Pure white is what it is (depends on context), but pure black can be considered as missing signal or signal with zero amplitude - and dividing by it gives infinity.
Be scale finite or infinite, zero is starting point of it.
Sure you can say that black level corresponds to some value of signal - but I think in current thread context it is simpler to agree that black is just zero; at least that was basis of my claim.
In layman's terms exposure is the amount of light that strikes the sensor per unit area during a shutter actuation.
Some people incorrectly define exposure as being how light or dark an image looks.
Exposure and image lightness are two different things because for a given scene the image lightness can be altered in camera without changing the exposure at all.
For a new film or processing change, I always run tests to determine the maximum exposure to avoid blowing out highlights, and the minimum exposure to still show important detail in shadows. I've done the same with a digital camera. Depth of field and blur requirements are entirely different subjects, unrelated to exposure latitude (AKA dynamic range).
That's fine but what you call "exposure latitude" and what I call "exposure latitude" are two different things.
DOF and blur requirements are factors in "exposure latitude" as I described earlier.
You asked how I define it - I did so.
Isn't that entirely about video? For 4K video (8 MP), pixel count is one of the least important factors with regards to DR. In fact, pixel count isn't much of a factor with regards to DR, period. For example:
And I replied "That's fine" but pointed out why our definitions of "exposure latitude" are different.
Well, firstly, it confirms what I already knew, that TN doesn't know what the word 'exposure' means. Second, that he puts meaningless tests on his channel. That is, he was testing the 'DR', actually the coding range of log encoded (processed) video files, which isn't sensor DR - and didn't tell us enough about his methodology to know whether it was even a valid test of log encoded coding range. And finally, that you're happy to have 'facts' which alter according to your current talking point. If you remember, we previously agreed that the A7SIII sensor has 48MP, so it has the smallest pixels of the three.
Which no one has questioned. Certainly not I.
The only thing I questioned was your definition of "exposure" when you first mentioned "exposure latitude" because "exposure" means different things to different people.
You explained what "exposure" means to you and then I explained why "exposure latitude" means different things to you and me.
good point, but the 48 meg pixels are binned, the 12 meg single pixel sensor is better again. which brings us to the conclusion again, that 1 big pixel is better than 4 binned and 4 binned is better than more smaller pixels regarding the accuracy of the read voltage/charge.
since canon or nikon have not disclosed what they are doing with the pixels in video recording, from TN tests its clear that nikon are line skipping and not sub sampling from the whole sensor just so they can con buyers that speed is more important than image/video quality. sony are so far ahead of the competition for pro quality gear. i even discovered the other day that the whole AI focusing is a load of crap just like ive been saying for 2 years, i bought the a6700 last week and been testing it side by side to my a7iv. i still have more testing , its a lot of fun doing the comparrision, the a6700 is still very good, i just need to get the fine adjustments down pat before i make a definitive comment.
"does not affect"
It makes no photographic sense to compare noise or DR pixel-for-pixel two photos made with different numbers of pixels. We would compare the same proportion of the photo. For example, in the case of 48 MP vs 12 MP, this means we would compare the total noise/DR of 4 pixels from the 48 MP sensor to 1 pixels from the 12 MP sensor (or any multiple thereof).
So, if a 12 MP downsampled photo/frame from a 48 MP sensor is the same or less noisy (has the same or more DR) as the whole photo/frame from a 12 MP sensor, with all else more or less equal, then, no, more smaller pixels are absolutely NOT at a noise/DR disadvantage to fewer larger pixels.
You use the same false logic that so many mFT users use when they say that mFT has "more DOF" than FF, because they compare photos taken with the same f-number, as if the FF user is somehow forced to use the same f-number that they are using. In fact, I heard a former mFT mod on DPR make the claim that using f/4 on mFT is so much "easier" than using f/8 on FF. And when I asked how it was "easier", well, we all know what kind of "entertainment" ensued. Yet, here you are doing the exact same thing:
It brings us to no such conclusion at all, Don.
which part ? www.teamwavelength.com/photodiode-basics/
why do you think manufacturers sell different sizes ? www.osioptoelectronics.com/products/photodetectors/monitor-photodiodes
I don't know, Don. How about you quote the portion that states that fewer larger pixels result in a photo with more DR than more smaller pixels, and we'll go from there and/or explain why the Z6.2 isn't mopping the floor with the Z7.2:
because nikon is manipulating the raw files in all there cameras, you wouldnt know what was correct with that company. they have been doing it for years. also i must eat my words on the subject .i tested my new a6700 yesterday for dynamic range against my crop mode on my a7iv and i can tell you the dynamic range of the a6700 was amazing, noise was finner but the detail was still there in all its glory. havnt done any low light comparisons yet, but have done some sun back lit head selfies yesterday and im impressed with the shadow skin tone quality from the a6700, and wouldnt hesitate using it professionally. the final test will be my extreme macros as the stacking programs stack noise into blobs and so far the a7iv can produce the cleanest detail to date.
Two questions:
1) How do you know Nikon is manipulating their RAW files?
2) If true, then why is Nikon doing a better job manipulating the RAW files for their Z7.2 than for their Z6.2?
Should have led with that. 😉
Well there you go: crop mode on the A7.4 vs the whole photo of the A6700 uses the same sensor area, thus records the same amount of light for a given exposure, yet the A6700 puts a lot more pixels on the subject.
My money is on the A7.4, unless you're cropping. For the same number of stacked exposures using the whole of the photo, the A7.4 photo will have the advantage. The A6700 will be able to match the A7.4 by stacking more exposures, but then you could have just used more exposures with the A7.4, so using more exposures on the A6700 is artificially putting the A7.4 at a disadvantage.
its going to be interesting. the a6700 turns my 4x finite objective into 6x mag where atm im using a tube lens on my 10x infinity objectives to achieve 6x . the 4x objective even though is the cheapest objective i own it has amazing detail right to the edges of the frame. so im hoping the noise at iso 100 is fine enough to not interfere with detail. at f16 the pixel size of the a6700 is in diffraction territory so im hoping it doesnt degrade the fine detail. im heading home tomorrow and have lots of work to catch up on so it will be a couple of weeks before i can truly test the new a6700.
