D5 - long term experience?

[Jack Dahlgren]

Also dynamic range is just one parameter, a full characterisation of the SNR vs. luminosity reveals more about the differences between sensors than just looking at DR.

Indeed, I find the previously referenced chart to be thin justification for choosing one camera over another.

[David H. Hartman]

Indeed, I find the previously referenced chart to be thin justification for choosing one camera over another.

I've read criticism of the linked site and it's findings. I believe the chart linked is good enough for a practical evaluation of various cameras. I believe dynamic range is only one criteria to consider.

From practical experience that I found the Nikon D2H deficient in dynamic range under many situations. In particular I found it near impossible to record shadow detail in a groom's black tuxedo while holding diffused highlights in the bride's white gown. I was shooting a cousin's wedding by request in a photojournalist style in deference to the official photographers. I was shooting with on camera flash as mandated by light in the church and reception hall. A better photographer probably would have dealt more constructively with the challenge.

I would have preferred to have shot the wedding with a Nikon F5 using Tri-X. The family wanted color and I could not find but 2 rolls of color negative film with in miles.

Dave

I'm not a Nikon D5 owner. I made the post above to stimulate discussion believing D5's dynamic range is not deficient but rather consistent with the primary use photographers put the D5 to.

[Jack Dahlgren]

I've read criticism of the linked site and it's findings. I believe the chart linked is good enough for a practical evaluation of various cameras. I believe dynamic range is only one criteria to consider.

From practical experience that I found the Nikon D2H deficient in dynamic range under many situations. In particular I found it near impossible to record shadow detail in a groom's black tuxedo while holding diffused highlights in the bride's white gown. I was shooting a cousin's wedding by request in a photojournalist style in deference to the official photographers. I was shooting with on camera flash as mandated by light in the church and reception hall. A better photographer probably would have dealt more constructively with the challenge.

I would have preferred to have shot the wedding with a Nikon F5 using Tri-X. The family wanted color and I could not find but 2 rolls of color negative film with in miles.

Dave

I'm not a Nikon D5 owner. I made the post above to stimulate discussion believing D5's dynamic range is not deficient but rather consistent with the primary use photographers put the D5 to.

I think that since the D2H is nearly 18 years old, that most people (who could afford a D5) could decide between D2H and D5 without using that site.

As for your experience in that wedding, a minor suggestion... if the groom would have considered perhaps a baby blue tux the contrast would not have been so grave. 

[David H. Hartman]

The D1 in the chart and the D2H which was marginally more capable in dynamic range compared to the D1 shows us where we came from. Yes these cameras are about two decades old. 

My point is the D5 isn't deficient even though it doesn't have quite the dynamic range of the D850 at low ISO(s) and the D850 isn't deficient even though it doesn't have quite the dynamic range of the D5 at higher ISO(s). The D5 and D850 overlap some in their capabilities but each is well suited for their intended purpose.

Dave

I didn't look up the release dates of the D1 and D2H.

If someone has a D5 that they are dissatisfied with please send it to me!   

[Hugh_3170]

Those that like older Dx series sensors, but can't afford such a body, the D700 (tweaked D3 sensor) and the Df (tweaked D4 sensor) are now quite affordable on the second hand market.

[Jack Dahlgren]

Those that like older Dx series sensors, but can't afford such a body, the D700 (tweaked D3 sensor) and the Df (tweaked D4 sensor) are now quite affordable on the second hand market.

Yes, Df is quite capable in low light.

[Hugh_3170]

If I recall correctly, the D700 had the same AF and exposure systems as the D3, so for its day it was a really great value proposition.

The Df has borrowed heavily from the D600/610, albeit without the oily shutter issues - thank god!

Yes, Df is quite capable in low light.

[Ilkka Nissilä]

If I recall correctly, the D700 had the same AF and exposure systems as the D3, so for its day it was a really great value proposition.

Yes and no; when using screwdriver AF lenses, I noticed the D3 focused them much faster than the D700. Marianne Oelund at dpr forums also reported having measured the focusing speeds (I believe with AF-S lenses) and the D3 was also measurably faster than the D700 in her tests.

[Hugh_3170]

Well they both certainly shared the same Multi-CAM 3500FX autofocus sensor module featuring 51 AF points. 

However I do accept that the parameters that were in play for the D3 AF versus those for the D700 may have resulted in faster AF performance for the D3.  Without having checked out the trials you refer to, I do know that the perfomance of the D700 in other areas very much depended on whether or not the handgrip with the additional battery power was fitted.

Yes and no; when using screwdriver AF lenses, I noticed the D3 focused them much faster than the D700. Marianne Oelund at dpr forums also reported having measured the focusing speeds (I believe with AF-S lenses) and the D3 was also measurably faster than the D700 in her tests.

[chambeshi]

Thanks for this suggestion

Does PhotonstoPhotos have a page for SNR vs. luminosity? Is not luminosity another term for Dynamic Range? I looked at the DXOMark site but any analyses are too well hidden to excavate. I also recall this thread here in 2017, just as the D850 had been measured: in which Bill Claff's underscored his explicit methodology he uses to measure sensors

https://nikongear.net/revival/index.php?topic=6649.msg107054#msg107054

Also dynamic range is just one parameter, a full characterisation of the SNR vs. luminosity reveals more about the differences between sensors than just looking at DR.

[Frode]

Yes and no; when using screwdriver AF lenses, I noticed the D3 focused them much faster than the D700. Marianne Oelund at dpr forums also reported having measured the focusing speeds (I believe with AF-S lenses) and the D3 was also measurably faster than the D700 in her tests.

That was my experience also.

Same regarding D5 vsD850 as well.

[Ilkka Nissilä]

Thanks for this suggestion

Does PhotonstoPhotos have a page for SNR vs. luminosity? Is not luminosity another term for Dynamic Range? I looked at the DXOMark site but any analyses are too well hidden to excavate. I also recall this thread here in 2017, just as the D850 had been measured: in which Bill Claff's underscored his explicit methodology he uses to measure sensors

https://nikongear.net/revival/index.php?topic=6649.msg107054#msg107054

Luminosity refers to the luminous power of the light source. That wasn't quite a good choice of word from me, as what I meant is really the pixel brightness in the image or the number of detected photons for each photosite. I'll try to give a more precise explanation. SNR of the signal from each photosite is dependent on how many photons are detected (leading to photon shot noise n_ph = sqrt(eta*N), where eta is the quantum efficiency and N is the number of photons), the thermal and other sources of noise. As the photon shot noise depends on the quantum efficiency and the other noise sources depend on the quality of the analog-to-digital conversion and electronics, different sensors have different SNR vs. number of detected photons. This can be measured for R, G, and B photosites.

DXOMark do publish "Full SNR" graphs; for some reason I'm not able to see them in the mobile version, only when viewing their site in a desktop browser. However, without the numerical data these are not easy to utilize. The comparison between the full SNR across camera models is not directly featured. I'm not sure if it is possible to extract the full SNR data from the site in numerical form. They use "gray value" in the X axis of the full SNR graphs, I'm not sure how they calculate the gray value from R, G, and B.

The protocol they use and the parameters are explained here:

https://www.dxomark.com/dxomark-camera-sensor-testing-protocol-and-scores/
https://www.dxomark.com/glossary/color-depth/

What they do present in an easy-to-compare way are the following:
- SNR for 18% gray (this represents the signal-to-noise ratio of midtones)
- Engineering dynamic range (which is the (log) difference between maximum luminosity that doesn't introduce clipping and the luminosity at which the SNR equals 1)
- Tonal range (essentially the log number of distinct gray tones that can be distinguished from noise)
- Color sensitivity (log number of distinct color values that can be distinguished from noise)

Basically, SNR for 18% tells us how good the SNR is for midtones, dynamic range tells us how wide a range of tones (in terms of image luminance value, I wasn't quite able to see how dxomark calculate this from R, G, and B SNR graphs, is there visually perceived sensitivity weighting of the colors?) can be separated from noise, and it reflects mainly the quality of the deepest shadows or how deep those are. Tonal range tells us how many different gray values can be distinguished from noise and this is dependent on the whole SNR curve. Color sensitivity tells us how many different colors can be separated from noise, again it is dependent on the whole SNR data.

A lot of online discussion revolves around dynamic range but I believe all of these parameters are important to characterise sensor image quality. Dynamic range is imporatant if one photographs a very contrasty scene, such as one where the sun is in the photograph and foreground objects are in shadow or if one underexposes the main subject heavily and requires brightening of the subject in post-processing. In such cases the sensor dynamic range can be, and is, important. However, in a normally-exposed image which doesn't need shadow-lifting, the other image quality parameters can tell us more about the differences in image quality. I guess photographing contrasty scenes is very common and is perceived as a limitation of photography vs. how the human eye and brain see scenes (processing a lot!), and thus the dynamic range gets talked about a lot.