Tuesday, December 19, 2006

K10D Interview with Mr. Tatamiya of Pentax Japan

Recently, DC Watch Impress published an interview in Japanese about the K10D with Mr. Tatamiya of Pentax. With translating help from Ken Takahisha, I'm posting it here at OK1000 with English formatting. Special thanks go out to Ken for his input, the PDML list, and DC Watch Impress for their interview. It's very lengthy, but you'll find hints about the internal discussions that created the K10D at Pentax, why they feel the RAW format is so integral to a mid-level camera that it deserves its own button, and why the camera is heavier than its predecessors. So dig in!

The Reason Pentax has changed

After the announcement of the K10D, the market anticipation built up to the point that at first release the initial stock was sold out too quickly. In fact, they had delayed the release in order to “stock up” -- a drop in the bucket.

Although Pentax products have been enjoying quiet and lasting popularity, there has never been a moment like this. With various articles and reviews about K10D itself everywhere, we interviewed Mr.Tatamiy about the background of the development of K10D in order to explore why they could produce popular products, beginning with K100D.

Off the mark sales forecast

DC Watch Impress: Because of the attractive specs for the K10D, it was easy to anticipate the popularity in Japan domestic market, but it was also very popular at the Photokina booth in Europe which was traditionally a weak market for the Pentax brand. In Japan, people were already anticipating a possibility of stock shortage even before the pre-booking stage. How Pentax has been viewing these market reactions?

Mr. Tatamiya: There was the K100D this year which was very popular, and we naturally expected that K10D sales would be good too. In spite of this, we never expected the pre-booking situation before the release. We were truly surprised at the difference “by digits.”

DC Watch Impress: You said the forecast was wildly different “by digits”. Is “by digits” a metaphor, or, were the digits really different from the forecast?

Mr. Tatamiya: We thought it was going to be popular, but not as popular as the K100D, because the K10D was still a sub $1,000 camera. However, although we cannot give you any specific figures, we were overwhelmed by far more than twice the pre-booking quantity of K100D orders. For example, a big chain store who pre-booked 300 units of K100D now gave us a pre-booking order of over 1,000 units. This is just in Japan, but the market reaction in the overseas market is much larger than that.

DC Watch Impress: We had an impression that the Pentax brand was not particularly strong in Europe, but is it popular there too?

Mr. Tatamiya: The Pentax brand has been establishment in Japan, and there are many long time Pentax fans in North America. On the other hand, in Europe it was not particularly well recognized. However, the pre-release reception was excellent and in a certain country, an order of 1,000 units came in when we forecasted only 50 units for their initial order based on our history with them. Right now the order level in Europe is as good as in Japan or North America, which frankly caught us off-guard. Additionally, as far as the K10D is concerned, the reaction from the Asian market, particularly Hong Kong and Korea, is really on fire.

Organizational changes produced the K10D

DC Watch Impress: There was a comment by Mr.Torigoe in his interview at Photokina that the atmosphere in the development front line has significantly changed. As a leader of the project, what are your feelings about this?

Mr. Tatamiya: The excellent market acceptance of the K100D really changed the atmosphere. Engineers are human too, and they were very happy to find that the products they produced were so well received at market. That became a strong motivation, and the last spurt in wrap-up phase of the K10D was very “contents rich” and we are proud of it.

DC Watch Impress: What do you mean by “contents rich”?

Mr. Tatamiya: Development staff for both the K100D and K10D is partly shared and overlapping. During the 3 months since completion of the K100D until the last push for the K10D, they voluntarily offered their utmost effort to polish up the feature details and performance specs, even though a advanced-amature camera has a lot of features already and the work load and contents are considered heavy.

For example, relative to the 3 fps burst rate, it was barely achievable in the final stage. However, we still explored to see if it could be further improved by reevaluating, tuning, etc. The development coordinator specialized on this feature came up with an innovation only he could come up with, reconstructing the tuning sequence, and now at least 3fps is assured even under the most difficult exposure conditions. In actuality, you can measure it to confirm that there is slightly more than 3 fps, particularly under better conditions. If you trip the shutter, you can feel the speed of it.

Also, initially in the RAW burst mode, it stopped when buffer was full, after which writing to memory media continued. However, we designed a parallel writing in order to create as much buffer room as possible so that an additional 1 to 2 fps would be possible. This attention to detail was given by every staff member spontaneously, which resulted in accomplishing the design specs, and in some cases, exceeding catalogue specs.

DC Watch Impress: The *ist D and *ist DS also were good products, but they were not as popular as the K100 and K10D. What has changed before and after the K100D, and what part of the product concept do you think was appreciated by the market?

Mr. Tatamiya: It’s true that our working motivation has boosted tremendously, but the development staff never changed nor did their capability suddenly improve.

On the other hand, the organizational structure has changed markedly. The production planning department for the development team and marketing and sales staff were operating independently but were consolidated into one in April 2005. This means that the group which takes the pulse of the market and those who take initiative in product development work together. The first product to bear fruit from this arrangement was the K100D.

DC Watch Impress: In other words, reorganization to quickly grasp market needs and reflect them in products led to success?

Mr. Tatamiya: We were always grasping market needs, but the problem was our inability to consolidate them into one central location. By directly connecting marketing and development, it became possible to accurately target user demographics for each product. There is traditionally an opinion that SLRs are for shutterbugs and there is no marketing required. However, in reality, people of varying needs are buying DSLRs. We thought we should develop products for each target market, so that they would be satisfied.

Taking advantage of the smal company’s merits

DC Watch Impress: I see. Then, in the case of the K100D, how are those design concepts exhibited?

Mr. Tatamiya: Regarding the K10D, it means no “middle of the road” to satisfy everybody. The need to satisfy most everyone is we intended for the K100D. The K10D an upgraded model by one notch, with the easy picture modes eliminated, priority on features, weatherproofing, and the physical ergonomics, over just a compact package.

DC Watch Impress: When we first encountered the K10D, it felt rather heavy for Pentax, who have been advocating lightweight and compactness. Weight by itself is not necessarily a bad thing but it is certainly different from past path.

Mr. Tatamiya: In the very early stage of the development, we already decided to incorporate what we thought were necessary features for a mid-class body and associated parts without unduly shaving anything off. Weatherproofing was an important factor too. Then, all the parts to be squeezed into this body were lined up and the final size/weight were determined. There was indeed a concern raised about size and weight by the marketing people.

However, a model that prioritized small package and lightweight was the K100D. Also, for the scope of features incorporated in this body, we believe it is still rather compact compared with the competition. The K100D opts for features and easier user interface over miniaturization.

DC Watch Impress: Traditionally, Pentax always have produced unique features to improve operability and user interface, such as Hyper Manual, etc. This time, a new exposure mode, ISO priority mode, and other new exposure modes were added. How were these features created?

Mr. Tatamiya: Pentax’s development staff all love cameras and photography, and these engineers have been constantly suggesting a number of innovative features. But they were not particularly well adopted into new products in the past. However, the aforementioned organizational change enabled the development staff to have frequent discussions directly with the core staff of product planning and marketing, which in turn encouraged submission of new ideas. We always speculated what type of people would purchase the K10D, they must be those who prefer “real stuff”, so then decide what the real stuff is, etc.

In this context, we discussed a possibility of changing the basic concept of exposure control, so that users can choose what they think would be the best exposure mode depending on situations. It would be dumb if we could not maximize advantages of DSLR over film SLR. We examined various ways of operating a DSLR, taking advantage of features of digital camera, which resulted in the addition of two new modes.

DC Watch Impress: I gather that in addition to the effect of the organizational change, the closer communications and freedom typically associated with smaller companies are producing flexible ideas.

Mr. Tatamiya: When market needs are conveyed to front line development, designers and engineers often bring solutions immediately. However, in large organizations wherein many layers have accumulated, fast communication becomes difficult. In the case of the K10D planning, marketing, product planning, all the designers and engineers had discussions from the users’ viewpoint on an equal level, which is reflected in the detail of ideas and specs. As an example, with the RAW button there have been various debates among us. Some say most users would prefer JPEG, yet others say they only shoot in RAW setting. In reality, there may be many users who usually shoot in JPEG, selecting RAW as required. Just adding a button affects a lot of areas like mechanical design, firmware and user interface etc. However, as long as it is a conclusion based on detailed evaluation of true needs for user interface, the plan gets approved smoothly. Organization change actually derived from Pentax’s advantages as a smaller company. We could align all the necessary disciplines, i.e., mechanical design, electronics design and the development team, focusing what users really need.

Dust Reduction by “knocking” and coating

DC Watch Impress: With regard to DR, what innovations, besides coating, are involved?

Mr. Tatamiya: The expression “dropping the CCD” is often used to describe our DR system, but “knocking” is a more accurate description, as our DR utilizes the acceleration of SR actuator.

DC Watch Impress: We, of course, assume that durability testing has been conducted…

Mr. Tatamiya: Yes, anybody would be concerned about it. We absolutely exchanged notes with the CCD device maker regarding durability against shock up to so many G’s, etc. Then we added a sufficient safety margin to determine the mass acceleration. Therefore, there is no concern on CCD deterioration or break of wire bonding etc.

The idea of dropping dust by knocking the CCD element was patented a long time ago, but that was not enough. So we adopted a hard coating technology which our eyeglasses people had. This coating repels even oily substances by treating the surface at the micro millimeter level. Then we discovered that the combination of hard coating and physical knocking was very effective for dust removal, because if any dust contains moisture or oil, it is absorbed onto the surface and physical energy alone won’t remove it.

DC Watch Impress: Any improvement in 2nd generation SR?

Mr. Tatamiya: The CCD actuator was powered up and its control program was modified. At this point, the actuator was enlarged, the size of the unit and weight have been substantially increased due to a new CCD which is larger and heavier. Since the dimensions of the complete CCD unit has increased anyway, we thought we might as well take advantage of it and strengthen SR. This also resulted in the accuracy of SR.

The algorithm of our control program has been tuned to increase stability. In the past, there were certain situations wherein the SR movement pattern was not quite satisfactory, but we tuned it by trial and error. By increasing actuator power, activation is fast and braking is easier, which led to a stable SR effect.

Cost as a true necessity rather than specifications alone

DC Watch Impress: Relative to image rendering, the default setting is now “natural”, nevertheless, overall image/colour rendering now appears to be more straightforward. What has changed?

Mr. Tatamiya: The basic concept never changed since the *ist D. However, the performance of the image processing engine has been improved and a higher degree of edge compensation has become possible. In edge processing, it can’t appear too thick, rather it must be thin which might be making images look more refined. By using a new processing engine, different image processing from before, it’s now easier to tune.

DC Watch Impress: The combination of new a analog front end (AFE) by Newcore Technology and the latest image processor by Fujitsu seems to be very effective, but what are the merits of this?

Mr. Tatamiya: We are not disclosing the details of our image processor. The firmware was also developed by us. As you indicated, it is true that we now have a lot more elbowroom in image processing because the AFE now has a 22 bit AD converter. The improvement of image processing boils down to speed. Memory is now a high speed DDR2 with broader bus bandwidth which is sufficiently increasing image processing speed by a very good margin.

DC Watch Impress: Some are not too happy with the 1/4000 shutter speed or 3fps burst rate, but I personally believe that it was the right approach to suppress these parts and redistribute the cost over weatherproofing, etc. What is the reason why you were not particularly concerned about these figures (shutter speed, 3 fps etc)?

Mr. Tatamiya: Consider 10 years ago when people bragged about 1/8000 shutter and synchrospeed 1/250 which were considered the norm for a decent body. In addition, lenses had to be F/2.8 constant zoom. Today, these don’t seem particularly important.

Users became wiser and they are now able to judge for themselves what’s really important for getting the images they want. Then the digicam enabled sensitivity control at will, and more people really prefer compact lenses rather than those that are just one stop faster. They like easier operation over unnecessarily higher specs. Users' needs have now become very diverse.

Knowing this, if we still push for increased shutter speed and burst rate etc, it requires a more robust mechanical design with a higher speed shutter charge, mirror dumping and even increased SR reaction etc. Moreover, the CCD readout has to be 4 channels and AFE has to be doubled. Technically, all these are completely feasible but it was obvious that the cost would increase, as does the size/weight. So, we dropped these requirement at a very early stage of development.

DC Watch Impress: In other words, cost has been redistributed to something other than shutter speed or burst rate etc?

Mr. Tatamiya: This severely scrutinized cost distribution is the result of a well thought-out selection of what’s really required for a mid-level body. Not a $3000 to $4000 body. We consciously set a price point which is affordable for the target market, whereby it became possible to create high performance and usability under conditions in which the target users would use the K10D.

In the future, we will continue to analyze what users really want and develop models that will be truly useful in actual use, not just for a spec sheet that's easier on the eyes.

Original interview copyright 2006, DC Watch Impress. Translated by K. Takashita and M. Gaudet.

Link: DC Watch Impress Japanese Intervew with Tatamiya

No comments:

All Labels For This Site (Alphabetical)

© 2006-2011 by OK1000 Pentax Blog. All rights reserved. OK1000 Pentax Blog is not affiliated with Pentax.