"Petzval" Photoshop Filter or Action?

[Daniel_Buck]

Right, but first you have to generate (or worse yet, capture) the depth map, and then you have to know what to do with it. Applying blur based on z-depth is nowhere near enough to get you to the "petzval look."

Just because it's on a computer, doesn't mean it's one click easy But that said, photoshop is pretty limited to what it can do. A heavier compositing program is capable of doing alot more.

[Jim Galli]

Not only is it not convincingly duplicated, but consider that each petzval has a slightly different personality than the next one, the old timey ones especially are more 'petzvally' than their more modern equivalents sometimes, and that any given petzval will look different as you change formats with it. For instance an 8 1/2" Petzval used on 4X5 is quite conservative. Use it one 5X7 and you're into the edge coma far further with some exagerated fall off and softness, then put it on a 6 1/2 X 8 1/2 camera and you've got massive coma, swirls, fall-off and darkened corners, each look different from the previous.

[cowanw]

Not only is it not convincingly duplicated, but consider that each petzval has a slightly different personality than the next one, the old timey ones especially are more 'petzvally' than their more modern equivalents sometimes, and that any given petzval will look different as you change formats with it. For instance an 8 1/2" Petzval used on 4X5 is quite conservative. Use it one 5X7 and you're into the edge coma far further with some exagerated fall off and softness, then put it on a 6 1/2 X 8 1/2 camera and you've got massive coma, swirls, fall-off and darkened corners, each look different from the previous.

Which makes it easier to claim that a constructed effect is just like that petzval you have not seen
Regards
Bill

[Struan Gray]

I've always wondered why stereo pairs were so popular in the early days of photography. Obviously they needed the depth information so that their Difference Engines could calculate the required swirl.

FWIW, photogrammetric software will automatically generate depth maps from stereo pairs. The maths is tedious, but not hard, and computers don't get bored.

Ben: it occurs to me that you don't have to compute and store hi-res spot diagrams for all possible positions in image space - although I can see it might make sense if you are doing a lot of simulations of the same lens. The strength of aberrations tends to vary in a smooth analytic way, so I suspect you could get away with a coarse grid of values for each of the main aberrations, which you then interpolate and convert into a point spread function as needed. I suppose that when things get as ugly as with a plastic singlet the brute force approach may be more efficient though.

[goamules]

What I've learned in my experience is nothing is impossible. I've worked with space programs, telecomm, stealth, and other "miracle" programs, and we've often done what someone says cannot be. I'm no programmer, but "one day" someone will be able to duplicate most lens effects with software. I also know I want nothing to do with it.

[nathanm]

It's fascinating how much advanced technology, labor and cost it can take to simulate or reverse engineer something which occurs in the natural world for cheap. Using Photoshop tools to simulate the look of the old photographs I was inspired by was exactly what I was doing before I decided to shoot LF film. But if you just use "old" tools to start with, instantly you get 75% of what you want for almost no cost. Some things are just not worth "faking" when the real thing is just so much easier to do. I agree with goamules in that we will be able to duplicate all these things eventually which is cool and all, but dang - if you like the look of light passing through some ratty old chunk of glass then just shoot through some ratty old chunk of glass instead of using $10,000 of computer gear to muck up a photo shot through pristine glass! Like, *duh*!

[Kerik Kouklis]

But if you just use "old" tools to start with, instantly you get 75% of what you want for almost no cost.

And just where are these "almost no cost" Petzvals?

[Brian Ellis]

interesting image of the base of Steptoe....

If it was a photograph made at the base of Steptoe it would indeed be extremely interesting. : - )

[jb7]

Maybe you could adapt the program to be able to add flare, fingerprints and dust?
That'd be the whole package complete...

[Marko]

It's fascinating how much advanced technology, labor and cost it can take to simulate or reverse engineer something which occurs in the natural world for cheap. Using Photoshop tools to simulate the look of the old photographs I was inspired by was exactly what I was doing before I decided to shoot LF film. But if you just use "old" tools to start with, instantly you get 75% of what you want for almost no cost. Some things are just not worth "faking" when the real thing is just so much easier to do. I agree with goamules in that we will be able to duplicate all these things eventually which is cool and all, but dang - if you like the look of light passing through some ratty old chunk of glass then just shoot through some ratty old chunk of glass instead of using $10,000 of computer gear to muck up a photo shot through pristine glass! Like, *duh*!

Generally, I agree with the idea of original being better than a copy.

What I disagree with is the cost part, and I don't mean just financially. In order to "just use the old tools" or "just shoot through some ratty old chunk of glass", one needs to have the ratty old camera to hold that ratty old chunk of glass and the film holders to go with it. One needs to know how exactly to make the three work (camera, lens, film) and one needs to make a myriad other choices such as the choice of film and developer, provided that one is already familiar with either film or development and that one already has everything that's needed to properly process that film and turn it into an image. Be it wet darkroom, enlarger and whole nine yards or just the scanner.

If you already have all of those things and know how to use them, then yes, it is much cheaper to "just use them". But for people who don't have those tools and who practice a different type of photography and are equipped accordingly, but still want " the look", mimicking it in software could make more sense and would be much cheaper in the end.

Of course, this is from a perspective of someone for whom The Image is both the beginning and the end and the process is just a tool.