How to make sense of, and compare DxOMark Low Light ISO ratings between bodies

Zarch

Suspended / Banned
Messages
1,005
Name
Mick
Edit My Images
Yes
DxOMark.com said:
Sports Score is based on Low-Light ISO performance (values in ISO index). Low-Light ISO indicates the highest ISO sensitivity to which your camera can be set while maintaining a high quality, low-noise image (based on a Signal-to-Noise-Ratio [SNR] of 30dB, a dynamic range of 9EVs and a color depth of 18bits). As cameras improve, the highest ISO setting to produce 30dB, 9EVs, 18-bit images will continuously increase, making this scale open. Low-Light ISO performance is of primary importance in photojournalism, sports and action photography.

I've seen a lot of talk recently about DxO marks trying to quantify ISO performance between various Nikons models. I'm interested in how my D3100 stacks up against other models with a view to a potential upgrade, in particular the D7000. But don't really know how to read these DXO ISO ratings and how to compare them with regard "real world" performance.

I took a look at my Lightroom catalogue over the weekend and it says that just 15% of my shots are ISO 1600 or 3200 and only another 15% between 800 and 1599 (crazy Auto ISO) :)

So with 70% of my shots at ISO 799 or less, would I really gain much from a D3100 > D7000 upgrade purely on ISO performance?

In real terms, what does D3100 (919) v D7000 (1167) really mean? I suppose I would see better images at 1600/3200 taken with the D7000, but by how much?


Here's the full table to Nikon cameras that I pulled from DXOmark and Snapsort.

3253: D3s
2980: D600
2979: D800E
2965: D4
2853: D800
2303: D700
2290: D3
1992: D3x
--------------- distinct jump in ratings
1284: D5200
1183: D5100
1131: D3200
1167: D7000
977: D90
919: D3100
868: D5000
707: D300s
679: D300
583: D200
563: D3000
562: D60

Is it naive to base image quality comparison's like this solely on these ISO figures? If so, what else do you need to take into account?

I think its an interesting subject and would like people's opinion on it. :D
 
The dxo figure is a pretty vague number and using there figures its half a stop difference between the 2. Don't upgrade if its purely for ISO performance increase as you will likely be disappointed.

To get a much better idea try this studio comparison, where you can adjust the iso and view raw and jpegs and see with your own eyes the difference.
this example is both shot in raw at 1600. You can move the curser around the image to examine different areas. http://www.dpreview.com/reviews/studio-compare#baseDir=%2Freviews_data&cameraDataSubdir=boxshot&indexFileName=boxshotindex.xml&presetsFileName=boxshotpresets.xml&showDescriptions=false&headerTitle=Studio%20scene&headerSubTitle=Standard%20studio%20scene%20comparison&masterCamera=nikon_d3100&masterSample=dsc_0013.acr&slotsCount=4&slot0Camera=nikon_d3100&slot0Sample=dsc_0013.acr&slot0DisableCameraSelection=true&slot0DisableSampleSelection=true&slot0LinkWithMaster=true&slot1Camera=nikon_d7000&slot1Sample=dsc1_1888.acr&x=0&y=0
 
Last edited:
The dxo figure is a pretty vague number and using there figures its half a stop difference between the 2. Don't upgrade if its purely for ISO performance increase as you will likely be disappointed.

To get a much better idea try this studio comparison, where you can adjust the iso and view raw and jpegs and see with your own eyes the difference.
this example is both shot in raw at 1600. You can move the curser around the image to examine different areas. http://www.dpreview.com/reviews/studio-compare#baseDir=%2Freviews_data&cameraDataSubdir=boxshot&indexFileName=boxshotindex.xml&presetsFileName=boxshotpresets.xml&showDescriptions=false&headerTitle=Studio%20scene&headerSubTitle=Standard%20studio%20scene%20comparison&masterCamera=nikon_d3100&masterSample=dsc_0013.acr&slotsCount=4&slot0Camera=nikon_d3100&slot0Sample=dsc_0013.acr&slot0DisableCameraSelection=true&slot0DisableSampleSelection=true&slot0LinkWithMaster=true&slot1Camera=nikon_d7000&slot1Sample=dsc1_1888.acr&x=0&y=0

Basically if the numbers double that's an improvement of one stop, if you are looking for an improvement in ISO then anything less than 1 stop will probably leave you disappointed

So in reality, there's very little difference in ISO performance between any lower end Nikon released in the last 2 years? (D3100 > D5200)

You'd assume the D7100 would top the 'lesser' list next month though?
 
So in reality, there's very little difference in ISO performance between any lower end Nikon released in the last 2 years? (D3100 > D5200)


You'd assume the D7100 would top the 'lesser' list next month though?

There is very little difference in any aps-c sensor lower end or higher end in the last 2yr.

Yeah should do if it has no low pass filter. This is said to gain you up to a stop improvement alone.

I doubt you will see the type of gains in iso performance we have seen in the last 5yrs again with current tech. This is why both the mp and iso war is over and the FF marketing has begun. :)
 
Last edited:
I stand to be corrected but I think the easiest way to describe the result is that is the figure you will expect to see noticeable noise. As others have pointed out, up less it's a stop difference AT LEAST, you'll be looking and pixel peeping to see the difference.

Obviously the jumps come when FF is compared to but remember, back in it's day the D200 was supposedly good for ISO noise - no doubt in a few years time there could be another stop difference in the future current DX bodies. We could see the same 1 stop difference in a few years but also will most likely be the same difference in FF bodies.

Anyway, unless you're on a really old model (which you aren't) it's not worth upgrading just for ISO: spend your money on faster, sharper lenses. This is the reason I need my lenses to be pin sharp wide open, because I often shoot at the highest end of my ISO and you can always get away with it and clean it up IF you have a very sharp lens.
 
Last edited:
The headline DxO mark numbers really don't tell you the whole picture, a much more useful tool thats easier to quantify is their sensor comparisons , here it is with the D3100 and D7000 loaded in...

http://www.dxomark.com/index.php/Ca...brand)/Nikon/(appareil2)/680|0/(brand2)/Nikon

Click on the measurements area and you'll have the option of picking several graphs that test various different aspects agenst each other.

SNR 18% measures the noise performance at various ISO's, the higher on the verticle axis the lower the noise at a chosen ISO. The D7000 looks to me to be just under a stop better(so ISO 200 on the D7000 will look similar to ISO 100 on the D3100) than the D3100 at low ISO dropping to more like half a stop at around ISO 800 then back to closer to a stop as you go higher.

Dynamic Range(DR) is the ability of the sensor to record a strong contrast at different ISO's without noise and loss of detail becoming an issue and is measured in stops on the verticle axis. This is really where the D7000 sensor offers a big upgrade to the D3100, at ISO 100 the D7000 is offering almost 3 stops more DR but this advanatage declines as you incrase ISO, by ISO 800 there pretty similar.

If pure noise performance in non dymatic lighting situations is what your after I'd say you'd be best of saving for a used D700 rather than a D7000, you'll see a much more noticble difference there compaired to your D3100.
 
Last edited:
Remember that is it not just differences in sensor, but also other very important parameters, such as AF, fps, metering, build quality - which is where D7000 would trash D3100 under more demanding conditions. If you were to compare the sensors alone - these numbers in a linear scale pretty well suggest that D7000 would give a slightly cleaner image, but nothing groundbraking. A D600 would make much bigger difference. You pay your money and you choose the features.
 
Devil's advocate time. Certainly not directed at anyone in particular, just a general observation...

If you truly want fantastic low light performance Nikon DX body then I know a secret that could save you all a grand or even more.... Buy an SB700. Want an even cheaper method? Use a cheap tripod!

Seriously, there's an awful lot of hypothetical low light situations on forums where one iota of grain is completely unacceptable right through to the (understandably enormous, probably large format) final print, flash photography is absolutely disallowed and we need to shoot high speed unpredictable action. If that is genuinely your photographic life, then you probably already shoot full frame and own fast glass, because you probably realised long ago that you're pushing the boundaries of available light photography (and probably pouring out for each new gen of camera on the nose as the sensor tech improves).

If not, then a little CLS or pocket wizardry alongside an acceptance of some grain could bring your ISO's right down. I'd just like to know how much of this is middle-aged science application (and as a middle-aged scientist I'd truly understand), or whether perhaps it's just a little gearlust looking for justification to go out and spend the slush fund ( and as a confirmed slush fund spender I'd empathise a lot more than my wife would!).

I was swayed by the allure of ISO madness, too - bought a D700. It was good, but not night-and-day great if I'm honest. Nowadays, if there is a scientific application of photography I'm still interested in, it's dynamic range. That's best at base ISO and there's nothing like lighting the scene to get there.
 
Last edited:
Devil's advocate time. Certainly not directed at anyone in particular, just a general observation...

If you truly want fantastic low light performance Nikon DX body then I know a secret that could save you all a grand or even more.... Buy an SB700. Want an even cheaper method? Use a cheap tripod!

Seriously, there's an awful lot of hypothetical low light situations on forums where one iota of grain is completely unacceptable right through to the (understandably enormous, probably large format) final print, flash photography is absolutely disallowed and we need to shoot high speed unpredictable action. If that is genuinely your photographic life, then you probably already shoot full frame and own fast glass, because you probably realised long ago that you're pushing the boundaries of available light photography (and probably pouring out for each new gen of camera on the nose as the sensor tech improves).

If not, then a little CLS or pocket wizardry alongside an acceptance of some grain could bring your ISO's right down. I'd just like to know how much of this is middle-aged science application (and as a middle-aged scientist I'd truly understand), or whether perhaps it's just a little gearlust looking for justification to go out and spend the slush fund ( and as a confirmed slush fund spender I'd empathise a lot more than my wife would!).

I was swayed by the allure of ISO madness, too - bought a D700. It was good, but not night-and-day great if I'm honest. Nowadays, if there is a scientific application of photography I'm still interested in, it's dynamic range. That's best at base ISO and there's nothing like lighting the scene to get there.

One of the most refreshing posts I've read on here in ages :clap:
 
Devil's advocate time. Certainly not directed at anyone in particular, just a general observation...

If you truly want fantastic low light performance Nikon DX body then I know a secret that could save you all a grand or even more.... Buy an SB700. Want an even cheaper method? Use a cheap tripod!

Seriously, there's an awful lot of hypothetical low light situations on forums where one iota of grain is completely unacceptable right through to the (understandably enormous, probably large format) final print, flash photography is absolutely disallowed and we need to shoot high speed unpredictable action. If that is genuinely your photographic life, then you probably already shoot full frame and own fast glass, because you probably realised long ago that you're pushing the boundaries of available light photography (and probably pouring out for each new gen of camera on the nose as the sensor tech improves).

If not, then a little CLS or pocket wizardry alongside an acceptance of some grain could bring your ISO's right down. I'd just like to know how much of this is middle-aged science application (and as a middle-aged scientist I'd truly understand), or whether perhaps it's just a little gearlust looking for justification to go out and spend the slush fund ( and as a confirmed slush fund spender I'd empathise a lot more than my wife would!).

I was swayed by the allure of ISO madness, too - bought a D700. It was good, but not night-and-day great if I'm honest. Nowadays, if there is a scientific application of photography I'm still interested in, it's dynamic range. That's best at base ISO and there's nothing like lighting the scene to get there.

Dan,

Thanks for the fantastic response. :clap:

"I'd just like to know how much of this is middle-aged science application , or whether perhaps it's just a little gearlust looking for justification to go out and spend the slush fund"

LOL, I had to giggle. I do love to try and justify these upgrades in my own mind, be it phones, lenses or any other tech kit. In the end, I 99% of the time cave in and get the new gear.

I actually already own a SB-700 flash and i'll openly admit, I don't use it enough because I don't fully understand how to use it properly.

Its on my to-do list to make the flash more discreet than the "BANG, LIGHT" way I seem to be using it at the moment. :eek:

I think I appreciate that when it comes to image quality, especially ISO performance that i'm not going to see night/day improvements if (probably when) I upgrade my D3100 to a D7000.

But after two years of owning my D3100 (my first dSLR) i'm getting the stage where I 'think' I need all controls to hand rather than having to dig into the menus. Even though I can't really recall many times where i'm digging in the menu. :|

So I'm still finding it hard to 'justify' the £300 it will cost to go from D3100 to D7000 as there isn't really anything wrong with it.

I just see the D7000 as my long term camera. If I get that, then I don't see any need to go any higher for my amateur needs. Famous last words I know. :lol:
 
Devil's advocate time. Certainly not directed at anyone in particular, just a general observation...

If you truly want fantastic low light performance Nikon DX body then I know a secret that could save you all a grand or even more.... Buy an SB700. Want an even cheaper method? Use a cheap tripod!

Seriously, there's an awful lot of hypothetical low light situations on forums where one iota of grain is completely unacceptable right through to the (understandably enormous, probably large format) final print, flash photography is absolutely disallowed and we need to shoot high speed unpredictable action. If that is genuinely your photographic life, then you probably already shoot full frame and own fast glass, because you probably realised long ago that you're pushing the boundaries of available light photography (and probably pouring out for each new gen of camera on the nose as the sensor tech improves).


Neither flash or a tripod is going to be any use if you're shooting a gig. Or an event, like a wedding, where you're told you cannot use flash, and a tripod is useless when there's movement.

I'm sure anyone buying a higher end body knows how to use CLS/tripod when needed, but they may well need good ISO performance on top. And why not?
 
Neither flash or a tripod is going to be any use if you're shooting a gig. Or an event, like a wedding, where you're told you cannot use flash, and a tripod is useless when there's movement.

I'm sure anyone buying a higher end body knows how to use CLS/tripod when needed, but they may well need good ISO performance on top. And why not?

The second bit of Dan's post that you quoted says exactly that. :)
 
Yes it does! If you're serious enough that you shoot regular gigs with an SLR (rather than an iPhone video, like most joe public would) then you'll go with a fast, long lens like an 85/1.4 and a high ISO monster like the D3S. You're exactly the sort who is 'pushing the boundaries of available light photography'.

However, if you're looking to buy a D3S and an 85/1.4 because you've got an evening out coming up, then you're doing things exactly backwards.
 
I have a Nikon D700 , was considering changing to the D600 for the higher ISO Performance but from looking on amazon and other reviews there seems to be a few issues with quality control so i may give it a miss for the moment
 
Back
Top