CP+ 2015 Sigma Interview - "small office, big factory"
At this year's CP+ show in Yokohama Japan we made time to sit down with several senior executives from major manufacturers, including Sigma. In this interview with Kazuto Yamaki, CEO of Sigma, we spoke about the challenges of making lenses for ever-increasing pixel counts, the company's 'small office, big factory' philosophy and why the company is continuing to make cameras.
|Kazuto Yamaki, CEO of Sigma, pictured at the CP+ show in Yokohama, Japan.|
Your Art lenses, optically, perform better than equivalents from - for example - Canon and Nikon, but at a lower asking price. How is that possible?
There are several reasons why we can do that. First of all we're a small company, comparitively. So we have a much smaller administrative staff than bigger companies. That's my father's philosophy - “small office, big factory”. What that means is that we have a very small, lean team for administration, but a big engineering team.
I've kept that concept and as a result Sigma has much lower administration cost. That's the first reason. The second reason is that we do most of our manufacturing at out own factory. If we used additional suppliers we'd have costs to pay to them. We try to do everything ourselves as much as possible.
Are you more focused on high-end products than mass market?
For economical reasons we need to focus on higher-end products. We make all of our lenses in Japan. The yen is weaker now than it was so our production cost is higher than if we made our lenses in other countries like China or elsewhere in Asia. But the second reason [that we're focused on the high-end] is that it's our pleasure to make high quality products, rather than high-volume products.
So you make more profit on the higher-end lenses?
It depends on the product. Even some high-end products have small profit margins because they're produced in smaller volumes, for example. The biggest difference from mass-market products is that if we add value to our high-end products, the customers appreciate it and will pay for it. In mass-market products the price is the most important part of the specification.
Do you see Sigma as a luxury, premium brand?
My dream is not to be a luxury brand as such. I want to produce high-quality products at a reasonable price. But in terms of quality, yes - we want to be seen as a premium brand.
Is there any particular technical innovation that you can think of which has made it possible to produce higher-quality lenses than you could in the past?
Compared to decades ago, of course computers help a lot. The turnaround time for ray-tracing calculations is much faster, so our engineers can try out many many ray traces. When I joined the company it took more than 30 minutes to make a single calculation but today we can do it in less than one minute. That's the biggest difference, but that's the same for all companies.
|An engineer in Sigma's prototyping department in Aizu, Japan, working with a computer on the design of a potential future lens.|
The other thing is our factory. Our factory has a very high capability when it comes to polishing very difficult glass elements. Optical design is always in conflict with [engineering] conditions. Conflict for example with the mechanical specifications, such as size and weight, and other things like the need for optical stabilization, as well as manufacturing capabilities.
If engineers in a factory cannot polish a particularly difficult piece of glass that is called for in a lens, the optical designer has to give it up. Our factory has a lot of ability when it comes to complex designs so they can [ensure] high yield [even of complex designs].
When did Sigma start using computers?
I don't know, a long time ago. Maybe 30 years ago - something like that. At the time, computers were so slow that they didn't change everything overnight. Optical designers still needed intuition and experience to design good lenses, and they had to stick to specific ways of constructing different types of lenses [they were more restricted to known lens designs]. But today, calculations are so fast, engineers can try all sorts of lens designs [and experiment].
Also, modern optical engineers can benefit from past designs. There is a huge database of lens designs in the industry, so engineers can check in advance what kind of design might be suitable for a specific lens.
How does Sigma deal with focus calibration issues with different DSLR manufacturers?
We cannot change the camera firmware, but we can adjust the firmware of our lenses. If customers send us their camera body and lens, we can calibrate them. That's the number 1 service request from our customers. But a lot of people don't like to do that because it's time consuming, so as an alternative we provided the USB dock for fine-tuning. If you use micro-adjustment in the camera body, it shifts focus by the same amount for all subject distances and regardless of focal length, in the case of a zoom lens. But with the USB dock you can shift AF according to subject distance and focal length.
It works very well. But if you have four focal lengths and four subject distance ranges, that's 16 values that the user has to enter. That's a lot. Are you interested in automating or helping to automate that process?
That's the challenge. We need to provide a better solution because it's not perfect for some customers. For someone with focus issues, first of all we recommend the customer to send in their camera and lens, and we'll do it for them.
What's the biggest challenge to lens design, in your opinion?
There are many challenges, but the biggest challenge is increasing resolution in cameras. Cameras will have even more pixels in the future, and not many manufacturers can make lenses that match the resolutions of those cameras.
|Canon's new EOS 5DS and 5DSR offer 50MP sensors, which according to Mr Yamaki, will really challenge a lot of lenses.|
Are you confident that your Art-series primes can resolve enough detail on the new Canon 50MP sensors?
I think so - of course it depends on the aperture value, but we believe that our new 24mm F1.4 is the best lens of this focal length on the market. So if it isn't usable on the new Canon cameras, there isn't another 24mm lens which is.
Can you tell us more about your Foveon A1 lens testing tool?
We developed our own MTF measurement instrument, which we call the A1. We use our own Foveon sensor inside it. Before we developed the A1 we had MTF measurement instruments which used conventional sensors, but the sensors weren't as sharp as we needed. So we switched to our own Foveon sensor, and now all of our Global Vision lenses are tested on the A1.
Is it difficult to correct for certain formats like Sony's full-frame Alpha series which have very short flange-back distances? Is that one of the reasons why we're not seeing Sigma lenses for those systems?
No - we want to increase our lineup for mirrorless cameras, but it's just a matter of priorities.
Can you give us an idea of your current priorities?
DSLRs first, particularly Canon and Nikon, since most of our customers use those systems. And after that mirrorless. Sony FE-mount.
You mentioned Foveon - you just released a new DP-series camera, the dp0 Quattro. It's fair to say that Sigma's cameras are not mass-market products. Are you happy that they're fairly niche cameras or do you want them to be more popular?
Business-wise I'd like to be more mainstream, but we know our customers. We need to differentiate our cameras from others. If we made conventional cameras like other companies there's no benefit to our customers. They're not a huge group but we know what our customers want, and the dp0 is a perfect camera for them. So our first priority is to please our existing customers, and then slowly increase and expand the customer base.
Can you give us a general idea of how profitable your cameras are compared to your lenses?
The camera business doesn't make money at all. It costs a lot to develop the sensors, so including these costs we're always losing money. So in strictly business terms it's not a good business. But I want to continue the camera business for as long as possible for two reasons. First of all it's our dream to be a camera manufacturer. That's a long-time dream of my father's. Secondly, in making cameras, we're accumulating a lot of experience and technical knowledge which helps us make better optics. We have some of the sharpest cameras in the industry, which means that in order to support our cameras, we have to make really sharp lenses. Which is good for all cameras. Also we had to develop our own processor to process images from our sensor, which allowed us to develop our new tools for MTF measurement.
So your latest lenses are better as a consequence of your camera development?
Yes, I think it has contributed to us being able to make such good optics. So in business terms maybe it doesn't make sense but in terms of technical experience and knowledge it makes perfect sense.
There were some rumors going around that your 24-105mm zoom lens was discontinued. Obviously the rumors were untrue, but has that lens performed as well as you'd hoped?
I'm happy with the optical performance of the lens as a standard zoom. This kind of lens is among the most difficult to design because it covers wide-angle, through standard to telephoto. The ideal construction of a wideangle lens compared to a telephoto is very different. So we need to compromise to some extent. And the 24-105 has good performance for that type of lens.
In the market, its performance hasn't been excellent, but it's OK. The rumor about discontinuation came about because we can't supply it in enough volume. We forecast demand for this lens as being very low, so we didn't plan to make many. But at some point they all sold out, and we then suddenly had a lot of orders placed - but then it takes four months[to restart production].
With lenses like this, which you expect to be relatively low-volume products, do you routinely manufacture them in batches?
Yes. Some popular lenses like our 35mm and 50mm primes and some popular zooms are produced continuously, but some more niche products are produced in batches every three or four months.
|The 'Art' series of Sigma F1.4 primes now comprises three focal lengths - 24mm, 35mm and 50mm (shown here). The 35mm and 50mm are optically superb and it looks like the new 24mm will follow suite.|
Obviously Sigma focuses on quality, which means big lenses, but are you interested in creating optics which are a compromise between quality and size?
That's the Contemporary line. Lenses in that line are a compromise between size, weight, price and performance. They're not low-end products but we're trying to create a balance. For the ‘Art' series lenses we're prioritizing performance over size, weight and cost.
A lens like Sony's Zeiss 55mm F1.8 for the FE mount is optically superb but quite small and light. If you were to design 'Art' lenses for that system, would you try to strike a different balance between quality and size, and try to make the lenses smaller?
Yes - that's also one of the segments we're interested in. Another motivation for us to make new lenses is to differentiate between other manufacturers. Sony already has a 55mm F1.8 which is nicely sized and reasonably priced, so why make another? But F1.4 compared to F1.8 is a huge difference, so we'd either make a F1.4, or something different.
What do you want to do next, in the 'Art' lineup?
The base concept is the best possible quality. The other challenge is to make something innovative which has never existed in the past. Actually, the new Canon 11-24mm is very impressive. That's a really good lens. It's expensive though.
|Canon's 11-24mm is the world's widest zoom lens for full-frame, and it seems that Sigma is itching to get into this market segment again.|
Do you think there's an opportunity for Sigma in the ultra-wide zoom segment?
Yes, we need to [get involved]. We're a pioneer of wide-angle zoom lenses. In the 70's we developed a 21-35mm manual lens, which I believe was the first real wide-angle zoom lens. Then we developed an 18-35mm, then a 17-35mm, then a 15-35mm, and a 12-24mm. We've always developed wide-angle zoom lenses. Canon has broken our record! We need to refresh this product line.
Many wide-angles don't accept filters because of their large front elements. Do you think this is an issue, or are filters becoming less relevant?
No filters are important. We want to make lenses that accept filters, but physically it's very difficult. Because light comes in at very wide angles. But customers do still believe it's nice to have [the ability to use] a filter.
What are your customers asking you to do next in the Art-series lineup?
The biggest request is for a 24-70mm F2.8. Then macro lenses, and also a wideangle zoom lens, something like a 14-24mm.
Where do you think Sigma will be in ten years?
It's very difficult to predict the future. But photography has a history of more than 100 years, so the market might shrink even more but enthusiasts will still exist in ten years. So Sigma will focus on those enthusiast and professional photographers and we will provide high quality lenses for those customers.
We speak to Mr. Yamaki at every possible opportunity, because his insights into the industry - and his candor about the challenges that it faces - are rare in a senior executive.
Partly this is because he is CEO of a rare type of company. Sigma is a family business, relatively small, and all of its lenses are made in a single factory in northern Japan, from materials mostly sourced domestically. Previously when we've spoken to Mr. Yamaki he has been quite open about his lack of interest in expanding his business beyond the point where it might outgrow the qualities that make it so special. I saw for myself his father's 'small office, big factory' philosophy in operation when I visited the Aizu factory a few days after we spoke at CP+. The practical advantage of a very localized supply chain is evident in the fact that the factory was able to return to limited production within only two days of the devastating earthquake that hit Japan in 2009. Companies with more complex supply chains struggled for much longer to return to normal operation.
In many ways in fact, Sigma operates a bit like a luxury, boutique manufacturer. Except that some of its lenses are in such high demand that the company can't make them fast enough. We see no reason not to believe Mr. Yamaki's confident claim that his new 24mm F1.4 is the best lens of its type on the market, and at $849, it's likely to be in very high demand. Assuming that Sigma continues to create and sell more lenses, will the need for increased production end up in conflict with Mr. Yamaki's desire to preserve the 'small office, big factory' philosophy? Only time will tell.
We do know that Mr. Yamaki tends to take the long view. To be honest, we weren't surprised to learn that Sigma loses money on its camera business, but we were impressed at his reasons for continuing to pursue it. One of the strengths of the dp-series of compact cameras especially is their superb optical quality. It makes sense that the experience gained in creating these products feeds into the design of the company's interchangeable lenses. Likewise their sensors. Although Foveon sensors have never challenged conventional Bayer designs in mainstream use, their specific advantages are appreciated by a small number of loyal users, and - crucially - they form the heart of Sigma's 'A1' MTF testing machines.
So what did we learn from this conversation? Firstly, that Sigma puts a lot of store in its core philosophy, which values engineering over administration (and indeed marketing). Secondly, that Sigma is continuing to focus more on producing high-quality lenses aimed at enthusiasts than on budget and OEM products. And specifically on unusual lenses which don't already exist in the market. Thirdly, we learned that Sigma is content to lose money on cameras for the sake of gaining knowledge that can be used to enhance its other, more profitable products.
Finally, it seems like Mr. Yamaki is keen to challenge Tamron, Canon and Nikon in the standard and ultra-wide zoom categories. Canon's 11-24mm specifically has obviously got his attention, and hypothetically, an Art-series 10-20mm for full-frame would be quite something, wouldn't it?
Note: This interview contained a mixture of on and off-record disclosures. As such, both questions and responses have been edited for clarity and readability where necessary.
|US Version, Canon|
|US Version, Canon w/ Filter Kit|
|US Version, Nikon|
|US Version, Sigma|
|US Version, Sony|
|Sigma 50mm F1.4 DG HSM Art Lens for Canon Cameras -...|
|Sigma 50mm F1.4 DG HSM Art Lens for Nikon Cameras - Fixed...|
|Sigma 50mm F1.4 DG HSM Art Lens for Sony Alpha Cameras -...|