[Class A]

I took the RAW images and converted them to JPEG in Photoshop, then resized them.

Photoshop's raw converter, ACR, is not ideal to make this kind of comparisons.

Even if the ACR sliders show the same value, this does not mean that the same conversion parameters are applied. ACR does a translation under the hood and uses different default values depending on the camera. One source of the difference you are seeing -- clearly the K200D images look brighter -- could be differences in the tone curves that are embedded in the camera profiles used by ACR. If the shape of these tone curves is different then the images will look different. This probably explains why you see more detail in the K200Ds shadows.

Another potential source for differences could be the effective ISO values used by the cameras. A set value, such as "ISO 400" rarely translates into effective "ISO 400", but typically something lower, say ISO 350. Have a look at DxOMark to see how much the K200D and Kx differ in their effective ISO values and that could contribute a little to explaining the differences you are seeing.

The K-x clearly has higher dynamic range then the K200D at all ISO levels, so I'm rather sure you are not seeing sensor differences but a combination of the above. You may want to try a neutral RAW converter such as dcraw. I guess RawTherapee is also neutral.

[Darley]

Photoshop's raw converter, ACR, is not ideal to make this kind of comparisons.

Even if the ACR sliders show the same value, this does not mean that the same conversion parameters are applied. ACR does a translation under the hood and uses different default values depending on the camera. One source of the difference you are seeing -- clearly the K200D images look brighter -- could be differences in the tone curves that are embedded in the camera profiles used by ACR. If the shape of these tone curves is different then the images will look different. This probably explains why you see more detail in the K200Ds shadows.

Another potential source for differences could be the effective ISO values used by the cameras. A set value, such as "ISO 400" rarely translates into effective "ISO 400", but typically something lower, say ISO 350. Have a look at DxOMark to see how much the K200D and Kx differ in their effective ISO values and that could contribute a little to explaining the differences you are seeing.

The K-x clearly has higher dynamic range then the K200D at all ISO levels, so I'm rather sure you are not seeing sensor differences but a combination of the above. You may want to try a neutral RAW converter such as dcraw. I guess RawTherapee is also neutral.

Well I did say it was unscientific! I just converted them in RawTherapee (neutral) and the differences are even more marked. I don't understand the science but to me the K-x images are clearly not accurate reproductions of the scene, whereas the K200D gets it right, bar the white balance inaccuracy.

[Stone G.]

I think the older camera shows more detail especially in darker areas. The range of light levels captured is greater, which I assume is what dynamic range refers to. However the K200D stops at ISO1600 and in very low light the extra sensitivity of the K-x is not just better, it blows the older camera away

I have - kind of - the same experience. Obviously, the spatial resolution of my K-5 is better than that of my K200D; and a camera's High Dynamic Range capability is normally undertood as the ability to capture large differences between maximum and minimum intensities of light within one and the same exposure, and I don't really find much of a difference here.

But when it comes to representing subtle differences in shades of light, I think my K200D is the winner. Here's just one example (400 pct. crop) of a very high contrast and high dynamic scene: The Lunar terminator.

PEF files have been converted in Pentax Digital Camera Utility 4. In order to secure(?) identical conversions, I have copied the parameters from one, selected file to all images and then converted all of them in a batch process.

[Darley]

I have - kind of - the same experience. Obviously, the spatial resolution of my K-5 is better than that of my K200D; and a camera's High Dynamic Range capability is normally undertood as the ability to capture large differences between maximum and minimum intensities of light within one and the same exposure, and I don't really find much of a difference here.

But when it comes to representing subtle differences in shades of light, I think my K200D is the winner. Here's just one example (400 pct. crop) of a very high contrast and high dynamic scene: The Lunar terminator.

PEF files have been converted in Pentax Digital Camera Utility 4. In order to secure(?) identical conversions, I have copied the parameters from one, selected file to all images and then converted all of them in a batch process.

It does seem to fit in with my findings - and my theory that CCDs are more sensitive to subtle differences in shades of light/dark. The *ist of course is hard to compare because of the difference in image size. I've had a number of CCD cameras including two Sonys (who of course make a lot of sensors) and have always preferred the images from them. I guess it's a matter of taste and what you want to use the camera for, but I'm still of the opinion that CMOS sensors are inferior and used mainly because of cost and the demand for live view and video.

[Class A]

It does seem to fit in with my findings - and my theory that CCDs are more sensitive to subtle differences in shades of light/dark.

To the best of my knowledge both CCD and CMOS sensors have a linear response to light.

It is just not the case that one type has better shadow definition than the other.

But when it comes to representing subtle differences in shades of light, I think my K200D is the winner.

I am convinced that what you are seeing are different tone curves. Even though you used PDCU for both cameras that does not mean that they received the same push in the shadow regions. Maybe Pentax changed the default black levels. IIRC, earlier models did not have truncated RAW values and later models changed that. This could make a big difference.

Also, your K200D images seems have higher exposure; the highlights look brighter. Again, this may have happened in the RAW development.

If you make sure that all shooting parameters are the same and the same effective ISO (not the manufacturer ISO) has been used and made sure that the RAW conversion was exactly the same -- ignoring embedded tone curves, for instance -- then I'm confident that you won't see a difference in shadow detail anymore.

If you don't want to make a flawless comparison, you could try to push the shadows in the K-5 image in order to reveal the same detail. Provided you have used the same aperture and shutter speed both cameras, the K-5 shadows should contain as much detail as the K200D shadows (and more). I'd be very surprised if you weren't able to push the K-5 shadows much more than the K200D shadows before they fall apart.

[wombat2go]

I was doing some indoor portraits of the boys today with the ist ds and thought of this thread.
A good chance to directly compare the K-01 and the ist ds to see why I prefer the ist ds.

The lens was the same - SMC Pentax -M 1:3.5 28mm ( I used 2 other lenses too)
The flash was the AF360fgz in slave mode down low to the left side, pointing to the ceiling in A mode.
I just changed the bodies onto the same tripod.
Both iso 200 and both f/11, both cameras in manual of course.
So these photos have a range from the blown tablecloth next to the flash, over to the darkness on the right side.


istds jpg & raw
https://app.box.com/s/xt9wzgvkkbkhjxcupcty
https://app.box.com/s/gcerfsttli1wqhn6kdmj

K-01 jpg & raw
https://app.box.com/s/l9u0f78ti2bzdwzon8ic
https://app.box.com/s/kf6gr0rtndypkheji93j

To me, it is definite that the ist ds colors on the Eizo are more accurate and realistic to what I see in daylight coming through the window particularly in the lower light area, both in the camera jpgs and the raws as they come into ufraw with the camera color temperatures.
For example the curtains are really a gold color like the ist ds shot, not gray tinged like the K-01.
The awb color temps from the cameras were almost identical (4850K +/- 15K). The K-01's raw EV needed a 1.5 EV boost to be same brightness as istds raw .
I suppose I could pull the curves of the K-01 into color, but why bother, I think I will just use the ist ds while it still works fine.

Finally, although not related to this thread, I can focus the ist ds sharply, but see that the K-01 is not focussed correctly, it is front focussed.
I have a 2010 model Olympus M43 and I also get the feeling the ist ds is better than it, although I have not done a direct comparison.

[Darley]

To me, it is definite that the ist ds colors on the Eizo are more accurate and realistic to what I see in daylight coming through the window particularly in the lower light area, both in the camera jpgs and the raws as they come into ufraw with the camera color temperatures.
For example the curtains are really a gold color like the ist ds shot, not gray tinged like the K-01.
The awb color temps from the cameras were almost identical (4850K +/- 15K). The K-01's raw EV needed a 1.5 EV boost to be same brightness as istds raw .
I suppose I could pull the curves of the K-01 into color, but why bother, I think I will just use the ist ds while it still works fine.
.

I'm with you on both those points. I did some more tests today and am convinced my K200D is picking out a lot more detail in shaded areas - I'm also convinced it's a CCD thing. Like you I want images that I can use more-or-less straight from the camera for the most part, not having to mess around with tone curves to get the detail that another camera gives me with SOOC JPEGs. I think high-MP CMOS sensors and the latest processors are capable of producing great images but there are compromises - designed for lower power consumption and video, for example. I guess it's like comparing different types of 35mm film: some will prefer one type and some another. I have decided I prefer CCD

[Darley]

To the best of my knowledge both CCD and CMOS sensors have a linear response to light.

It is just not the case that one type has better shadow definition than the other.


I am convinced that what you are seeing are different tone curves. Even though you used PDCU for both cameras that does not mean that they received the same push in the shadow regions. Maybe Pentax changed the default black levels. IIRC, earlier models did not have truncated RAW values and later models changed that. This could make a big difference.

Also, your K200D images seems have higher exposure; the highlights look brighter. Again, this may have happened in the RAW development.

If you make sure that all shooting parameters are the same and the same effective ISO (not the manufacturer ISO) has been used and made sure that the RAW conversion was exactly the same -- ignoring embedded tone curves, for instance -- then I'm confident that you won't see a difference in shadow detail anymore.

If you don't want to make a flawless comparison, you could try to push the shadows in the K-5 image in order to reveal the same detail. Provided you have used the same aperture and shutter speed both cameras, the K-5 shadows should contain as much detail as the K200D shadows (and more). I'd be very surprised if you weren't able to push the K-5 shadows much more than the K200D shadows before they fall apart.

I don't have the K5 - that was another poster. I can only go by the ISO figures in the camera, and as both are Pentax I assumed they would be similar. The JPEGs certainly came out very similar, it is the RAW files that differ greatly even though they were converted using identical settings. Maybe the K5 does hold as much or more detail but why waste time fiddling with software edits when you can be taking photographs?! I know some people like to do that, but I expect the camera to be able to produce images the way I want them i.e. by accurately reproducing the scene, without needing a lot of post-processing. Maybe that's the difference between CCD and CMOS, but for me I'm happy to stick with the older technology