D600-D800 high ISO RAWs compared

Started Sep 15, 2012 | Discussions
Mato34 Contributing Member • Posts: 920
D600-D800 high ISO RAWs compared

(Edit: You can see original screenshots in my gallery better than in this post. Seems to be some kind of processing when posting here, kind of resizing or reduction in JPG quality or something like that...).

Downloaded from Focus Numerique:

http://www.focus-numerique.com/test-1390/reflex-nikon-d800-bruit-electronique-12.html

http://www.focus-numerique.com/nikon-d600-jpeg-raw-toutes-sensibilites-news-3605.html

Processed using Raw Therapee, using same parameters (well, had to give an slight +20 on saturation on the D600). Then applied some contrast on Photoshop using curves and resized D800's picture to 24 MP.

I show three 100% crops comparison from the ISO 6400 (D600's shot is exposed 1/3 EV less than the one from D800 but exposure level in the images is almost the same, perhaps due to different lens used):

I can't see any meaningful difference between. The same for ISO 3200 and ISO 12800 (which I compared also). A good result regarding noise to me, given the D800 is at the same level than the D3s below ISO 12800 when resizing.

Regarding noise in the shadows at high ISO, I've tested it using ISO 1600 picture pulling up +4 EV in Raw Therapee. Looking at some shadows:

The D600 seems to perform equal, perhaps slightly better in second screenshot but nothing to write about I think. Anyway I won't rely too much on this: exposure parameters are the same on both cameras but exposure levels are very slightly darker on the D800.

Not to mention the caution we all must have regarding these first tests and comparisons we are seeing.

But... right now I think that the D600 it's up to the D800 regarding high ISO performance.

Saludos!

-- hide signature --

Please, excuse my poor english...

 Mato34's gear list:Mato34's gear list
Nikon 1 J5
Nikon D3S Nikon D600 Nikon D800
If you believe there are incorrect tags, please send us this post using our feedback form.
panos_m Senior Member • Posts: 1,350
Re: D600-D800 high ISO RAWs compared

Thank you for putting this up!. I guess the need for more saturation for the D600 comes for the missing color profile in RT.
--
Panagiotis

HSway
HSway Veteran Member • Posts: 3,147
Re: D600-D800 high ISO RAWs compared

Details (and that will be argued here most of the time) will depend on exact setting. Screen and print results will bring another variable.

-- hide signature --
rhlpetrus Forum Pro • Posts: 24,860
Excellent!

Confirsm the expectations about the Sony sensors, about same, maybe a slight advantage in favor of D600, but nothing that can be seen in a print or at regular internet sizes.

Kudos to Nikon/Sony!
--
Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon 1 V1 Nikon D7000 Nikon 1 Nikkor 10mm f/2.8 Nikon AF Nikkor 35mm f/2D +4 more
coudet Veteran Member • Posts: 3,974
Thanks! Re: D600-D800 high ISO RAWs compared

Mato34 wrote:

But... right now I think that the D600 it's up to the D800 regarding high ISO performance.

I did the same comparison yesterday (dcraw + ISO 25,600 files) and reached the same conclusion. Though, there's a difference in exposure and also since D600 isn't yet supported, color is off..

rhlpetrus Forum Pro • Posts: 24,860
Base ISO shadow recovery

Do they have base ISO images? That's where the shadows recovery matters the most.

This is looking very promising!
--
Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon 1 V1 Nikon D7000 Nikon 1 Nikkor 10mm f/2.8 Nikon AF Nikkor 35mm f/2D +4 more
SergeyGreen
SergeyGreen Contributing Member • Posts: 582
This is getting suspicious :)

I do not see any difference in detail either, surprise.

-- hide signature --

-sergey

clarnibass Senior Member • Posts: 1,987
Re: D600-D800 high ISO RAWs compared

Thanks. Not that it necessarily means anything, but wondering how come especially in the first photo the D600 is sharper (e.g. the coke bottle)?

rhlpetrus Forum Pro • Posts: 24,860
Re: This is getting suspicious :)

SergeyGreen wrote:

I do not see any difference in detail either, surprise.

The D800's were downsized, otherwise you'd have the impression of more noise from it.

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon 1 V1 Nikon D7000 Nikon 1 Nikkor 10mm f/2.8 Nikon AF Nikkor 35mm f/2D +4 more
rhlpetrus Forum Pro • Posts: 24,860
Downres effect

Downres'ing has an impact on detial, due to pixel merging. You need to do a little sharpening, which would increase perception of noise though.

clarnibass wrote:

Thanks. Not that it necessarily means anything, but wondering how come especially in the first photo the D600 is sharper (e.g. the coke bottle)?

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon 1 V1 Nikon D7000 Nikon 1 Nikkor 10mm f/2.8 Nikon AF Nikkor 35mm f/2D +4 more
clarnibass Senior Member • Posts: 1,987
Re: Downres effect

rhlpetrus wrote:

Downres'ing has an impact on detial, due to pixel merging. You need to do a little sharpening, which would increase perception of noise though.

I guess that's possible, but I don't know... I checked that before and the reduced photos looked pretty much as sharp or even sharper, without any sharpening after reducing the size. The difference here looks too big for that IMO but I am not sure. Maybe a difference in DOF from using a different aperture, or maybe slightly different focus? Just guessing since I have no idea what lenses/apertures they used. In addition the eletronic board at the bottom looks pretty much identical which is another reason I don't think it is the size reduction.

SergeyGreen
SergeyGreen Contributing Member • Posts: 582
You are right

rhlpetrus wrote:

The D800's were downsized, otherwise you'd have the impression of more noise from it.

In the second image, it is the right frame that should be larger in frame to frame comparison (if the downsizing to be equal), but the noise looks almost the same as it is in both. So downsizing could also explain why the detail looks almost equal as well, it is simply hidden.

-- hide signature --

-sergey

dethis2 Regular Member • Posts: 301
Re: D600-D800 high ISO RAWs compared

Marcos, thanks allot.

As PetrosM wrote the input color profile is missing from RT at the moment.

Please try this profile https://rapidshare.com/files/2253876622/NIKON_d600_VFA00050_crop_GM_WP11.icm .

Take a look at this thread for updates ..
http://rawtherapee.com/forum/posting.php?mode=quote&f=2&p=30130

Jack Hogan Veteran Member • Posts: 6,206
Re: Thanks! Re: D600-D800 high ISO RAWs compared

coudet wrote:

Mato34 wrote:

But... right now I think that the D600 it's up to the D800 regarding high ISO performance.

I did the same comparison yesterday (dcraw + ISO 25,600 files) and reached the same conclusion. Though, there's a difference in exposure and also since D600 isn't yet supported, color is off..

Many differences: lenses, Exposure, WB, lighting, angle, etc. The images all open properly iin VNX2 and CNX2 so all the details are there - profiles should not be an issue here. As a result of the differences the D800 image is a lot brighter in the shadows, which makes direct comparison a little tough. But I tend to agree, the comparison comes out well for the D600.

rhlpetrus Forum Pro • Posts: 24,860
Re: Downres effect

Yes, one cannot use these samples, or DPR's, for detail analysis, they often change aperturs and also plane of focus (or they forget to AF FT the lenses used). The same is true for lighting, so I'd take these samples only as rough indicators of performance. But they tend to follow other more accurate analysis.
--
Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon 1 V1 Nikon D7000 Nikon 1 Nikkor 10mm f/2.8 Nikon AF Nikkor 35mm f/2D +4 more
u007 Senior Member • Posts: 1,681
Re: D600-D800 high ISO RAWs compared

Nice comparison, but why don't you apply some noise reduction?

Lightroom and most other software applies colour noise reduction by default with pretty much zero side-effects, so you might as well do that. Obviously it's the final image which matters, and some sensors "clean up" better than others because of the noise patterns they make. Would be worth a try to see what the best image you can get from each sensor looks like!

-- hide signature --

My travel photography blog - http://www.frescoglobe.com

inasir1971
inasir1971 Veteran Member • Posts: 3,611
Why not compare by downsampling them both to the same size

here are the two files downsampled to D700 resolution using VNX. My CNX won't open it, but the settings that are active are vignette (normal), Hish ISO NR (normal) which are the same for both.

By downsampling both in the same program and with the same settings - they both have the exact same output sharpening so neither has an advantage. Small pixel vs smaller pixel I guess.

  • the 1/3rd EV exposure difference is probably partially explained by the use of the 50 1.4G with the D800 and the 50 1.8D with the D600 - the lenses may have different light transmission levels.

  • the D800 shows better detail - see for example the printing on the circuit board, or the contour lines on the map.

  • disturbingly the D600 shows more of the magenta cast than the D800 also suffers from at high ISO (what limits the D800 is this amp glow/sensor noise) - see the grey color patch. This is consistent with other sample images on the net.

  • the ISO 6400 performance of any camera in this sort of setup (with bright lighting) is very different to real shooting where high ISO is used mostly because of low light levels

Click (original) to download.

 inasir1971's gear list:inasir1971's gear list
Sony RX1R II Nikon D4 Nikon D850
Emacs23 Regular Member • Posts: 454
Re: Why not compare by downsampling them both to the same size

I noticed their D800 ISO 6400 raw has 1/250 exposure, while the D600 only 1/320.
So, it's enough not to use their shots for any kind of estimation.

 Emacs23's gear list:Emacs23's gear list
Sony Alpha NEX-5N Leica Summilux-M 35mm f/1.4 ASPH Leica Summilux-M 50mm f/1.4 ASPH Leica Super-Elmar-M 18mm f/3.8 ASPH Leica APO-Summicron-M 90mm f/2 ASPH +3 more
moving_comfort
moving_comfort Veteran Member • Posts: 8,205
There is a focus/lens issue affecting perceived acuity

SergeyGreen wrote:

rhlpetrus wrote:

The D800's were downsized, otherwise you'd have the impression of more noise from it.

In the second image, it is the right frame that should be larger in frame to frame comparison (if the downsizing to be equal), but the noise looks almost the same as it is in both. So downsizing could also explain why the detail looks almost equal as well, it is simply hidden.

Doesn't quite explain the 'softer' look, though, downsampling itself wouldn't bring that to that degree. It's definitely a lens or focus issue making most of the difference there.

Too bad, exact focus matching would allow a downsampled-36MP vs native-24MP comparison with reagrd to acuity.

.

-- hide signature --

Here are a few of my favorite things...
---> http://www.flickr.com/photos/95095968@N00/sets/72157626171532197/

 moving_comfort's gear list:moving_comfort's gear list
Nikon D800 Pentax K20D Nikon AF Nikkor 180mm f/2.8D ED-IF Nikon AF Nikkor 20mm f/2.8D Nikon AF Nikkor 50mm f/1.8D +10 more
Raul Veteran Member • Posts: 7,960
Re: Why not compare by downsampling them both to the same size

Emacs23 wrote:

I noticed their D800 ISO 6400 raw has 1/250 exposure, while the D600 only 1/320.
So, it's enough not to use their shots for any kind of estimation.

The difference in exposure affects what, so as to make comparison invalid?
--

"There is not a thin line between love and hate. There is - in fact - a Great Wall of China with armed sentries posted every 20 feet between love and hate." (House)

Keyboard shortcuts:
FForum MMy threads