Sony Q&A, late Summer 2022: APS-C & RX futures, the BIONZ-XR processor, lens tech and more
posted Thursday, October 20, 2022 at 6:26 PM EST
This is now the fifth in my series of articles about meetings and interviews I had with photo industry executives during my last visit to Japan in late July of 2022. A myriad of distractions have contributed to my taking a very long time to get them all out (I still have one more interview and a couple of other tech articles yet to go), but fortunately the information has aged very well, and I know our readers appreciate the depth of detail I present.
I interviewed three managers from Sony’s camera and lens business; Masaaki Oshima, Senior General Manager, Camera System Business Division Imaging Products & Solutions Business Group, Masanori Kishi, Senior General Manager, Lens System Business Division and Atsushi Ueda, Senior Manager, Technology Communication Sect., Product Communication Dept., Imaging Products & Solutions Marketing Division. As their titles suggest, Oshima-san is in charge of all of their camera business, Kishi-san is in charge of their lens business, while Ueda-san is in charge of product marketing for the Imaging Products & Solutions business as a whole.
Oshima-san and Ueda-san were both present in the meeting, while Kishi-san provided written answers because he was unable to attend in person.
It was as always a wide-ranging and interesting conversation, but I think this time I came away with a bit firmer picture of Sony’s plans going forward, despite their understandable reticence to discuss them in any detail. Combining their answers with my own observations of the current camera market, I left with the feeling that I have a pretty good idea in at least general terms of what to expect from Sony over the next year or so.
Read on for all the details, and let me know your own thoughts in the comments below!
Just as I was about to post this, Sony released a teaser, saying that a new camera was going to be announced on October 26th! Stay tuned, I think this is going to be an interesting one, especially given some of Oshima-san's comments below!
How is the photo market doing these days?
(And some musings of my own...)
DE: My first question is about the market, and I’m interested to hear your perspective, given your position as a market leader. How do you see the current market and what do you predict over the next 12 months. My impression is that the photo market took a serious hit along with everything else, but at least the consumer side is coming back some now. Sony has a significant pro business, and I think that segment was particularly hard-hit, how is it looking currently? What do you think the next 12 months hold?
Oshima-san: So far, the mirrorless market is growing, even in this tough situation. Fortunately, we have been leading this market, although as you said, we have been affected some by the COVID 19 impact, especially for still shooting. But when it comes to video shooting, video recording demand keeps increasing even in the current situation so that’s why we’re continuing to lead in this market. We assume that this trend is going to continue, so that’s why we never stopped our development.
DE: So the still market has been declining, but video is increasing?
Oshima-san: Yes, because of the COVID 19 impact, the customer couldn’t go out to shoot, so that’s why in this period the still-shooting demand has been decreased. We think once people can go out again, the still-shooting demand will also return. And video demand is continuously growing, so we believe the mirrorless market will not shrink, but will continue to grow.
DE: That’s very encouraging, and I think you’re right. People couldn’t travel for a long while, but now US airports are *very* busy.
As I said, this is encouraging, and echoes what I heard from other manufacturers on my trip. Product demand for still photography has decreased (although some manufacturers have reported it holding level or even slightly increasing, depending on where they are in their product cycles), while video seems to have continued to grow unabated.
What’s my personal view? I think the camera market is going to be strong over the next period of time. Still photography is going to continue to be a strong market, built around the long-term core base of enthusiast shooters. Viewed separately, I think interest in still photography is actually increasing modestly from year to year these days, although the rate of increase is nowhere near that of video. In practical terms though, there are no longer such things as “still cameras” or “video cameras”, only hybrids, as there’s essentially zero incremental cost in making a camera capable of doing a very good job of both.
There are important ergonomic and feature differences required for the two types of shooting, though, and to that extent, there’s room for cameras to be differentiated based on their primary intended use. On a practical basis, this is more a factor at the two extremes of the video market, the low end and high end than in the middle, “enthusiast” area. At the low end, you want a camera that’s simple to use, which means not confusing the user with features, controls and complicated menu options that don’t apply to their usage. At the other end of the spectrum, high-end video shooters need a ton of very specialized features, that go far beyond anything a typical “hybrid shooter” would need, so it doesn’t make sense to prioritize these in the user interface.
This leaves a broad happy middle, occupied by enthusiast and professional hybrid shooters. These include still-photography enthusiasts and pros, many of whom also want strong video capabilities, without giving up still-capture features and ergonomics. It also includes pro video shooters who don’t need to go all the way to “cinema”-level capability, but still need very competent video recording. (I’m counting things like on-camera XLR inputs and sophisticated audio recording, extreme log settings, high-bandwidth RAW video recording, etc as “Cinema” level capabilities.)
Bottom line, I think the camera business is going to be stable and perhaps even grow modestly over the next several years. While there’s a big move towards video, I don’t think enthusiast still shooters need to worry about not getting the cameras they want and need from Sony or any other manufacturer.
Where are you at with supply chain issues?
DE: My second question is about supply chain issues. Things seemed to be starting to get back to normal, but then China fell into what seems like an endless series of lockdowns. Every company is different, how is Sony Digital Imaging being affected by supply chain disruptions, how have you been managing them, and do you have a longer-term plan to mitigate such problems in the future?
Oshima-san: Yeah, unfortunately, we had some impact due to supply chain issues; as you know, everybody has been [having difficulty]. We had some impact [along with everyone else], so we are continuously discussing with our suppliers how to minimize that impact. Fortunately, we have been doing our best to manage to cope with the severe situation to meet customer and dealer demands so far. From now on, we think the situation will get better, because we’ve already discussed with suppliers how to meet our demand in time for shipment from factory to dealers. So while we have had some impact previously, from now on, we’re not so worried about it.
DE: And even with the impact, there weren’t really product shortages?
Oshima-san: Unfortunately, we did have some impact [in terms of shortages], not only in one country but globally, but things are getting better.
DE: And even with the recent China shutdowns, you don't see a problem going forward?
Oshima-san: Yes, fortunately, we can manage that.
This also jibes with what I heard from other manufacturers; pretty much everyone seemed to feel that supply chain issues were all in the rearview mirror. Personally I think we haven’t entirely seen the end of them, due to the ongoing severe lockdowns in various parts of China resulting from their “zero-COVID” policy. While Shanghai seems to be staying clear of restrictions of late, critical technology hub Shenzen was locked down for weeks beginning in late August, with some areas still under restrictions. Given the infectiousness of the Omicron variants (which are also far less dangerous than earlier variants), it seems almost guaranteed that there are going to be more lockdowns going forward. How much these will affect the photo industry is an open question. One positive thing that’s happened over the last two years is that manufacturers now have much better maps of the entirety of their supply chains, so are much more able to plan intelligently to minimize impacts.
Where do APS-C cameras fit in? Will we eventually see a new high-end model?
RDE: Apart from a couple of APS-C models clearly aimed at YouTubers and other vloggers, Sony’s camera development efforts have been almost entirely focused on your full-frame product line. That said, three of your four lens announcements so far this year have been for APS-C lenses, but their specs and performance characteristics also seem aimed at the vlogger market. What’s your view of the APS-C space? Are the days of high-end APS-C camera bodies over? I’m sure many of our readers will be shouting “hell no!” at their screens as they read this, and both Nikon and Canon have relatively new APS-C bodies, but what does Sony see happening in the market? Is there just not as much demand for higher-end APS-C cameras, or is it more a matter of your very successful and highly profitable full-frame product line demanding all available resources?
Oshima-san: We think that the APS-C models are also important for us to meet the market demands. Our one mount is a strength, so you can use [the same lenses] on APS-C and full-frame cameras, and be used by pros, amateurs and video shooters. That’s why we have to focus more on APS-C bodies, not just full-frame bodies. Unfortunately, I can’t mention future products in detail, but yes, we understand the market demand for APS-C models as well, and we’ve never stopped development of them.
DE: So the last sort of high-end APS-C was back in 2019 [the Alpha 6600], so it’s fair to say that there will be more high-end APS-C models at some point, you just can’t say anything about detailed features or when they might happen. Is that something I can tell our readers?
Oshima-san: Yeah, as I mentioned, I understand the market demand in that segment, for APS-C bodies. I can’t mention anything about details,
DE: Maybe I can phrase it another way. There’s a lot of market demand for video for YouTubers. Are you also feeling market demand for high-end still cameras?
Oshima-san: Market demand is a little bit changed from only still cameras to still and movies, hybrid usage. That is why we have to consider how to meet both demands.
DE: Yeah, but when I think of the ZV-E10, that was really the last APS-C body, I think that’s very specifically a video camera, for YouTubers, and it doesn’t really have what people would consider high-end capabilities for still shooting. [Sony themselves actually refer to this as an “interchangeable-lens vlog camera.]
Oshima-san: Yeah, we prepared the ZV-E10 for the customer who wants to join our mirrorless world, and to start their “video life”. So that’s why we offered that model. Once they notice that shooting is fun, they’ll think of upgrading their gear. So that’s why if… once they want to upgrade their gear, some APS-C high-end model might be an opportunity for them.
DE: Ah, that’s interesting: The ZV-E10 was really developed as a way to draw people onto the platform. There’s a large movement of people who want to do vlogging; they aren’t traditional photographers, but if you can capture them, they’ll look for more advanced cameras in the future, adding to the demand from your current users.
It seemed that Oshima-san was itching to be able to tell me about a shortly-forthcoming high-end APS-C camera model, but wasn’t able to due to the embargo on it. It’s taken me so long to finally get this article posted that we now know what it was that he was chomping at the bit to talk about: The new Sony “Cinema Line” FX30, with a list price of $1,799 body-only, or $2,199 equipped with the optional XLR Handle Unit (the same as used with the FX3), which provides three audio inputs, two with pro-standard XLR/TRS connectors, and one with a standard 3.5mm stereo mic jack.
While the FX30 is certainly capable of shooting great stills, it’s clearly designed as a video camera first and foremost; rather than a high-end stills camera, it’s an affordable cinema body. (If the video focus wasn’t already clear enough, Sony doesn’t even list a spec for continuous still-shooting speed.)
It makes sense that Sony would extend their line of high-end cinema cameras down to a lower price point and smaller size, but I confess to being disappointed as a stills shooter not to see Sony’s latest technology make it into a high-end stills-oriented body.
That said, I can at the same time see that there might not be that much of a market for it. The costs of full-frame bodies have come down enough in recent years that the cost of a high-end APS-C body would overlap with a significant portion of Sony’s full-frame series. - So lower cost wouldn't be a reason a photographer would choose an advanced APS-C Sony body over a full-frame one. (Except in the case of a product like the FX-30, hitting the market at less than half the price of the next-lowest model in the Cinema Line.)
When it comes to size and weight, the difference between the smallest of Sony’s full-frame bodies and typical APS-C ones is noticeable but not terribly significant on its own. What matters far more for enthusiast users is the overall size and weight of the system as a whole, when looking at a gear kit consisting of a couple of bodies and some number of lenses.
In their lens development, Sony has very wisely focused their efforts on full-frame E-mount lenses, with full-frame image circles. They’ll have full-frame size and weight, but can be readily used on all of Sony’s mirrorless bodies. APS-C E-mount lenses by contrast can be smaller and lighter but will only work on a subset of their bodies. (And when you consider who it is that’s buying and using the two types of bodies, full-frame users as a group are far more likely to buy multiple lenses than the vlogger types who probably constitute a majority of Sony’s APS-C business these days.)
Given all this, there’s not really that much of a size/weight advantage for an enthusiast shooter when it comes to an APS-C vs a full-frame body within the Sony system.
When it comes to the competitive landscape, if size and weight are primary considerations for someone, they’d likely find systems from OM Digital Solutions and Fujifilm appealing, and the simple availability of a high-end APS-C body from Sony might not have much impact. (Of course there may be plenty of other factors affecting the choice of system, but my point is that Sony wouldn’t gain much in this sort of competition by offering a high-end APS-C body.)
Bottom line, as Oshima-san said and the FX30 proves (not to mention last year’s ZV-E10), Sony isn’t remotely stepping away from the APS-C market. At the same time though, I don’t see a market for a high-end APS-C body, other than the likes of the FX30. I do think that we’ll continue to see Sony’s product decisions revolving around video. The ZV-E10 covers the lower end of the range for the vlogger crowd, and the FX30 offers a significant step up in capability for ZV-E10 users looking to take the next step. With more than a $1,000 price difference between the ZV-E10 and FX30, I could imagine there being another model coming to fill that gap, but I don’t see still shooting on such a camera being more than a secondary feature.
How about advanced compact cameras, like the RX series?
DE: Then there’s the advanced compact cameras like the RX100 and RX10 models. Are you continuing to work on development in that space, or is there not much demand for capabilities beyond what the RX100 VII and RX10 IV offer? It’s been 3 years since the RX100 VII was announced, and 5 years since the RX10IV was new. Have those product categories reached the end of the line?
Oshima-san: Yeah… It’s difficult to answer, but originally, I was the father of RX series - so that’s why…
DE: Yeah, we must have met before, maybe that’s why you look familiar. If you were involved with the RX series, I’m sure we’ve talked before.
Oshima-san: Yeah, I think so, yes. From RX100 originally, RX1, RX10… I was the leader of the development for them. That’s why I love the RX series <laughs>. So now we are continuously producing the RX series, and while we can’t say anything about future strategy, but we don’t think we’ve quit the RX series yet. So that’s the answer…
DE: And in terms of the current models, it’s still a very viable product, people are still buying them in good quantity?
Oshima-san: Yeah, yeah, yeah, many customers are asking us about new models! <laughs>.
DE: So those continue to be very viable and popular products, and while you can’t talk about specifics, it seems there’s clearly a future for the product line.
Overall, I’m happy to hear that the RX models continue to sell well. The RX100 V is my go-to for an all-in-one pocket camera, and its capabilities are very impressive. (I’ve taken advantage of its 960 fps ultra slow-mo capability on a number of occasions, both for measuring small time intervals and for seeing the details of how mechanisms operate at high speeds.)
I’m not surprised that the RX cameras are enjoying a long life, as they don’t have any true competitors in the market, which is to say that for some of their functions, they’re the only way to get the job done. Still, it was good to hear from Oshima-san directly that they’re continuing to be popular, and that the line will continue into the indefinite future.
What’s Sony’s general strategy for AF development, particularly AIAF?
DE: While sensor technology has reached a level of sufficiency for most applications, autofocus technology is currently a hotbed of activity, with AI-based AF modes for specific subjects being rolled out by more and more manufacturers. Sony’s superb eye-detect AF was well ahead of your competitors when it was first released, and you’ve extended it to handle animals with great success as well. What about specific modes for things like motorsports, birds-in-flight, planes, trains and other subjects? What’s Sony’s general strategy or philosophy for autofocus development, particularly as it relates to AI technology?
Oshima-san: So obviously, image sensors and AI and autofocus are totally different technologies, but we combine them into our cameras. Originally, scene recognition was the most important to improve our cameras, that is to say “what’s this object” or something. Now we use our AI technology for the AF, but originally we had to consider whole object recognition. That was the most important point we focused on. So yes, our AF technology and AI technology are our base… strong technology, so that’s why we’ll continue this area of technology development. Unfortunately, we can’t announce any products yet, but definitely we will be announcing more advanced technology in the future, in terms of AIAF.
DE: So there’s definitely things coming in that area obviously…
Oshima-san: Stay tuned <chuckles>
DE: It’s interesting, you said even from the beginning, much before you were looking at using AI for autofocus, you were using it to recognize the kind of objects that were in the scene, to capture the best images.
This was some very interesting background on Sony’s AI efforts, and an exciting tease that more will be coming.
It’s interesting that Sony’s use of AI first began with whole-object identification for scene recognition, later expanding on that ability to use it for autofocus. This makes sense, but it hadn’t previously occurred to me as the way that AI technology’s use in cameras would develop.
Most intriguing is Oshima-san’s mention of more coming in the AI area in the future. He of course didn’t give any idea of what or when, but his manner suggests that there’s something significant right around the corner. As he said, stay tuned!
(At this point in the interview, Oshima-san had to leave for another meeting, so Atsushi Ueda took over, answering a number of more technical questions)
Some details about the BIONZ XR processor…
Ueda-san: Yeah, so each BIONZ XR chip is different, they have separate roles and they work together to achieve up to 8x the processing speed than the previous BIONZ X.
DE: Ah, so it’s not like there are two chips that are the same, doing the same things, it’s that the BIONZ XR’s functions are partitioned over two chips. Interesting. Olympus used two identical chips in their E-M1X, but you’re dividing the work based on different functions. How are the tasks divided, is one more video and one more still, or is it one does a lot of AF, and the other does more image processing, or…?
Ueda-san: We can’t provide more detail, but the total performance comes from the combination of both chips, working as a system. For example, the BIONZ XR in the Alpha 1 is able to handle a huge amount of data, like 50 megapixels 30 frames per second. It also performs the 120 fps AF calculations during continuous shooting. Those are examples of the BIONZ XR working as a total system.
DE: How much more power-efficient is the BIONZ XR than the previous chip? Do you have a metric for processing power per watt or some other measure of its efficiency?
Ueda-san: Yeah, it’s very powerful and power-efficient, but it’s very difficult to comment on specific numbers and we can’t say about the future. The BIONZ XR evolved the movie recording capabilities in particular, like 4K 120p and 8K 30p, together with the real-time autofocus processing. So it’s very powerful and power-efficient. And it also provides improved processing, for the new color science, for better color reproduction. Also to achieve lower noise and high dynamic range and high resolution in both still and movie recording. In order to bring out the full potential of the image sensor, the BIONZ XR processing engine is always working together with the image sensor as a total system.
DE: As you continue to evolve the full-frame product line, will we see performance increases in lower-end models as well?
Ueda-san: Yeah, we will be able to provide a better shooting experience and a unique shooting experience by the combination of Sony’s unique image sensors with the powerful BIONZ XR processing engine. - That’s kind of the unique advantage, because we are the number one image sensor maker, so we have Sony’s own unique image sensor, and we also have our own original BIONZ XR processing engine, so we can combine them to give a unique shooting experience to users
Ueda-san mentions the BIONZ XR as being Sony’s own design, which is unusual in the camera market. Long-time IR reader and commenter DrJon recently noted that most manufacturers use off-the-shelf SOCs (Systems-on-a-Chip) for their camera processors. DrJon named chip provider Socionext as the source for almost all of them) The two exceptions are Sony and Canon, who use their own self-developed proprietary chips. While we don’t know anything about how they’re making use of that capability, it presumably gives them a lot more flexibility in how they design their systems. That said, iterating silicon chip designs is a very expensive and time-consuming process, and the high up-front costs make it difficult to justify fully custom chips unless you’re cranking out millions of them. The economics obviously have worked for Sony though, and the control it gives them over their overall system architecture allows them to innovate in ways that companies using off-the-shelf chips wouldn’t be able to.
What led to the development of the very fast mechanical shutter in the Alpha 1?
RDE: The 1/400 x-sync speed of the Alpha 1 is very impressive, going well beyond any full-frame shutter we’ve seen in the past. I have two questions. First, what was the motivation for pushing further on mechanical shutter technology, vs dropping the mechanical shutter entirely, as Nikon did with their Z9?
Ueda-san: Yeah, although the flash sync speed has been an area that has not been evolved yet over the years. So we wanted to provide a new flash photography shooting experience, by using the newly developed dual-drive shutter to capture motion in a moment with flash.
DE: Yeah, I guess you can do flash photography with an electronic shutter, but the issue is that the electronic shutter can’t scan as fast as a really fast mechanical one.
Ueda-san: Yeah, that’s part of the performance of our newly-developed stacked CMOS sensor in the Alpha 1. The new stacked sensor has very high-speed readout, as you said. And the BIONZ XR’s very high-speed processing performance also enables high-speed flash photography with the electronic shutter. So both the sensor and the processor enable flash with electronic shutter.
DE: But is the x-sync speed with the electronic shutter longer than 1/400 second?
Ueda-san: Yes, 1/400 second is with the mechanical shutter for full-frame mode, and 1/500 second for APS-C mode, but it’s 1/200 with the electronic shutter.
DE: That’s a big advancement indeed; the fastest x-sync speed I was aware of previously with a focal plane shutter was 1/250 second.
This is unfortunately in error, thanks to my faulty memory :-/ The original Nikon D1 had an x-sync speed of 1/500 second, but its shutter design was dropped after that model. The D1’s shutter also only had to traverse an APS-C sensor, so an x-sync of 1/400 on the full-frame Alpha 1 is even more impressive. (If any readers have any insight on why the D1’s shutter mechanism was abandoned, please let me know in the comments section. Ditto if you’re aware of other focal plane shutters with higher x-sync speeds.) Regardless of historical precedent, as far as I know, the Alpha 1’s 1/400 second x-sync beats every other camera currently on the market.
What technologies let you make the Alpha 1’s shutter so fast?
The second question has to do with the technology in the shutter itself. This is the first time I’ve heard of a dual-driven shutter mechanism, using a mechanical spring along with an electromechanical actuator. Can you talk a little bit about how the new shutter works and what led to its development? (Does the spring basically just provide more “kick” to get the shutter blades moving at the beginning of the exposure, or does it continue to provide energy throughout the stroke?)
DE: So the second half of this question has to do with the technology itself. You describe it as a dual-driven shutter mechanism. So there’s a mechanical spring and also an electromechanical actuator. Can you talk a little bit about how that works, and what led to this development? Is it that the spring basically gives it a kick to get started, and then the electronic drive takes over? What can you tell me about that technology?
Ueda-san: Unfortunately, we refrain from explaining too many details about that technology. But the dual-drive shutter uses an electronic actuator in addition to the conventional spring, to enable the high speed drive of 1/400 second, which was previously unattainable. The use of carbon fiber for the blades in the shutter curtains also contributes to the light weight of the mechanism and its robustness.
DE: Ah, yeah, so you used carbon fiber to reduce the moving mass that had to be accelerated and decelerated.
Ueda-san: Yes, so it’s not only the dual-drive that makes it faster, we also have the carbon-fiber shutter curtains.
DE: Yeah, I (sadly) understand that you can’t describe too much detail about the dual-drive mechanism itself.
Was there any trade-off between still and video stabilization in the Alpha 1?
When it comes to image stabilization, your CIPA performance figures for still shooting (5.5 stops in the Alpha 1) seem a bit modest compared to some of your competitors, with some claiming as high as 8 stops of shake reduction. On the other hand, the Alpha 1’s active-mode stabilization is arguably the best I’ve seen for handheld video. What can you say about your approach to image stabilization? Were there tradeoffs between still and video stabilization that led you to favor video use cases?
Ueda-san: We don’t think there are tradeoffs between still and video stabilization. We understand that there are different needs for still stabilization and movie stabilization, and we have provided correction methods for both. Since we received many feedbacks regarding image stabilization, we would like to continue to provide unique shooting experiences with image stabilization. So we will continue our development of IS for both still and video shooting.
DE: So the 5.5 stops is just a separate limitation…
Ueda-san: Yes, it doesn’t relate to a tradeoff between movies and stills.
I had a number of questions about Sony’s latest tech and plans for their lenses as well and Masanori Kishi (Senior General Manager, Lens System Business Division) was kind enough to provide written answers to my questions even though he wasn’t able to attend in person. I unfortunately wasn’t able to engage in my usual back-and-forth with him, but I’ve added my own comments below to highlight various points. (As usual, notes I’ve directly inserted into his text for clarity are set off in square brackets, to distinguish them from his own words.)
What were the advances that let the new 24-70 and 70-200 f/2.8 be so small and light?
DE: You’ve recently released updates to two important f/2.8 lenses, the 24-70mm and 70-200mm. Both new models offer better optical performance in smaller and significantly lighterpackages than their predecessors. I’d be interested to hear a bit about how they were developed, and what let you make such significant gains in both performance and weight.
Masanori Kishi: Both [previous] lenses have been used by many customers in various situations, and we have received requests for higher image quality, smaller size and lighter weight, as well as faster AF speed and better focus tracking capability. And we are aware that customers’ needs have changed and been diversified recently, especially the need for movie shooting has increased a lot.
Masanori Kishi: To meet a wide range of customer needs for both still and movie shooting, we decided to release the second-generation G Master zoom lenses, fully equipped with the latest technologies including key optical elements and floating focus design with Sony’s original XD Linear motors. By combining the evolution of Sony’s unique related technologies with the latest optical technology, we have created a breakthrough in optical design, and realized a fusion of epoch-making performance and compactness and light weight. We would like to accelerate the fusion of these SONY’s unique peripheral technology and optical technology continuously [going forward].
This probably doesn’t need much interpretation; technology has advanced since the original lenses were designed, and Sony could take advantage of these advances to deliver better optical and AF performance while at the same time reducing size and weight. While not giving away anything specific about future plans, Kishi-san clearly indicates that we’re going to see more products coming based on these advanced technologies.
Sony’s latest “XD Linear Motor” focus-actuator technology is a key part of this, not only enabling faster focusing and focus tracking, but removing constraints on lens designers as well - which in turn can translate into improved optical performance.
Because autofocus speed is so important for photographers, it can often drive the choices lens designers make at very fundamental levels. With a given level of actuator technology, the lighter a lens’s focus group is, the more quickly it can be moved back and forth, and the less you have to move it the better as well. This can force tradeoffs on things like image quality or the size of the non-moving lens elements, all in the interest of keeping the focus group as small and lightweight as possible.
A lot of different ways of driving focus elements have been tried over the years, including tiny DC motors, ultrasonic actuators of various kinds and form factors, stepper motors and now linear motors.
It’s not just about moving the lens elements quickly though, they need to be moved very precisely as well, which requires extreme-resolution position sensors and fast algorithms and digital processors to control it all.
A linear motor is basically just a brushless DC motor (the same tech used in drones) “unrolled” into a flat arrangement. The powerful magnets needed to drive them mean that linear motors themselves take up more space in the lens barrel than some other options, but the power they deliver can be an order of magnitude greater.
I’ve just hit the high points here, but Sony’s XD Linear Motor focus actuators actually require an astonishing amount of technology to make them work with the speed and precision that they do. First developed for the 400mm f/2.8 lens released back in June of 2018, they’ve been at the heart of a number of lenses since then, and have allowed Sony’s lens designers to create optics they could only have dreamt of a decade ago.
How did you arrive at the optical formula for the new 24-70mm f/2.8?
DE: Looking at the optical configuration of the 24-70mm first, the front half (left side in the diagram above) of the new lens looks like essentially the same formula as the previous model, but an XA element has been introduced as a thinner “corrector” element, vs making the large convex/concave element aspheric. The back half of the lens looks very different though. What was the design process to arrive at this overall configuration? The new design uses twice as many ED and XA elements than the original. What are some reasons why those elements were either impossible or impractical when the first version was designed?
Masanori Kishi: To achieve the smallest and lightest F2.8 24-70mm zoom in its class, a completely redesigned optical design features five aspherical elements, including two high-precision XA (extreme aspherical) elements. The FE 24-70mm F2.8 GM II also includes two ED (extra-low dispersion) plus two Super ED glass elements. Combined with a new floating focus mechanism, the lens achieves not only outstanding resolution throughout the image area at all zoom and aperture settings but also improved minimum focusing distance. And for movie shooting, the focus breathing is reduced for expressive focus transition, and XD Linear Motors enhance AF and tracking performance even for high frame rate recording. This is an ideal match for high-performance compact Alpha system bodies, and will cover a wide range of shooting situations.
The question I was trying to ask was one that always intrigues me: How do optical designers go about a new lens design? What’s the process that leads them from new technology such as a new ultra-low dispersion glass or the ability to mold aspheric lenses from ED glass to a particular arrangement of lens elements and characteristics?
That intent of my question wasn’t clear in the way I wrote it, and I didn’t have a chance to follow up with additional questions in person, but Kishi-san does call out the key features of the new lens’s design.
Comparing the new design with the old one, it’s almost as if they said “let’s just use twice as much of all our cool stuff and see what happens”. I’m kidding of course, but it’s easy to get that impression when you look at the optical formulas of the two lenses. What’s significant is that despite the drastic increase in the number of ED, super ED aspherical and and extreme aspherical elements, the list price of the new 24-70mm is only $100 more than the original when it was announced in 2016.
What I find interesting in this design evolution is just that: It’s clearly an evolution of the previous lens’ design, in that the types of elements (plano-convex, convex/concave, etc) and their general arrangement are fairly similar between the two designs. The second element from the front is the same shape as previously, but now is made from ED rather than conventional glass. Interestingly, the fourth element from the front changed from an XA (Extreme Aspherical) to a conventional one, with a fifth XA element added behind it as a corrector optic, accomplishing roughly the same thing optically, but with a thinner aspheric element that’s probably easier to manufacture. (Aspheric lenses are made by pressing hot glass into tungsten molds. The softened glass takes on the shape of the mold, but significant differences in thickness across the element can cause problems as the glass shrinks when it cools. The thinner corrective element in the new 24-70 design is still characterized as an “extreme” aspheric, but I suspect it's easier to manufacture than the previous XA element that was so thick around its perimeter.)
As I mentioned above, what made the new design possible wasn’t so much technologies that hadn’t been available six years ago as it was that the manufacturing costs for those technologies came down enough that Sony could use them in almost twice as many of the new lens’s elements, while keeping the retail price nearly the same.
[These diagrams will be combined into a single illustration…]
The new 70-200mm f/2.8 is very different from the previous version. Why?
DE: There’s a much more radical difference between the old and new models with the 70-200mm lens. The new lens has far fewer elements (17 elements in 14 groups, vs 23 elements in 18 groups), and looks like a completely different optical formula, apart from the general similarity of the three front elements. What led the optical designer to essentially start with a clean sheet of paper for the new lens, vs taking a more evolutionary approach? Again, what were the advances in technology in the six years since the original was released that allowed such a completely different design? Does the reduced number of elements improve local contrast, or does the number of elements not matter so much in that regard, given modern nano-coating technology?
Masanori Kishi: To achieve the world’s lightest F2.8 70-200mm zoom, the new optical design reduces the number of lens elements in the front of the lens, and the use of an ED aspherical lens and the XA lens helps to reduce the number of lens elements overall, which also helps to reduce weight. Also, Super ED and ED glass are placed in key locations to thoroughly suppress chromatic aberration. Additionally, four of Sony’s original XD Linear Motors with its controlling technology deliver high thrust efficiency that boosts AF speed by up to 4x compared to the previous model. So, this lens is designed to provide a new shooting experience unique to Sony camera systems in both still and movie shooting of a wide range of subjects, including sports, portraits and landscapes.
Compared to the 24-70mm, the new 70-200mm looks more like a “clean sheet” redesign, although you can still see the earlier model's heritage. As Kishi-san said, they’ve significantly reduced the number of lens elements, especially those towards the front of the lens body. The more mass there is that’s further away from the camera body, the more unwieldy a lens will feel in the hand. - And of course, the less glass you need overall, the less the whole lens will weigh.
I of course have no idea, but suspect that the aspheric lens element made from ED glass is a key factor in the new design.
As I mentioned above when talking about the 24-70mm’s design, aspheric lenses are made by pressing hot glass into tungsten molds and then cooling them *very* carefully. This is challenging but do-able with conventional glass, but ED glass is much more temperamental. It softens at a higher temperature than ordinary glass, and its thermal expansion characteristics are more problematic. The bottom line is it’s very difficult to mold aspheric elements with ED glass; we’ve only recently begun to see aspheric lens elements made with ED glass, and the new 70-200mm f/2.8 prominently features such a lens in its 8th element from the front.
I’d love to know the full story of how the new 70-200mm f/2.8 came to be, but I suspect Sony wouldn’t be willing to answer my questions about that, even if I could figure out how to properly ask them. It’s intriguing though, that in contrast to the new 24–70mm f/2.8, this 70-200mm uses advanced technologies for fewer of its elements. This suggests to me that there’s something much more radical and innovative about the new lens’s design. The ED aspheric element is certainly part of the story, but it’s got to be more than just that. How did the designer arrive at such a radical departure from the previous design? Were there other things besides just the availability of ED aspherical lenses that led to the new design, or was it more just down to the creativity of the designer, to come up with a new approach that no one had previously thought of?
Whatever the process for having arrived at it, the new 70-200mm f/2.8’s design is a testament to what can be achieved with modern glass and manufacturing techniques.
The original 70-200/2.8 also still being sold. Will we see more of this with other updates?
DE: You decided to leave the original 70-200mm f/2.8 in your lineup, alongside the new model. This is reminiscent of Sony’s approach with compact cameras, leaving older models in the lineup even as new ones are introduced. (This always made a lot of sense to me, to continue selling older models at reduced prices, once the engineering costs had been recovered.) Is this something we can expect to see in your lens lineup, as more earlier designs are updated?
Masanori Kishi: We refrain from [discussing our] future lens lineup strategy, but Sony is always listening to our customers. Thanks to feedback from users around the world, including leading professionals, Sony continues to develop the lens lineup and has achieved the 70 E-mount lens selections. We would like to provide wider choices for a wide variety of customers.
This wasn’t much of a surprise; I never really expect manufacturers to talk about their future plans, but I always feel obligated to our readers to ask, and sometimes I do get answers :-)
We’ll see how things evolve, but a significant part of Sony’s strategy in high-end compact cameras (the RX10 and RX100 series) as well as the various generations of their full-frame mirrorless bodies was to leave earlier models on the market when new ones were introduced, just reducing their price points once the R&D costs had been recovered. This is perhaps less applicable to lens designs than it is to cameras, but the same general principle applies: Once you’ve recovered the original engineering and development costs, the incremental cost of just cranking out more units is significantly less. Why not simply drop the retail price and continue selling the products at the same profit margin, just on a lower cost basis?
Lenses have longer life cycles than cameras, and there are fewer obvious features to differentiate between newer and older models in the marketplace, but it could very well make sense to continue selling the original 70-200/2.8 alongside the 2.0 version. It’s possible that having both in the market at the same time is just a temporary situation as inventory is cleared from the channel, but I don’t entirely discount the possibility that we might see both versions selling alongside each other for some time to come.
What’s your lens roadmap look like from here?
DE: It’s been a while since I last saw a lens roadmap for Sony, what does your current plan look like? Will we see more redesigns of earlier models? Are there any new focal length/aperture combinations coming? (Any chance for a tilt-shift lens?)
Masanori Kishi: We cannot comment on our future lens roadmap, but we have a lot of new lens ideas which we would like to release in the future. Please look forward to it.
Again, this wasn’t an entirely unexpected response. It came out in the interactive discussion that lens roadmaps aren’t really something that Sony does.The last roadmap I’d seen turned out to have been all the way back at Photokina 2018. At that time, photographers were concerned about how sparse Sony’s lineup of full-frame lenses was, so they wanted to let people know that there were a lot of new lenses coming in a relatively short time frame. So they released a lens roadmap to reassure the market that their full-frame platform would be well-supported by a robust selection of lenses in a relatively short time frame. Once they’d managed to convey that message though, that was the extent of their roadmap disclosures, through the present.
In the physical meeting though, the other participants noted that Kishi-san wants to release as many new lens designs as possible. The impression I was left with was that of a mad genius, constantly bubbling with more ideas for unique lens designs than any one company could ever implement. - So it’s a given that there’ll be no shortage of new designs for the FE mount in the coming years :-)
As I said at the top, I came away from this meeting feeling like I had a better idea than usual of how Sony’s products will evolve over the next period of time. As with all manufacturers, we’ll see more and more advanced AI algorithms applied to autofocus. Sony was a very early leader in applying AI technology to camera functions, and Oshima-san seemed to indicate that we’d be seeing the next generation of their AIAF technology sooner than later.
It’s clear that Sony is very strongly focused on video shooting, with both of their most recent camera bodies being designed specifically for that market, one at the low end, and the other at the “affordable” high end. Enthusiast still shooters are very well-covered by their full-frame bodies and full-frame E-mount lenses, and I think we’re going to see a mix of redesigns of older lenses and new models aimed at specific use cases. (Where are the tilt-shift lenses, by the way? Sony now has a very robust lens lineup, I’m thinking it’s about time for them to make a tilt-shift lens. They’ve shown no inclination towards that market in the past, but I’ll go out on a limb and say that we could see a tilt-shift optic in the next 12-18 months.)
With the very recent release of the FX30 Cinema-series body, it’s clear that Sony has no intent of backing away from the APS-C format. That said,for reasons I outlined above, I’m not sure that we’ll see a high-end APS-C body that isn’t primarily video-focused. We might see a video-oriented model priced somewhere in between the ZV-E10 and FX30, but a high-end “Alpha 7000” model that’s more aimed at still shooters? I don’t know; I’d very much like to be proved wrong, but I don’t see a marginally smaller body being enough reason for someone to buy an Alpha 7000 over a full-frame model with roughly equivalent features at about the same price.
What do you think, though? I’ve included more of my own speculation in this piece than I normally do, and would really enjoy hearing whether you think I made sense or am all wet. Let me know in the comments below, and share any other thoughts you might have about what was said (or not said ;-) in this interview.