Why 100MP+ Medium Format

This comparison of the medium format Fujifilm GFX 100 and Sony A7IV says otherwise: www.dpreview.com/reviews/image-comparison/fullscreen?attr144_0=fujifilm_gfx100&attr144_1=fujifilm_gfx100&attr144_2=sony_a7iv&attr144_3=sony_a7iv&attr146_0=100_0&attr146_1=100_5&attr146_2=100_0&attr146_3=100_5&attr178_0=2&attr178_1=1&attr178_2=2&attr178_3=1&normalization=compare&widget=587&x=0.18635354933829643&y=0.5017597853197943

By definition, the comparisons shedding light on this discussion are of non-equivalent images.

The fact that it is possible to make equivalent photos using different format systems doesn't negate the potential of a larger format to capture more light.

By definition when discussing equivalence, the lens on the larger format camera delivers a lower exposure to its sensor than delivered by the lens on the smaller format camera. If sensor size weren't a factor, the smaller format camera working with the greater exposure would record more total light. However, equivalent systems record the same total light. That's possible because of the larger format's greater surface area.

Shot noise is proportional to the square root of the signal (total light) delivered to the sensor. This is a fact of nature.

Equivalent photos are, by definition, made with the same total light by different format cameras working with different exposures. The greater surface area of the larger sensor allows it to capture as much total light as the smaller format camera working with the greater exposure.

By definition when discussing differences in total light captured by different format systems, we talking about non-equivalent systems and photos being made.

That it is possible to make equivalent photos with different format systems does not negate the fact that a larger format sensor will capture more total light than a smaller sensor when both work with the same exposure.

It's true that one of the limitations of a smaller format is the availability of lenses matching either the ultra wide angles of view or the shallow depths of field of lenses available to a larger format system.

However, it's still possible to choose exposure settings allowing medium format and micro four thirds systems to make equivalent photos.

"Stops below what?" is an important question. "Stops below raw data saturation", and "stops below middle grey for the intended exposure for an ISO setting", are two very different things. It is up to the manufacturer how much they split the "DR" into headroom and footroom, relative to middle grey. For example, many people were impressed when some m43 cameras started sporting very impressive DR at a given ISO setting several years back, but the fact is that the manufacturers were overrating ISO compared to other cameras to provide more automatic headroom. It was basically like HTP on Canons, which take a digitization and gain that would otherwise be ISO 100, and use it for ISO 200. Neither Bill Claff nor DxO measure (P)DR in HTP mode, but those metrics would have a hollow improvement if they did. This is why quoting DR or PDR for an ISO setting can NOT be used to gauge exposure-referred noise.

What rational person judges images by pixel views when different numbers of pixels make the same image?

The fact is, the post-gain read noise of the A7R4, aggregated over the entire sensor, or aggregated per unit of sensor area, is not leading-edge, at all. It is a little worse than most current sensors except the remaining few that still use off-sensor ADCs. So, only FF cameras like the Canon RP and the Nikon D6 currently have much less base-ISO DR at the image level.

Generally expected, and it would always be true with photon-counting sensors of the same QE, but post-gain read noise is a wild card that can creep in when it is particularly strong and/or the pixel count is low. It really is quite detached from sensor size.

never look at DPR test images the quality of there testing is reflected by the skin tone quality in the test scene 😉

we are discussing the a7iv not the a7r4

the compression factor is 4x on the a7s3 so why isnt it 4x better than the a7s2 😉

I didn't realize that DPR was testing for skin tones

I didn't realize that DPR was testing for skin tones
[/quote]

i love it when im always right . dpr test scene ( how to hack the actual real world results) took both raws into ACR and push the exposure slider to the far left . and see what happens to the blown highlights.
its been that long i cant remember how to upload an image 😂

[quote="@IanSForsyth"]

Well how does this relate to skin tones ? What skin tones are being shown? prove to me any part of the DPR test is a correct test for skin tones?
It has taken over 1000 post across 3 threads and you still don't understand

didnt you look at the image i posted 😂😂 the a7iv has the gfx on toast and i even magnified the a7iv image past the gfx that has 3 x the megapixels 😆 look at the graduations smooth against blob.
how can i insert the image not just the link ? there no infor on the faq of this site.

I have success in loading the image 🤣🤣

Don's got his tapdancin' shoes on, today 😆

Please, share a link to the original raw files.

the files are on dpr. the 200 meg gfx file would take forever to load to my one drive.
its the paint pallet dishes

Oh, so the photos where the A7IV photo is obviously noisier when both are lifted 5-stops in post.

we are discussing DR . noise has nothing to do with DR , recording accurate colour and tonal graduation. when did raw files have a name "capture one session 000020 🤔