This just in... Nikons rock at high ISO...

Shoot the D2h outdoors at night at 1600 and you will see plenty of noise in the dark shadow tones.

Shoot it indoors at 1600 in contrasty light and you will see plenty of noise in the dark shadow tones.

Shoot it in full sunlight at 1600 and there will be little noise.

It is not a low light camera, unfortunately.
 
If you let me color-correct it properly, sure. And properly means
optically, not digitally. You need an 80A to fix the blue channel,
just as though you had daylight-balanced film loaded but were
shooting in blue-impoverished incandescent light. Same issue.
Same solution.
And lose 1 1/3 stop? So the 1/20 @ f2 becomes 1/15 @ f1.4. And you've
cut your viewfinder brightness in half. And you've reduced your AF
system's effectiveness.
These things are true, but you cannot argue that I have in so doing
also dramatically improved the quality of the data. I have more
"blue" photons, so I know more about them and no longer have
to fabulate fictions. This leads to improvedimage quality. You can't
argue otherwise.

You should gel your flashes to incandescent to match the incandescent
ambient when you're using the 80A. Much more natural looking. High
quality data. Better pictures. While I have use f/1.4 lenses with an 80D
and no flash for indoor shots, I much prefer gelling all the speedlights.

Also, an 80A might be a little dark. You can probably get by with one
half that dense. It will certainly improve matters, and is a realitistic
solution to a real problem.
Sorry, no sale.
Sale? Sale?? What are you talking about? Sounds like you're a bit
money-obsessessed. Well, men supporting large families may have
that problem, and I can't really say anything against their needs. I'm
just glad that I do this for pure pleasure, not for pecuniary profit.
Radix malorum cupiditas est, and all that.
I've got a better solution: the 5D.
I'm afraid you've posted that to the wrong forum. For your penance,
go to a Canon forum and wheedle them about some scenario for
which the better solution is the D2X. Go on, go do that. Now.
Because what you're doing here is just as useful, just as uplifting,
just as relevant to the forum's focus and content. So go annoy
the Canonites, Greg, by telling them they need to buy a Nikon rig.
I'm sure you'll be perfect at it. You've clearly the experience to
really stick it to them. You'd make two forums a better place, too.

--tom
 
Shoot the D2h outdoors at night at 1600 and you will see plenty of
noise in the dark shadow tones.
Do you forget to have your hot mirror on? And what's the true
signal data in all three channels under that light? Is this starlight
away from the blight of "civilization"'s light polution, or do you have
poisonous contamination--and if so, of what color? Are you
compensating for this?
Shoot it indoors at 1600 in contrasty light and you will see plenty
of noise in the dark shadow tones.
Did you forget to use a CC20B with your hot mirror?
Shoot it in full sunlight at 1600 and there will be little noise.
And there will be even less is you use a CC20M with your hot mirror.
It is not a low light camera, unfortunately.
At some point, technical matters require technical solutions,
not point-and-shoot "just do it" hopes and prayers. Understanding
how digital sensors emulate (but do not provide) higher sensitivies,
what white balance does to you when you squeeze 7 stops of push
from this poor thing at high pseudo-iso, and how this interacts with
sharpening and dodging are things you really do have to learn if you
want get the most out of your shot. The disparity between how one
raw processor handles noise compared with another one is also
a surprise to learn. If you don't want to learn things, just click away.
You won't get many photons that way, nor good ones. It'll probably
be good enough for most people who are just looking for snapshots
or who don't keep throwing a 50x loupe at their 8x12 prints. Clicking
away without learning is especially cheaper if the cost requires
doing research. There's another solution, too, but well, let's just say
that it's a darned pity tripods are so last-century tech, eh? If a
cell phone doesn't need it, neither do you.

Moral: Don't push too hard, you'll break your picture.

--tom
 
I'm talking about luminance noise, not chroma noise.

Tell me, what do you accomplish by adding that cc filter in daylight?
 
made a very good point.

Has anyone bothered to read the incredibly detailed reviews on imaging-resource.com???

Anyone that thinks the newer Nikons are "noisy" up to ISO 1600 needs to look at that site. Then come back here and start trying to spout the nonsense that Canons have "significantly" better high ISO noise performance.
 
Kind of depends on what you value

I've been following this thread and others like it.

I can go with the notion that it's a little harder to get a good Nikon high ISO photo under certain conditions or that grain doesn't matter too much when you print at certain sizes - no problem

But I don't think anyone is going to convince me that I actually like noise or grain; sure for 1 in about 1,000 plus photos grain might produce an interesting effect - but it's far from my standard preference.

Perhaps some people love grain and apsire to develop a grainy style; and perhaps many other people just don't notice grain, but I also think there is a farily large pecentage of shooters who aspire to have sharp photos and who don't like grain - any grain.

In any event, regardless of how many people like or don't notice grain, I'd prefer a capability to shoot with as little light as possible and still produce sharp photos. Perhaps other people would like to inject more grain in their photos when lots of light is available - but I'm betting it will be a while before that feature catches on

I don't have to be sold on Nikon vs. Canon; I made the decison to go with Nikon about 5 years ago and I'm very happy with Nikon; but I'm not ready to buy "grain is beautiful". I can think of a lot of attributes that could be ascribed to good IQ, but "grain" isn't one of them. YMMV
 
The Canons are better at hi ISO. This is just a fact.

Yes, I shoot Nikon.
 
I've done ISO100 shots with a little visible noise, which was too much for more than one stock agency.

If you plan making money on your photos, noise can be an issue (even for Canon users).

claus
 
dont get carried away either way I guess. The point is that nikons are not completely useless at low light and canons dont magically turn awful light into sparkling beauty at iso 3200.
 
I agree fully with the OP. Nikon cameras aren't that bad @ high iso. Due to a missed opportunity for a D2X from a friend of mine(I already sold my D200& my S5pro to finance the D2X...), I bought in a stupid impulsive moment a 5D with a couple of lenses and sold the rest of my Nikon gear after that. I expected very high IQ @ high iso, but I am very disappointed. When you're shooting in really low light situations @ iso 3200, there is horrible noise as well with the 5D. Underneath the unprocessed samples, only resized for the web. I would say that my D200 would have done about as well in this situation. This whole 5D superiority myth is created by Canon PR people. Canon 5D is better than D200 @ high iso, but the difference is not a fraction of what Canon PR want to make us believe...



and a crop



Very soon I'll be shooting exlusively with Nikon again...
--
Kindest regards,
Stany
I prefer one really good picture in a day over 10 bad ones in a second...

http://www.fotografie.fr/
 
The noise in that photo isn't that bad. A d2h is unusable at 3200
 
If you let me color-correct it properly, sure. And properly means
optically, not digitally. You need an 80A to fix the blue channel,
just as though you had daylight-balanced film loaded but were
shooting in blue-impoverished incandescent light. Same issue.
Same solution.
And lose 1 1/3 stop? So the 1/20 @ f2 becomes 1/15 @ f1.4. And you've
cut your viewfinder brightness in half. And you've reduced your AF
system's effectiveness.
These things are true, but you cannot argue that I have in so doing
also dramatically improved the quality of the data. I have more
"blue" photons, so I know more about them and no longer have
to fabulate fictions. This leads to improvedimage quality. You can't
argue otherwise.
No argument about that. I'm sure it does, and you've piqued my curiosity enough that I'll probably do some experimenting to find out how much.
You should gel your flashes to incandescent to match the incandescent
ambient when you're using the 80A. Much more natural looking. High
quality data. Better pictures. While I have use f/1.4 lenses with
an 80D
and no flash for indoor shots, I much prefer gelling all the
speedlights.
You'd almost have to gel the flash I would think - other wise the foreground would be way too blue, wouldn't it?
Also, an 80A might be a little dark. You can probably get by with one
half that dense. It will certainly improve matters, and is a
realitistic
solution to a real problem.
Sorry, no sale.
Sale? Sale?? What are you talking about? Sounds like you're a bit
money-obsessessed. Well, men supporting large families may have
that problem, and I can't really say anything against their needs. I'm
just glad that I do this for pure pleasure, not for pecuniary profit.
Radix malorum cupiditas est, and all that.
Come on now, you have to know that's a figure of speech! Apologies if you're joking.

--
http://www.pbase.com/gzillgi
http://www.pbase.com/gzillgi/wedding_portrait

 
I agree with you on the BS associated with measurebaters on 100% crop and noise. If it looks good on an 8x10 at relaxed arms length then it's fine by me. I NEVER look at stuff at 100%. I always look at the print. Composition, timing, and focus are much more important.
--
Cheers.

...Please don't rub up against my glass...
 
Can you point to anywhere on that website that tackles this issue in any kind of scientific way?

The bottom line in all this discussion is, if you're happy with what your camera gives you, then more power to you! Note, I love my D200, and am encouraged by Nikon's stated intentions to improve in high ISO noise differences with clearly superior Canon cameras.

Well, Nikon won't couch it in those terms, but being the user of a 1Ds Mark II at work I can tell you the difference is night and day at anything over 400 ISO. To the point that I won't crank my D200 up to 1600 ISO for any reason at all. Being a professional, my standards are way beyond what looks good at arm's length in an 8x10. When I make prints, they are rarely smaller than 16x20 and they are often shot in far from ideal light. Thus my standards are likely different that for many people here. No big deal. Just different.
made a very good point.

Has anyone bothered to read the incredibly detailed reviews on
imaging-resource.com???

Anyone that thinks the newer Nikons are "noisy" up to ISO 1600 needs
to look at that site. Then come back here and start trying to spout
the nonsense that Canons have "significantly" better high ISO noise
performance.
--
Eric

Ernest Hemingway's writing reminds me of the farting of an old horse. - E.B. White
 
Come on now, you have to know that's a figure of speech! Apologies if
you're joking.
Yeah, that was too extreme.

Greg, try a cyan filter instead, like CC30C or such. This will only
quell the over-represented red but still let in the green. You won't
lose at all the stops you would with a pure blue filter like the very
strong 80A.

--tom
 
You should gel your flashes to incandescent to match the incandescent
ambient when you're using the 80A. Much more natural looking. High
quality data. Better pictures. While I have use f/1.4 lenses with
an 80D
and no flash for indoor shots, I much prefer gelling all the
speedlights.
Yes, you always have to gel a flash to the color of ambient light,
or else you'll get a mismatch, unless you're shooting at 5400K
ambient, of course. AutoWB tried to make the best of this, especially
depending on flash strength and the precise mix of flash vs ambient
that each shot actually produces, but there's only so much you can do.


The odd thing here that I think people seldom do is that you
gel the flash to incandescent to match room lighting, but then
suppress the incoming red signal with a cyan filter, or the incoming
red and green wiht a blue filter. I'm not sure how the light-balancing
filters like an 80A or 82A actually affect all three wavelengths as
compared with color-correcting filters like CC20M which block out
one or two specific frequencies.

http://www.aeimages.com/learn/color-correction.html

I think you'll be surprised by how much better data you can
get out of your camera if you fix the color in analogue.
Here's a long postings I just put up that tries to regurgitate
stuff that people who know much more about it than I do have
been saying:

http://forums.dpreview.com/forums/read.asp?forum=1021&message=24057261

--tom
 
The thing I notice on this forum is that some people see grain in EVERYTHING. I have to laugh over and over again at some of the noise comments in this forum where some people seemingly think that if there is any grain or noise you might as well have not taken the shot.

I prefer to start with a clean image and add grain if I want it. I would still rather capture a grainy moment and deal with it than to not capture the moment at all.

--
Ed C.
 
That's because D2H has the worst high ISO.
Try the upgrade D2Hs instead. It outperforms 1D MarkIII at high ISO.
 

Keyboard shortcuts

Back
Top