You mix "resolution" and "image quality" in your response. The two are different. My statement was about resolution. Image quality can improve even when resolution goes down. But your statement about resolution being multiplicative is false (for resolution and for every other measurement chain).
www.dxomark.com
The term "image quality" is an inherently arbitrary and subjective term, so I'm not going to waste time debating why everyone has to adhere to your definition of it. I don't define "image quality" in the way the author does either.
We know that is how their data works. I just don't believe they are actually testing every single lens + body combination that they list.
While I get your point about the number of tests, there were examples in this thread about numbers being all over the place with different [but same mount] cameras
All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
Every improvement helps - was the meaning of the original claim.All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
There is no such thing as "resolution MFT". MFT is the product of resolution and contrast. Here is a good primer. One quote (boldface is mine):
Yes, but improving any of the terms in the chain increases the end result. So even an average lens will benefit from a better sensor and vice versa.All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
You are correct. My point was the "image quality" was used to describe MTF in the paper.
Yes, but improving any of the terms in the chain increases the end result. So even an average lens will benefit from a better sensor and vice versa.All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
No, in layman's terms, it is the contrast at a given resolution.
That might have been me. Panasonic lenses being tested on Olympus bodies yielded different results than Panasonic lenses being tested on Panasonic bodies. Taking that further, different Panasonic bodies yielded different results from each other. Too many variables and the only constant was the lens?
That depends. The 'MTF's there are functions, not numbers. The multiplication is function multiplication. Whether the multiplication 'reduces' or 'helps' the outcome depends on the nature of the function. However, if we take the position that a perfect system would be diffraction limited', that is have the diffraction MTF only, then each additional function there puts us in a worse position than that perfect lens. The point being mate is that MTFsensor() fro a high pixel count sensor puts us in a 'less worse' position than MTFsensor() for a low pixel count sensor.All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
It's a major contributor, the other being pixel aperture. See Jack Hogan's article . But, if we assume 100% microlenses, the a higher MP produces a 'better' MTF.
I don’t think what Olympus have written here is helpful in the context of this thread. It’s meant as helpful marketing stuff. I note the quote’ doesn’t make any attempt to quantify the term resolution or say how it could be measured.
Give me a theoretical example of increasing a sensor's MP not increasing system MTF. Practically speaking it doesn't jive- for example an AA filter hurts MTF, but camera makers have been weakening them or getting rid of them altogether with increases in MPs. Anecdotally I have seen increases in detail and sharpness with mediocre glass on increasingly high MP sensors.If "better" means a sensor with higher MFT, I agree with you. But just increasing the sensor resolution does not guarantee that. That was the point I was trying to make.Yes, but improving any of the terms in the chain increases the end result. So even an average lens will benefit from a better sensor and vice versa.All those numbers are less than 100%. Every additional multiplication term reduces the value, it does not "help" it.
JACS, I acknowledge that you are a layman. Mathematically, "at a" means "multiplied by" (actually, convolved with, but never mind).
Thanks for the lecture. I am a math professor.
