Petzval DSLR Lens Is A Portrait Hero

[2WheelPhoto]

Or is it!?


Lomography?s Kickstarter-Backed Petzval DSLR Lens Is A Portrait Hero | TechCrunch

 

[tirediron]

Interesting; I wouldn't mind trying one, but the project seems woefully under-funded to me... I would think you would need to be in the millions of dollars of investment to make this work.

 

[Derrel]

Interesting. I watched their video. Huh. Brass barrel. Waterhouse stops. I wish them luck.

 

[Gavjenks]

Interesting; I wouldn't mind trying one, but the project seems woefully under-funded to me... I would think you would need to be in the millions of dollars of investment to make this work.

Uh... millions? What? You could almost certainly do this with 10,000 dollars.

Lenses are cheap! In all but the most expensive Canon and Nikon lenses, the actual glass elements are probably around $10-30 each cost to them (not including the odd expensive "ultra low dispersion" or diffractive or aspheric). And there are a few different manufacturers around who can produce new lenses to spec for around that much as well, as long as you order like 50 or more. Definitely on the low end if you are doing an old timey lens without crazy coatings or aspherics, etc.

Lens = brass tube ($5) + mount to camera ($5) + some sort of focuser (maybe $10-20 depending how fancy) + maybe $40 for a Petzval (1 achromat + 1 meniscus + 1 biconvex is 4 uncoated simple elements at $10 or so each) + let's say $5 for assorted screws and mounting stuff + $10 labor to assemble + $5 stops + $10 shipping

= $110 costs. Then you have to do a production run of about 50 for that price of lens element = $5,500 multiplied by whatever your profit margin is, let's say 100% = $11,000 kickstarter would do it. Larger amounts than that are either more units than 50, higher profit margins attempted, or slightly more luxurious components, or a combination thereof.




Remember:
1) No autofocus. Looking at the images online, it looks like their focus is either a simple rack and pinion (like the type you see in cheapo $30 macro bellows) or a Crayford focuser, both of which are cheap and easy.
2) No image stabilization.
3) No diaphragm, let alone an ELECTRONIC diaphragm. Just a hole you put sheets into.
4) No zoom.
5) No need for crazy well corrected optics.
6) No cost for hiring a top notch or even a professional lens designer. They're using a well known, existing design, which can be optimized for this particular lens in about 2 hours using free software (look up Qioptiq if interested) by an amateur who has a decent idea what they're doing.
7) No need for any fancy or difficult to align retrofocus designs. Since it's a portrait lens, it's longish anyway, and the short flange to focal distances of modern DSLRs isn't a barrier. This makes manufacturing easier, and also automatically makes it about 4x cheaper at least to achieve any given level of optical quality.

 

[Derrel]

I chuckled when I heard that Russia's (in)famous Zenit company would be building this thing!!!

zOMG...when I was a boy, I bought a Zenit-made 35mm SLR, the Cosmorex SE, sold by New York's old Cambridge Camera...wow...it had a POS (no, not Positively Old-School, but the other kind of POS) 58mm f/2 thread mount lens with shiny, tinny diaphragm blades and primitive anti-reflection coating--the lens flared like a bastard whenever it was pointed in the same direction as the sun, or a desk light, or a lightbulb....the camera's back latch's so-called lock would not stay locked...a film rewind knob that fell off and was lost...a film advance mechanism that broke after a year's light duty use...an utter POS camera and lens combo, all made by the fine folks at the USSR's Zenit outfit.

One thing though, I was always impressed with the Zenit company's motto: "Zenit--where 1970's Soviet technology and alcoholic Russian assembly-line workers combine to make the shoddiest camera and optics gear this side of a Pakistani sweatshop camera factory. When it has to be second-rate, make sure it says Zenit."

 

[Gavjenks]

I chuckled when I heard that Russia's (in)famous Zenit company would be building this thing!!!

zOMG...when I was a boy, I bought a Zenit-made 35mm SLR, the Cosmorex SE, sold by New York's old Cambridge Camera...wow...it had a POS (no, not Positively Old-School, but the other kind of POS) 58mm f/2 thread mount lens with shiny, tinny diaphragm blades and primitive anti-reflection coating--the lens flared like a bastard whenever it was pointed in the same direction as the sun, or a desk light, or a lightbulb....the camera's back latch's so-called lock would not stay locked...a film rewind knob that fell off and was lost...a film advance mechanism that broke after a year's light duty use...an utter POS camera and lens combo, all made by the fine folks at the USSR's Zenit outfit.

One thing though, I was always impressed with the Zenit company's motto: "Zenit--where 1970's Soviet technology and alcoholic Russian assembly-line workers combine to make the shoddiest camera and optics gear this side of a Pakistani sweatshop camera factory. When it has to be second-rate, make sure it says Zenit."

It's being made by Zenit in Russia? In that case, I revise my estimate for the cost of one unit from $110 to about $50-75. Which gives them another $20,000 or so from their funding goal to go toward overhead legal expenses and such, if for some ridiculous reason, they wanted to try and patent something about this, or bully some people or whatever. Sigh.

 

[Derrel]

Yeah...Zenit!!! I know!

They make a mighty fine sardine tin!!! Best in the business,some people say. Robust, old-school, Soviet-style key-opening lids!

And I hear Zenit is pretty doggone good at toy kaleidoscope assembly as well!

 

[amolitor]

Presumably they're building some tooling and so on. Just because you can make one for $100 BOM cost doesn't mean.. anything, really. Manufacturing isn't taping **** together in your garage.

Although with Zenit there may be more overlap..

Cheaper than the Cooke, though!

 

[Gavjenks]

Presumably they're building some tooling and so on. Just because you can make one for $100 BOM cost doesn't mean.. anything, really. Manufacturing isn't taping **** together in your garage.

Although with Zenit there may be more overlap..

Cheaper than the Cooke, though!

Why would you NOT tape **** together in your garage, if you could? Not literally your garage, necessarily. Perhaps your existing warehouse/whatever if you're an extant company like Lomography, but still. Set up some tables with some cheap jigs and hand tools, maybe a couple pieces of basic shop equipment like a lathe, drill press, maybe even a mill. Then design an efficient flow and hire local people to assemble your stuff. They could probably put together a brass tube with some mounted lenses and a handful of drills and screwing in half an hour, tops, even if you have a terribly designed assembly line. So my $10 labor covers up to $20/hr (wages + employer taxes + whatever insurance, etc.) = probably $12, plenty to have a good supply of workers in most places in America right now, let alone Russia.

And even if you own none of that manufacturing equipment yet (Lomography probably does though), that all only adds a few thousand to your costs. So 15,000 instead of 10,000. Whatever. Still not millions, or even hundreds of thousands being necessary here. It's not like they a dozen rubber molds and 6 laser cut pieces, electrofocus clearances and 0.01mm tolerances on every unit, or anything. It's a tube with another tube inside of it. Come on.



I know for a fact that Lensbaby, for instance, does pretty much this. They assemble stuff with decently paid American assemblers at a bench of mostly hand tools and low grade shop equipment, in a studio space in Oregon, and still offer everything easily at 75-300 dollar price points or so. Anything much fancier than that, and you're probably better off just outsourcing to a CNC shop or whatever you need, unless you're selling tens of thousands of units, or such.

 

[amolitor]

Lensbaby's approach to QA, just to pick one aspect of an actual manufacturing process, is "the less, the better". This is, um, different from more widely used processes.

 

[tirediron]

Interesting; I wouldn't mind trying one, but the project seems woefully under-funded to me... I would think you would need to be in the millions of dollars of investment to make this work.

Uh... millions? What? You could almost certainly do this with 10,000 dollars.

I suppose it all depends on their approach. If they're going to make a few dozen units using off-the-shelf components, maybe, but if they're actually going to be produced on a commercial scale, not so much. Even Zenit is going to have to tool up to fabricate the barrel and the lens elements, etc. If I could buy the elements off-the-shelf, I could turn one of those out in a day for $50 in materials, but to do it on a big scale is going to mean a substantial investment in materials, tooling, CNC/CAM time, etc.

 

[Derrel]

I think the needed R&D budget for an existing camera and lens making company would be more than $10,000, but less than "millions" of dollars. Especially in a country like Russia, where American dollars are a desirable currency to be payed in, and where wages are not astronomically high, even in the sciences and manufacturing segments.

I looked up Petzval lens: Petzval lens diagram - Google Search

This is pretty doggone low-tech, and as mentioned, the focusing system seems like a pretty simply rack and pinion drive, and the stops are simple Waterhouse-style metal plates...no complicated mechanical iris diaphragm,no instant-return automatic diaphragm, but OLD-school Waterhouse stops...this is 1840's technology.

I would wager than any one of a thousand optical engineers could take a Petzval design from any era and scale it to cover 24x36mm in 85mm focal length.

As far as the optics...Edmund scientific sells grab bags full of 20-25 lenses for $11.94 in quantities of 5 bags or more...this thing only needs what? One cemented front group, and two rear group elements? I mean, it's like a barrel, four spacers, and a locking ring on the front...this thing is SIMPLE, as far as a multi-element lens goes...and some aberration in this kind of an "art lens" is no big deal, and in fact may very well be desired by many buyers. seems like with CAD design and C&C machining, the mechanical aspect of this thing ought to be like a one-week design project, at most...if it's being done by a company that already makes cameras and lenses and has people whose jobs center around building lenses.

 

[amolitor]

I think you still want some precision in the Petzval design. Isn't it supposed to be a "good" lens except for a tragic lack of field flatness?

Wikipedia suggests that you might be able to get all the Petval lenses you can eat out of olde skool slide projectors. I know the wet plate guys use these things a fair bit, didn't know they were Petzval designs. Like all things Lomography, this project is mostly about extracting money from hipsters. Doesn't mean you don't need some cash to start up a manufacturing operation, though.

 

[Gavjenks]

I think you still want some precision in the Petzval design. Isn't it supposed to be a "good" lens except for a tragic lack of field flatness?

Wikipedia suggests that you might be able to get all the Petval lenses you can eat out of olde skool slide projectors. I know the wet plate guys use these things a fair bit, didn't know they were Petzval designs. Like all things Lomography, this project is mostly about extracting money from hipsters. Doesn't mean you don't need some cash to start up a manufacturing operation, though.

My random single achromat I found in my cupboard and made into a lens the other day with blue painter's tape (not black velvet. not bulkheads) was already good enough for serious portrait work on the artsier side of things. it's sharper than almost any lensbaby, lomo, diana, etc. for instance (I think they intentionally make theirs crappy, not just out of cheapness). If I were using 4 elements and could choose all their characteristics, I could make it easily 5x better, on par with the best of the best of actual non-intentionally-aberrated 19th century or early 20th century lenses, for $50-$100.

It's not actually very hard to design lenses! Yes, it is hard to design 10x ZOOM lenses that are 1/3 or 8x the size of their focal length and have almost zero defects wide open at f/2.0 and have everything electronically optimized with zero sound and embedded gyroscopes, sure. But it's not hard to design a normal, mid range focal length manual lens. Anybody could learn in a day, seriously.

WinLens: Install <-- Free, and you can put in any parameters you want, and it will tell you how much coma, astigmatism, CA, SA, vignetting for your sensor size, blah blah anything you want. Even with camera movements included. From my experience plugging in some ones I've actually bench tested, it's pretty damn accurate.

As you mentioned, amolitor, I think most of the business acumen in that company is coming up with ways to attract the interest of hipsters. Like the stupid brass barrel schtick. Brilliant. I think that represents 10x more cleverness than is required to actually manufacture the thing.

If I could buy the elements off-the-shelf, I could turn one of those out in a day for $50 in materials, but to do it on a big scale is going to mean a substantial investment in materials, tooling, CNC/CAM time, etc.

Why would you use a CNC machine to cut brass tube in a straight line? That takes 20 seconds on a $200 bandsaw by hand. Similar for all other assembly steps for a product of this type.

Sure you COULD cut a brass tube with a 100 watt laser cutter instead, and pay for industrial robots and programming. Or you could build a wooden pre-measured form that sits in your bench vice and let some $12/hr schmuck push them through a bandsaw instead, finishing an entire 1,000 unit production run of tubes in like 2 days, and costing you probably less than the shipping alone would be to the CNC facility.

CNCs are not just something you use by default whenever you make more than 12 of something. Even the largest scale manufacturers in the world will still use humans instead, if humans are cheaper per unit. And in this case, they probably are, for anything less than tens of thousands of tubes being cut. Lomography is not going to be selling 80,000 hipster petzval lenses... or if they are, they don't know it yet. Else they wouldn't be on kickstarter, raising money for 300 units.

Hell, I probably could have cut the tubes for half that many units in the time I've spent writing posts on this thread. Or in the amount of time I would have spent on the phone getting a quote from the CNC people.

 

[vintagesnaps]

I love the idea of it, using something modeled after what was used during the early era of 1850s photography in the 21st century. But then I have a vintage Corfield lens on my digital camera pretty often so I don't see why it couldn't work. I guess it will depend on the finished product.

I'd been on this site not too long ago looking at the page on Voigtlander lenses which led me to the page on Petzvals. Antique and Classic Cameras
Antique and Classic Cameras

I don't think all the info. in the original story is completely accurate (which seems to happen often enough on many sites); the Petzvals were not popular in the 19th century as much as they were some of the earliest lenses, used often on daguerreotype cameras I believe. By the later 1800s I think they were already being replaced by newer lenses, and similar lenses were made by other companies as well.

And Lomography has been around since the early 90s; the article which might have benefitted from a little more research seemed to imply it hadn't been much more than 10 years, it just probably wasn't well known in the US til more recent years.