I was researching further the D3x noise issue reported by Thom Hogan and I came across this test from popphoto. I do not know how the tests were performed, but here is a comparison between the noise levels of Nikon D3, Nikon D3x and Canon EOS 5D Mark II:
Canon EOS 5D Mark II noise levels
The Canon seems to be better than the D3x at any noise level. Also, the D3 has better noise level (200 ISO) than the D3x? Did they also got a defective D3x? Maybe those numbers don’t really mean anything. Maybe this is all normal and I just need some more sleep. What do you guys think?
Related posts:
43 Comments
I dont know which charts you were reading, but the 5DII scores much better than the D3x. However, the numbers are pretty much useless since there is no explanation on how they came up with those numbers.
Agree. This site is nonsense. They just make up numbers with no explanation as to how they were derived. In a way it is even more dishonest than simply stating it as an opinion. At least with a personal opinion the reader realizes they are getting the thoughts of one person. By applying these bogus numbers they are implying there was some kind of objective measurement when in fact there was not.
I was looking at the noise level charts on the right.
2.98 for ISO 6400 on the D3x vs 2.0 on the 5DmkII for the same ISO. Less is better. On what did you base the sentence “At high ISO, the D3x is a clear winner (against Canon)”?
I meant the other way around
What about the DxOMark site?
it seems reliable, and their numbers show different.
Or maybe it isn’t right ?
Its natural and expected that the D3 has better noise levels than the D3X. Larger photosites. As to the Canon, neither Know nor care
Not really, if you take into account number of pixels. Then, you have to resize and normalize images size, then difference is much less. In DxO Mark, that’s what the PRINT option will do, for charts.
Eh?… no. You have failed to understand the very basics of how the sensor size, photosites/MP count affect the noise levels. Had they increased the MP count on the D3, I would not be taking photos in the dark with it with no tripod or, as last week, in the theatre lighting during Swan Lake.
the numbers are absurd.
All numbers are my friend, all numbers are.
There are better sites…….
It seems that the readings are taken from jpeg
It’s a different camera than the D3! It’s built for a different purpose, stop comparing the 2!!!!
I am just exploring the possibility that they may have gotten a defective D3x as well, that’s all.
actly. The D3 is a PJs tool. Haven’t they even brought out or thought about bringing out a version like one version of the old F3 specifically for PJs – no tripod socket and other bits that they don’t use and can cause probs on border crossings etc… tripod sockets collect dirt if you are in a real PJs environment.
D3X on other hand is really for studio work. Mind you, this is prob with high res digital… if I wanted to shoot real landscapes (for commercial gain) I would take medium format film gear. Similarly, I was recently invited by a friend to a fashion shoot (Lilly Allen was there!) for an international glossy mag and was surprised to see it all done on film – not a digital camera in sight.
There really is not a huge market for the likes of D3x and Canon 1Ds MKIII – except for rich amateurs who enjoy playing at being photographers and one or two so called professionals… but then I never really rated their work.
Exactly. A D3X is meant to really never go above say 400 ISO. They should have just made the range 100-400 to prove a point
1. They base their noise test on jpg output. 5D Mark II is well known for applying relatively strong NR at even the lowest ISO.
2. They are measuring noise and not noise vs detail. They ranked D200 iso1600 noise to be “low” and for it to have “shockingly low noise”. Yeah right!
http://www.popphoto.com/cameras/1844/camera-test-nikon-d200-dslr.html
High pixel count offsets noise, supposedly but smaller pixel pitch creates more noise. DxO Mark does Raw test as well as JPEG. Their noise test is done at Signal to Noise ratio 18%. The higher the dB levels, the less noise. Pop Photo test are mostly done JPEG. It all just varies on how they perform the test. Just compare them all and average them out.
Yeah ok by their measurement the 5dmk2 is better than the d3 at iso 6400 I don’t know what they are smoking
5DII jpegs uses heavy NR, even at base ISO, detail is washed out as well.
my d3 is at nikon now for this problem.
only in 14 bit mode.
have to crop it my wife she doesnt want to be in the pic.
http://gallery.me.com/gilk1#100262/DSC_2954&bgcolor=black
its good to know before we buy.
Guys,
Use this review below.
http://www.imaging-resource.com/PRODS/E5D2/E5D2RAW.HTM
Raw comparation (Canon 5DMII, Nikon D3x and Sony A900). What’s better?
Thanks
Leandro – Brazil
Thanks a lot for the link. Very nice site
I’m sure there are lot of things to be considered when comparing cameras, Noise levels is one of them, but there are also some other Very important that are not so widely checked by some reviewers.
If I’d just forked out upwards of £5000 for a camera body, I’d think it was faulty if it didn’t carry me home at the end of the day . . .
I never really looked at the noise levels of my D300, but could it be that there are also defective D300’s out there?
If you are happy with your camera, than it is not defective.
I think that Hogan and his band of warriors are being slightly sensitive to the tiniest of problems due to the high cost of the D3x. At the end of the day Nikon is going to replace any faulty product so get over it I say.
A good analogy is First Class air travel. If you ever want to see an example of how sensitive people become to their surroundings/products when they are paying large sums for it, this is it.
Between 02 and 06 I lived in Melbourne, Australia and would make the return flight to London twice a year. The company paid for First Class (or rather, they used some of the comps they got for working on the Qantas account). No idea the cost today but it was about $15,000 for the trip then.
Hilarious fun, except that these people would get stressed over it… You would have people telling the stewardesses… (really hot Aussie stewardesses… First Class is where the hot ones work on all the airlines!)
…that their Champagne was “about half a degree too warm”,
or “doesn’t have enough bubbles in it”.
or “I didn’t see you open the bottle, its not from the inbound flight is it… I think it is”.
“There is a fingerprint on the window and if you lie down, you can see it in the sunlight”
“The lettuce isn’t very green”
“The lettuce is too green”
My favourite…. “There is alot of turbulence, one does not expect turbulence in First Class, I am not paying to feel like the home help has just put me in my own washing machine”. (for full effect you have to hear the very proper English accent that went with the words – and she wasn’t that old!)
This is so true.
People by nature are whiners.
And that is a result of being a first world country and these people not having experience in getting some humbling perspective on what being happy with what you have is.
Yeah, they ought to read Kipling’s Captains Courageous.
Don’t you think it’s enough joke for a post ?
Hmmm … let me see.
First gen Nikon sensor … Canon … not first gen as it is taken from 1DMIII and improved.
Nikon sensor … more megapixel, therefore more noise ….
It doesn’t take a rocket scientist to expect the least noise to be D3, then 5DMkII and then last D3x due to sensor size and life cycle of the product.
That’s not what RAW shooters are seeing. The D3x has better noise performance at low ISOs than 5DII, especially in darker areas. The Imaging Resource RAW tests confirm that as have other people with careful RAW analysis.
Check these tests at Digilloyd:
http://diglloyd.com/diglloyd/2009-01-blog.html#_20090126D3x_vs_5DM2
Take Pop Photo ratings with grain of salt. They are far too subjective (what the heck does “extremely high” mean??). Compare the charts. One camera will have the red bar significantly lower than another but both will be “extremely high”.
Their ratings show that my D300 has same image quality as the D3. These charts show the d300 as better color accuracy and far less noise at ISO 800. Tyah…whatever.
D300:
http://www.popphoto.com/cameras/4911/camera-test-nikon-d300-vital-statistics-test-results-page3.html
D3:
http://www.popphoto.com/cameras/5047/camera-test-nikon-d3-vital-statistics-page3.html
I know you were just checking for defects. That’s totally cool. But in general, even if they showed the D3x as better on all accounts, I still wouldn’t trust their ratings.
I think the point here is that either pop photo has a completely ridiculous rating system, or they got a bad sample of the d3x – because according to them, the 5D MKII has less noise at ISO 3200 than the D3x at ISO 100.
I don’t see any way in hell that this can be true.
this was exactly my point and the reason for this post – thank you
As already mentioned, they use jpegs and D3x has a very light approach to jpegs, while the 5DII has the heaviest NR jpeg machine around. Even Canon shooters are saying the jpeg engine is pretty bad, with very poor details because of the noise red applied (even at low ISOs).
As mentioned above, I’d trust DXOmark’s noise comparison over pop photo. At least they give an explanation of their numbers.
Last night I set up a test at various ISOs (100-1600) to compare my D3 and my D3x when the output is scaled to the resolution (i.e. D3 resolution). Foll results with sample images are at http://www.jonmccormack.com/d3x.html . The net is that when you compare apples to apples the D3x is better at lower ISOs (100-200), both cameras are pretty much the same at mid ISOs (400-800) and the D3 is better at higher ISOs (1600+). Not a huge surprise but with all the spin on this subject it’s good to see that the physics do actually hold up.
Cheers,
J
If you pay the outrageous price for a D3x you have every right to compare it to whatever you want.
Hello, this is a very interesting site. I think I will stay a while.
The numbers are not absurd, they are scientific. And those people who say numbers are nothing, go back to them a second later in their defense.
The reason why Dxolabs is so reliable, is because they produce one of the best raw converters out there, so these guys know what they are talking about when they do these scientific tests. When scientific tests showed that Fujifilm S5 had the best DR our of all digital cameras, the field tests facilitated those results as well. When Canon used to have the best high ISO images from any camera, the tests all so proved that. I don’t think anything has changed, except some trolls go around blowing smoke yelling “numbers don’t mean anything”.
Out of all other camera tests usually performed by untrained eye, or by bloggers, I trust these results the least, but I will trust Dxo labs more.
You still shooting RAW matey? Haven’t done that since the D200 was introduced and I earn over $100k per year on the back of my work.
Really, you should shoot film anyway, but if slumming it in the digital world for a bit, the gear available now negates the need for RAW. My D3 is so spot on – without fail, every time, – that I don’t need RAW – even though I put it in some extreme situations.
Creating work for yourself – still I guess that’s what amateurs do.