Hacker News Reader

Comment by k310 | Hacker News Reader

I looked for a simple explanation of why dispersion matters, and this seems helpful:

https://www.targettamers.com/guides/apochromatic-lenses/

in the context of apochromatic lenses: those that are optimized for three different wavelengths of light, not just two, which an achromatic lens does.

In the old days, calculations were done by hand, not that this is a big deal, but the big deal is that some really outstanding lens designs were made this way. Computers make the optimization extremely fast these days, but the calculations rely on the properties of the elements, which also vary in cost, durability and so on. So the computer can’t optimize for a continuous range of refractive indices and dispersions, only discrete real-world ones that the glass makers list, and which are specified (or not) to the program by the designer.

Fluorine also has special properties as a coating.

https://www.digitalcameraworld.com/features/this-is-why-your...

View on HN

> Computers make the optimization extremely fast these days...

A lot of computer design is now aimed at optimising for tolerances in lens elements and mechanical housings to reduce precision necessary when assembling them: they can drop elements into the tube and ship them off with little or no calibration: this is done particularly with kit lenses which are price sensitive.

I did error budgeting for environmental effects on optical systems at Itek. Glad to see so many optics folks here. I wonder who they are, but won't ask. ;-)

I assume that's the same Itek that made lenses for the Corona Spy Satellite program?

That's cool as heck. I am just a photographer, but I collect aerial lenses and own 2 Pacific Optical 18" f/3 lenses in 70mm format, serial numbers 6 and 13. I'll keep my eyes peeled for a surplus Itek 24" f/3.5 though :)

Yes, same one. I find de-classified articles and books now, and share them with former colleagues. All the designs I worked on were meant for space. And I have no idea if any were actually deployed.

As I noted in the “handle” post, k310 is the office number I had at Itek, as well as a nifty sonata by Mozart.

I later worked for Lockheed. More space optics and other cool stuff. I was in the R&D division.

I hope you can make good use of those lenses. I stuck with commercial optics (Nikon and Hasselblad). My optical hacking never exceeded making some adapters with a Unimat SL that I got. Surprise, computers got my attention, and my first “real” computer was an IMSAI that I built.

I read a lot. Kingslake, Smith and so on. A nice book, if you can find it is “Photographic Optics” by Neblette. It goes over all the classic designs, including all the familiar camera lenses from the 70’s and 80’s. Nowadays, computers do all the designing, and you can’t recognize “classic” designs like the tessar, Sonnar and so on, in them. But TBH, I finally got a new 105 macro lens with unrecognizable (to me) design from Nikon to augment the old 55mm macro that I started out with in 1970 or so. A simple double-gauss design.

Did you know what those “P” and “H” and other suffixes meant on Nikon lenses? They are the number of elements, in Optics Latin. H for hex, or 6 elements, P for penta or 5, and so on. The old lenses, without the “AI” aperture index gizmo, fit the new mirrorless body with the FTZ adapter. I gather that they were no-go on the DSLR’s. I skipped the DSLR generation entirely, since I used a Coolpix with 24-1000mm effective focal length for many, many years. Heck, it worked and got me great photos.

That's awesome! Thanks for taking the time to reply.

I am a young guy - just turned 30 last month - but I exclusively shoot and develop film and have my own mini darkroom. I enjoy the challenge and physicality of it all, plus I already spend enough time around computers and digital tech as a software engineer so it's nice to have some analog pursuits. An IMSAI is a bit before my time, but I have a soft spot for older CPUs like the similar z80 or 6502 from writing emulators.

For 35mm I mostly shoot on Canon FD, so while I recognize some of those classic design names I haven't shot on them. I haven't been able to shoot on the behemoth PO lenses as much as I'd like since they weigh nearly 60 pounds a piece and are a hassle to use, but I've jerry rigged a Graflex Crown Graphic to the back and can do a manual shutter with some ND filter and a quick on/off of the lens cap. They cover 4x5 or more at the distances I use them and are incredible lenses for portraiture.

My next projects are to mount both PO lenses together and put Graflex Speed Graphic 4x5 cameras behind each in order to make a massive binocular camera. I also have a Goerz 47" process lens that's got an image circle of nearly a meter, so my next shoot with that will be a 1:1 self portrait on xray film, since it's the only semi affordable option for such a massive exposure.

Great! You’re having fun. I got a few odd surplus lenses but since I started photography by taking flower and landscape photos, the “optics lab” part never took flight.

Trying to keep things simple here.

Itek made semiconductor lithography stuff way back when right? I did a stint at one of its competitors. I didn't know about the spy satellite stuff. Neat!

I'm not a lens designer, but want to play one on TV.

> in the context of apochromatic lenses: those that are optimized for three different wavelengths of light, not just two, which an achromatic lens does.

Achromatic lenses (corrected for blue and green) was acceptable for black and white orthochromatic film and plates which are only sensitive to blue and green. Even with panchromatic b+w film, achromatic lenses are usually satisfactory, but the chromatic aberration becomes visible with colour film. Aprochromatic lenses are corrected for blue, green and red.

Lenses can also be corrected for broader spectrums that include UV and IR: Nikon made such a lens – the UV 105mm f4.5 – for technical/scientific applications.

Telescopes for astronomy are often corrected well into the infrared regime due to some of the fairly interesting emission lines that can be imaged using an appropriate camera.

Zeiss and Hasselblad sold some superachromat lenses as well.

Mamiya did as well, for the RZ67. Or at least, their charts claim they are corrected well into 700nm+: https://ianbfoto.com/downloads/Mamiya%20RZ67/Mamiya%20RZ67%2...

They say no focus adjust is necessary for the APO lenses for IR, B&W, and color film.

The linked article perfectly explains it though no?

Never bother someone who has just came out of a rabbit hole.

I laughed at the fact that you seem so familiar with this state and are sympathetic to it.

I mean, dispersion matters simply because people generally don't want colored fringing and general haziness in their photos?

There are two types of chromatic aberration (CA), both caused by dispersion.

* Axial: perfect focusing is impossible because different wavelengths focus at different distances. If green light is focused correctly, then red and blue light is out of focus. This affects the entire image and is very difficult to fix in post-processing.

* Transverse: there's no unique image because magnification depends on wavelength. Blue light forms a slightly larger image than green, and green larger than red. This manifests as color fringing, most apparent near the edges and corners of the image, far from the optical axis. This can be alleviated algorithmically, by resizing the red and blue sub-images to match the green one.

Future improvements in simulations and manufacturing will probably enable micromanaging light in 3 dimensions. Ive seen some very strange fractal and composite lenses.

Reply on news.ycombinator.com