2
\$\begingroup\$

I am looking for a high resolution camera for fine art prints and the Sony A7R seems great since it has the same Sensor res as the D800 in a smaller enclosure.

For some reason when I look at sample shots of the A7r at 100% compared to samples of the D800 at 100% and realized that the A7r looks a lot more artificial with it's noise pattern almost looking like bad compression artifacts, where as the D800 noise looks more film like to my eye when judged at 100% full resolution view.

So my question is:

Is the difference in Noise represented by these two sample sets, because of differences in the Sensor or Noise rendering or are there other circumstances that contribute to the different in outcome in this case (Iso / Shutter Speed / Light / Lens / Apparture)

I think it's pretty obvious what I mean. Especially if you compare Faces, Skin and fine Textures, the A7R has a sort of artificial almost phonecam like artifact like noise, while the D800 seems to be very smooth imo.

Would be great if somebody could enlighten me, since I am just judging by Eye and I don't want to go for the D800 if the difference is just because of the lens or some other factor.

Sample Sets in Question below, please look at the original resolution file at 100%. especially the horse ones of the A7r and the ones of the boy and the man with the gaping earring on the D800.

A7R Sample

http://www.dpreview.com/products/sony/slrs/sony_a7r/sample-photos

D800 Sample

http://www.dpreview.com/galleries/reviewsamples/photos/1933469/?inAlbum=nikon-d800-review-samples

Improve this question
\$\endgroup\$
3
  • 2
    \$\begingroup\$ Aren't all of those images JPEG? So you are comparing two different JPEG processors(or at least RAW converter implementations) and not the sensors directly/or the RAW output. \$\endgroup\$
    – dpollitt
    Dec 5, 2013 at 22:34
  • 1
    \$\begingroup\$ Although not the primary difference in the case of the dpreview comparison, differences in lens should be a consideration. There are some things you can do with some lenses available in the Nikon mount that you can't do with the Sony simply because comparable lenses just aren't available in the A7R's mount. \$\endgroup\$
    – Michael C
    Dec 5, 2013 at 23:45
  • \$\begingroup\$ I know we're comparing jpgs here, but the difference in noise texture is quiet big in my opinion with the A7r looking very digital in the fine details. I doubt the overall noise texture would look a lot different if it was raws right? \$\endgroup\$
    – Roland
    Dec 6, 2013 at 2:10

2 Answers 2

Reset to default
6
\$\begingroup\$

You're comparing JPEGs here - therefore you're not just comparing the sensors, but whatever processing parameters the manufacturers and/or reviewers have chosen to use when creating those JPEGs. From a quick play with DPReview's studio comparision, it's fairly obvious that Sony have stronger noise reduction / sharpening / contrast enhancement in their pipeline than Nikon do. Comparing the same shots as RAW gives the two looking much closer to each other. You can argue about it a bit, but you're now into the realm of "small differences".

Moral of this story: don't compare JPEGs unless they've both been processed via the same pipeline.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ Was thinking the same thing. Need RAWs! I'm sure we could find some if we looked hard enough... \$\endgroup\$
    – dpollitt
    Dec 5, 2013 at 22:35
  • \$\begingroup\$ thanks, I realize that the A7r is supposedly better when judged with fine line resolution charts and all that, but still in real world examples the d800 looks so much better when it comes to noise and skin imo, but as I said, since I am not a professional in this area I can't be sure it's because of the sensor. also while I understand that raw holds more editing power. I think high quality jpgs shouldn't look so much worse for just inspecting right, or could jpg really introduce this mushy digital noise artifacts in the a7r samples? \$\endgroup\$
    – Roland
    Dec 6, 2013 at 2:06
  • \$\begingroup\$ Converting a lossless image to a high-quality JPEG certainly won't introduce significant artifacts (that's almost the definition of the "high-quality" bit!), but the noise reduction and sharpening that's applied as part of creating any image from a RAW file certainly can - it's the significant amounts of noise reduction and sharpening applied by smartphones that cause the characteristic artifacts, rather than the fact they're JPEGs. JPEG artifacts have a very different character - clasically, they're shown as "blocks" in the image. \$\endgroup\$
    – Philip Kendall ♦
    Dec 6, 2013 at 20:28
2
\$\begingroup\$

Is the difference in Noise represented by these two sample sets, because of differences in the Sensor or Noise rendering or are there other circumstances that contribute to the different in outcome in this case

While others have insisted that you should only be looking at RAW, in answer to the above, it's all of them.

When comparing jpgs, it's the sensor, the camera's processing engine and then it's photographic settings.

When comparing RAWs, it's just the sensor (and lens/photographic settings).

For the most part, when comparing jpgs, it's the processing engine. All this does is adjust parameters and compresses the image.

If you want to purely compare sensors, you look at RAWs and ensure you use the same software to process them.

In regards to the images, the ones you are comparing to are under very different conditions. One is in a studio and the other is outdoors.

It will be better to compare these ones:

SONY Alpha 7r

Nikon D800

And even then, it's hard to tell zoomed in.

You should really test both under the same conditions, in a studio environment where light can be controlled (as suggested by Phillip in the DPR links).

In summary, noise contributes; lack of light, sensor and jpg engine. But to compare noise hanlding, it should be under the same environment, same software and shooting in RAW.

Improve this answer
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.