An in-depth tour of Nikon’s Hikari Glass factory | Hacker News Reader

An in-depth tour of Nikon’s Hikari Glass factory

imaging-resource.com

225 pointssohkamyung39 comments

> ...the general shape of this diagram calls to mind the shape of the Japanese archipelago, so Japanese optical engineers will often refer to a type of glass as a region of Japan. For instance, if an engineer is looking for a glass with low Abbe number but high refractive index (the extreme upper right of the diagram), they'd say they're looking for a "Hokkaido" glass. (Hokkaido is Japan's northernmost island.)

I found this a wonderful little detail that shows the humanity of the factory.

> It looks a bit like a map of the Japanese islands, so optical engineers will often talk about "Hokkaido", "Tokyo" or "Nagasaki" glass.

It is interesting; I wonder if (especially since so much of the glass in optical devices comes from Japan) these terms are pretty universally-applied in the lens-making industry. It's one of those things that makes perfect sense once you've seen both the chart and a map of Japan... but if not, you'd be quite puzzled.

Interesting the comments about the bubbles being detrimental to the optics. For most of last century, having bubbles in the lens elements was actually a sign of a great lens! It was eventually narrowed down on how the glass was cooled (confirmed by this article) and it seems the slight change of refraction was largely compensating for the bubble(s).

I've got some early, pre first world war lens that are absolutely amazing optically, bubbles and all. I've also got one Tessar 4 element lens in particular, on a pre 1936 (uncoated) Zeiss 6x9 Folder that is just out of this world in sharpness.

High quality glass making has a long history and it's only been the last 50-100 years that Japan has climbed up the tall ladder to become a world-class glass manufacturer.

Some of the most important science ever done was done using German lenses by French scientists (the germans didn't have the scientists, the french didn't have the lensmakers) - 2+ meters lenses used to concentrate light to burn metal in evacuated chambers. Eventually, Germany developed its own industry and science center around glass (Jena) and became the world's leader in high quality lenses (late 1800s).

In the US, it was Corning that did this work. They did melts and cools that lasted months to make the blank for the great Caltech telescope in the 20s or 30s.

it all comes down to the frit.

The image that struck me is the tree senior executives comfortably dressed not in golf shirts or suits, but in practical work fatigues.

What struck me is how everyone is clearly very senior (in age) looking, or at least how nobody looks young. Are there no apprentices/journeymen type of up&comers? I haven't had a chance to read the full in-depth article, so maybe that is mentioned, but this looks like one of those jobs that probably doesn't attract a lot of interest from the younger generation.

I think there are (young) apprentices. But some parts of the process probably requires years of practice and experience before they are allowed to work with minimal (or no) supervision that only 'older' people appear at those stations.

An example from the article:

I asked Arai-san how the press-operator knew how long to apply the forming pressure, and he replied that there were three factors: 1) the softness of the glass, based on its appearance, 2) how the glass feels while pressing it, and 3) how the glass feels when it drops from the mold. I can imagine that it takes a lot of experience, to be able to take all these cues into account, to produce perfectly-pressed preforms!

The two people in one of the first pictures look younger - but Japanese people are old in general. The median age is 47 and the population is actually decreasing.

Their economic system of having everyone pretend to work 12 hours a day may have had some issues.

There are major cultural difference between America and Japanese corporate life. Among them are uniforms for workers, and not just factory workers; many companies have standard-issue uniforms for every employee from janitor to CEO.

Not sure that's the case here, but it's got to be much more common in Japan than the US.

Unsurprisingly toyota lean got managers go to the factory lines to actually see and talk.

It's interesting to see how manual the process still is.

I would've thought for something like manufacturing glass, you would be able to make the process more or less entirely automated.

For cheap glass with big error bars on refractive index, dispersion, and defects, the process is entirely automated - think window glass or beer bottles. Quality control is still hard, and a lot gets recycled as cullet or discarded, but that isn't too much of a problem, since it's cheap. When you're working at that sort of scale, the furnace/melt behavior tends to be quite predictable and stable as well. Once you've dialed it in, it's good for quite a bit.

For such small batches of glass (500kgs is nothing!) with very tight tolerances across lots of variables, and considerable compositional diversity (Nikon isn't melting just one glass, it's melting quite a few, to do clever things with dispersion), there are so many variables that you can't really automate everything.

(As a glass melting nerd, I'd really love to see all of the proprietary information that they left out - but I guess that's why it isn't there.)

Curious: would making big pieces, like sliding glass doors, be automated, manual, or a mixture of both?

Entirely automated - if it's glass, big, flat, and smooth, it was almost certainly made via the float glass process, or an equivalently automated continuous-flow system.

Any ideas on the process used to make the glass Apple are using on their new campus buildings?

Is it mainly automated or more manual?

Also does it have any differences like more iron or other additives to lend strength or conversely less to lend clarity?

Sliding glass doors are typically one or two panes of hardened 'float' glass. I put the 'float' between quotes because these days hardly anybody makes float glass by floating molten glass on plates covered with molten tin but instead the glass is cast continuously between rollers. The cooling regime has a large effect on hardness (as does composition), and for extra strength there are other tricks such as lamination.

Nope! The tin-bath process is still dominant these days - the only people rolling glass are small-scale stained glass manufacturers.

For lower volume (or higher tolerance materials), the overflow downdraw method and updraw methods are used, mainly for display glass.

You're right, I had the volume mixed up, thank you for the correction.

But:

> the only people rolling glass are small-scale stained glass manufacturers.

What about:

- the glass used for solar panels

- the glass used for wire bonded glass

- patterned glass

AFAIK those are all made using rollers.

I'm guessing this is the main reason lenses like the 600mm f/4 FL VR are $12,000+ (probably every element is handmade/inspected), while cheap plasticky lenses like the 18-55mm f/3.5-5.6 VR are $100 or so.

My guess is the cheap lenses are fully automated, using fewer actual glass components, without features like 'ED' glass which requires all these extra manual processes and firings.

There's a lot that separates a cheap 18-55mm kit lens from the super-expensive telephotos. But probably the single biggest factor is that the 600mm f/4 requires much more glass and in much larger pieces, so even with the same target for optical quality you would expect to have much lower yield and higher manufacturing costs for the massive telephoto than the pocket sized kit lens.

It's probably more interesting to compare the differences between two lenses of similar focal length and aperture. With that kind of comparison, the kind of differences that you'll find are that the more expensive lenses tend to have metal instead of plastic frames, may have optical stabilization in the lens, fancier or faster autofocus systems, better sealing against water and dust, and more advanced coatings on the optical elements. In terms of the glass itself, you will also generally find that the more expensive lenses use more optical elements with potentially more complex shapes (eg. aspherical elements) to reduce distortion, so they tend to have more glass than the cheaper lenses. Given how many other factors contribute to the price differences, I doubt that you actually find order of magnitude differences in the price of unground lens blanks of similar size destined for cheap vs high-end camera lenses.

FWIW: the 18-105 has an ED element for "only" $400: http://imaging.nikon.com/lineup/lens/zoom/normalzoom/af-s_dx...

[The 55-200 also has an ED element for $150, but that looks like a sale effect, with an original price of $350]

Nikon seems to have moved a lot of their high-end manufacturing to Thailand. The D810 body is the first "pro spec" Nikon body with the circular viewfinder to be built overseas AFAIK.

They still have two plants in Japan where high-spec lenses and bodies are built. I think the criteria is "how hard is it for us to build", not ED (etc.) glass: The new 105mm f/1.4 and 200-500mm f/5.6 both have a lot of ED glass but are made in China.

There is still enough demand and value attached to "made in Japan" for Cosina to make vintage throwback "Voigtlander" lenses in Japan, and for Converse to make high-quality Chuck Taylor sneakers there.

A couple of years ago I was at Leica's booth at NAB talking to a rep about one of their Sumicron-C lenses. For the nominal fee of $100,000.00US, they will make you a set of lenses by hand. Of course, it takes 18-24 months to make them. For an additional $75,000.00US, they will nearly double the number of lenses in the kit. Apparently, if there is the slightest defect in the glass, they do not attempt to polish/grind it. They simply crush it and start with a new one. Hand made seems to be the way it's done. Maybe they have already learned what Tesla is learning. Automation is not always the best way.

https://leicastoremiami.com/collections/leica-summicron-c-le...

I'm not sure this is the case, to me it feels more like Tesla does premature optimization. They should first make it right, then make it efficient.

Not to rain on Leica too much (I love and had a whole M system) but I experienced some serious backfocus issues with their Summilux 50mm which despite two visits to their factory (which took two months) they were unable to fully rectify.

I’m not the only one, photographer Ming Thein has complained about the sample variation of Leica’s hand made lenses too.

The hand crafted nature of their range finder system is far from perfect and for the price it really should be better.

Japan still greatly values the "small scale, hand made" market to a degree effectively entirely lost in most of the rest of the world. https://www.youtube.com/watch?v=vYzQNgIwqgo was a great showcase of that, too.

Do you have other example of "small scale, hand made" products from Japan?

High end mechanical watches?

Satellites?

As has already been suggested - high-end watches; Grand Seiko and Credor.

Fountain pens - Sailor, Nakaya, Namiki for instance.

Oh, and knives. Definitely knives.

In their higher end items, they tend to employ traditional craft techniques (zaratsu polishing; urushi lacquer to name but two) which are incredibly labour and skill intensive.

Thank you. I'm currently wearing a Made in Japan Seiko watch, but not a high-end one ;).

Oh, don’t knock the Seiko - Seiko is a true manufacture - they do EVERYTHING in-house - movements, cases, dials, mainsprings - even lubricants.

This in stark contrast to some of their Swiss, in general more prestigeous counterparts.

(Do yourself a favour and have a look at the Grand Seiko website - they have some quite interesting info on their craft techniques. Not to mention that it is watch porn squared.

Oh, I agree that Seiko is doing great watches, I've had mine for years and I truly appreciate wearing it.

Sake is often made by hand on a smaller scale, here's a documentary on the brewing process: https://www.youtube.com/watch?v=quBO1QTlroI

Admittedly, that documentary is made by the marketing department for the sake factory, but it's still an interesting video.

Terada Gretsch guitar factory:

https://www.youtube.com/watch?v=GUB6QICAS5I

> It's interesting to see how manual the process still is.

This reminded me of a programme I saw a couple of years ago about making a perfect sphere. From what I can remember they said that was hand finished. I've just done a quick search and it looks like it would have been Achim Leistner making the 1Kg sphere.

From http://www.abc.net.au/catalyst/stories/s2038283.htm and more info at https://www.newscientist.com/article/dn14229-roundest-object...

Narration: The sphere started life as an ingot of pure silicon 28, brought here from Germany. Silicon crystal was chosen as the material because of its uniform atomic structure. The silicon ingot is cut and ground into a rough ball shape. Then Achim starts the painstaking process of transforming it into a near perfect sphere.

Achim Leistner: And this can only be done by hand because you actually have to feel what is happening. When you’re taking off a nanometre or 5 nanometres, or 10 nanometres or 20 nanometres, some of the errors we’re trying to correct are in the vicinity of only a few nanometres.

Narration: That’s a few billionths of a metre! If this sphere was the size of the earth, that accuracy would be equivalent to changes in height of only 4 or 5 metres over the entire surface of the planet. But that’s only half the challenge. To redefine the weight standard as a universal constant, the team also have to count the number of atoms in the sphere.

This was really interesting -- not a process I've ever realy thought about.

You might enjoy this, Ben Krasnow talking about the challenges of making photochromic glass:

https://www.youtube.com/watch?v=mUcUy7SqdS0

I've enjoyed reading this article, showing the process to fabricate just a glass lense.

The site has other articles like that one, but only related to cameras. Does anyone know of similar article, but for other industries?

> The finished preforms are packed in vinyl trays, with a number of preforms in each carrier - at least for smaller lenses like these. I imagine that the huge front elements for the like of a 600mm f/4 might call for more robust packaging.

It looks just like a muffin tray, though each muffin is a lens element!

At first i was like: "In-depth, yeah, right ...". Then half way in the material i found myself confused with how much information was delivered. Great work!

Reply on news.ycombinator.com