad is starting to overlap with "let's use really old lenses on ridiculously high resolution cameras."
This is Zack Snyder with a Red Monstro 8K camera.
He is using a "rehoused" vintage 50mm f/0.95 Canon "Dream Lens" which was first manufactured in 1961.
This old lens is put inside a fancy new body that can fit onto modern cameras.
Which means Zack is getting nowhere near 8K worth of detail. These lenses are not even close to being sharp. Which is fine. I think the obsession with detail can get a bit silly and sometimes things can be "too sharp."
But it is a funny juxtaposition.
The dream lens is a cool lens. It has character. It has certain aberrations and defects that can actually be beneficial to making a cool photograph. It's a bit like vinyl records for photography.
[ Peter Thoeny ]
It has vignetting and distortion and a very strange swirly background blur.
[ Gabriel Binder ]
Optical engineers have been spending the last 60 years trying to eliminate these defects. And I sometimes wonder if they are confused by this fad.
"I WORKED 70 HOURS PER WEEK TO GET PERFECT CORNER SHARPNESS!"
And whether you prefer to work with a perfect optic or a vintage one... it is a valid aesthetic decision either way. I think vintage glass can really suit candid natural light photography. You can almost get abstract with these lenses.
[ Peter Theony ]
Personally I like to start with as close to perfect as possible and then add the character in later. That way I can dial in the effect and tweak how much of it I want. But even with modern image editing tools, some of these aberrations are difficult to recreate authentically.
That said, it can be very easy for the "character" of these lenses to become distracting. And just like when someone first finds the lens flares in Photoshop, it can be easy for people to overdo things.
Zack Snyder decided to be his own cameraman and used only vintage glass in his recent movies and it has led to some complaints about the imagery.
I mean, Zack Snyder overdoing something? I can't even imagine it.
Non camera people felt Army of the Dead was blurry and a bit weird but they couldn't quite explain why it felt that way.
The dream lens has a very wide aperture and it lets in a lot of light. But it also has a very very shallow depth of field. Which means it is very difficult to nail focus.
[ Peter Thoeny ]
Her near eye is in focus and her far eye is soft. You literally can't get an entire face in focus.
There is no reason you have to use the dream lens at f/0.95 at all times. But just like those irresistible lens flares, Zack couldn't help himself.
Here is a blueprint that you can't really see.
Extreme close ups of faces without autofocus at f/0.95 is nearly impossible to pull critical focus on.
Looks like Zack nailed the area just above the eyebrow here.
Let's try to find the point of focus in this one.
Ummmm... she is just... blurry. Missed focus completely.
But Zack isn't the only one going vintage. I've been seeing this a lot recently.
Shogun is a beautiful show. And for the most part, I really enjoyed the cinematography. But they went the vintage lens route and it kept going from gorgeous to "I can't not see it" distracting. And perhaps because I am familiar with these lens defects I am more prone to noticing. But I do think it hurt the imagery in a few spots.
Vingetting is a darkening of the corners of the frame.
Light rays in the corners are much harder to control. A lot of modern lenses still have this problem, but they create software corrections to eliminate the issue. Some cameras do it automatically as you are recording the image.
Vintage lenses were built before lens corrections where a thing—before software was a thing. So you either have to live with them, try to remove them with VFX, or crop into your image and lose some resolution.
It's possible this is the aesthetic they wanted. They felt the vignetting added something to the image. But I just found my eyes darting to the corners and not focusing on the composition.
And then you have distortion.
In this case, barrel distortion.
This is mostly prominent in wide angle lenses. In order to get that wider field of view the lens has to accept light from some very steep angles. And that can be quite difficult to correct. So you kind have to sacrifice any straight lines.
And sometimes this was a positive contribution to the image.
I thought the curved lines matched the way they were sitting here.
But most of the time I just felt like I was looking at feudal Japan through a fish's eye.
It's a bit more tolerable as a still, but when all of these verticals are bowing in motion, I start to feel like I am developing tunnel vision.
I love that this is a tool that is available. Rehousing lenses is a really neat process and I'm glad this old glass is getting new life.
This documentary shows how lens rehousing is done and is quite fascinating if you are in to that sort of thing.
But I think we are in a "too much of a good thing" phase when it comes to these lenses. I think a balance between old and new can be found.
And I also think maybe Zack should see what f/2.8 looks like. He might like having more than an eyebrow in focus.
255 notes
·
View notes
Doing It Wrong On Purpose: Episode 1 - The Un-Ship
Today's experiment: What happens if I prompt for something, and then negative prompt all the main keywords, plus various synonyms and related words?
The answer: Some gloriously weird stuff.
For example, let's look at a negative cat:
Positive prompt: A cat on a windowsill during a storm
Negative prompt: Cat, feline, felidae, kitty, kitten, animal, pet, windowsill, window, glass, pane, house, storm, rain, water, lightning, thunder, clouds, torrent, downpour, snow, blizzard, wind, windy
Interesting! Let's get a little more fantasy with it and try for an anti-deer:
Positive prompt: A deer in a peaceful flowery meadow, crystals, midnight, fantasy, colorful
Negative prompt: Deer, cervidae, animal, elk, moose, stag, doe, fawn, reindeer, antelope, cervid, antlers, flowers, night, dark, trees, foliage, bloom, stars, night, tranquil, fantastic, vibrant, cool, magic, blue, moon, sky, crystal, stone, statue, topiary, floral, blossom
Between these two experiments, including a few dozen other generations that remain unposted, one thing I can say for sure is that for living subjects, it's a great way to get the kind of anatomical wonk that older models are (in)famous for - and it makes sense why, the model is trying to make something that looks like a certain subject...but once it starts to look too much like it, well, shit, we told it NOT to do that! Break something up! Given that I love that kind of wonk, I think I've found a useful tool for myself.
One more living subject, and let's get even more abstract with our direction here:
Positive prompt: mind horse
Negative prompt: horse, equine, colt, filly, mare, stallion, bronco, pony, mind, brain, thought, essence, psyche, intelligence, consciousness, imagination, dream, soul, visualization, intellect, wit, cognizance
Now let's try something that isn't alive. One thing I love AI for is surreal settings and landscapes - lets try one now!
Positive prompt: A magic palace garden made of crystal and gold
Negative prompt: Palace, magic, crystal, gold, fantasy, castle, estate, stronghold, temple, garden, flowers, plants, blossoms, bloom, blooms, trees, grass, stems, foliage, leaves, greenery, branches, bush, bushes, hedge, hedges, metal, luxury, stone, glass, brass, rose, polished, jewel, prism, courtyard
I then tried to see if, learning from the animal subjects, I could make it more likely to return one of my favorite "mistakes" - making it impossible to discern the point where a water area ends and a sky area begins. I wasn't immediately successful, but I came up with some results I found pleasing regardless-
Positive prompt: Secret hideout in a cave behind a waterfall in the foggy forest on a floating sky island in fluffy clouds
Negative prompt: hideout, camp, campsite, home, abode, house, dwelling, rest, shelter, waterfall, water, cave, grotto, forest, woods, woodland, trees, fountain, cascade, pond, stream, lake, river, brook, puddle, creek, pool, beach, ocean, sea, cloud, clouds, sky, cumulus, cirrus, nimbus, fog, storm, rain, sunshower, falls
It seems that with landscapes it's got a much clearer and more specific "idea" of what a [SUBJECT] without [SUBJECT] looks like; it's more inclined to invent very specific, very consistent unasked for related elements. With the animals, I was tweaking the weight on the positive prompt to avoid getting straightforwardly just what I had positive (and negative) prompted, but with landscapes, I just get... almost something else entirely.
So how about inanimate objects? Let's try a ship, perhaps?
Positive prompt: A huge sailing ship with brilliant prismatic crystal sails on a stormy, turbulent sea of sunset clouds
Negative prompt: ship, boat, sailboat, sailing ship, pirate ship, galleon, ketch, schooner, sloop, cutter, sail, sea, ocean, storm, wind, rain, water, waves, cloudy, clouds, fog, sunset, dusk, dawn, sunrise, twilight, evening
...okay, I'm in love with the un-ship. It truly does manage to consistently give me results that look like, yet entirely unlike, a ship. It is everything I love about AI as a medium. More than that, it is my friend.
At lower positive prompt weights, they only get even more beautifully chaotic.
I want to live on one of these (in an alternate universe where they're geometrically possible and structurally sound, that is).
Failing that, I will be featuring them a lot from now on.
All images generated using Simple Stable, under the Code of Ethics of Are We Art Yet?
133 notes
·
View notes