The K3 III in the comparison tool

flektogon wrote:

Well, 15 min walking to the Metro store was enough to cool my head off . OK. everything what I (and few others ) wrote, is just BS! It is irrelevant what parameters any given photo editor uses to open the raw file. Once you see the bitmap image, this is your starting point (base level) for any further modifications. You apply this and that modification, again it is irrelevant to know it (or save those numbers, like sharpening of 45). What counts is, how your modified bitmap image looks. If you are happy with what you see, just save it! Whether you save it as a TIFF or JPEG (except as a replacement of your original raw file), it doesn't matter. Whenever you open it again, and regardless on which editor/viewer you use, you will see again the same image.

So, I think that the entire discussion here was, well a good exercise with (our aging) brains .

I'm not sure what you're trying to say

My POV is that the raw converter you use (and your monitor etc) is part of the process that leads to a "developped" photo, exactly as before the paper, film, iso, chemicals, were part of a process to develop your photos. And as before, if you use a different paper you get something different. Which is the point in fact

To me it's very relevant what parameters a raw converter applies by default, exactly because as you said it's the reference you start from. If it's f**** because the raw converter applied crazy sharpening generating awful artifacts, then it's a problem and exactly what you want to avoid when you shoot RAW.

The amazing advantages of digital:

- you can have your photos developped automatically (and for free) by your camera, and you can even tweak rather precisely how they should be developped (miyabi akurayi whatever saturation etc)

- you can also develop your originals (raws) yourself with whatever RAW converter you like, how many times you like ! And your original RAW is never destroyed or damaged during the process ! Hallelujah

- even worse, now you can also have your camera develop the raws on the camera itself, after the capture

Then I often read people saying that a raw is not an image, this is true of course, but it's almost an image to be honest. The only difference is the demosaicing (usually you can also choose the algorithm). And the algorithm can sometimes have great impact when you pixel peep (so it might be important for a studio scene comparator).

My own process, I use darktable but I'd do the same in any:

- first I switch off any default processing done by the tool (sharpening etc)

- if I can't do that I garbage the tool

- then on DT I automatically select the demosaicing algorithm depending on the capture (for ex I use a different one for high ISO) thanks to profiles

- usually I compare between camera white balance and default WB proposed by DT. But in general I use the one from DT and start from there

- I do other PP stuff freely

- the last module in the chain adapts DR to output DR which is much more limited in general (monitor, print, etc). Everything in between demosaicing and this is linear and DT is great to avoid colors moving all around without logic. Color science is a fascinating domain. When you tell people about red green blue it seems nothing very complex can happen, but this is so very wrong.

- optionally if I use something like topaz tools I export temp images as uncompressed tiff, process them in topaz, then DT to export the final jpegs. I found it to be better than processing jpegs in topaz. But I don't process the raws in topaz because it can't compare with DT for that

- then I export the output image(s), and I keep them too

If another tool like viewers browsers or other raw converters etc have a different way to display the images: I don't care. I control what I can, calibrate my monitor, then it's not my problem anymore what people see.

That's also why sometimes I like to print because real people really coming at my real home can really see the real images as I wanted to (or let's say, as I could) have them rendered.

I also keep away from ACR or other popular tools because in my understanding you don't control everything. No in fact it's because I think they don't run on Linux. Topaz neither, and it's a real pain ...

If I remember well there was a comment from DPR about the fact that different versions of ACR applied different defaults, leading to different settings applied in the studio scene comparator sometimes.

For such a use case, or for "science" studies, it is very important to control the whole chain very precisely. And if you miss something after you painfully collected thousands of samples/photos, it can be a big problem you don't want to solve.

But I disagree that such a tool is useless by itself. It's a great concept, and even with its shortcomings the DPR tool is a very nice effort. There are things not only regarding sharpness, but also color processing etc, depending on brands and cameras. And sharpness should be checked in the center anyway, it's not a lens comparator !

I'm glad they eventually made the K3III samples more meaningful by reshooting them with electronic shutter (they also made some pixel shifted samples which is very cool to see that K3III high ISO noise is better than so many cameras even FF mirrorless...)