I couldn't find many High ISO comparisons online between these cameras and so decided to do my own.
Both were shot raw under 4800K LED lighting, using the same Voigtlander Ultron 40mm lens at f/5.6 (X-E1 used this adapter). Manually focused, tripod, timer delay, processed in LR4 using all defaults except white balance adjustment. The X-E1 produced a lower brightness image at every ISO even though the exposure and lens were identical; to account for this and make the comparison fair I made the following LR4 exposure adjustments to bring the X-E1 brightness up to the D7100's levels (matched histograms):
ISO 3200 +0.60EV
ISO 6400 +0.55EV
ISO 12800 +1.50EV (X-E1 shot max ISO 6400, +1EV base adj in LR4)
ISO 25600 +2.46EV (X-E1 shot max ISO 6400, +2EV base adj in LR4)
Below are 2 side-by-side comparisons at each ISO, one to demonstrate color+detail in midtones and the other to demonstrate shadow noise. Both were downsampled to 8MP to normalize for resolution differences and make this a sensor-area equivalent comparison. I can make the raw files available if there is any interest. X-E1 is on the left, D7100 on the right.
I'd give the edge to the D7100, but....I'd rather carry the XE-1 around me with all day.
Which is exactly why I just bought a XE-1 and 14mm f2.8 lens. My D3s certainly has the edge in high ISO over it, but the Fuji 14mm is just unmatched as far as compact ultra wide primes go and at just $700 (after rebate)
Always interesting to see how different aspects between camera's stack up, but usually they don't give that good of "big picture" indication of what a given camera is like in everyday usage.
FlyPenFly wrote:
What's the dynamic range like at base and low iso?
aleksanderpolo wrote:
If the underlying sensor is really Sony, it should be similar to the one in 5N/6.
The sensor is probably the same Toshiba sensor as in the D5200. It sets a new standard for sensitivity in larger sensors, but it is not EXMOR. While the low ISO DR is very high, we are back to pre-EXMOR low ISO shadow banding amd I have seen shadow banding in samples.
I am looking forward to the EXMOR-R sensor in the RX100 II.
FlyPenFly wrote:
I thought one of the key benefits of xtrans was very high DR?
I've seen that advertised for it, but all the samples seem to show slightly less DR at lower ISO than with the stock EXMOR 16MP. DR looks better than the EXMOR at high ISO though.
X-trans has very good DR in a downsampled image. It has worse DR "per pixel" in a full-size image than any given Exmor of the same generation.
Bayer based images have the highest available resolution physically possible in a one-layer full-color image. This also means they have the highest DR in a raw-converted image (not necessarily in each raw pixel value before conversion though!).
X-trans has higher accuracy of value in a 3x3-size average cluster, better than if you average a normal Bayer in 3x3 clusters. This means that DR[X-trans] / DR[Bayer] goes up as noise increases.
Compare a high-ISO (downsampled) 1800x1200px image from the X-E1 and an equivalent image from any of the current "same sensor base" competitors, and the X-E1 retains a bit more shadow "detail". But at low ISO, it just can't resolve detail as well as the Bayer due to the numerical instability of the subsampling, so DR at high-resolution, low-ISO image suffers.
Images are more than their base raw per-pixel performance, you also have to apply quite a lot of processing to get a full-color image. In X-trans you lose some per-pixel detail at low ISO, in Foveon you lose color accuracy and in Bayer you lose high-ISO stability (small image size coherence).
FlyPenFly wrote:
What's the dynamic range like at base and low iso?
It's very good actually. Here are three exposures; the first is at Base ISO (200), the second at Base ISO underexposed by 4 stops and pushed 4 stops in LR4, and the last at ISO 3200. The IQ delta between the second and third demonstrate the base ISO DR (shadow noise) and color fidelity retention, which depicts DR and ISOless'ness of the sensor.
Here are three 100% crops of an 8MP downsampled image from each of the 3.
To me, I don't see a huge difference between the noise performance of the two. What concerns me is the color bleeding evidence from ISO6400 onward, and given how the ISO6400 is really around ISO4000 from other manufacturer's camera, this is going to be a problem. Do you see it from the ooc jpeg too? Or is this again an Adobe problem? I took the center crop from your test, snapsy, hope you don't mind:
aleksanderpolo wrote:
PS. The banding is pretty bad, is it in Sony's 24MP sensor too?
No banding in the Sony 24MP sensor, but the D7100 uses a completely unrelated Toshiba 24MP sensor anyways, there are no non-Sony cameras with the Sony 24MP sensor currently.
aleksanderpolo wrote:
To me, I don't see a huge difference between the noise performance of the two. What concerns me is the color bleeding evidence from ISO6400 onward, and given how the ISO6400 is really around ISO4000 from other manufacturer's camera, this is going to be a problem. Do you see it from the ooc jpeg too? Or is this again an Adobe problem? I took the center crop from your test, snapsy, hope you don't mind:
Here's an ISO 25,600 OOC JPEG vs an ISO 6,400 RAW (same exposure as JPEG but pushed 2 stops in post since the XE1's max raw ISO is 6,400). I see the color smearing in both.
