Originally posted by Class A Comparing two different cameras is irrelevant to a discussion about the "accelerator unit" (which neither the K-1 nor the K-3 have).
Never have I claimed that one camera cannot be more accurate (have higher fidelity) than another.
It's a logical conclusion based on your statement above. My point is that some cameras are better at maintaining colour accuracy than others. Accelerator chip or not is moot. Sometimes I think you miss the point on purpose just to argue. It is quite possible that the cameras with the accelerator chip are more accurate. And it's quite possible the accelerator chip contributes to that. But as afar as I know that's never been demonstrated, so anecdotal evidence from owners is all we have.
Originally posted by Class A The fact that the input DR has to be compressed to an output DR is irrelevant.
Take any input or output DR you like, e.g. 7EV. You can artificially increase that to say 8EV by using image processing. That's what the "accelerator unit" does as well. It performs a post-processing step.
You do understand a K-1 or K-1ii will capture almost 15 EV DR given the opportunity? Why would you even consider a case where the camera only produced 7 EV? The fact that you can, doesn't mean anyone does or should.
The only element you need to concern yourself with, is capture DR. There is one fact you are leaving out. 7 EV increased to 8 EV in post, will have less colour depth, and fewer increments than 15 EV reduced to 9 EV. Taking a 7 EV image and increasing it to 9 for your monitor or TV, produces an inferior image, with all kinds of possibilities for artifacts like banding etc., the two processes are completely incompatible. One gives you the best possible image, one gives you a noticeably inferior image. When you add DR in post you cannot increase under-exposed shadows or blown out highlights. It adds nothing to the image visually. However in carefully reducing EV you can make visible all the original captured information, added detail in shadow and highlight.
I know you know all this, I'm not sure why you're even arguing this.
I guess you're pointing out that DR added in post is not the same as DR captured by the sensor. And that's a perfectly valid observation. Even if the accelerator chip increases the DR, it adds no new information beyond what the camera captured. The accelerator has no ability to help the pixel wells hold more photons nor does it change the lower limit at which the sensor can detect light. Nor can it add more values to the colour depth. All those things are functions of the sensor's capture capabilities.
Did I get it?