[Fenwoodian]

.
I have a 10MP Pentax K10D digital camera with a CCD sensor. I also own some newer Pentax cameras with CMOS sensors.

To my eye, I see a subtle difference in the way CCD cameras render color as compared to the way CMOS cameras render color.

I know, that the sensor itself is not the difference maker, rather the CFA that sits on top of the sensor itself. So, let's not get into a technical discussion on that.

For those of you who shoot both CCD and CMOS cameras, do you too detect a difference in color? If so, how would you describe this difference, and which do you prefer CCD or CMOS colors?

[photoptimist]

My personal perception of color accuracy and color balance totally sucks. I don't even notice any color differences between CCD and CMOS images unless they are side by side. Even with side by side shots I don't know which one is "better." Maybe my brain's automatic white balance is turned up to 11.

However, my desire to get decent images in suboptimal lighting (dusk, forest floor, indoors, night sky, etc.) means I love the high-ISO performance of modern CMOS. I do love my K-10D which was my first DSLR and first Pentax. But I love my K-5 and K-1 even more because they can get good images in even lower light.

[Dartmoor Dave]

I wonder if it would be possible to add a poll to this thread, asking something like:

Do you see no difference at all between CCD and CMOS colours?
Do you see a small difference between CCD and CMOS colours?
Do you see a big difference between CCD and CMOS colours?

Personally I see a huge difference, to the extent that I've never been able to force myself to move on to a CMOS as my main camera. I'm always fascinated and slightly amazed when people say that they don't see much difference, so it would be interesting to see some poll results confirming how many people actually do see it.

My guess is that the majority would say that they can see some difference, but not enough to outweigh things like higher resolution and better high ISO performance. While a small but stubborn minority of us will say that the CCD/CMOS difference is a major factor in our camera choice.


Edit: I'm going to say something now that I've wanted to say for a long time, and I don't care if I get flamed to a crisp for it. The colours in most of the K-1 shots I've ever seen are horrible.

[acoufap]

Lets have quick look at how sensors of the usual bayer sensors are built - that's independent of CCD or CMOS technology.

The light sensitive sensor cells of a sensor for usual digital cameras are color blind. Colors are obtained by using color filters ordered in a special pattern (Bayer Pattern) above the sensor cells. The raw converter algorithm (software) interprets the captured signals of each cell and the color filter that covers it. Pixels of a visible image are computed by looking at neigboured cells and their captured brightness values while considering "their" filter color (red, green and blue), the cameras white balance measure and some other parameters set in camera (remember pentax custom color profiles?).

To my imagination this works the same for CCD and CMOS sensors. Let's suppose I'm right.

Then I'd say: if there really are differences in colors between CCD and CMOS chips then it's mostly related to the software raw convertion algorithms and the rules that are used to determine the colors - indepentently from in-camera raw conversion or raw conversion on a computer in post.

But I may be wrong and everything depends more on the kind of color filters above the sensor cells and they may be different for CCD and CMOS sensors?!

---

Last edited by acoufap; 04-01-2020 at 09:55 AM.

[biz-engineer]

To my eye, I see a subtle difference in the way CCD cameras render color as compared to the way CMOS cameras render color.

CMOS defaults are cooler than CCD defaults. The difference can be corrected via raw files. I can see how the same sensors render differently depending on camera brand using them. The difficulty is how deep our own "color competence" is.

I've been able to produce some kind of Canon color rendering or Nikon color by playing with combinations of sliders and Pentax JPEG style, but it was more like trial & error cooking rather than fully knowing what I was doing.

My knowledge of color science is near ground zero so I have a huge potential to learn it. The good news is learning more about color composition is free, except spending the time to sort out educational material and taking the time and focus to learn it.

Edit: I'm going to say something now that I've wanted to say for a long time, and I don't care if I get flamed to a crisp for it. The colors in most of the K-1 shots I've ever seen are horrible.

You are not the only one not to like Pentax default colors. Personally, I find that Pentax default JPEG styles are too bold, just lacking refinement. In camera high contrast filters and bold monochrome just give too much contrast even at minimum settings and the default monochrome style lacks contrast. For me those in camera settings are unusable, I have to use tone curves in raw dev to produce decent B&W photos.

[stevebrot]

There is a subtle difference...a certain additional richness with CCD, but I shoot with what I have and am fairly enamored of having live view available for manual fine focus and vintage lens metering. FWIW, didn't @BigMackCam create a recipe for replicating CCD look with CMOS a few years back?


Steve

[kypfer]

FWIW, my CCD cameras have greater infra-red sensitivity than my CMOS cameras, at least a stop, as measured through the same lens and Wratten 89B (R72) filter on each body, the Wratten 87 is also useable on the CCD's, but not on the CMOS, therefore my inference is that the CCD cameras may produce more saturated deep reds in the visible spectrum. As to how this is interpreted in "post" is entirely up to you. For me, I feel a lot of the pictures from my CCD cameras need no further "help", whereas I often need to "tweak" the output from the CMOS cameras ... I'll admit that the end result is near the same, but the CCD cameras need less work to get there!
Obviously there are many other factors to consider, such as subject, weather, time of day and lens fitted

---

Last edited by kypfer; 04-01-2020 at 03:13 PM.

[photoptimist]

Regarding sensors: although CCDs and CMOS both use silicon, they don't have identical spectral response. The difference is because silicon absorbs different wavelengths at different depths and the two technologies have different semiconductor structures at different depths. In general, CCDs are less blue sensitive than are CMOS sensors.

Regarding people: the genetics of vision mean that different people do see colors differently. This is true even for those who are not color blind. CCDs may be a better match for some people's retinal response. CMOS may be a better match for other people's retinal response. And maybe some people see both sensors as a little bit off. There's also a lot of neurological processing in color vision that is required to see saturated colors. The human retina has much poorer color separation than does the typical Bayer CFA and it's the brain that makes sense of it. Perhaps some people grew up in households that favored more or less acute color discrimination which then affects how they see the world and photos of the world.

[Dartmoor Dave]

I think it varies by manufacturer too. For example, I find it quite hard to tell Leica M9 shots (CCD) from M10 shots (CMOS). But I think I can tell a shot from any Pentax CCD camera from any Pentax CMOS camera every time. Perhaps the K20D could fool me, but I don't think any of the later cameras could.

[mikesbike]

Regarding sensors: although CCDs and CMOS both use silicon, they don't have identical spectral response. The difference is because silicon absorbs different wavelengths at different depths and the two technologies have different semiconductor structures at different depths. In general, CCDs are less blue sensitive than are CMOS sensors.

Regarding people: the genetics of vision mean that different people do see colors differently. This is true even for those who are not color blind. CCDs may be a better match for some people's retinal response. CMOS may be a better match for other people's retinal response. And maybe some people see both sensors as a little bit off. There's also a lot of neurological processing in color vision that is required to see saturated colors. The human retina has much poorer color separation than does the typical Bayer CFA and it's the brain that makes sense of it. Perhaps some people grew up in households that favored more or less acute color discrimination which then affects how they see the world and photos of the world.

This makes a lot of sense to me. Today's cameras are very complex, and human perception is yet more complex. The result being, there is no "right" answer. With film, the various formulations produce predictable results in terms of color palette and saturation. But even then, although there have been general agreement regarding those differences, there are still perceptual variations among people as to the applications and which is the most desirable.

---------- Post added 04-01-20 at 12:41 PM ----------

I think it varies by manufacturer too. For example, I find it quite hard to tell Leica M9 shots (CCD) from M10 shots (CMOS). But I think I can tell a shot from any Pentax CCD camera from any Pentax CMOS camera every time. Perhaps the K20D could fool me, but I don't think any of the later cameras could.

I was going to say- first, I have noticed some degree of color difference for each DSLR model I have had, from the *ist on, and that during the years I was shooting with both my K200D (CCD) and K20D (CMOS), the K20D's output demonstrated every bit the color richness of the K200D, at least!

---------- Post added 04-01-20 at 12:44 PM ----------

CCDs may be a better match for some people's retinal response.

I believe this is true, since not everyone is in agreement over this issue- far from it. I noticed some of my Pentax DSLR models have had a sometimes noticeable, slight magenta cast, going back to my CCD models, whereas the more recent ones, especially my KP seem more neutral. (to my retinal response) But that is not to say I find the KP's colors to be "cold", indeed I find colors to be vivid and vibrant.

Then there is the interaction of any post processing software with the camera's output.

[luftfluss]

IIRC Hasselblad claimed they were only willing to move from CCD to CMOS once they were satisfied with the color response they could wrangle out of the CMOS sensor.

Or, that is simply marketing spin.

[adamaitken2020]

.
I have a 10MP Pentax K10D digital camera with a CCD sensor. I also own some newer Pentax cameras with CMOS sensors.

To my eye, I see a subtle difference in the way CCD cameras render color as compared to the way CMOS cameras render color.

I know, that the sensor itself is not the difference maker, rather the CFA that sits on top of the sensor itself. So, let's not get into a technical discussion on that.

For those of you who shoot both CCD and CMOS cameras, do you too detect a difference in color? If so, how would you describe this difference, and which do you prefer CCD or CMOS colors?

I had a K10 D and the colour it produced could be great. Now the modern CMOS sensors are much better and there's no not much difference between a K10D CCD and a up to date Pentax CMOS. The big difference is that that K10D has narrow Dynamic range compared to anything after K5. so you would only get the best colour depth on a low ISO.

---------- Post added 04-01-20 at 10:36 PM ----------

CMOS defaults are cooler than CCD defaults. The difference can be corrected via raw files. I can see how the same sensors render differently depending on camera brand using them. The difficulty is how deep our own "color competence" is.

I've been able to produce some kind of Canon color rendering or Nikon color by playing with combinations of sliders and Pentax JPEG style, but it was more like trial & error cooking rather than fully knowing what I was doing.

My knowledge of color science is near ground zero so I have a huge potential to learn it. The good news is learning more about color composition is free, except spending the time to sort out educational material and taking the time and focus to learn it.



You are not the only one not to like Pentax default colors. Personally, I find that Pentax default JPEG styles are too bold, just lacking refinement. In camera high contrast filters and bold monochrome just give too much contrast even at minimum settings and the default monochrome style lacks contrast. For me those in camera settings are unusable, I have to use tone curves in raw dev to produce decent B&W photos.

I agree, looking back at my old Pentax jpgs that I shot on default. I decided to do RAW processing and got much better results.

[Digitalis]

To my eye, I see a subtle difference in the way CCD cameras render color as compared to the way CMOS cameras render color.

It is a very subtle thing to notice, but gradations between complimentary hues seem to be more saturated and the shading between analogous hues appear smoother. My hue perception is extremely high and this is probably a result from an inexplicable defect in my vision, my left eye sees colors differently from my [dominant] right eye.

The differences are mostly seen in the green hues which unsurprisingly, is where 50% of colour information is derived on Bayer mosaic sensors. The remaining colour channels take up 25% each, the rest of the data is interpolated.


Pentax K10D - Sigma 180mm f/3.5 APO EX

IIRC Hasselblad claimed they were only willing to move from CCD to CMOS once they were satisfied with the color response they could wrangle out of the CMOS sensor.

This was back in the day where CMOS foundries weren't quite up to the grade in terms of quality and thus colour fidelity was lacking on early CMOS sensors, specialized and thus more costly and carefully manufactured CCD sensors were better, for a time.

[acoufap]

Regarding sensors: although CCDs and CMOS both use silicon, they don't have identical spectral response. The difference is because silicon absorbs different wavelengths at different depths and the two technologies have different semiconductor structures at different depths. In general, CCDs are less blue sensitive than are CMOS sensors.

Regarding people: the genetics of vision mean that different people do see colors differently. This is true even for those who are not color blind. CCDs may be a better match for some people's retinal response. CMOS may be a better match for other people's retinal response. And maybe some people see both sensors as a little bit off. There's also a lot of neurological processing in color vision that is required to see saturated colors. The human retina has much poorer color separation than does the typical Bayer CFA and it's the brain that makes sense of it. Perhaps some people grew up in households that favored more or less acute color discrimination which then affects how they see the world and photos of the world.

some interesting issues, thanks mentioning!

[Dartmoor Dave]

The differences are mostly seen in the green hues which unsurprisingly, is where 50% of colour information is derived on Bayer mosaic sensors.

Now that really is interesting. Green is the dominant colour in my part of the world, and the ability to render the greens of the Dartmoor landscape accurately is one of the major reasons why I've stuck with CCD. No matter how hard I've tried with custom profiling and adjusting colour curves in processing, I've never been able to get the greens quite right with any CMOS I've tried.