[barondla]

Interesting discussion and answers. Haven't read the article linked above... yet. Found another article explaining reasons ccd has different color. There are reasons besides the color filters.

Read a good way down to find the color part.
CCD vs. CMOS Image Sensors in Machine Vision Cameras - Adimec
Hmm, perhaps I should pit my K10D against my K-1 in a color test.

Thanks,
barondla

[Ian Stuart Forsyth]

Interesting discussion and answers. Haven't read the article linked above... yet. Found another article explaining reasons ccd has different color. There are reasons besides the color filters.

A great deal of the time its the profiles that are found within the raw converts that will give you the greatest variance in color between the 2 different sensors. Often times the default color profiles that the converts use are more or less designed around what is pleasing and really has nothing to do with color or even color accuracy. Within these profiles tonal range, how the colors appear, contrast are all based on what is pleasing. This can vary based on what was pleasing at the time for the developer of that profile or raw converter.

One has to look no further than what variations you can find in how raw converters give you in colors over a given time with the updates to the converter, this varies as much greater that what you see between CCD and CMOS.

If you take away the profiles that are based on what is pleasing for that given time and use custom profiles that are based on profiles made for color reproduction then 99% if the time people cannot guess as to what is a CCD image

---------- Post added 12-27-2019 at 02:10 AM ----------



One if from a CCD and the other from CMOS

There is some differences in the final colors but this could be due to the profiles created for the 2 cameras, the profile for the 645d used the SMC A 75mm ƒ2.8 and for the 645 Z used the DFA 25mm ƒ4.
and not to mention that the 2 IR test images are taken with 2 different lenses also in between the 645z and the 645d

[i_trax]

that's why still have K10D and will never sell it.

[Rondec]

I think it is generally a matter of taste. If CCD "beats" CMOS in terms of color, it only happens at low iso settings. I have some good photos from my K10 up to about iso 400, but then everything went south in a major way. But I've been pretty pleased with the colors I get from the K-1.

[ChristianRock]

There were multiple shots taken. before and after. K10D still had the same color in the sky

That is important information.
Thanks for the reply, it shows that the CCD does pick up colors differently than the CMOS sensor.

[BigDave]

In this case just missing entirely.

If missing entirely, were the colors missing in a certain area of color space (e.g. Reds, Greens, light Reds, Dark Reds)?

[ChristianRock]

I think there's another aspect to CCD, or at least the Sony CCD sensor that I'm used to (K10D) versus the CMOS sensors I've used from Samsung (K20D) and Sony (K-r, K-50, K-S1).

There seems to be a more "compressed" dynamic range already baked into the files - I think that since the sensor capture itself in CCD is analog and only the readout is digital (and someone can correct me if this is wrong), that it happens during the analog capture.

I first noticed it right away after getting my K10D which substituted my K20D. Part of my photography hobby has always been to go out during my lunch break - when light is at its worse! - on "photo walks". I did this for a couple years with the K20D and I have to say that it was always a pain to process the pictures that had sun + shade in bright days because I had to expose for the sun-covered area (Sunny 16) and the shaded areas would usually be almost black. And since the K20D's dynamic range is pretty poor, getting a couple stops out of the shaded areas always came with a lot of that ugly Samsung-sensor noise.

After I went out with the K10D in similar conditions, I found out that getting 2 stops out of the shadows was not nearly as painful at ISO 100 - the dynamic range in the shadows was already better than the K20D. But most importantly, I didn't really have to dig into the shadows that much - there was already more shadow detail in the picture itself. In other words, the shadows weren't as dark compared to the bright sun-covered spots. I know this is a very non-technical explanation but it's the best that a non-technical person like me can attempt to explain...

Here's the picture that really brought me the "a-ha" moment...



With the K20D, the shadow areas would be much darker straight out of the camera. With the Sony CMOS sensors that would also be true but recovering detail would be much easier due to the excellent dynamic range at ISO 100. But just seeing this straight out of camera made for much more enjoyable walks with the K10D (of course I did lift the shadows a little bit for a more natural look, but it wasn't bad to start with).

Every once in a while I keep looking at a Sony A290 for the 14MP CCD sensor - the latest and best CCD sensor ever created in APS-C format. That and the Zeiss 16-80mm f3.5-4.5 lens that was a thousand dollars new and now sometimes can be found really cheap, would probably make an amazing CCD-based setup... but I keep thinking, I'm already good with the K10D.

---

Last edited by ChristianRock; 12-27-2019 at 04:35 PM.

[Piotrek K]

I could imagine that CCD with limited dynamic range could create better out of the box pictures compared to much later Sony sensors which have amazing dynamic range, but they need some post work or clever in-camera processing for as punchy result as was normal K-10D/K-200D.

[Leumas]

I think there's another aspect to CCD, or at least the Sony CCD sensor that I'm used to (K10D) versus the CMOS sensors I've used from Samsung (K20D) and Sony (K-r, K-50, K-S1).

There seems to be a more "compressed" dynamic range already baked into the files - I think that since the sensor capture itself in CCD is analog and only the readout is digital (and someone can correct me if this is wrong), that it happens during the analog capture.

I first noticed it right away after getting my K10D which substituted my K20D. Part of my photography hobby has always been to go out during my lunch break - when light is at its worse! - on "photo walks". I did this for a couple years with the K20D and I have to say that it was always a pain to process the pictures that had sun + shade in bright days because I had to expose for the sun-covered area (Sunny 16) and the shaded areas would usually be almost black. And since the K20D's dynamic range is pretty poor, getting a couple stops out of the shaded areas always came with a lot of that ugly Samyang-sensor noise.

After I went out with the K10D in similar conditions, I found out that getting 2 stops out of the shadows was not nearly as painful at ISO 100 - the dynamic range in the shadows was already better than the K20D. But most importantly, I didn't really have to dig into the shadows that much - there was already more shadow detail in the picture itself. In other words, the shadows weren't as dark compared to the bright sun-covered spots. I know this is a very non-technical explanation but it's the best that a non-technical person like me can attempt to explain...

Here's the picture that really brought me the "a-ha" moment...



With the K20D, the shadow areas would be much darker straight out of the camera. With the Sony CMOS sensors that would also be true but recovering detail would be much easier due to the excellent dynamic range at ISO 100. But just seeing this straight out of camera made for much more enjoyable walks with the K10D (of course I did lift the shadows a little bit for a more natural look, but it wasn't bad to start with).

Every once in a while I keep looking at a Sony A290 for the 14MP CCD sensor - the latest and best CCD sensor ever created in APS-C format. That and the Zeiss 16-80mm f3.5-4.5 lens that was a thousand dollars new and now sometimes can be found really cheap, would probably make an amazing CCD-based setup... but I keep thinking, I'm already good with the K10D.

There was something I read about CMOS and CCD having non-linear, vs linear falloff of sensitivity in shadows and highlights. Don't remember the details, but there was a difference between the two