D500 ISO question

Joined
Jul 25, 2017
Messages
956
Location
Pittsburgh, PA
Real Name
Ian
I don't understand why there's so much discussion around the D500 vs. FF camera bodies. The D500 is a 20MP APS-C sensor, of course it's going to be noisier than a 16MP or 20MP FF sensor. It's not until you get to around 45MP on a FF camera (cough cough, D850 cough cough) where the noise performance between the two is about equal.

People who think the D500 is cleaner than any modern Nikon FF camera are smoking crack. It's not. But, sometimes there are tradeoffs that cause the D500 to be used. AF system, size/weight, 1.5x reach, etc..

I like my D500's because they have fantastic AF, and give me 1.5x more reach, and I tend to shoot in decent light. I don't have the money for a 100-300 f/2.8 and 500/4, but I do have the money for a 70-200 f/2.8 and 300/2.8, which gives me pretty much the same reach.

By all means, if you shoot in sh|t light and can afford to spend $5.5k on a D5, $10.5k on a 500/4, and $2.5k on a 120-300/2.8, be my guest. Will you get cleaner images? Yes. Will you have shallower DoF? Yep. But you'll also be $18.5k poorer, and that's money that many people can't afford to spend on photography equipment, especially if it's something where you're not making any money.

The D500 isn't as good as the D4, D4s, or D5 when it comes to high ISO performance. But it's by far the best APS-C camera ever made, and I love shooting with mine.
 
Joined
Feb 18, 2015
Messages
1,104
Location
New York State
Because of this thread, I went back and read the original B.Claff thread from 2016 on DPR.

I noticed something which had escaped me previously: namely that the original images from which Claff compiled his data were all shot with the camera set for -0.33 EV.

Many of the sample photographs (possibly taken on a pre-release D5) were shots of nocturnal exteriors and dark-wood furnished interiors.

Because of the Nikon meters' tendency to under-expose, coupled with the minus EV settings which were actually employed, those sample photographs were underexposed by a full EV.

That would surely have skewed the PDR values which were obtained from those images and were posted in the charts; and might explain why my real-life experience with the D5 has been so different from what those charts seem to indicate?
 
Last edited:
Joined
May 5, 2005
Messages
21,659
Location
SW Virginia
Because of this thread, I went back and read the original B.Claff thread from 2016 on DPR.

I noticed something which had escaped me previously: namely that the original images from which Claff compiled his data were all shot with the camera set for -0.33 EV.

Many of the sample photographs (possibly taken on a pre-release D5) were shots of nocturnal exteriors and dark-wood furnished interiors.

Because of the Nikon meters' tendency to under-expose, coupled with the minus EV settings which were actually employed, those sample photographs were underexposed by a full EV.

That would surely have skewed the PDR values which were obtained from those images and were posted in the charts; and might explain why my real-life experience with the D5 has been so different from what those charts seem to indicate?
I think Bill Claff knows what he's doing.
 
Joined
Feb 18, 2015
Messages
1,104
Location
New York State
He may do but the photographer who provided the RAW nefs (on which Claff based his data) shot them on a pre-release D5 and, apparently, underexposed the photographs as well.

All that I know is that the charts do not replicate the way in which a D5 images in real life when proper exposures are employed.

That is why I pay scant attention to those kinds of charts and had already pre-ordered my D5 (based purely on Nikon's published specifications!) long before those charts were published.

My D5 was delivered, from Nikon's first released shipment, about one week after the charts were published. This camera has performed more wonderfully than I had ever imagined.
 
Joined
Jan 2, 2009
Messages
4,750
Location
Alaska
I don't understand why there's so much discussion around the D500 vs. FF camera bodies....
Simple. If we discuss religion or politics the mods will shut the thread down. We can get away with discussing to no end any irrational opinions we have about things photography related :D
... it's by far the best APS-C camera ever made...
Now who's smoking crack? In sunny 16 lighting IQ from the D500 won't touch a D7200 ;)

And don't even get me started on the VF blackout time... :mad:
 
Joined
Jan 2, 2009
Messages
4,750
Location
Alaska
By the way, Photographic Dynamic Range(PDR) is not the same as the Dynamic Range(DR) that is typically discussed relative to sensor performance. When this was first posted I didn't notice the P ahead of the DR. What I gather from the Photons to Photos site, PDR is a parameter that was created in an attempt to come up with a standard measure of IQ at various ISO settings and a S/N ratio of 20. However it is based on certain assumptions one of which is that an uncropped image is printed at 8x10 to be viewed at "arms length". So as soon as one starts up/down sampling, cropping, using different formats, etc, this falls apart. To their credit, for some of the full frame sensors they also produce curves for images shot in DX crop mode.

IMO we're typically concerned about noise in low lighting/high ISO situations. Under those circumstances the dominant source of noise (with 1s or shorter ss) is read noise. The curves indicating relative performance of read noise are hard data without all of the various assumptions involved.
 
Last edited:
Joined
May 5, 2005
Messages
21,659
Location
SW Virginia
IMO we're typically concerned about noise in low lighting/high ISO situations. Under those circumstances the dominant source of noise (with 1s or shorter ss) is read noise. The curves indicating relative performance of read noise are hard data without all of the various assumptions involved.
My only problem with that is that image quality is more dependent on signal-to-noise ratio than on absolute noise. Comparing only noise levels between a full-frame camera and a crop sensor camera would not be meaningful since the bigger pixels of the full frame would collect more light and produce a higher signal-to-noise ratio.
 
Joined
Oct 9, 2005
Messages
23,850
Location
Moscow, Idaho
At the risk of sounding like a broken record or a shill, here is how Steve Perry sees the PDR vs EDR (DXO #) issue:
Subscribe to see EXIF info for this image (if available)
 
Joined
May 5, 2005
Messages
21,659
Location
SW Virginia
At the risk of sounding like a broken record or a shill, here is how Steve Perry sees the PDR vs EDR (DXO #) issue:
View attachment 1632227
I'm still working my way through Steve's book, but I think I would agree with that. All we're really looking for is a way to compare cameras that is consistent across formats, and I think that is what Bill Claff's PDR is. Based on my experience with quite a few Nikon cameras (and one Olympus) his PDR is quite consistent with my experience.
 
Joined
Oct 9, 2005
Messages
23,850
Location
Moscow, Idaho
^^ That's where I'm at too. Besides wallowing through Bill's site, reading Steve's book, and now (thanks to you) I've embarked on a marathon session of Marc Levoy's lectures!!!
I attempted to watch Marc back when his lectures were first released to the public, but I was still working and had little free time and my brain was full. Now, I have time and room for new ideas--even if I don't quite understand it all. But boy, is he ever a good teacher!
 
Joined
Jan 2, 2009
Messages
4,750
Location
Alaska
My only problem with that is that image quality is more dependent on signal-to-noise ratio than on absolute noise. Comparing only noise levels between a full-frame camera and a crop sensor camera would not be meaningful since the bigger pixels of the full frame would collect more light and produce a higher signal-to-noise ratio.
Noise measurements indicated in DN(digital number) are realized noise levels. Noise is noise. Comparing a FF and APS-C sensor with the same pixel count and same noise reading will produce identical images.

On the other hand, at low light levels where shot noise is not dominant, given a scene with a given level of light a sensor with larger pixels(regardless of format) should have a better S/N ratio and therefore record less noise(measured in DN). IMO S/N ratio is the best information to compare across the entire range of EV. I'm not sure why Claff chooses to plot PDR and read noise but not plain S/N ratio. Presumably they came up with PDR for the broader audience of users. IMO in low light situations the DR capabilities of the sensor are way over rated. DR in low light scenes is very small/narrow and a sensor capable of recording more DR is not highly relevant.

At the risk of sounding like a broken record or a shill, here is how Steve Perry sees the PDR vs EDR (DXO #) issue:
View attachment 1632227
Given those two parameters to compare, I completely agree(see comment above re.DR).
 
Joined
Jan 2, 2009
Messages
4,750
Location
Alaska
But directly under Claff's plots of "Read Noise in DNs versus ISO Setting" he says:

These raw values are not appropriate for comparing camera models because they are not adjusted for gain or area.
He makes those comments for the same reason that he came up with PDR. It does get complicated when you try to compare cameras with significantly different pixel counts or FF vs APS-C. Read the material on where the read noise numbers come from and make an informed decision. We each make decisions based on our own understanding of available information relative to how/what we shoot.
 
Joined
Jun 1, 2009
Messages
7,449
Location
Texas (KSKF)
OTOH,
Not everything is about IQ, charts, and graphs,....For example, I could NEVER own another camera that did NOT have the joystick that the D500 and Z series have....

I don't care what all the "numbers" say....
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2005-2019 Amin Forums, LLC
Top Bottom