Quote Originally Posted by MikeG2012
Simply because when in photoshop, you seta genericfilter to sharpen the image as a whole and it applies it based on the differences in the pixels,

But what if DPP uses USM sharpening, whereas Lightroom uses deconvolution sharpening? Or what if DPP uses the demosaic algorithm for sharpening and Lightroom uses USM? Well, one solution might be to copy the exact same sharpening code from one program to the other (either break into Canon and steal it, or reverse engineer it). But even that is not enough, because the demosaic and color processing is going to affect sharpening as well.


So even if you have the *exact* same sharpening code and controls in both programs, and process the same raw in both programs at the same setting (e.g. "+3"), you're still going to get images thatlook like they were sharpened differently.And that's just *sharpening*, never mind the tons and tons of other features.


Quote Originally Posted by MikeG2012
, one would think that the camera has to write the data to explain how DPP needs to apply the filter.Therefore, another program would be able to read it as well. Otherwise, DPP wouldn't know how to apply the setting either.


<meta content="text/html; charset=utf-8" http-equiv="content-type" />



That's not how it works. If you tell your BFF "let's do our secret handshake", you don't have to spell out the dozen steps involved, from high fives to fist bumps, because you've already worked it out together in advance, in secret. But if you tell a stranger "let's do our secret handshake", they wont know what to do.


Same thing with Canon. The camera designers and the DPP programmers get together in secret and decide to use "ALO: medium" in the metadata. Then, in DPP, that will actually cause a thousand different complicated processing steps and secret blend of herbs and spices.


So there are several issues here:
  • Do third parties even want to try to read Canon metadata and emulate DPP image controls and rendering?
  • If they want to, are they able to see the metadata, or does Canon prevent them by encrypting the data?
  • If they are willing, and the data is available, are they *able* to figure out the secret sauce that Canon puts behind the controls?



From my understanding, the second issue is never a problem. The "encryption", if any, is always amateur-hour stuff that's easy to crack. The most difficult step is figuring out Canon's secret sauce. But even if they could do it somehow with no work involved whatsoever, I think many raw converters wouldn't even want to. First, it would require a whole separate set of internal processing -- essentially two raw converters in one program -- and that kind of complexity can ruin a programming project. For example, if a third party used a different color space than DPP, users wouldn't be able to get the same color in some cases, so the program would have to allow emulation of DPP's color space processing.Furthermore, differentiating themselves from DPP is how they sell products, so they may not feel that "can emulate DPP perfectly" is a valuable feature.


I think the ideal solution would be for all camera manufacturers to open source their raw converters.