The CP+ 2019 trade show in Yokohama, Japan, gave us the chance to speak to most of the major camera makers. Panasonic put forward an extensive team to discuss the company's move into the full-frame market.
Michiharu Uematsu - Technical PR Adviser, Merchandising Dep. , INBU
Tetsuya Uno - Group Manager, Optical Engineering Dep. , INBU
Koji Shibuno - Manager, Software Engineering Dep. Key (responsible) engineer of AF engineering, INBU
Takayuki Tochio - Senior Coordinator, Product Engineering Dep. , INBU Key engineer of Picture Quality
Hidenari Nishikawa - Senior Coordinator, Merchandising Dep. , INBU
Shiori Kitaoka - Senior Coordinator, Communication Dep. CMD
Please note that this interview was conducted partly via an interpreter, and has been edited for clarity and flow.
How different is it to design a lens for a much bigger sensor? What are the challenges?
When we're designing a lens, we look first at the most important properties: sharpness, transmission, distortion, vignetting, CA. But there are also some things are not measurable, quantitatively, such as bokeh. These are the 'taste' of the lens. So we need to understand this.
We assess bokeh at various distances. With Micro Four Thirds, the depth-of-field is deeper, relatively speaking. It's obvious where the tack-sharp range is and where is the 'big bokeh' [significantly out-of-focus area] range is. But now, for the full frame we need to consider the intermediate range: the 'small bokeh' area. It could be around the focal plane: that may already begin to be defocused with a full frame sensor.
We're constantly developing this to ensure the perfect 'taste' to our S-series lenses
We segment with depth: for the big bokeh area we want the beautiful bokeh with no onion ring effect and no unpleasant vignetting [of the bokeh: the 'cats' eye effect']. We need to have smooth vignetting from the center to edges, with circular bokeh in the center and a smooth progression to non-circular bokeh near the corners.
With the small bokeh [transitional] area we think about the double-line bokeh: which is considered a 'busy' bokeh effect. Around the focal plane area the transition between tack-sharp and out-of-focus is very important.
We needed a way to quantify and evaluate these usually qualitative aspects. This feeling when we look at the picture, we need to calculate or transfer those parameters to the quantitative evaluation.
We're constantly developing this to ensure the perfect 'taste' to our S-series lenses.
How well can you model bokeh or do you just have to build a sample lens and test it?
Thanks to very good simulation system we can roughly predict what the bokeh will look like beforehand. We create a prototype model but there can be gaps between what we expected from the simulation and exactly what we made, so we iterate with creation of the prototypes. So we model, create pre-production samples, evaluate and adjust before making another prototype model.
We firstly simulate a typical value of what we'd like to develop, then we can simulate the differences made by manufacturing tolerances, so we can predict those as well.
What are the challenges of making larger lenses with the high speed AF performance we've seen in Micro Four Thirds?
In terms of the mechanical part of the lens, we have a newly-developed double focus system and ultrasonic assist system. We have the double focus system in the 50mm lens and ultrasonic assist in the 24-105mm and 70-200mm.
In our double focus system, we have two groups of lenses which move independently for autofocus. Because we've separated the two groups, we can broaden the freedom of the development to give the best image quality and fast autofocus. Because we divided the AF into two, each group ends up being lighter, so they're faster, too.
We have a newly-developed double focus system and an ultrasonic assist system
For the ultrasonic assist, it's a new addition, on top of the linear focus actuator. The system keeps the lenses moving constantly, so that they don't have to overcome 'stiction' [the friction that needs to be overcome to start moving] before being driven to the correct location. This means the force needed to drive the lens is much less. So we have an ultrasonic system moving the lens a matter of nanometers, to ensure the lens isn't having to move from a dead start each time. It's a small enough vibration that you don't see it as AF wobbling.
Have you shared your DFD technology with Sigma and Leica?
The autofocus system protocol between the body and the lens is standardized in the L-mount alliance. So the other alliance members know what information the autofocus system needs. Beyond that it's up to the individual lens businesses whether they include that information to be fully compatible. It's up to them what they implement.
How do Leica L-mount lenses behave on the S1 and S1R?
Because this [DFD information] is already included in the protocol standard of the L-mount alliance, they already know how the body can command the lens to work.
The protocol hasn't changed: L-mount lenses that already exist should work in the same way as our new lenses. The protocol was already fixed when Leica started it as the SL system. Of course we're working on it all the time so it may be upgraded at some time, but for now it stays the same.
How does Panasonic plan to balance its resources between L-mount and Micro Four Thirds?
Because this was the launch of the S-series, we put the maximum effort into the S-series, including lenses. But you already know that we're developing the 10-25mm F1. 7 lens for Micro Four Thirds, that's coming soon. So of course we are developing both S-series and G-series at the same time.
The forthcoming Panasonic Leica DG Vario Summilux 10-25mm F1. 7 is supposed to show the company's continued Micro Four Thirds ambitions.
Has the expansion to a new system involved any expansion of engineering or manufacturing capacity?
We have optimized our human resources internally so that we can develop both S series and G series to the highest standard.
The S1 video spec is good but not up to GH standard. Do you see L-mount making sense for pro/semi-pro video?
First of all, let us confirm the concept of these cameras: the S1R is for the professional photographer while the S1 is for the hybrid stills and video photographer, so the concepts of those models are a little bit different. This might be why the spec is not beyond the GH series.
In the future, we'd like to consider the users' needs: it's open to users' feedback.
Do you feel there's an advantage to Micro Four Thirds, when it comes to video?
There are many categories of video shooters from broadcasting or run-and-gun, through to cinema and creative videography. If we are thinking about broadcasting or run-and-gun videographers, there may be some advantage to the small, compact and lightweight combination. That's the first benefit, compared to the larger sensor system. Image stabilization and overheating will be less of an issue, compared to full-frame sensors.
The degree of [creative] freedom broadens if you use a larger sensor, so maybe the cinema type of videographer would prefer the larger sensor, such as full frame. Maybe for cinema it's better to take full frame.
Was EVA-1 level videography considered when developing the L system and its lenses?
Of course we have been discussing with professional camcorder team as a fundamental part of the engineering, but we haven't discussed deeply how we might create such models. We're not discussing which mount would be used for the coming product.
Of course we don't disclose any further product information. At this point we don't have any future information we can disclose.
Panasonic talks about the 'cameraness' of the S1: the degree to which it feels like a camera
You're arriving in a very crowded sector of the market. What is it you think Panasonic can uniquely bring?
Firstly, cameraness: the interface, in terms of both hardware and software [making the device feel like a camera]. If you look at competitor models, they're very innovative cameras but they're having some negative comments in the market that the grips are smaller than expected or that the interface is [too] cutting-edge, leaving people confused when they first try to use them. We listened to a lot of customers' feedback to create the S-series, so we designed the position of buttons, shape of buttons and also the menu settings accordingly. We're proud of this cameraness and the interface, first of all.
In terms of features, we have the high resolution mode, which is number one in the market and Dual IS that gives up to six stops of stabilization. 4K/60p is a feature only we have in the full frame market and the electronic viewfinder is really high resolution: that too is number one in the market. This isn't everything, of course, but we can offer all those functions and features, to even professional users so that they can easily come to our system and start using them for their professional work.
Also our HLG Photo mode: this comes from our background in videography and this allows for a new photographic style.
What does HLG mean for stills photography: what happens next? Editing tools for HDR images?
The editing, as you say, and the public awareness are important in the near future. Firstly we'd like everyone to try the HLG HDR photos, where people can have a [means] of expression beyond the standard definition.
In terms of editing, we are co-working with an alliance: we're trying to find a solution to edit HDR images. On the technology side we'd like to be at least one step ahead of our competitors, so we will keep improving the image sensor technology to be capable for the high dynamic range world.
Of course we're one of the few companies that can create HDR TVs as well as cameras, so of course we're working with [our] TV business group. That way you're not only shooting but can also display and view HDR images in the best possible environment.
For the HDR Photo style we're working to have a wider [DR] option so that you can be really creative when shooting HLG photos.
Editor's note
This meeting was a mixture of briefing and interview, which is why some of the answers are quite long and detailed. The upshot of this is, perhaps, that you don't have to read quite so closely between the lines to see the message Panasonic wants to send. My own interpretation of it would be 'we're not just an electronics company: we care about even the most subtle aspects of image quality. ' Or maybe that's just my response to hearing the word 'cameraness' again.
The thing I haven't been able to capture in the text is the slide Panasonic showed me about the autofocus hit-rate they measured when shooting their cameras side-by-side with phase-detection-based rivals. The testing protocol wasn't fully disclosed, but it showed their products delivering a hit-rate comparable to some pretty good cameras, and even out-performing phase-detection systems when the subject gets close to the camera.
My own interpretation of it would be 'we're not just an electronics company: we care about even the most subtle aspects of image quality. '
This is something we'll be looking at as we test the S1 and S1R, because the general perception of DFD is that it's flat-out inferior to phase-detection. This isn't helped by the visually disturbing 'flutter' as the cameras try to maintain focus on moving objects (an effect made more dramatic by the high res viewfinders and shallow depth-of-field of the lenses on the S cameras).
Overall, though, it's clear that Panasonic wants its S1 and S1R to appeal specifically to professional stills photographers. When it comes to video, the company's plans seem less well-developed. For now, at least, it seems that Panasonic sees the GH series as its main video/stills camera platform.
. dpreview.com2019-5-13 16:00
