[dylansalt]

Yeah even in jpeg the Sony 16mp sensor captures sunsets just fine

[northcoastgreg]

Skies and sunsets were less good [with K-30] - less gradation in the sky and shadows.

Mostly probable explanation here is exposure. The K-30 sunset shots were perhaps a tad over-exposed compared to the 5DII. Or perhaps the K-30 shots needed more work in post to bring out the data in the raw files.

[normhead]

I am not sure you are right with your interpretation of bit depth is. It seems that you are talking about dynamic range when you say pulling out shadows and shades of grey. They are not the same thing. The bit depth is ONLY supposed to be related to colour and gradation. Either way, when the OP is saying that the colour gradation is not good on the k30, with 12 vs 14 bit, you couldn't tell on a computer screen as screens have a lower bit depth than cameras are capable of. The same for JPEG's.

You will only ever see it when you do major retouching and you have something with the exact same image at the exact same time from 2 different cameras.

You would be unlikely to "just happen to notice the difference" without having the exact same image from another camera with a different bit depth.

I don't have a k-30 so I can't say for sure. I'd love to compare though. We can see a difference between K-x and K-5, and K20D and K-5, so whether or not you can see a difference between K-30 and K5 remains to be determined. I'm not going to prejudge that one. Except to say I'm betting on the K-5. The K20D doesn't do as well as the K-5, so, is it colour depth, or is it the age of the camera? I'm probably not going to buy a k-30 just to find out.

You will only ever see it when you do major retouching and you have something with the exact same image at the exact same time from 2 different cameras.

If you consider something as simple as boosting a shadow to be major retouching then yes, but if a slight movement of a slider bar isn't major retouching, then no. If your exposure is bang on and there are no contrasty shadows, then yes. If you need to do a little shadow boost or a little highlight recovery, then 14 bit is your friend.

Lets say you have the 16 shades of gray black you get in a jpeg, in 8 bit and you boost a shadow. Say you slide 8 bits from your black range up into the gray range. That leaves you 8 shades of gray to fill 16 bits in the black/grey range and you get banding. If you start with 256 shades of black/gray (12 bit) and shift 240 of them up into the gray area of your gradient, you still have 16 shades of black grey to fill the 16 shades of grey in the 8 bit profile and you don't get banding, producing a much more natural look. In 14 bit you get 1024 shades of black/grey), you can slide 1008 bits to the grey area of the profile and still have a natural looking 8 bit output. The room you have for shadow recovery is significant, 750 more bits to play with in 14 bit depth, and it shows in your PP software and in your finished product.

If you do a lot of sunsets like I do, you rescue shadow and you are going to need this are going to use this, in every shot. Using a K-5 I can underexpose a stop to give me richer colours in the sky, knowing i can pull up 2 stops of shadow detail. But, until someone mails me a free K-30 d I won't be able to compare the two cameras.

If you go to this link and look at the trees on the far shore in the finished version, and compare that to the trees in the above jpeg sunset, I think you'll understand what I'm talking about. Details in the dark areas in one, a mass of blackness in the dark areas of the jpeg, despite the fact my K-5 raw was taken at -1 EV and seriously under-exposed the shadows.


https://www.pentaxforums.com/forums/photographic-technique/208706-sunsets-technique.html

[sbroadbentphoto]

Here is the not so big difference between 12 and 14 bit demonstrated. This is further reduced by jpeg compression. No real practical or useful difference for most people
Photomatter Reviews - Nikon D300 Review Part IV - 12 bit versus 14 bit color - James Markus Photography LLC

[Nuff]

I think the issue might be that the 5DII handles highlights better and K5 needs to be underexposed.

Not to mention it's easy to blow red channel...

[normhead]

Here is the not so big difference between 12 and 14 bit demonstrated. This is further reduced by jpeg compression. No real practical or useful difference for most people
Photomatter Reviews - Nikon D300 Review Part IV - 12 bit versus 14 bit color - James Markus Photography LLC

I have seen a scientific paper that suggests there is no theoretical difference between 12 and 14 bit data. It was a while ago I read it, and my optics etc. isn't at a high enough level to understand what he was saying, but I do have this note in the back of my mind "it has yet to be shown that 14 bit is better than 12 bit, more study needed". So I'm not going to argue too vehemently... the difference between jpeg (8 bit) and raw (even 12 bit) however is firmly established and not at all in question.

The problem with these kinds of comparisons is there are many different things that can be happening. In some cases 14 bit (K20D) is not as good as 14 bit (k-5). SO how do you tease out which comparison might be relevant to the subject? Unlike cameras with AA filters, Pentax doesn't make two identical cameras, one 12 bit, one 14 bit. You'd have to test two cameras with the same sensor and the same processing engine, etc. and if it's that close it's probably not worth worrying about. Just based on what I've read, there could be theoretical reasons why 14 bit is a pile of nonsense. At least one person who knows a lot more than I do has said so.

All I know is we still have a K-x, (12 bit) because it works for us. The K20D (14 bit) is on the auction block and looking for a new home.