This is for Ted to get his photographic fix.
Photos to Photons ves DXOmark
should DR be based on noise or actual captured images.
eg: DXO the sony a7s sports iso rating 3700 vers sony a7s3 iso 2500 knowing that the sony a7s3 is binned 48 meg to 12 vers a7s 12 large single pixels. Bills site shows a different result.
Hello, Donald: I am biased for Bill Claff and against anything DxO but, in any case, there's an ISO Standard for Dynamic Range, so noise.
so adding noise reduction increases DR 🤔 as it clearly shows in bills site but is the exact opposite to cine-d testing where the z6ii tested 2 stops less DR 😊to other FF cameras.
Numbers are nice, but photos are better, especially with regards to DR (in my opinion, DR is the most misunderstood of the IQ metrics). To that end, I really like the DPR "DR comparometer". The main issue with this method is how strongly the results vary with different RAW converters.
That said, I don't think DR is an issue with any cameras of the past decade unless you're someone who does a lot of heavy shadow pushing. It's the same with pixel count. Yes, 48 MP is better than 24 MP, but unless you're heavily cropping or displaying super large and viewing super close, the differences really don't matter all that much. In fact, depending on the photo, the differences won't matter even then. : )
That's not to say that more DR (or more resolution) can't be useful, but the best way to achieve this, methinks, is the way smartphones do it: merging multiple photos. If modern cameras were able to take even two exposures and merge them into a 16 bit (or even 24 bit!) RAW file, that would be something special. For example:
If the camera had taken even two exposures, one with the same exposure taken, and the other with, say 1/4 the exposure time, and merged them, then there would still be detail in the moon sliver. Could I have done it manually? Sure. Set the camera to shoot an exposure bracket, convert the photos, then merge with the desired levels. Would really like one stop shopping, here, and, honestly, I don't think it's that much of an ask. Sure, if you're firing off 20 fps, but if you're just taking that one shot, and it takes the camera even three seconds to process it into a high bit high DR RAW file, that would be awesome.
But with regards to sensor tech for single exposures, yes, there's a difference, but, for the most part, I don't think the difference really matters as any "meaningful" difference would require a massive increase in DR which would be best served, by far, by merging multiple exposures like smartphones.
im the same and photos are the only way to exactlly measure DR. given bills site says that the a7iv and a7r5 are the same but from images taken on Dustin Abbottts site the a7iv is clearly in front at 3 stops over exposed and clearly infront at 5 stops under exposed, so how do sites get it so wrong 🤔 i also came to the same conclusion that to actually reach 15 stop DR the testers must shoot a landscape using a 6 stop graduated ND filter 😁
As we continue to ignore the existence of the ISO Standard for Dynamic Range of a Digital Camera:
define 12 volts 🤔
too abstruse for me - no comment
its the same as asking to read iso standards 😁im more into actuall photographic results that we see with our eyes. we listern to audio quality with our ears we dont measure it with a meter.
In which case why were you talking about "which site has the more accurate testing protocol" and "the only way to exactlly measure DR" when human vision is the least accurate way to measure anything?
Human: 3"
Vernier Caliper: 3.231"
can you not see the comparrision images i posted and not see the difference 🤔
Please answer my question. If you don't, I'm done with our conversation.
the original post was a fairly basic question. my cameras measure the DR more accurate to the sites which claim correctness, its like which tool measures your car speed the most accurate. car speedo, radar gun, gps . or time and distance.
The A7.4 and A7R.5 look the same to me:
Terminology: "measure the DR" ... the units of DR are normally EV or sometimes dB. So what is the DR of your Sony cameras as measured by your good self? (another question for you to evade as usual).
All of which are more accurate than just looking at images on your screen.
'bye ...
9 stops. if dustin abbott can recover more than 5 stops under exposed and 3 stops over than thats 9. i read the other days richard says the canon r6iii has 8 stops of ibis with lens and gordon lang said he tested it at 4 stops, so my guess is your answer is based on richards even though you would need to hand hold the camera for 6 secs to acheive 8 stops , who is correct Richard, canon, gordon or me 🤔
which camera shows the most DR i over exposed the raw files by 2 stops
If a photo is truly overexposed, then pulling it back will not recover anything -- no matter what. That's what it means to be "overexposed". Now, true, it's possible that the A7.4 can handle more exposure than the A7R5, but I truly doubt it. That is, it's possible, I suppose, that one pixel on the A7R.4 can take in more light than two pixels on the A7R5 before oversaturating. But, again, I really doubt it. Here's BC's data on the A7.4 and A7R4 (BC didn't do the A7R5, but they're both the same pixel count):
Note the black bar -- that's how much the pixel can take before oversaturating. Note how the A7.4 has exactly twice the FWC as the A7R4. But the A7R4 has twice as many pixels (like the A7R5). Therefore, they oversaturate with the same exposure, and I'm betting it's the same with the A7R5. In other words, they will both overexpose the same, and, as I said, there is no coming back from overexposure.
So, the differences in DR come entirely from how much of a noise penalty there is in the shadows can be pushed with less of a noise penalty. So, we can't "get rid of noise in the equation" because that's what DR is all about. If we look at the green bar, we can see the per-pixel noise is the same. Thus, the A7R4 will have more noise due to having twice the number of pixels, thus lower DR. However, it will also have greater resolution. In many instances, that greater resolution will allow for more aggressive NR which will often tilt the advantage in favor of the higher pixel count (all else equal) when resolution is normalized. Here's an example of exactly that:
In any case, the DPR "DR Comparometer" pretty much tells the story -- the DR for the A7.4 and A7R5 is essentially the same. But why do you care? If you're not thinking of getting an A7R5 to replace your A7.4, then it's just an academic exercise. Like comparing the horsepower of your car to another car you're not going to buy, anyway. Unless you're like me, and just like understanding things. But you've never seemed to show an interest in that. : )
