[Aristophanes]

The cost of changing a piece of hardware at that low level is not just the price difference in the piece of hardware itself (the cost of producing a FF sensor is at least 2.5x and probably closer to 3x or more than an APS-C sensor).

A change at that low level would require completely different PCBs, firmware, and if you are talking about a lot more pixels you are talking about more memory and faster hardware and all of these incur engineering expenses that have to be spread out across the expected total sales of a product.

The cost is artificial because Canon and Nikon use smaller runs to justify $6,000 bodies.

The real material cost is between $70 and $150 per sensor. Falk did some preliminary analysis here and from my end, the delimited market by Canikon is the prime factor.

Note how in my assessment I pegged an FF K-7 at $2,000. I added another $200 for the additional processing and memory cache. The real costs of these are now in the pennies per unit.

This still assumes a healthy margin. (With the Yen at record highs, I might add. That's an enormous factor here).

The more you make, the cheaper it gets. The longer you make it, the cheaper it gets. The more sunk investment Canikon have in sensors this large, the less they will be able to charge later down the road. It's simple amortization.

Nikon has brought sensor development in-house precisely to match Canon when the FF market goes broader and deeper and becomes commodified, like APS-C has. They did not want to be at the mercy of Sony.

Nausea? Take Gravol.

[Mister Guy]

There was no "demand" for credit default swaps either. But they were created, marketed to spur demand, took off, and crashed the international economic system.

There's always demand for fraud. The more clever the fraud, the longer it takes to realize it was fraud. Sometimes, the fraud is even technically legal, and laws needs to be changed.

[Class A]

I'm not a Nikonian, don't think I'll ever feel like one. There are things about the gear I don't like and I miss about Pentax designs.

What is it one should be wary of if one thinks that the grass is greener of the Nikon side of the fence? What is it that you don't like about Nikon gear and you like about Pentax gear?

If they offer a good Ff body, I'll sell what I just bought and get a new system again and I suspect many others will also.

Why would FF do the trick?
FF has no intrinsic low-light advantage. As a wedding shooter you'll probably appreciate the higher dynamic range, but I wonder if it is really FF you want or very high-ISO performance. The K-x is supposed to be a good high-ISO performer.

I think it is reasonable to expect that the two new Pentax models to be revealed in September will have very decent high-ISO performance.

[Class A]

.
DOF:....

You really should reference the source for the information you posted: Depth of Field, Digital Photography and Crop Sensor Cameras - Bob Atkins Photography

[ManuH]

Nikon has brought sensor development in-house precisely to match Canon when the FF market goes broader and deeper and becomes commodified, like APS-C has. They did not want to be at the mercy of Sony.

And it's a good decision. Before Sony entered the DSLR market there was no conflict of interest, Sony was just providing sensors. For how long will Sony share their best sensors? Right now it makes sense because they are below 10% market share and thus selling their sensors to Nikon and Pentax helps amortizing the costs. But if they get close to Canikon market shares, not sharing or delaying the availability to competitors is a possibility. Pentax tried to go with Samsung but now they are again relying completely on Sony, I'd guess that their next cameras will have a Sony sensor after the relative underperformance of the K-7.

[Peter Zack]

What is it one should be wary of if one thinks that the grass is greener of the Nikon side of the fence? What is it that you don't like about Nikon gear and you like about Pentax gear?


Why would FF do the trick?
FF has no intrinsic low-light advantage. As a wedding shooter you'll probably appreciate the higher dynamic range, but I wonder if it is really FF you want or very high-ISO performance. The K-x is supposed to be a good high-ISO performer.

I think it is reasonable to expect that the two new Pentax models to be revealed in September will have very decent high-ISO performance.

Pros with Pentax: Ergonomics. They have it right IMO. Canon with the rear wheel is aweful in my hands and I don’t like a lot of the button placement. Nikon is similar to Pentax with the E dials but the lens removal button is in the wrong place (opposite Pentax’s) and it’s just not as easy to change lenses quickly. It’s because the grip is on the Pentax release side. You can hold the grip and with one finger push the button. Then with the other hand, remove the lens. With it on the other side of a Nikon, there’s nothing to hold on to while pushing the button.

ISO button on a Nikon is on the top left. With it on the OK button or the top right on a Pentax, you can change ISO’s much faster without removing you’re eye from the VF.

Pentax menus are better but Nikon did a few things right. The help menu is excellent. The “My Menu” is great (make your own custom menu). Best of all, they have the menu set so if you leave the menu, it goes back to the last item you looked at when you access the menu again. So no starting at page one and scrolling to get back to where you were before.

SR and WS. Unless you buy the overpriced VR lenses (I only have one, the 24-120mm), you lose SR. WS, There are no Nikon WS lenses. They only claim dust proof. The bodies are WS (some) but Pentax has WS in several lenses.

AF speed and accuracy. Nikon is very fast. The light has to be near black to see it hunt. It’s fast and locks the subject when you need it. Plus it can track an object moving in any direction much better. I rarely used AFc on my K20D’s as I found it nearly useless. Not now.

Buffer speed. Huge difference. There's just no waiting at all. You can take 3-4 shots in a row and look at the histogram nearly instantly. I'm not sure if this is a difference with SD vs CF and or the processor. But it's much faster.

The advantage in low light with the D700 is huge. It’s partly because they stuck with a 12MP sensor. ISO6400 is the same as ISO800 on a K20D. That’s pretty large (3 stops). I a dark church, that’s the difference between a flash shot and natural light. I have a K-x and although very good, it’s not the same. The K-x seems to have a lot less detail at similar ISO’s and just not as sharp looking. Plus Chroma noise on the D700 is much better controlled.

FF has one other thing APSc just can’t do. Yes there’s a difference in DOF but more than that, a 28mm lens is a 28mm lens. Sure you lose the FOV crop at longer distances, but a lens like the 28-75mm f2.8 Tamron sings on a FF body. I got a Sigma 12-24mm FF lens and that’s wild on a FF. Corrected ultra wide with about 125 degrees FOV. I like the Sigma 70-200mm more on it as well.

Again, I don’t want to come across as trashing Pentax. Far, far from it. Give me this camera in a K-XXX and I’d sell what I just bought. Pentax has a lot right and for the average shooter, there’s nothing missing. Below ISO400, the way the sensor processes an image is superior in every way. For me and what I do, it just wasn’t enough. I just think APSc in all brands is maxed on high ISO. The D300s can’t touch the D700. The 7D is worse than the K20D.

[Winder]

There is a lot of new tech that is going to be revealed next month. If the new Pentax has the Sony 16MP Exmor HD sensor it will take a big step forward. No it will not match a D700, but 3200 could very well look like the current 800. There seems to be some high expectations for this new sensor an architecture.

What is more important to me is a faster more accurate AF system and better DR. The K-7 is very close to being a serious competitor for the 7D and D300s. New sensor and better AF and the next K-7 will be a class leading camera.

[Christopher M.W.T]

Look what Pentax really needs more than FF is better accessories.

Today I was playing with a Canon 580EXII flash, which here in Australia is at a similar price point to the Pentax AF540GZ and wow what a difference.

The AF540GZ feels like flimsy plastic whereas the 580EXII feels like a professional unit well worth the money, has environmental seals, feels solid, tight and well worth its money.

The 540GZ looks pathetic by comparison.

[Class A]

Thanks, Peter for the comprehensive explanaition.

It’s because the grip is on the Pentax release side. You can hold the grip and with one finger push the button.

I do that as well, but did you notice that the lens index marker is on the wrong side (at the grip side)? When I hold the camera in my right hand to change the lens, I cannot see the index marker. I always thought I must be doing something wrong. With experience, I learned to put the lens to the mount with almost the correct orientation but with the hidden index marker, it remains a flight in the dark.

Perhaps there is a good way to handle the Nikon system as well (holding the camera in a way you haven't thought of yet?).

The advantage in low light with the D700 is huge. It’s partly because they stuck with a 12MP sensor.

We need not to get into that in this thread but the often made assumption that fewer pixels are better for low light isn't correct (-> Contrary to conventional wisdom, higher resolution actually compensates for noise).

ISO6400 is the same as ISO800 on a K20D.

Have you tried LR3's noise reduction? It might do wonders for the K20D whereas the Nikon image has received a lot of treatment by the camera already. I might be wrong, but I know for sure that LR3 has quite impressive (chroma & luminance) noise reduction that retains a lot of detail. Unfortunately, even LR3.2RC is pretty buggy and poorly programmed in parts but it starts to get usable.

[gawan]

We need not to get into that in this thread but the often made assumption that fewer pixels are better for low light isn't correct (-> http://www.dxomark.com/index.php/eng/Our-publications/DxOMark-Insights/More-pixels-offset-noise!).

I am sure Peter meant larger pixels (pixel pitch) actually, not fewer pixel count per se...

[...] LR3.2RC is pretty buggy and poorly programmed in parts but it starts to get usable.

The final release of LR3 is quite stable, IMO, and yes, the the sharpening/NR is really good.

[falconeye]

I am sure Peter meant larger pixels (pixel pitch) actually, not fewer pixel count per se..

Please, read ClassA's link.
Whatever Peter meant, it is wrong. Noise is ONLY a function of sensor size (when considering f-stop and exposure time to be constant). It is not a function of pixel surface.

If the new Pentax has the Sony 16MP Exmor HD sensor it will take a big step forward. No it will not match a D700, ...

I expect Exmor and Exmor HD to have similar performance. Exmor HD was optimized for more versatile video readouts and may have advantages for contrast AF too.

APSC Exmor (as showcased by the K-x) matches the D700 in dynamic range, but not in low light performance. Dynamic range is a function of full well capacity and read-out noise. Sony beats Nikon in this department, probably due to the column A/D architecture. Low light performance is a function of quantum efficiency and overall sensor surface. All current Bayer micro lens CMOS sensors currently seem to be very similar in this department. So, no substitute for surface here.

Something which sort of stupefies me though is that a full frame sensor's potential biggest advantage is almost always dismissed: a doubling in usable pixels! We all admire a 645D's resolving power. But full frame is 2/3 down the road from APSC to a 645D. Strange, really strange ...
BTW, that's part of the reason why the D700 is severely crippled for a full frame camera.

[ManuH]

Look what Pentax really needs more than FF is better accessories.

Today I was playing with a Canon 580EXII flash, which here in Australia is at a similar price point to the Pentax AF540GZ and wow what a difference.

How true. The flash system needs to be revamped, if Hoya cannot dedicate the resources for this, why not make it compatible with Canon (or preferably) Nikon flashes?

Something which sort of stupefies me though is that a full frame sensor's potential biggest advantage is almost always dismissed: a doubling in usable pixels! We all admire a 645D's resolving power. But full frame is 2/3 down the road from APSC to a 645D. Strange, really strange ...
BTW, that's part of the reason why the D700 is severely crippled for a full frame camera.

I noticed the same thing and that's why a D700 doesn't appeal me that much. A D700 shooter needs a D300 when wanting to shoot above 300mm. Look at Thom Hogan latest picture on his website, shot with a D300 + 200-400mm, not with his D3s. The Sony 24MP is more interesting as you can crop when needed to a respectable 11MP. Unfortunately we all know their relatively bad noise performance (compared to others FF, it's still better than any APS-C).

[Pål Jensen]

And it's a good decision. Before Sony entered the DSLR market there was no conflict of interest, Sony was just providing sensors. For how long will Sony share their best sensors? Right now it makes sense because they are below 10% market share and thus selling their sensors to Nikon and Pentax helps amortizing the costs. But if they get close to Canikon market shares, not sharing or delaying the availability to competitors is a possibility. Pentax tried to go with Samsung but now they are again relying completely on Sony, I'd guess that their next cameras will have a Sony sensor after the relative underperformance of the K-7.

The sensors are sold by Sony semiconductor; they do not make cameras. I would expect the semiconductor department to be both very sucessful and profitable. Sony probably loose money on the DSLR division.
I believe it is more likely that Sony stop selling DSLR's before they stop selling sensors!

[Pål Jensen]

Something which sort of stupefies me though is that a full frame sensor's potential biggest advantage is almost always dismissed: a doubling in usable pixels! We all admire a 645D's resolving power. But full frame is 2/3 down the road from APSC to a 645D. Strange, really strange ...
BTW, that's part of the reason why the D700 is severely crippled for a full frame camera.

I find that very strange too. FF starts to make sense for me at 30+ mp....
But at that time we may have an FF 645 camera with 60+ mp.....

[Peter Zack]

Whatever the reason Pixel size, pixel density etc, there is a huge difference. I read so much conflicting info on the subject, I have no idea what's true. But considering the 7D isn't much good at high ISO and that's a new camera, I would have assumed that Canon would have made it a High ISO king with its release. It's got just about everything else. It's also interesting that the K-7 (15MP) is beaten easily by the K-x (12MP) at high ISO's.

The Sony 24MP FF models aren't much good at high ISO's either. In fact Ned Brunnell said once that approx 12 MP is the sweet spot for APSc. Interesting that's where the K-x is and it's a high ISO APSc champ against any brand.

I really do think there's a connection between pixel density and digital noise. Since I have no engineering background, I can only go by what I see. Look again at Nikon. D3s is the ISO champ (12MP). D3x is the resolution champ (24.5 MP). Similar cameras built for different shooters. The X is the sports, wildlife shooters choice and the S is the wedding, portrait and event shooters choice.
The wedding camera from Canon is similar but uses an H chip at 16 MP, so the density is higher but much lower than a 7D. It's their high ISO champ.

So Faulk I really think there's a lot more to pixel density that your link would suggest. These company's can't be making this choice for lower density chips in high ISO performers without cause.

As to the flash issue. Absolutely. scrap the current models. They are nothing even close to the competition and even cost just as much. it's not just the flash but the flash and the camera exposures. No where near what is available.