Tariq Gibran wrote:
The DpReview base ISO raws from the NEX--7 still look damn noisy to me.
This portion of the image is from the corner, and I think there is some limitation from the lens there. However, if you look at the rest of this image (you can download the RAW, or the ACR-converted-raw jpeg) and re-size the 5N image to the 7 image size (6000x4000), you see are very clear and noticeable advantage in detail, resolution and sharpness with the NEX-7 image over the NEX-5N image. Yes, there appears to be a touch more noise some of the shadow areas, but really it is barely noticeable compared to its very obvious advantages.
Lotusm50 wrote:
This portion of the image is from the corner, and I think there is some limitation from the lens there. However, if you look at the rest of this image (you can download the RAW, or the ACR-converted-raw jpeg) and re-size the 5N image to the 7 image size (6000x4000), you see are very clear and noticeable advantage in detail, resolution and sharpness with the NEX-7 image over the NEX-5N image. Yes, there appears to be a touch more noise some of the shadow areas, but really it is barely noticeable compared to its very obvious advantages.
I was wondering why the watch was so smudgy with the 7 compared to the 5N, didn't realize it was so far into the corner. Here's another comparison closer to the center, off to the left...
bluetsunami wrote:
I was wondering why the watch was so smudgy with the 7 compared to the 5N, didn't realize it was so far into the corner. Here's another comparison closer to the center, off to the left...
And against some other heavy weights...
ah, much better. forgot how much soft low contrast lenses make sensors look bad.
michaelwatkins wrote:
I wish it would be... but remember the series Rob Skeoch shot?
maybe you are ready to close the case based on such evidence, I'd prefer to see production cameras shooting proper test shots before I do.
Obviously you are delighted at the prospect that the n7 might not shoot that lens well. Hence you are hardly holding a disinterested opinion, but to be gloating over the triumph of the GXR, and at considerable pains to point out the demise of the n7.
Kharma dictates you will soon have your shots ridiculued by a like minded M9 shooter
, I'll be sniped off by a karma chameleon any second now I'm sure.
It'd be just lovely if the NEX-7 were to treat our lovely rangefinder lenses nicely all across the board. I suppose I could carefully pick and choose which lenses I use but if your example is representative, it'll cost me. A lot!
While I share your concern about noise and tonality, I think there is something funny about just this patch from the A900.
Look at the areas A and B below. The noise difference between the NEX-7 and the A900 is very small in area B. The funny thing is, the NEX-7 has more noise in area A than B, while the A900 has less noise in area A than B.
Tariq Gibran wrote:
I suspect there is excessive blue channel noise with the NEX-7 sensor. So as long as you don't shoot blue skies, blue watch faces or anything blue...
Is the blue channel usually the noisiest on a DSLR? I seem to remember that it was on my original 1Ds as well.
Beni wrote:
Is the blue channel usually the noisiest on a DSLR? I seem to remember that it was on my original 1Ds as well.
Yes, I believe it is. It's odd to see it so noisy with a current sensor at base ISO if this is what's going on but I suppose that's what happens when you cram so many pixels onto an APS sensor.
thrice wrote:
Damn the A900 looks nice at base ISO!
The NEX-7 test image is soft in the lower-right corner, at close distance. Everywhere else in the frame, it has remarkably higher micro contrast than the A900. If you process the two images to identical micro contrast, the noise difference will be less. I think we see an example of how the AA filter affects the effective noise, as a strong AA filter requires more sharpening which will in turn enhance the noise.
alundeb wrote:
The NEX-7 test image is soft in the lower-right corner, at close distance. Everywhere else in the frame, it has remarkably higher micro contrast than the A900. If you process the two images to identical micro contrast, the noise difference will be less. I think we see an example of how the AA filter affects the effective noise, as a strong AA filter requires more sharpening which will in turn enhance the noise.
The lens on the a900 looks soft in other areas vs on that watch face so I suspect detail and micro contrast differences are down to the lens and possibly even focus/ dof issues.
Tariq Gibran wrote:
The lens on the a900 looks soft in other areas vs on that watch face....
That is not the case as far as I can see. It is only the comparison with an even softer crop for the watch face that makes it appear sharp. The watch face is not sharper than other areas for the A900 IMO.
You can see it in the iso 100 pictures as well, been playing with the IR raw's some more, it's not just the blue channel, quite a lot of noise in the red channel and green channel as well at iso 100.
alundeb wrote:
That is not the case as far as I can see. It is only the comparison with an even softer crop for the watch face that makes it appear sharp. The watch face is not sharper than other areas for the A900 IMO.
I disagree and for me this proves it. This is close to the center and the a900 crop here is clearly not even close to resolving the detail one can see in the watch crop. Had I shot this with my a900, I would be convinced there was an issue either with focus point, dof or the lens because this does not reflect what the a900 is capable of in my experience.
Tariq Gibran wrote:
I disagree and for me this proves it. This is close to the center and the a900 crop here is clearly not even close to resolving the detail one can see in the watch crop. Had I shot this with my a900, I would be convinced there was an issue either with focus point, dof or the lens because this does not reflect what the a900 is capable of in my experience.