Can I suggest Bruno Latour's 'Science in Action'?
Seems a good way to prevent me from 'skewing' the results as you have been claiming I have been doing all along.
Er, you admitted to processing the samples differently. That skews the results quite nicely.
Dr Cat: if you apply the same sharpening to two different cameras with two different optical characteristics (primarily the AA in this case) the only scientific and objective conclusion you can draw is how that particular process creates different files from the two different cameras. There is no scientific basis on which to draw a general conclusion.
If a man is short of iodine, and adding iodine vastly improves his health, it is not scientific to conclude that all people need that level of iodine.
If a certain type of process creates sharp files in one camera, there is no basis to conclude that that process is suitable for another camera.
There is, in fact, only one measuring stick to cuts through all the different factors of the hardware, the software, the processing, etc: that is the final file itself.
You accept that the final final is a valid measurement for comparison when testing the same method on two different cameras, so why can't you accept that any two final files, made by the two different processes necessary to get the peak performance out of two different cameras (like no two people have the same optimal diet) can be the final measuring stick?
Are the Olympics invalid because the athletes have different diets?
The best way forward would be to ask others to perform the same tests -- attempting to maximise detail -- and see if they can obtain as much detail as you believe you can get from the E-3 samples. If they can't, then your perception is wrong.
How is this logical? What if one uses an unsuitable process on one of the cameras just as you have been doing with the E3?
Your present claim is that RawTherapee is not fair because it favours the E-5, and Olympus Viewer 2 should be used.
No, these are words in my mouth again. I said Raw Therapee works really well for the E5 in terms of detail and therefore would be a good choice if you're processing E5 files. However, I can get a lot more out of other processors in regards to the E3 than you have been getting using Raw Therapee. Therefore I concluded that either you are processing the E3 files incorrectly or Raw Therapee is not good for E3 files.
Therefore I concluded that a proper test of detail would require something different than what you are doing. That is all.
This is really simple. People make the best E5 file they can. Others make the best E3 file they can. Then we compare. Everything else is simply reductive detail. Is not the final image the most important consideration?
OV2 happens to suppress detail a lot -- more than any other raw developer I tested -- and thus makes it harder to perform a comparison of detail. Your reasoning behind this appears to be: if you aren't obtaining the result you want, then you just have to change the variables (any and all variables) until you come up with a result you want.
No. You keep the variables that make for the best E5 file. I change the variables of the E3 file only. You're the one that keeps posting E3 files using your process of choice and then tell me that that is conclusive proof that the E5 is better. There is no consideration in your method of preparing E3 files at all. This is not very scientific, by the way. (well, bad science perhaps).
If you are an academic, I have issue with your method and logic.
Here is where things stand, as I see it. The myriad crops, processing, sharpening, re-cropping, and changing from one raw developer to another has not yielded results from the E-3 sample that come close to what you get by performing the same on the E-5 sample.
Nothing wrong with this. It is your opinion.
The same lenses used on an E-5 yield much more resolution (roughly 30% more) than when used on an E-3 -- this will translate into a lot more detail. People with extensive experience of both cameras believe that the E-5 yields much more detail. Olympus big selling-point for the E-5 is to deliver more detail.
Opinion and marketing. Not very scientific.
You have failed to prove your claim, and this should surprise nobody. The E-5 does yield a lot more detail -- considerably more than the MPixel increase should. (That 30% increase in lens MTF50 comes from an 11% increase in linear pixel count.)
MTF is affected by sharpening according to the people who do it.
But please don't start yet another thread claiming that we're all deluded about the E-5 performance until you have a result that looks at least plausibly close, and is attained using the same methodology.
The last phrase in this sentence artificially constrains the comparison. Answer me this, Einstein: if Raw Therapee is the best for the E5 but Viewer allows me to get more detail out of the E3, then should I force this comparison to go to Viewer? Should I also sharpen the E5 files (and get tons of artifacts) as much as I sharpen the E3 files?
If you really really think this is how it could be done, I'll process the E5 file (it will be the first one I've processed in my comparisons with you) exactly the same way I processed the last E3 file I posted and we'll see what it looks like. I can tell you it will look like SHlT because what I did to get detail out of an E3 file will practically destroy an E5 file. This would likely result in the E5 having considerably less detail than the E3.
Would that be a fair comparison then? Is that what you are really advocating?
