D3X and A900 comparison ISO 1600

[David Kilpatrick]

Using a light level typical for tungsten light shots at 1600 (1/60th at f8 = 1/500th at f2.8 for action, 1/250th at f4 for concerts etc) I have made a quick first day comparison which, in the process, also evaluates two lenses (24-120mm VR Nikkor and 28-105mm RS Minolta) and two anti-shake systems.

Nikon image:

http://www.pbase.com/davidkilpatrick/image/106952293

Sony image:

http://www.pbase.com/davidkilpatrick/image/106952572

There are caption notes to each image. The point of targeted focus was the end of the fingerboard with the spattered rosin on it. Feel free to dowload and print. The red rosin cloth is interesting, so is the sharpness of the point of focus.

David

[Winston]

For me, the apparent exposure is too different for me to compare anything. All I can determine is that I like the darker one (A900) better.

[InTheSky]

David ... is there something that Minolta eat in the past the make them building this kind of good lens ???

by all your observations since you are working with A900, this make me feel that Even old glass of Minolta (no needed to be G lens ...) are able to challenge lens for 5 or 6 times the price other brand ...


P.S. where are the Canon shot ?

REgards,


Frank

[A1000]

Remember, it is the same Sony sensor, developed by Sony, not Nikon, only Nikon got the later version with live view and video function. Sony rushed with A900 to get it in the market, so it happened before the new sensor production was at full speed.

[Dr. Harout]

David, could you make the same comparison with same lens (same brand, I mean). For example the use of a specific Sigma lens on both and repeat the procedure. Thus we could omit lens performance.

[David Kilpatrick]

In this case, I wanted to see lens performance. I was interested in seeing exactly how the VR and SSS systems coped with approximately 2 stops of camera shake risk (1 stop in focal length equation terms, plus 1 stop for close-up subject on 24 megapixels). I have done more tests, not just this one pair of shots, and the amount of image detail loss caused by the VR system in the Nikon 24-120mm would rule it out as a useful lens for me. I have also been working for a couple of weeks with the 5D MkII and IS 24-1045mm - same result. I have to assume that some of the very poor results, with sharpness loss towards one extreme of the image, are caused by the way in-lens stabilisation works.

I know it is considered heresy to suggest that in-body stabilisation actually works at all, let alone that it might be better than in-lens - but with these wide-to-tele lenses, I'm seeing far better results from the A900 with SSS.

Sigma is sending me a pair of 28-70mm f2.8s in Nikon and Sony, to be able to make a direct comparison. I also just handed them over the money for the 12-24mm - not sending that lens back!

David

[aster]

Thanks David,

Please keep pouring in on us any findings you think are worth sharing on any instant of your tests, be them about various lenses or the camera systems!
It feels good to know...

Yildiz

ps. You definitely are responsible for my putting the Sigma 12-24mm on my list of lenses to find one good copy as well.

[01af]

In this case, I wanted to see lens performance. [...] the amount of image detail loss caused by the VR system in the Nikon 24-120 mm would rule it out as a useful lens for me.

David, are you sure the detail loss is due to VR only? From several Internet reports I'm getting the impression the AF-S Nikkor 24-120 VR has the reputation of being not a very good performer generally. See here, for an example.

See more about the A900 and its competitors from the same author (Mike Johnston of The Online Photographer) below. Sorry for the longish list but it's worthwhile reading---educating, thought-provoking, and even funny at times. But beware, it's "just" personal impressions and opinions, not really reviews or even formal tests:

http://theonlinephotographer.typepad.co ... kicks.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 00-20.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... merci.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 900-b.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... e-par.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 900-s.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 900-f.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 120mm.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... ett-1.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... -d700.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... s-ima.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 900-a.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... -d700.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 5d-ma.html" onclick="window.open(this.href);return false;
http://theonlinephotographer.typepad.co ... 5d--2.html" onclick="window.open(this.href);return false;

I know it is considered heresy to suggest that in-body stabilisation actually works at all, let alone that it might be better than in-lens - but with these wide-to-tele lenses, I'm seeing far better results from the A900 with SSS.

It's heresy for those who are more into believing than into knowing (i. e. mostly Canon and Nikon fanboys). Those who actually tried AS or SSS know that it not only works but works really well. I am convinced in-body image stabilisation is just as good as in-lens stabilisers at long focal lengths and significantly better at short and medium focal lengths.

-- Olaf

[braeside]

David please let us know when your review will be in the BJP.

[David Kilpatrick]

I'm doing a quick first impression which they need by Weds. That is an unusual day for the BJP, normally Monday is their deadline, so I guess it will be holiday timing changing this. I would assume they do a double issue appearing in the week before Christmas and it might be this.

There will be a later test with comparisons, and that will appear sometime in January; I'll try to flag up when. Right now my main problem is Capture NX2. If anyone knows exactly which file must be found and removed to erase Capture NX2's memory of an earlier trial period, let me know. The Uninstallers are broken on Mac OSX Leopard. I have installed a trial on my Dell 2.0GHz PC but it's so slow it is impossible to use. Nikon does not issue proper raw conversion software with the D3X, and although they kindly let me have a review copy of Capture NX boxed, NX2 is a separate paid item not an update and they have never sent me an unlock code or a boxed copy - even with the D700, which was the first camera to need it.

This puts me in exactly the same position as a real Nikon owner - stump up, or shut up, and I can tell you it does not feel good. If I had just bought a D3X and discovered no proper software in a box with a £5500 camera, plus no support through any third party yet, I would be livid. It adds another £140 to the true cost of any Nikon pro DSLR, unless they make a free offer (which they currently do with the D700). This is KM did with Dimage Master, it was a bad policy to create a fake item which could be given free as a promo, by charging for something which should have been part of the camera's cost originally.

Sony has been far more sensible and you get all the software you need, including remote tethered shooting, in with the price.

David

[David Kilpatrick]

I have just had to process the files using the three converters - Canon's crashed twice, can't handle a folder with lots of images. Nikon's is even slower than IDC2. Actually, IDC (though lacking some of the bells/whistles stuff) emerges as the fastest of the three and with the least non-intuitive interface. Amazing. It builds 100% views faster, and its interface with the Lightbox app works well. Nikon's workspace is very confusing and gets mixed up with other apps in Mac Leopard so that their windows can appear on top of NX2 windows... help! It turns out we did get an NFR copy of NX2, but Richard has it, so he just let me have the serial number and that solved the problem legitimately, able to use the Mac.

But - IDC can not recover unclipped colour data in highlights anything like as well as ACR 5.2, and all three own-make progs fail to drill down to low bit values, which ACR does well (hence the lousy noise, but incredible shadow detail).

David