Q&A: Canon helps usher in the video SLR era

The camera industry and photographers, having just gotten accustomed to the arrival of video in point-and-shoot cameras, just now are beginning to grapple with its arrival in the more serious SLR realm.

Chuck Westfall, technical adviser for Canon's professional products marketing division and a 26-year veteran at the Japanese company, is in the thick of it. Nikon was the first to market with a single-lens reflex camera equipped with video, the D90, but Canon offers video in two SLRs: the high-end EOS 5D Mark II, with a large sensor the size of a full frame of 35mm film, and the Rebel T1i, a more affordable, mainstream model.

These cameras combine high-definition video--1900x1080 pixels at 30 frames per second in the case of the 5D Mark II--with SLRs' advantages when shooting in dim conditions and with a broad variety of lenses. But even though today's video SLR features offers hold some appeal to enthusiasts and professionals, they're something of an awkward afterthought. SLRs and those who use them that haven't yet had much time to adapt.

Welcome to the world of digital photography, where change is incessant. In an interview with CNET News, Westfall talked about not just video, but also OLED displays, the arrival of rival full-frame SLRs from Sony and Nikon, changing flash card and file format standards, wireless networking, and more.

Question: The age of the video SLR has begun. A lot of people in the high-end camera market are set in their ways, and video is a radical difference for a lot of them. How does that change the camera design, the marketing, and everything you have to do to sell a camera?
Westfall: Right now we're at an early stage for all this. Actually, the camera design itself has not changed that much--yet. If you look at a camera like the 5D Mark II, it's really been intentionally designed as a still camera first and foremost. The video feature is definitely a very important addition to the camera, but it really is an option as opposed to the main purpose. One of the things we're hoping to accomplish with this approach is to give our still photographers--the bread and butter of our business--the opportunity to try something different without necessarily taking them out of their comfort zone--in terms of the camera shape, size, layout. I'd say that a very key element, at least in terms of 5D Mark II, is that it does have the full-frame sensor. This has a big impact in image quality in terms of low noise and angle of view issues, but it also has tremendous importance to a still photographer in the respect that the look and feel of the movies is very similar in terms of perspective and angle of view and depth of field. The familiarity photographers already have developed in terms of what lenses they're going to pick can be carried forward.

We pretty rapidly went from "Wow, freaky, video!" to "How come it can't do all the things I want it to do?" For example, setting the aperture or locking it at a certain ISO--why are those features not there?
Westfall: In the beginning, we had to take into account that there are going to be a fair amount of users out there who are into the customization aspect, but the overall customer profile on this camera (the 5D Mark II) includes a lot of amateurs as well. Anything we can do to give those customers automation on the basics of video, including exposure, shutter speed, aperture, and ISO--those are things that will make it easier to produce high-quality footage without that much experience.

To be honest with you, one of the other issues is that adding the full range of manual controls on this camera makes it a much more complicated instrument. It's not necessarily that we're never going to do it, but it's generation 1.0. We'd like to get some market feedback, which we've already received now, before we start making any serious changes to the overall feature set or design.

Do you think video is going to be pretty much standard in every SLR henceforth?
Westfall: We don't want to guarantee that. It is going to depend on the overall market strategy. But at this stage, the image processors we're using, especially the Digic 4, are powerful enough that it really makes it very easy for us to add that feature without increasing the cost.

It's fascinating to me watching the arrival of video. It's not as if it hadn't existed before, but when it started showing up in the 5D Mark II and the Nikon D90, there was this novelty aspect. It was the video snapshot that people were rediscovering. You could even call it a moving picture. It was like a century ago where people were excited by the possibility of watching a horse gallop. It seems like there are a whole lot of people who hadn't considered video beyond the kid blowing out the birthday candles--the equivalent of point-and-shoot photography. There's some cultural shift going on. I'm not sure if that's a short novelty phase or whether it's going to be a deep shift in how people use cameras.

Westfall: There's a perfect storm of technology developments that has occurred over the last few years that is really making this much more accessible. HD television penetration into the market is much deeper than it was even a year ago. HD cameras and camcorders are able to connect directly through HDMI to play back movies and still photos, so it couldn't be much easier. At the same time, the cost of recording media storage capacity is drastically reduced even compared to a year ago. 8GB or even 16GB cards are going for less than $100. That gives people the capacity to experiment and produce lengthy feature films if they wanted to.

How's the low-light performance compare to reasonably high-end HD video cameras?
Westfall: It's actually a lot better. In part it has to do in part with the size of the pixels on the sensor. Compared to a camcorder, they're huge. We're looking at 6.4 micron pixels on this camera (so each pixel is 6.4 millionths of a meter on edge). By comparison, on the CMOS sensor in the Vixia series' 3 megapixel CMOS we did, that was only a 1.7 micron pixel pitch. The difference between 6.4 and 1.7 is quite substantial. As a result, you're going to see an improvement, especially in the noise quality.

The 5D Mark II debuted a notch cheaper than the 5D, and everybody wonders what the trajectory is for full frame. We have Sony and Nikon providing some competitive pressure in the full-frame market. Is the low price because of the competition, because the technology is cheaper, or some other factor?
Westfall: I think there are elements of all those things involved in pricing. You've got to be competitive in the market. On the other hand, the manufacturing issues that have been developing over time have worked out in our favor as far as lowering costs. We've got a lot more experience than any other company in terms of developing full-frame CMOS since we were with the first one out there back in 2002 with the EOS 1Ds original. All that experience including the design as well as the manufacturing has really worked out in our favor to make the costs more reasonable.

What's the future trajectory? Are prices going to come down further? Are the people buying a 50D today--the higher-end enthusiast market--going to be making the jump to full-frame, or is there always going to be big quantum leap between APS-C (the smaller sensor size used in mainstream digital SLRs) and full-frame?
Westfall: We're going to evaluate the market overall in terms of the pricing issues and try to offer the best combination of features for the money. Value is very important to everybody, especially in this economy. That doesn't necessarily mean we would avoid a full-frame sensor just to cut the cost, but on the other hand we can't really say that a low-cost full-frame camera is going to be something that we're definitely going to pursue. We generally take it as it comes to see if we can find a spot in the market for the lower-priced full-frames.

With 21 megapixels on Canon's high-end SLRs, are we reaching practical limits for resolution? There's discussion that image quality now is limited by lens optics, even with pro-grade L lenses, regardless of the technical challenges of higher sensor resolution.
Westfall: As always, Canon's bottom line for digital imaging is the ongoing improvement of overall image quality, especially for printed output. With that goal in mind, Canon does not consider 21 megapixels "a practical limit" for sensor resolution in the 24x36mm imaging format. Higher sensor resolution is only one element of overall image quality. However, the evolution of the EOS-1Ds series from 11.1 megapixels in 2002 to 16.7 megapixels in 2004 and 21.1 megapixels in 2007, together with advances in digital image processing has clearly shown improvements in terms of lower noise and finer detail in printed output at any comparable ISO speed setting. The larger the print, the more noticeable the improvement. I would expect this trend to continue in future products. I would also expect to see further improvements in the optical performance of Canon EF lenses over time.

Nikon has been making market share gains in the last year. How does that change things for Canon in terms of pricing, marketing strategy, or product development?
Westfall: I'll say this: It's been very good for Canon as well as the rest of the industry and customers especially to have a very vigorous competition between the two companies. It works out to better products and very competitive pricing. We're in this business to stay. We don't back down from challenges. Our situation is pretty good. We are very well positioned in terms of our investment in R&D to be able to continue developing new products and technology to keep us at our No. 1 level. Based on the market research we've seen, although it is true that Nikon made some gains on us last year, we still ended up in the No. 1 position in units and dollars for both the SLR and compact camera categories.

How about Sony? I think they're up to No. 3 in the SLR market. Do they strike terror into your heart, or have they not made enough gains to make you lose any sleep yet?
Westfall: On the SLR side, they are making some headway, but they're still according to our information slightly below 10 percent of the market. So they are an important factor, but they are not as frightening to us at this point as they could be, theoretically. We're watching them and seeing how they're handling the issues they have.

Can you be more specific about the issues they have?
Westfall: One of the things that is an issue for any company besides Nikon or Canon is the huge installed base of owners that already own our lenses. Has Sony, being a relative newcomer, got enough user base to be able to chip into that share? The SLR market is not just camera bodies, it's also the lenses. For some customers it's even more so the lenses.

Yeah, as soon as you buy one lens beyond that kit lens, the barrier to switching becomes pretty high. The overall balance between APS-C and full-frame--it's probably way less than 5 percent are full-frame.
Westfall: I couldn't break it down for you, but full-frame is growing. Cameras like the 5D and 5D Mark II have really helped out a lot, and that's just for us. The other full-frame cameras coming into the marketplace from Nikon and Sony are also making a dent.

So the proportion of full-frames is increasing?
It is. The fact that there several products out there below the $3,000 price point is really helping a lot.

When you look at wireless networking support--there are a few cameras here and there that support it, but it really doesn't seem to be mainstream at this point. What's the barrier to getting 802.11 wireless networking build into a camera?
Westfall: There are a couple things. One is the existing 802.11b and g are almost at the end of their life cycle in the market. Although there's a lot of compatibility out there in terms of home networks for that technology, we're looking ahead to the next generation. At this point we have hopes that both the speed of data transmission and the range of transmission will improve substantially. Part of our issue is judging the market for the right time to bring that new technology in.

Are you talking about 802.11n or post-802.11n?
Westfall: N is one of the options. But there are other options out there. Canon is closely monitoring the situation to see which one of the options ends up being the de facto standard.

GPS support is another challenge. Nikon now has a widget that plugs into the hot shoe, but most people I know carry a separate GPS unit. What's the prognostication for built-in GPS for easier geotagging.
Westfall: It's going to be here sooner rather than later, but there are issues. One is the power consumption. The in-camera solutions we've seen so far are very thirsty for battery power, especially in an area where there is very low signal, and that is quite common when you're indoors for example. Another thing I think is going to have to be implemented to really make this fly properly is the ability to have a compass heading in your camera. It's not enough to know where the photo was shot; it's important to know which direction the camera was pointed.

Canon a few years ago demonstrated a 5D with an OLED (organic light-emitting display) screen. I just saw some cheap OLED keychain displays that cost $60 or so. What's the timeframe for building OLEDs instead of LCDs in the cameras? It's a much higher-quality image and I there might be battery advantages as well.
Westfall: It is a higher-quality image, no question about it, and the power consumption is lower because of the design. It's a very efficient design. Canon has been developing that technology for quite some time. We showed our prototype in 2005, and here we are four years later. I think the bulk of what's been going on in that interim is perfecting that technology--not just in terms of performance but also in terms of manufacturing process. I can't give you a timeline today, but we're a whole lot closer now.

One of the things I could tell you about is that in 2008, we purchased a manufacturing facility for OLEDs, so in Japan, we're creating the infrastructure to be able to bring this online. The issue for us is not just the perfection of the technology and lowering of the manufacturing costs, it's also being able to generate the quantity of these devices to satisfy our needs across an entire range of consumer products. If you look at our global production, we have in somewhere in the vicinity of 25 million or so compact cameras every year, and it's growing. Digital SLRs last year we did 4.4 million. And that's just cameras. Then you could start talking about the Pixma inkjet printers--a great many of them have LCD displays.

Since right now you don't build your LCD displays, that would give you a better profit margin, would it not?
Westfall: I think that's one of the issues for us. Probably the other issue is being able to control the time to market with a lot of our products.

Instead of having to wait for somebody else to develop it?
Westfall: Or, depending on their manufacturing capacity, how long it's going to take them to deliver an order. If we manufacture our own displays, we're in better shape to forecast our capacity.

The overall camera market, on the compact side, is flat or down overall. How's that affecting Canon specifically?
Westfall: We're still very optimistic. We've been No. 1 there since 2004, and we've been building our market share recently. We see that as opportunity for further growth, even if the overall industry is shrinking slightly this year. The fact is that even in these tough economic times, people still need to record their memories. The use of digital photography for that is not going to go away.

But there's a lot more growth in the SLR side. Why did that start to take off now? Is it pricing?
Westfall: It helped quite a bit that the price of SLRs has come down, but I think the other issue is that the penetration of SLRs is only about 10 percent compared to 80 percent for point-and-shoot. There's a lot of room to grow. And the installed base for point-and-shoot users is so huge--they are looking for something better. Whether it's the responsiveness of the SLR, or the image quality, or its interchangeable lenses--all those things are very attractive to point-and-shoot customers.

Why don't we have a snappy compact digital camera. One that focuses quickly and shoots very quickly. There's a big lag, still.
Westfall: It's actually gotten a lot better. Every year, it keeps on getting better. The one thing that is always going to be an issue is the autofocus system being part of the actual image sensor instead of a separate module. The so-called contrast-detect autofocus simply takes a long time to execute. We're making some inroads on that from the processer speed and the intelligence of the focusing algorithm to cut that down to the minimum level.

It's also an issue once you have live view in an SLR, especially with video. That seems to me a major disconnect when people move from the point-and-shoot world. You can use live view, but it's really not very satisfying because of the focus problems. When are we going to see that fixed?
Westfall: It's clearly a need that's going to be addressed. My guess is that in the next stage the product design engineers will have to think about a dedicated ASIC-type chip (application-specific integrated circuits are tailored for specific functions) that is really geared primarily to that task so you can speed up that operation without affecting other camera performance issues.

In storage, CompactFlash and SD seem to be the winners. You support both, and all your compacts use SD. What's the timeframe for the new SD standard, SDXC, arriving?
Westfall: I'm pretty sure it will be available around the industry by the end of next year.

What are the big advantages of that besides just capacity?
Westfall: There is really an advantage in terms of data writing speed. That has a big impact on performance issues such as shooting movies. If you're going to be shooting high-resolution, 30 frames-per-second HD video, you really need speedy data transfer. And I'm very interested from a personal perspective when this standard comes out if they will get past this existing file format issue, where the maximum individual clip is limited to 4GB. I think that should go away, and hopefully the SDXC standard might offer us a path to accomplish it.

Is SDXC good enough that you could get rid of CompactFlash and move solely to XC and its successors?
Westfall: Theoretically there's no reason why you'd have to stick with any of the legacy formats if SDXC is all that it's hyped to be, but we don't necessarily want to change our entire strategy at this point.

There's a new version of CompactFlash, the SATA version. Does that have some appeal to you?
Westfall: We're always going to watch to see what's available, whether it's SATA for CF or XC on the SD side. We like to get the right technology for the product. However, we're not in a position to confirm the details of future products.

There's a break in CompactFlash compatibility with the SATA version. As soon as you hit that discontinuity, that's a reason hat SDXC looks comparatively compelling to me if you can hit the capacity and data transfer speed you need.
Westfall: SDXC uses different (electrical) contacts, so the interface is not going to be the same. On the other hand, the form factor is basically the same. That bodes well for camera design.

There's a new version of Digital Negative (DNG) that Adobe Systems is kicking around that has the ability to store more data such as editing setting settings such as peripheral illumination correction or chromatic aberration. Does that make DNG potentially more appealing to you, or you sticking to CR2.
Westfall: As far as I know there's been no change in our status as far as keeping our raw image data proprietary. On the other hand, we are signed up with the Metadata Working Group, and we will be very active supporting those initiatives to make metadata much more effective for a variety of different software applications.

Any new thoughts on JPEG XR, which is destined to become a standard this year?
Westfall: I continue to be a strong proponent of that technology on my own. I'm not saying this is Canon's direction, but I've always felt that the JPEG XR technology that's being developed really has a lot to offer, both in terms of image quality and storage size. Those are both major issues where I think JPEG XR provides a huge improvement compared to existing JPEG.