Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
I'm in tooIf that is the case I´ll get it FOR SURE!
I´ve filmed a wedding this weekend and I kept thinking how awesome it would have been ot have had the A7s with me in certain shots.
The A7s DxoMark results are out. As predicted its base ISO DR is below the higher-density D800/A7r, for reasons relating to large pixels described earlier in the thread (here). For High ISO though the A7s has monster DR - it's 1.7 stops better than the current leader (Nikon Df) @ ISO 102,400.Thank you! I don't think that D3S can be used as good example because it is three to four years older than say D800/D4S/Df but even if we ignore that we are not comparing same generations of technology D3s has better tonal range and color sensitivity per pixel than any of them, it is only DR that it has worse by a bit and only until ISO 400, after which it matches D4s and both seem to be significantly better than D800.
That is with 12 MP technology that is 3-4 years older than sensor with more MP. Which makes me very curious what the results will look like with 12 MP technology that is newer than 36 MP one. Having A7S by the side of A7R will satisfy my curiosity. Now lets keep our fingers crossed that price is within reason.
Are we looking at same data or I am missing something? When I look at DXOMark's _non-downsampled_ data I see A7S having exactly same DR as A7R.The A7s DxoMark results are out. As predicted its base ISO DR is below the higher-density D800/A7r, for reasons relating to large pixels described earlier in the thread (here).
Per-pixel metrics have no significance since we render images (on a per-area basis) rather than individual pixels.Are we looking at same data or I am missing something? When I look at DXOMark's _non-downsampled_ data I see A7S having exactly same DR as A7R.
Maybe we need to discuss this differently. What is a dynamic range of a sensor to you? And why exactly we should ignore per pixel metrics? If DR on per pixel basis is exactly same for smaller and larger pixel can DR be increased at all just because we downsized the picture? And if answer is yes a) when does that stop, can DR become infinite?, b) will scene whose DR can't fit in range of "per pixel" fit once I take picture and downsize it? Will same scene taken with A7R and A7S with same exposure not have clipped highlights and shadows on A7R after downsizing if it had them clipped before and had them clipped on A7S?Per-pixel metrics have no significance since we render images (on a per-area basis) rather than individual pixels.
We should ignore per-pixel metrics because we render in terms of sensor area rather than pixels, whether we're rendering an image at a given PPI or cropping a subset of the image. The number of pixels contained in that area for a higher-density sensor will be more vs a lower-density sensor but the total area used from the sensor is the same (in millimeters from the sensor), thus the relevant sensor performance is the total light/noise in that area rather than the light/noise in the individual pixels from that area. The usual metaphor is buckets of water - one large bucket vs 3 smaller buckets but where the total volume of water held is the same.Maybe we need to discuss this differently. What is a dynamic range of a sensor to you? And why exactly we should ignore per pixel metrics? If DR on per pixel basis is exactly same for smaller and larger pixel can DR be increased at all just because we downsized the picture? And if answer is yes a) when does that stop, can DR become infinite?, b) will scene whose DR can't fit in range of "per pixel" fit once I take picture and downsize it? Will same scene taken with A7R and A7S with same exposure not have clipped highlights and shadows on A7R after downsizing if it had them clipped before and had them clipped on A7S?
Will same scene taken with A7R and A7S with same exposure not have clipped highlights and shadows on A7R after downsizing if it had them clipped before downsizing and had them clipped on A7S?We should ignore per-pixel metrics because we render in terms of sensor area rather than pixels, whether we're rendering an image at a given PPI or cropping a subset of the image. The number of pixels contained in that area for a higher-density sensor will be more vs a lower-density sensor but the total area used from the sensor is the same (in millimeters from the sensor), thus the relevant sensor performance is the total light/noise in that area rather than the light/noise in the individual pixels from that area. The usual metaphor is buckets of water - one large bucket vs 3 smaller buckets but where the total volume of water held is the same.
Highlights will be the same since total exposure is spread across the same area (ie, each individual A7r pixel will hold less light but in total hold the same amount of light as the fewer pixels of the A7s). The same is true for shadows except for the very deep shadows, where the A7r will hold the advantage at base ISO due to the lower read noise from the shallower pixel wells (even accounting for the higher theoretical read noise from more pixels having to be read) - the situation is reversed at High ISO where the A7s benefits from lower aggregate read noise from fewer pixels. This is why DR is better at base ISO for the A7r and better at High ISO for the A7s, and why this relationship holds true for most sensors of similar technology generations.Will same scene taken with A7R and A7S with same exposure not have clipped highlights and shadows on A7R after downsizing if it had them clipped before downsizing and had them clipped on A7S?
I feel we are not on same wave length here but I will try again: If DR of scene does not fit within DR of single pixel on A7R will it fit once picture is taken and downsized? Yes or no?Highlights will be the same since total exposure is spread across the same area (ie, each individual A7r pixel will hold less light but in total hold the same amount of light as the fewer pixels of the A7s). The same is true for shadows except for the very deep shadows, where the A7r will hold the advantage at base ISO due to the lower read noise from the shallower pixel wells (even accounting for the higher theoretical read noise from more pixels having to be read) - the situation is reversed at High ISO where the A7s benefits from lower aggregate read noise from fewer pixels. This is why DR is better at base ISO for the A7r and better at High ISO for the A7s, and why this relationship holds true for most sensors of similar technology generations.
For the highlight end, no. For the shadow end, yes (due to noise averaging from downsampling).I feel we are not on same wave length here but I will try again: If DR of scene does not fit within DR of single pixel on A7R will it fit once picture is taken and downsized? Yes or no?
Then the answer is no, downsizing won't make DR of a scene fit if it didn't fit before downsizing. Thus IMHO downsizing doesn't "buy me" more DR and consequently A7R will not / does not have more DR at base ISO once image is downsized.For the highlight end, no.
If your image consists of only a single pixel, sure. There aren't many creative photographic possibilities in a single pixel though.Then the answer is no, downsizing won't make DR of a scene fit if it didn't fit before downsizing. Thus IMHO downsizing doesn't "buy me" more DR and consequently A7R will not / does not have more DR at base ISO once image is downsized.
Thanks for the explanations re mpix & DR & ... .The A7s DxoMark results are out. As predicted its base ISO DR is below the higher-density D800/A7r, for reasons relating to large pixels described earlier in the thread (here).
I don't see this sentence adding any value to / providing any insight into discussion, all I see it doing is avoiding straight topic about fitment of DR of scene. So it's obvious this will not take us anywhere. Maybe you could please show me / us with real world example how shot of scene whose DR was not fitting one of sensor would end up better DR wise from downsized A7R file than shot of same scene taken with A7S without any downsizing?If your image consists of only a single pixel, sure. There aren't many creative photographic possibilities in a single pixel though.
That is one of the things I myself am always very interested to see about any new camera because having used number of cameras over the years left me with belief that there is much more to will I be happy with files from some camera or not than just what test numbers imply about sensor. I don't know what it is, I can't put my finger on it or describe it, like you said it is not quite tangible, but I do believe it is there as I believe I do "feel" it (I have had cameras whose sensor performance test numbers were implying I should have been equally, or more, happy with files from them as I was with ones from some other cameras I owned yet I never was nowhere near).... I am interested to see a range of images at lower ISO from the a7s ... to get some appreciation of what MR was seeing in the files.
If the new sensor delivers more of something not quite tangible in data that is nevertheless apparent in images ...