Low Light ISO Setting D750

NikonShutterBug

Senior Member
[FONT=Arial, sans-serif]Good afternoon Nikonites. My D750 came in Friday. I after making all basic recommended settings I decided to do an ISO test, D750 and D7100. Both cameras using the same Nikon 20mm 1.8g lens, same distance from subject using a tripod and same light source. All fairly equal other than the sensors. [/FONT]


[FONT=Arial, sans-serif]
ISO Shutter Speed f5.6
[/FONT]

[FONT=Arial, sans-serif]
D750 D7100
[/FONT]

[FONT=Arial, sans-serif]100 2.5 [/FONT]

[FONT=Arial, sans-serif]1000 25 30[/FONT]

[FONT=Arial, sans-serif]2000 50 60[/FONT]

[FONT=Arial, sans-serif]3200 80 100[/FONT]

[FONT=Arial, sans-serif]4000 100 125[/FONT]

[FONT=Arial, sans-serif]5000 125 160[/FONT]

[FONT=Arial, sans-serif]6400 160 200[/FONT]





[FONT=Arial, sans-serif]My first thought was, this is backwards??? Shouldn't the D750 have the faster shutter speeds at the same ISO, do to the better sensor? [/FONT]


[FONT=Arial, sans-serif]Second, the D7100 images got grainy in the shadows at about ISO 2000. The D750 at about ISO 5000. I had hoped the images of the d750 would get grainy at a higher ISO level than ISO 5000. So, I seem to have gained about twice the shutter speed in the same light, [/FONT][FONT=Arial, sans-serif]over the D7100[/FONT]



[FONT=Arial, sans-serif]Am I comprehending the test results correctly?[/FONT][FONT=Arial, sans-serif] One big reason I bought the D750 was to shoot in low light situations.[/FONT][FONT=Arial, sans-serif] What ISO's are you Nikonites using in [/FONT][FONT=Arial, sans-serif]hand held[/FONT][FONT=Arial, sans-serif] low light situations, such as a Night Club scene, or Night Street Photography. Thanks[/FONT]
 

NikonShutterBug

Senior Member
I seems the chart did not come out correctly. I'll try again.

ISO Shutter Speed f5.6

D750 D7100
1000 25 30
2000 50 60
3200 80 100
4000 100 125
5000 125 160
6400 160 200
 

Moab Man

Senior Member
The D7100 is grainy well before ISO 2000. A better evaluation is the usability of the image for what intended purpose. I shoot the D7100 and D750 (amongst others) for sports photography. At night the D750 ISO is usable up to and sometimes past ISO 10,000. The D7100 is retired before night due to grain.

However, using your analysis, a 150% increase is quite substantial.

Not a hostile response by me, but I see how it might be construed that way so please don't take it that way.
 

WayneF

Senior Member
I seems the chart did not come out correctly. I'll try again.

ISO Shutter Speed f5.6

D750 D7100
1000 25 30
2000 50 60
3200 80 100
4000 100 125
5000 125 160
6400 160 200

I don't think there is any evidence of any problem.

Your chart is showing a constant 1/3 stop difference in metering, regardless of ISO (i.e., it is instead a metering issue). If all is the same, technically the two cameras should show the same settings.

However, all is Not the same. With the same 20mm lens on the FX and DX cameras, the view being metered is simply different, the FX sees a view 1.5x wider than the DX sees. That different view would normally be expected to affect metering a little, due to simply metering different areas, and different things in those areas. I think it is is a very normal result. And other little differences could also affect it too, possibly like different Matrix or Center Weighted metering on the two cameras.
 
A lot of shooting in low light with the D750 is getting the exposure correct to begin with I shoot super high ISO as needed to get the shot I want and I do my post processing to take care of any noise I might have. This photo is shot with the NIKON D750 300 mm 1-1000 sec at f - 5.6 ISO 51200

03-18-16_0413 NIKON D750 300 mm 1-1000 sec at f - 5.6 ISO 51200 - Copy.jpg
 

TieuNgao

Senior Member
Talking about low light and high ISO, I have a little experiment to share with you:
I've taken the same picture at different ISO 6400, 3200, and 1600. Manual mode (S=1/40, A=7.1), on tripod and the same frame.
At ISO6400 the camera meter showed the "correct" exposure, and obviously under-exposure for the other two ISO values.
I used DxO OpticsPro 10 to automatically remove the noise for all three pictures, and added 1 and 2 EV for the pictures with ISO3200 and 1600, respectively, before converting to jpg.
As far as my eyes tell me they look the same! Please take a look at these 3 pictures.
_DVT3663_6400.jpg

_DVT3664_3200.jpg

_DVT3665_1600.jpg
 
Last edited by a moderator:

Moab Man

Senior Member
I wouldn't try to make a comparison based on a jpg, from a low resolution web image. Not a jab at you, but jpg drops a lot of data and the internet doesn't display at high resolution.
 
I shoot discos every week ..I get all the room lights off and use the DJs lights ,stand with your back to the lights (if they are not good I got my own set of battery powered lazers/led lights ) My camera is set to A mode min 1/60 auto iso 200-6400 f 8 and I just turn it to P mode which cancels the F8 and runs the lens wide open ( 3.5-5.6) .That's it ..no problems D 7100 with 18-140 and D810 with Samyang 14mm.
If you are talking street scenes with trailing car lights etc then base it on iso 100 F8 and 1 second and variants there of
 
Last edited:

Blade Canyon

Senior Member
I've taken the same picture at different ISO 6400, 3200, and 1600. Manual mode (S=1/40, A=7.1), on tripod and the same frame.
...and added 1 and 2 EV for the pictures with ISO3200 and 1600, respectively, before converting to jpg.
As far as my eyes tell me they look the same!

And they should look the same. The answer is in your post.

The amount of light hitting the sensor is exactly the same in all three shots. Same shutter speed, same aperture. By adjusting the camera's ISO, all you're doing is telling the camera what the final shot should look like.

If you take the RAW 1600 file and turn up exposure two stops in Lightroom, all you're doing is turning up the ISO to 6400, except you are doing that step in post processing. With the exact same light information as a picture shot at 6400 in camera, you should get the same result.

Paul says he sees differences, but I suspect they have more to do with your auto noise reduction.

J-See could explain it better. I did these exact same tests as you when he explained that he set his shutter speed and aperture where they had to be to catch birds in flight, then he did all of his ISO adjustment in post.
 

Blade Canyon

Senior Member
My first thought was, this is backwards??? Shouldn't the D750 have the faster shutter speeds at the same ISO, due to the better sensor?

They should have exactly the same shutter speed at the same ISO. The better sensor doesn't change the exposure. The better sensor will give you more light catching and/or processing abilities in low light (meaning you're using a higher ISO), exactly as you discovered. In fact, if all cameras are properly calibrated, all cameras and light meters should meter a constant scene the same. F16 and shutter speed of 1/100 should be the norm shooting ISO 100 in bright sunlight.

I think your differences in shutter speeds between the two bodies was explained by WayneF.

Hope you are loving that D750! Keep up the experiments.
 

TieuNgao

Senior Member
And they should look the same. The answer is in your post.

The amount of light hitting the sensor is exactly the same in all three shots. Same shutter speed, same aperture. By adjusting the camera's ISO, all you're doing is telling the camera what the final shot should look like.

If you take the RAW 1600 file and turn up exposure two stops in Lightroom, all you're doing is turning up the ISO to 6400, except you are doing that step in post processing. With the exact same light information as a picture shot at 6400 in camera, you should get the same result.

Paul says he sees differences, but I suspect they have more to do with your auto noise reduction.

J-See could explain it better. I did these exact same tests as you when he explained that he set his shutter speed and aperture where they had to be to catch birds in flight, then he did all of his ISO adjustment in post.

My point is that if they look the same (they do to my eyes) then I'd rather do it in the post; meaning I'd shoot at 1 or 2 EV under-exposure by using lower ISO to have, in theory, better DR, tonal range, and color sensitivity. Another bonus by doing that is less chance to blow the highlights.
 

Blade Canyon

Senior Member
My point is that if they look the same (they do to my eyes) then I'd rather do it in the post; meaning I'd shoot at 1 or 2 EV under-exposure by using lower ISO to have, in theory, better DR, tonal range, and color sensitivity. Another bonus by doing that is less chance to blow the highlights.

This highly technical article suggests your image quality will be better if you put the ISO close to where it should be, without clipping.

theory.uchicago.edu/~ejm/pix/20d/tests/noise/

And here's the most relevant quotation from that article:

"Bottom line: Read noise at high ISO is much smaller than read noise at low ISO, in terms of the error in photon counting that it represents. Thus, better image quality is obtained for using the highest ISO for which the signal is not clipped."




 

TieuNgao

Senior Member


This highly technical article suggests your image quality will be better if you put the ISO close to where it should be, without clipping.

theory.uchicago.edu/~ejm/pix/20d/tests/noise/

And here's the most relevant quotation from that article:

"Bottom line: Read noise at high ISO is much smaller than read noise at low ISO, in terms of the error in photon counting that it represents. Thus, better image quality is obtained for using the highest ISO for which the signal is not clipped."


Just below the "Bottom line" you quoted from that article, you'll see the graph 15a for Canon 1D3 where it clearly shows that at ISO 800-1300 the read noise is almost constant, and from ISO 1600 and beyond the read noise stays the same. That's why the next paragraph says:

"Note that the expansion on the bottom end of the range yields less and less as ISO is increased more and more -- going from ISO 100 to ISO 200 yields a big improvement at the lower end of exposure (at the cost of some latitude at the upper end); on the other hand, going from ISO 800 to 1600 doesn't make much difference at all in shadow S/N, and in addition one loses an entire stop of raw headroom. Above ISO 1600 there is no expansion of the shadow range whatsoever, just more and more lost from the top end. The ISO 3200 curve of the 1D3 isn't even plotted above, because it lies almost exactly on top of the ISO 1600 curve, apart from ending one stop earlier ot the top end. This is why it makes no sense to use absurdly high ISO's like 6400 if one uses raw capture -- it just throws away highlight headroom without getting anything back at the shadow end; it's better to underexpose by a stop or two at ISO 1600 if the shutter speed is needed, than to use higher ISO."
 

NikonShutterBug

Senior Member
Thanks guys for the education, I really enjoyed all of the comments. I am absolutely enjoying the D750. Wow what an amazing camera, and with proper direction from the Nikonites I am starting to create some nice photography. Nikonites, thanks so much for sharing.
 
Top