Could "K20D" actually be APS-H (1.27x) sized?

There are a lot of rumors flying around. I don't know if I believe any of them.

But here is some food for thought. If anyone can confirm these details (I'm

really searching Intel's site), this would be great info.

 

- Sony (designer?) and Intel (fab and support logic -- including

prototype/reference board?) are behind the newer 12MP CMOS sensor in the APS-C /

1.5x crop size (roughly 24mmx16mm), which is now available in the Nikon D300 and

Sony A700.

 

- Sony and Intel also have a 14MP CMOS sensor in development which will be

released soon. Several sites are actually reporting this is more APS-H / 1.25x

crop sized (roughly 29x18mm).

 

- The new Pentax DA* lenses are designed for up to a 1.25x crop, explaining some

of the added size (beyond just f-stop).

 

The current 6/10MP APS-C CCDs in the K100/10D are 1.53x crop. The current,

non-Pentax CMOS sensors in some other products (e.g., Canon) that are APS-H size

like (yes, I know, APS-H is actually 16:9, but not this sensor which is 3:2) are

around 1.27x crop I believe.

 

If there is a new, commodity CMOS sensor in development at 14MP and it was APS-H

/ 1.27x crop, then Sunsung-Pentax would have known about it via their

engineering "Gold Book" type vendor channels years ago. The design of the DA*

series and their 1.25x crop support could be well explained by this then. It's

not uncommon for designs to run late, and how the introduction dates line-up on

the DA* to the rumors of Pentax's new products make sense.

 

The 1.27x crop "K20D," with its required 1.25x crop DA* lenses, would command

that post-sale premium after-sale return, something that the new Hoya management

is probably looking for. Pentax still offers the "K200D" for more

cost-conscience buyers, but the "K20D" now gives Hoya more margins, especially

for moving that new DA* product stock. It's also far more of a mainstream,

commodity product, volume-wise, than the Medium Format (60mmx45mm) 645D would

ever be, whlie giving a solid upgrade path to possibly as high as 24MP with a

solid ISO in an APS-H sized CMOS sensor.

 

After all, the 645D started at 18MP, and was also rumored to be possibly 24MP at

one point. That was already a far cry from Hasselbad who is at 38MP. The volume

for such medium format is already constrained, and Hasselbad would be the call

over a solution that is only 45-60% the resolution, regardless of price (after

all, why go Medium Format unless for such fidelity?). But now, with a 14MP

sensor, offering both full-frame compatibility and a new line of premium DA*

lenses, Pentax gets far more volume and those margins to boot. And, again, by

moving to APS-H size, instead of sticking with APS-C, while using a CMOS sensor,

24MP should be eventually achievable without too much ISO hit (APS-C is probably

more at 18MP).

 

Now Pentax isn't going to abandon the 1.5x lens crowd altogether, that's where a

"sister product" like the "K200D" would fit. You repackage the K10D 10MP CCD

sensor, just like the *istD's 6MP CCD sensor was repackaged into the K100D

before, with updated SR, the higher-res 900K LCDs like on the D300/A700 (or

possibly that's only on the "K20D," and a cheaper, 460K version that is already

2x of today's 230K goes on the "K200D").

 

I'm sure the fab cost of the 10MP CCD sensor on the K10D is already little more

than the 6MP CCD of the K100D now, given the same APS-C die size. In fact,

engineering lifecycles are all about reuse and refinement, costs becoming lower,

hence the *istD into the K100D with more features, now the K10D into the "K200D"

with more features. Some might argue the K100D Super is "too new" to do that

too. But the K100D Super is little more than a firmware update (use SR dust

reduction, and new firmware so the slot can use the SDM in-lens), so it's just

"finishing off" existing 6MP CCD/control stock through mid next-year or whenever

they run out (even though sites will still have stock later than that, Pentax

won't supply them anymore much sooner). So any "K200D" is ready to take its place.

 

This strategy makes total sense to me, especially with the 645D being all but

officially cancelled and the new DA* lenses supporting up to a 1.25x crop

factor. But this is still 100% speculation, I have to admit, but it makes both

engineering and marketing sense to me.

Bryan, it does not make practical sense to me. The closer to full frame, the more valuable the older FF lenses become, most of which Pentax has discontinued. And, the more recent DA lenses would also suffer performancewise, in terms of compatibility. They are designed specifically for the smaller aps size sensor with 1.53x crop size. That, to me would put a serious monkey wrench into the idea of Pentax switching formats in mid-stream of their newfound momentum.

 

But Pentax has made some perplexing decision errors in the recent past- so one never knows.

Wow, you've really put a lot of thought into this, Bryan... :-)

 

I'm just going to sit back and enjoy the show in January.

 

And drool.

Hmm,

 

I don't think the DA* zooms are over priced. Actually, even at initial price I thought they were very fair.

 

I realize that people are confused, the want pro glass, but don't want pro glass price.

 

The DA* zooms are fantastic, well built, optically excellent, weather sealed, silent and fast focusing.

 

And they are stabilized (on a stabilized body which all current and future Pentax DSLR will be).

 

Compare them to the equivalent Canon/Nikon/Olympus offerings of pro glass and you'll not think there is a premium attached.

 

My philosphy has always been those that want/need pro end glass are willing to pay for it, those that don't say it's over priced.

 

 

As far as upgrading, there is this belief that Pentax CANNOT release new cameras without alienating current users.

 

For instance if I just bought a K100D Super and they release a K200D my camera is not less good then it was a few months ago.

 

The 1.5X crowd will never be abadoned with a 1.25-1.3X sensor. All DA lenses will work on a 1.25X camera.

 

As far as the 645D it was 18MP, then rumored to have a upgrade path (i'm assuming a back or a factory re-sensor) to 30MP. I believe Pentax was either using it at 18MP or giving an option.

 

Comparing it to Hasselbad is foolish. That system cost into the $50,000 range or more. The Pentax, like Pentax film medium format, would be value based (but high quality). I'm guessing it would have been in the $3000-$6000 range and competed with the Canon 1Ds rather than the Hasselbad.

 

Pentax needs to support older cameras with firmware updates but it should be releasing cameras not to please existing owners but to get new buyers.

 

I saw the same sentiment when the D1 was replaced by the D1H and D1X. People clamored that Nikon should buy back the D1's or give a large rebate.

 

I think of cameras like I do cars, a new model comes out more frequently than I need one. I get a new car not based on model year but when it reaches the point that I feel upkeep will exceed it's value. With a camera I buy a new one when specs and quality exceed what I have, or I feel I've justified the cost of the previous body. If I was making a living off this I might do otherwise, but my experience is most working pros don't immediately buy the next model either.

 

I've heard the sensor is a Samsung CMOS. Since it's specs are different from any Sony model, I'm guessing it probably isn't Sony. That would be a big chunk of change for pentax or hoya to buy into a propietary sensor for just Pentax. Samsung might be cheaper since it has a working colaboration with Pentax, and isn't known in the DSLR CMOS or CCD world.

 

BTW, your post reinforces the fact that engineers think about things way too much. Just relax, have fun with the camera you have, and eventually all this will be figured out. No amount of speculation on our part will have much impact. Yep, someone from Pentax might be reading these boards and taking notes, but in the end, Pentax is going to produce whatever it feels it can sell at a fair price and stay competitive.

BTW I thought the extra 2 lens contacts just inside the lensmount of the K10D was for operation of the SDM lenses. So the Super upgrade of the K100 would have to include installation of the contacts as well, I presume.

BTW, even if Michael is correct and those lenses won't work, adding a crop mode like Nikon has done on it's D2 and D3 series seems doable.

 

Actually, I'd love a 2x crop mode for sports. Smaller files, less post process cropping, etc.

 

However, my impression has been that MANY DA lenses work on film bodies, so it would seem reasonable that most, if not all DA lenses work on 1.25X bodies, and I know the DA* specs list they are designed for up to 1.25 (or 1.3x, don't remember which) sensor size.

Also, the Sony/Nikon LCD screen (produced oddly by Panasonic) is not 900K pixels...actually it's marketing genius unless I'm wrong (and someone will correct me if I am).

 

Actually it's a 900K DOT screen but the legit pixel count is a mere 300K, or only slightly nicer than the ones on the K10D. Actually factoring in that it's bigger, i'm not sure it's all that much nicer. Basically it looks like a small upgrade (although an upgrade) in screen resolution, not a massive upgrade.

 

What they did was took the RGB pixels and multiplied by 3 giving a 900K screen. I was duped initially myself thinking if a 250K screen looks nice, how amazing is a 900K screen. Glad something seemed iffy to me and I looked into it.

 

This is quite similar to Fuji and Sigmas inflated pixel count of the S5 and SD14, neither of which are actually the pixel counts they claim. Not in real world terms.

Justin, you mention an adjustable crop factor. I have long thought that to be a great idea. Like say a 16mp FF with an adjustable 10mp 1.5 crop. I could see a FF design with that feature, but I do not see the point to a design of 1.25 crop factor. Compared to FF and what we now have, it seems like some of the advantages of either would come up short, and no clear advantage to hang one's hat on.

Justin, I think the K10D's specs may also have featured that Nikon math that everyone was so shocked by. While the new Nikon LCD is 640x480x3, I suspect the K10D actually has something in the magnitude of 320x240x3. The difference in LCD resolution is roughly like the difference between older Pocket PC's and some of the latest ones with VGA (640x480) res screens. I wouldn't call it tremendously important, but I think there will be a visible difference. Roughly double the pixel density.

The Canon 1D is 1.3X and actually people I know who have both the 1Ds and the 1D actually (all of two people mind you) like the 1.3X as there primary camera better.

 

I was suprised by this because one of them is a FF devotee, or so I thought. He just said when I was talking to him that he found himself gravitating towards the 1DMII more than the 1DsMII in all but wide angle situations.

 

Anyway, there is a great advantage to me.

 

First SR in camera CANNOT be done with a true FF on existing glass in existing 35mm size bodies.

 

However, 1.25X would allow this.

 

Also, 1.25X is a nice compromise for FF people. Gives you closer to 35mm DOF and FOV but you keep the SR and can still use DX/DA lenses. And the lenses will still be slightly smaller and lighter.

 

While I never carried my 80-200 around for street shooting, I find my 50-135 comes with me often and it's desiged for 1.25X sensors.

 

Finally, short term, larger sensors are a path to lower noise. Long term, I don't see a huge advantage. Think about how film progressed, and it eventually got to a point that with a good grained film, and medium format technique, people were making MF quality prints on 35mm. Giving the extra few millimeters of sensor size should create a little head room either to add megapixels (not my choice) or reduce noise with larger photosites while keeping MP count high.

 

For my use, I'd love a 1.25X camera with a 2X crop mask. I'd love to knock those prospective 14MP in half (or to at least 10MP) or so to get me smaller files and give my lenses a little more reach in the camera, rather than PP crop all of them. And if a crop mode allowed a faster frame rate I'd be all for that.

 

Big files are great if you intend to print, but small RAW files are great if you want the best white balance and noise but don't intend to print.

 

Working with the 2.7MP photos out of the D1H is a dream, likewise when I move on to a D2H those sub 5MP RAW images are still going to be tiny compared to the K10/20D

There is every reason to increase sensor size at least to so called 'full-frame'.....it is not just a short term chase.

 

A bigger piece of film will ALWAYS beat a smaller piece and sensors will be bound by the same inescapable physical truths.

 

There have always been lots of people in photography who tell themselves fairy stories in order to soothe the hurt that they are unable/unwilling to pay for the real thing. "MF quality" from 35mm depends entirely on what you think "MF quality" actually is.

 

There will be plenty of benefits to FF sensors,....the Nikon D3 has iso up to 25,000...!!!! That's just a single feature.

I disagree. medium format was considered professional at one time and 35mm for amatuers, yet people like Galen Rowell took 35mm film into the realm that was held only by MF and LF photographers and produced images that rivaled those formats. They also produced images that couldn't be produced by those formats because they were too big and cumbersome.

 

Technology trumps size. Imagine in 1990 I told you we'd have micro SD cards that can hold GB's of info? You'd have laughed as you loaded your 3.5" floppy into the PC.

 

I realize it's hard for people to view into the future but if you look at just the past 15 years of advancements it's not hard to imagine a world of APS-C or 4/3 cameras getting amazing results at ISO 12,000. Actually, I'd be sold on ISO 3200, and coming from a strong film background even a killer ISO1600 would make me a happy camper.

 

Look at how horrible camera phones were a few years ago, not look at them. Some actually produce usable/printable images. Heck even my LG VX9900 produces a very decent 2MP image and it's barely in the upper tier of the camera phone line.

 

I truly believe people who think that bigger sensor will be better in a few years are short sighted. Sure a bigger sensor like a bigger piece of film will always yield a better image, but it's not solely about that, it's about all the things that come with that bigger sensor. Just like MF vs. 35mm it's not soley about IQ, but about being able to get the image in the first place.

 

I wouldn't be suprised if in 5-10 years there are HD camera phones that rival todays video cameras and DSLRs.

 

And if that is possible, only fools will be lugging around 35mm sized cameras with attached 35mm image cirle lenses that HAVE to have lense based IS.

 

Basically the future is in smaller sensors and in camera stabilization, both will trump 35mm systems just as medium format became a niche market when many professionals realized the size and quickness of a 35mm system was simply better. I expect some die hards to be shooting FF/35mm in a decade but I expect much of the market will have moved on to smaller 1.25X - 2X sensors for the smaller lenses, and in body IS/SR that doesn't compromise the optics for price.

 

Of course thats just my prediction. A decade ago had I told you phones would have cameras, and could hold an entire 1990 computers worth of programs and info on a card the size of a finger nail clipping you'd have laughed too. We'll see, but betting against technology is downright foolish!!!!!!

"and could hold an entire 1990 computers worth of programs and info on a card the size of a finger nail clipping you'd have laughed too."

 

It's actually more like a 100 1990 computers - 4Gt SD vs 40Mt HD. :)

 

 

"Prediction: The cost for 128 kilobytes of memory

will fall below U$100 in the near future.

Creative Computing magazine

December 1981, page 6"

Justin Serpico: "Actually it's a 900K DOT screen but the legit pixel count is a mere 300K, or only slightly nicer than the ones on the K10D. Actually factoring in that it's bigger, i'm not sure it's all that much nicer. Basically it looks like a small upgrade (although an upgrade) in screen resolution, not a massive upgrade."

 

Have you actually seen the Nikon screen? Going from 320x240 to VGA was one of the biggest visible changes in monitor technology.

I've never understood the way the manufacturers speak of 230K or whatever screens. Marketing I guess. Canon has a 320x240 screen and Nikon 640x480, that's a serious difference.

Nikon multiplies by 3 and so does Canon.

Actually Nikon's math for dots is 1920x480 = 900K, but resolution in the way people usually understand it is 640x480 because one pixel has three components, RGB.

- Charge Devices

 

I had professionally studied Charge Devices and their ability to hold hundreds of thousands of bytes, if not millions, in the near future. Understand the CCD approach began life as a memory chip, because the capacity of Dynamic RAM (DRAM), let alone non-wait state Static RAM (SRAM), technology at the time was still far, far less.

 

Now while we never imagined CCDs being used as a sensor memory, the second I first heard of it turn-of-the-century, it made utter sense. You want to temporarily store something, then transfer it into memory. CCDs were the perfect solution.

 

- CMOS Technology

 

Everything is logic these days. It only gets better. Sure, the density is a factor, but it can be overcome with better logic. Sensitivity will be absolute in another decade, well beyond anything we could have ever imagined. It will take a few iterations, but we'll get there.

 

Until the, with today's commodity designs in an APS-C/APS-H, 18MP/24MP should be possible with up to solid ISO 3200 sensitivity. You'll see some other, costly designs far beyond that soon, but it will be awhile before they are commodity. In fact, lens quality will probably become more of a factor before the sensor.

 

- LCD Technology

 

Also, LCD technology is going higher and higher, past 200dpi. It will not be long before 1080p will be avaialble in a 3" LCD on the back of a camera. Just give it 5-7 years and we'll be at 600dpi. I mean, they are already talking 30MP+ TV displays designs today, that should be feasible in less than 5 years.

Justin,

I've been a fan of Galen Rowell for years and remember his fantastic slide show years ago in

Severance Hall in Cleveland. However, in looking at his book on my shelf right now, I'd

disagree that he was able to make 35mm equal medium format. He did very well with 35 and

took it to the limit, but to me they simply don't compare to the other books on my shelf that

feature medium format.

Scott

up until now all we've had was pixel counts for screens, and to my eye more or less pixels has been a big deal.

 

For instance, I chose my wifes FX07 over many other cameras because of it's high res 2.5 in 230,000 pix screen. Sure she was jealous of the 3 in screens for a while but some of them actually had less pixels (put them side by side and she wasn't jealous anymore).

 

However, thanks for pointing out those are 640x480 screen resolutions, I didnt know that. I'm sure it is a big upgrade then.

 

Nevertheless, the 900K count is marketing. I'd have appreciated the basic specs just saying 640x480.

Scott,

 

I have every Rowell book (except for the new tribute one) and I'm sure there are differences, but have you seen his 20x30 gallery prints? I also have several good (the official printing company) of Ansel Adams. As well as several of my other photographic idol Tim Ernst (he shoots a mix of Pentax MF and 35mm) books. Plus, Tims photos (a mix of formats) graced the walls of the University of Arkansas School of Business.

 

Now, I'm sure you and I have a ton of mural size prints lining the house or gallery, but for me 20x30 (poster size) is plenty. Generally, I don't personally print larger than 12x18 and usually 8x12 or 10x13 or 11x14. Occasionally I print a pano as a poster.

 

My point being, even if cameras allow us 200x300 in prints, how many of us will ever print such are large picture?

 

So for me, 12x18 at ISO1600 in a more compact package is the holy grail. I barely complain now at my results but give me that, and I promise not a peep about camera improvements.

 

If they can downsize my lenses by even 15%, plus give me high quality optics and SR, I'll be thrilled. More portability than 35mm at better than 35mm results? Who wouldn't take that!!!

Do they make printers that big? 200 x 300 seems HUGE.

16' x 25' ... isn't that larger than billboard size?

It seems to me that once resolution of a certain very high level is reached at the FF 36mm level, and APS size as well, the visible difference compared to medium format will become less and less in practical terms. I have also read that the same mp level in an APS size sensor actually provides more resolution than a FF design because the say, 12mp is fit within a smaller frame. Thus taking a shot with the Nikon D300 would have higher mp of resolution than taking the same shot at the same distance with the D3 and cropping it to match.

 

If the K20D will use a 14mp sensor at 1.27x crop, it should be superior to the D3 as well, and about equal to the D300 in resolution. A slight alternate crop could, as justin suggests be employed for 1.53x, and still providing around 11mp. There is also more to resolution than just the amount of mp, as I read of a test showing the 10mp Nikon D200 registered slightly higher resolution than the 12mp D2x.

 

I am beginning to see the light in the possibilities as per the information and speculation offered in this thread!! For example, my very nice little Pentax FA 28-105mm f/3.2-4.5 used at a 1.27x crop would configure to about 35-130mm, and my pro Tokina 28-70mm lens would come in at around 35-90mm! So wide angle would now be restored with these nice lenses. If I need a little more tele and less wide, I could dial in the standard crop factor!

 

The DA* 16-50mm f/2.8, if designed to accomodate the 1.27x crop as Bryan thinks is the case, would come in as about 20-65mm. Pretty amazing when you think about it! It would have all the photo magazines raving. the new 60-250 would translate to about 75-320mm f/4.

BTW if all this pans out, and with the future indicated as envisioned, then I have to say this detailed information and speculation is extremely perceptive of you, Bryan!! Got me wondering if Pentax could use another insightful, innovative engineer.

 

All other input and discussion, particularly from Justin, have been very revealing and thought-stimulating as well.

 

Now if a new sensor were to emerge not having moire', etc. to filter, or requiring sharpening and other post-shot processing...

Rose, it was a point. Large format and medium format cameras allow for bigger prints. Generally, most people with poor technique and crappy lenses can only squeeze 8x10's or 10x13s out of their camera.

 

However, my point was most people never print bigger than that anyway.

 

20inches by 30inches is poster size, and unless you live in a mansion, most people (not selling there work) probably print a few of those in a lifetime, if ever.

 

So the idea that we need the ability to print giant prints is not necessarily true. My walls are covered with a few 20x30's but mostly 12x18 or smaller prints.

 

I'm not saying others don't have the need, I'm just saying for me, I'd be thrilled with great 12x18 at 1600 ISO and smaller lenses to boot.

 

Mike,

 

resolution isn't totally dictated by sensor size. the D200 actually had about the same lines as a FF 5D. So what you read might be true.

And, beyond all of these, you never look at a 20*30 as to a postcard, from length of arm !

You can have great posters even from 3-4M, but these megs must be real !