Is it safe to assume you were shooting aperture priority allowing shutter speed to adjust with ISO changes?

That would be a very safe assumption. Unless I have a good reason to --like lighting with monolights or pack and head systems, or shooting a series of shots to incorporated into a panoramic montage it is very rare that I do not use Aperture priority.

I would love to see the same series, done at the same time, with a D3 right next to the D3s.

For my purposes, it looks like I could use up to ISO6400 on the D3S, that's about 1-1/3 stop more than I'm comfortable using my D3 at for most purposes. But perhaps the D3s would show even better performance in a side by side comparison.

That kind of direct comparison testing will come when we do the full review of the D3s starting sometime next month. I think that if you aren't dealing with inky unlit darkness the D3s is actually at least a stop better than that (ISO 12800).

Don't forget that these are pretty much straight from the camera shots -- I haven't tweaked or sharpened or run them through a noise reduction plug-in filter or software.

Thanks so much for this test.

It appears that the 800 ISO image is sharper than the 400 and 100. Do you think that is camera movement? mirror vibration?

I also like the idea of a D3 control set as well but would you consider using a Prime Lens?

For all photos in the ISO sample gallery linked in my post the camera was not on a tripod - -which would have been ideal but circumstances did not allow for it, Rather I braced it against the wooden ring surrounding the area.
For the initial test shots in the gallery at http://www.ellisvener.com/data/web/Nikon_D3s_ISO_samples/index.html

which should only be used to judge basic image signal to noise qualities for files coming basically straight from the camera, the 14-24mm f/2.8G Nikkor lens was set to 14mm and f/5.6. Camera set to Tungsten WB and Aperture Priority. Shutter speeds were as follows;

LO 1.0 (ISO 100 equivalent): 1.3 seconds
ISO 200: 0.6 seconds
ISO 400: 0.3 seconds
ISO 800: 1/6th second
ISO 1600: 1/13th
ISO 3200: 1/25th
ISO 6400: 1/50th
ISO 12,800: 1/100th
Hi 1 (ISO 25,600): 1/200th
HI 2 (ISO 51,200): 1/400th
HI 3 (ISO 102,400): 1/640th*

* for the HI 3 image, either I changed the composition enough that the camera decided to open up by 1/3rd stop or they changed the lighting causing the exposure change.

Bill,

I think my post above explains the difference in sharpness. Atthe lower ISO settings the exposures were longer and I was bracing the camera against a wooden platform by holding my hand on top of the camera.

The difference in sharpness is not due ot the lens choice in other words and the Nikon 14-24mm f/2.8G is one of the sharpest ulra-wide to wide angle lenses, and certainly is one of the best resolving zoom lenses, of any focal length, that I have ever used.

But yes, I do some testing with one of the best "prime" (single focal length) lenses I own or that Nikon ever made: the 105mm f/4 AI-S Micro-Nikkor

Yeah, I agree about the 105 even though I don't have one right now - it's on my someday spare cash list.

Thanks for the explanations. I look forward to your future reviews.

Bill

Wow! An impressive showing. This does make a significant impact into the low light shooting capabilities for photographers. My 5D MKII is isn't usable beyond the 2400 ISA setting without major post production work. However, what would really be useful is if Nikon and Canon would start focusing on increasing latitude in sensor technology. If we could get a wider latitude it would be much more helpful with shadow details and preventing highlights from being blown out. Someday.....
Curtis
Wedding Photography Miami
Fort Lauderdale Wedding Photography

The Nikon D3X has just over 13 stops of dynamic range. The high end Canons, Sonys, Leicas, and other Nikons are roughly in the 12 stop DR range (depending on the camera. But with any camera, the higher you go with your ISO setting the usable DR definitely contracts over the best quality DR. By how much specifically I don't know.

In-camera settings such as noise reduction and Active D-lighting do not apply to RAW files (well, except maybe Sony's…). Why even mention it?

Just giving you a hard time. Thanks for these samples!

I would be interested in seeing how the same images would be processed by other conversion engines, such as CaptureOne, Bibble, Aperture, ACR6/Lightroom 3 (when available)

"In-camera settings such as noise reduction and Active D-lighting do not apply to RAW files (well, except maybe Sony's…). Why even mention it?"

Hi Cory,

There are three issues to address in your statement and question.

1) Noise reduction: At all ISO setting of 3200 and above Nikon programs the EXPEED processor in the D3s to automatically start using a noise reduction algorithm and applying it to NEFs as well as in-camera processed JPEGs and TIFFs. The amount of NR applied is keyed to the sensitivity setting (ISO up to 12,800 and equivalent ISO "HI" setting above that.)

2) Active D-Lighting processing is done on the camera sensor at the instant of capture. Based on the meter reading the gain on the signal on the individual photo elements -- pixels -- is adjusted as the data starts its lightning quick passage through the processing pipeline in the camera. That is straight from Nikon's sensor engineers in Sendai. If you have Active D-Lighting -- which I have found to be a terrific tool for general high contrast situations -- turned on it's effect is always utilized no matter which after camera raw processor you use, Although you can cancel it in Nikon Capture NX2 and Nikon View NX.

3.) I mentioned my settings in the interest of full disclosure.

Rob Galbraith was at the Big Apple Circus that night too and has published a gallery of his photos where he has treated his images the way most of us would use in the real world. Since my gallery of D3s samples is pretty much straight from the camera our two galleries complement each other, not to mention the fact that Rob is a better action photographer than I am!. The URL for Rob's D3s circus gallery is http://www.robgalbraith.com/bins/content_page.asp?cid=7-10045-10329

As concerns other raw processing engines that wil have to wait until those companies update their products.