[wahid_satu]

Nice work!

Thanks!

Originally posted by ElrondElensar
That will not explain anything here as Nikon cameras are using faster shutter speed.
In these examples, aperture is fixed, and nominal or 'manufacturer' ISO is too. So the accuracy of the Nikon ISO metering may indeed have an influence on the cameras calculations about which shutter speed to employ under the lighting conditions in the IR test scene ...

I hope this will clear things up a bit:

Still using Rawtherapee and Amaze I underexpose the K-3's image to match the Nikons' (still no other PPs). Not really fair but short of asking IR to reshoot the scene with matching shutter speed this is the best we can get. Overexposing the Nikons' would not be as fair since it will introduce more noise. And, guess what, I need -0.7 compensation to match the exposure of the Nikons'. Now considering that going from 1/800 to 1/1250 constitutes only -0.56 the additional -0.14 must come from the true ISO of the Nikons.

We can see from the result that the K-3 is less noisy than the D7100 but more noisy than the D600. The D7100 hold as much detail as the D600 in the brighter areas but not on the darkers where the K-3 beats it and nearly matches the D600 but with slightly more noise. IMHO, of course. Enjoy:

Attached Images

 

[IchabodCrane]

wahid_satu, thanks for all the work. As far as IQ goes, it looks like there isn't much that separates these cameras.

[rawr]

I underexpose the K-3's image to match the Nikons'

More interesting work. Thank you.

Selectively darkening noisy areas is a simple and effective NR strategy, as is crushing the blacks I often do it in Lightroom on high ISO shots.

I think your work continues to demonstrate the similarity of output from these three cameras, at least under bright studio lights.

It will be interesting to see how things work under lower light levels, where the sensor will have to work harder to assemble an image. Unfortunately, when Imaging Resource run their low-light tests (under the Exposure tab in their reviews) they only provide JPG's to look at. However there will probably be lots of user feedback about low-light performance soon, since I think it's a big reason many people want this camera. Myself included

[VoiceOfReason]

Photoshop CS 5.5 has no problems at all with them. Getting the detail is pretty easy too.

[Class A]

And, guess what, I need -0.7 compensation to match the exposure of the Nikons'. Now considering that going from 1/800 to 1/1250 constitutes only -0.56 the additional -0.14 must come from the true ISO of the Nikons.

The shutter speed difference is equivalent to -0.64 stops.

BTW, the comparison to the full-frame sensor is not fair, AFAIC, as f/8 on FF is ~f/5.3 on APS-C. That gives the full-frame sensor a bit more than a stop of an advantage (which it pays with shallower DOF). So bear in mind that we are not only comparing different sensor technologies, but also different lens speeds.

[Cynog Ap Brychan]

BTW, the comparison to the full-frame sensor is not fair, AFAIC, as f/8 on FF is ~f/5.3 on APS-C. That gives the full-frame sensor a bit more than a stop of an advantage (which it pays with shallower DOF). So bear in mind that we are not only comparing different sensor technologies, but also different lens speeds.

I'm not sure I understand this. In terms of light gathering, f8 is f8 on all lenses, no matter what size the sensor is. If the same lens is used on cameras with different formats, the depth of field must also be the same. The difference in depth of field comes about through either the use of a wider lens on the smaller sensor, or a greater camera to subject difference to obtain the same field of view.

[rawr]

a greater camera to subject difference to obtain the same field of view

Which is what they do on Imaging Resource. They use a 70mm f2.8 Sigma prime for their standard test scenes, and they move the camera backwards or forwards to fill the frame with the test scene(s) as exactly and consistently as they can manage. Which may have an impact on equivalence and DOF calculations.

---

Last edited by rawr; 11-01-2013 at 12:40 AM.

[Cynog Ap Brychan]

Which is what they do on Imaging Resource. They use a 70mm f2.8 Sigma prime for their standard test scenes, and they move the camera backwards or forwards to fill the frame with the test scene(s) as exactly and consistently as they can manage. Which may have an impact on equivalence and DOF calculations.

I understand - thank you. But do they use the same aperture for each test, or do they change it to get a comparative depth of field?

[rawr]

do they use the same aperture for each test, or do they change it to get a comparative depth of field?

From the EXIF, one can see that they use f8 all the time. So the aperture used is consistent. I guess they aren't that focussed on making sure the DOF is the same scene to scene, and chose f8 because it would probably provide enough DOF coverage of all the objects in these test scenes. It's probably the resolution sweet spot of the Sigma 70mm too.

[Class A]

In terms of light gathering, f8 is f8 on all lenses, no matter what size the sensor is.

That's not correct. The same f-ratio (e.g., f/8) implies the same light gathering independently of the focal length, but the size of the sensor matters.

F-ratio is an indicator for exposure. Exposure, however, is a relative measure (it indicates the amount of light per unit area) so if you have a larger sensor, the amount of total light gathered is larger. Unless you counteract that effect by stopping down.

Using the same focal length with both sensor formats, i.e., adjusting distance to subject, rather than using equivalent focal length, only influences perspective, but not the total light gathered (i.e., the noise level of the whole image).

Larger sensor do not have a low-light advantage per se. Only if you compare different sensor sizes by using the same f-ratio -- which isn't fair but even DxOMark does it -- then the larger sensor looks like as if it had ~1 stop better noise performance, if the sensor technology is the same.

It's probably the resolution sweet spot of the Sigma 70mm too.

The Sigma 70/2.8 EX DG Macro is much better than this. It peaks around f/4, hence at f/8 it already suffers from diffraction.

[Cynog Ap Brychan]

F-ratio is an indicator for exposure. Exposure, however, is a relative measure (it indicates the amount of light per unit area) so if you have a larger sensor, the amount of total light gathered is larger. Unless you counteract that effect by stopping down.

I agree the total amount gathered is greater, but it's spread over a greater area to give the same amount of light per unit area (more or less). I'm still not understanding what point you are making.

[ArneTorp]

Some pages that will explain DOF on different sensor sizes:

• Just for fun: Sensor Sizes compared for depth of field – Small, Medium and Large! | STEVE HUFF PHOTOS
• http://www.blogarithms.com/2013/05/15/sensorsizedof/
• Why does a bigger sensor lead to a shallower depth of field? - Photography Stack Exchange

[Class A]

I agree the total amount gathered is greater, but it's spread over a greater area to give the same amount of light per unit area (more or less). I'm still not understanding what point you are making.

Note that the images -- from the small and the bigger sensor respectively -- are ultimately compared at the same size. If we did not use different enlargement factors for the the different sensor formats then maintaining the same exposure would yield in the same noise being observed.

Imagine two prints of the same size, or two digital files with the same dimensions (say both 24MP). This implies that the smaller sensor requires a bigger enlargement factor. This is easy to visualise when you think about two negatives of different sizes (but the same exposure) being enlarged to the same photo paper size. The difference is still there, though, when dealing with digital images. If you compare two digital images at the same size (of the same scene) and one of them was produced with more photons (i.e., more total amount of light was used to create it), the latter will show less noise (strictly speaking: "will have better signal to noise ratio").

[Cynog Ap Brychan]

Imagine two prints of the same size, or two digital files with the same dimensions (say both 24MP). This implies that the smaller sensor requires a bigger enlargement factor. This is easy to visualise when you think about two negatives of different sizes (but the same exposure) being enlarged to the same photo paper size. The difference is still there, though, when dealing with digital images. If you compare two digital images at the same size (of the same scene) and one of them was produced with more photons (i.e., more total amount of light was used to create it), the latter will show less noise (strictly speaking: "will have better signal to noise ratio").

Thank you for that, and please excuse me if I seemed a little dense. I completely agree that the larger sensor should have a better signal to noise ratio if it has correspondingly larger photosites, all other things being equal (which they never are).

[rawr]

the latter will show less noise

You forgot to add the qualifier: "(all things being equal, in terms of the sensor technology in use). ..."