Why is D200 that much noiser than D80?

Started Jun 2, 2007 | Discussions
Muntz Contributing Member • Posts: 967
Re: Analog circuit

Nikon engineer:
No, the NR of D80 and D200 is almost same.
But the path from CCD to the analog front end is different.
It could happen that 4ch CCD picks much noise than 2ch CCD.

Not that I have anything less than complete faith in the engineer in question, but how convincing does "IT COULD HAPPEN that ...".

I realize stuff might get lost in translation, but if it IS in fact the analog circuit that has extra noise, I'd love to hear a figure like "the D200's analog circuit is 1.3% noisier than the D80's. I believe Nikon engineers probably would know this (not necessarily the guy in question, but someone must).

My opionion is that, while I'm sure a 4ch pathway may be noisier than a 2ch I think the difference would be negligibly minimal. Whereas I think it's obvious that the answer lies in NR. I've compared pictures from the two, and while the jpgs are a tad less noisy in the D80 I fiind the raws pretty identical. If it were the channel it would affect raws as well.

BUT, I could be wrong.

Joe0Bloggs Veteran Member • Posts: 5,391
Sample variation

Muntz wrote:

Not that I have anything less than complete faith in the engineer
in question, but how convincing does "IT COULD HAPPEN that ...".

I realize stuff might get lost in translation, but if it IS in fact
the analog circuit that has extra noise, I'd love to hear a figure
like "the D200's analog circuit is 1.3% noisier than the D80's. I
believe Nikon engineers probably would know this (not necessarily
the guy in question, but someone must).

There were reports of a banding problem with some early D200s. This could be put down to a mismatch between channels. (which were apparently interlaced) That this didn't happen for D80s probably means that there's a greater quality variation among sensors for the D200s. AndreasE claims to see a difference even in RAW, and you don't. You probably have a better D200 or worse D80 than he does.

Then too, RAW is not RAW--whatever converter you're using, it has probably done calculations to level out differences between the readout channels, otherwise I'd think you'd be practically certain to see banding in the RAW files.

kuma37 Regular Member • Posts: 130
Re: Analog circuit

Regardless of the Nikon engineer's interview,
the noise graph of dpreview review shows that
D80 is applying massive NR above ISO800.
It is obvious.

Ilkka Nissilä Veteran Member • Posts: 4,107
Re: Why is D200 that much noiser than D80?

Other people (one very knowledgeable pro in particular) have reported that D80 images at iso 100 are much noisier than D200 images shot under similar conditions.

Ilkka

jetstream Contributing Member • Posts: 632
Re: Why is D200 that much noiser than D80?

D80 and D200 use a similar sensor, I'm sure it's different, but the difference is more related to the I/O channels than anything else...

Absolute noise level of a sensor at a given ISO setting means nothing, the one thing that really counts is the noise level vs detail level... all other things being equal. The AA filter does affect the level of detail the sensor can extract from a signal, I asusme the D200 has a more aggressive AA filter...

It would be nice to see a RAW comparison of both cams with NR turned off, no sharpening, etc... I'm sure the D200 would perfom so close to the D80 that it would be extremely hard to see a difference...

(unknown member) Veteran Member • Posts: 4,734
It's the mirror.

Different mirror construction, the dampers in the d200 don't damp down the noise at ISO 1600 nearly as well as in the d80. If you got rid of the mirror, there'd be no noise.

jb_va2001 Veteran Member • Posts: 7,741
Must be. Never thought of that. /nt

Ednaz wrote:

Different mirror construction, the dampers in the d200 don't damp
down the noise at ISO 1600 nearly as well as in the d80. If you
got rid of the mirror, there'd be no noise.

sandy b
sandy b Veteran Member • Posts: 9,339
Nice comparisom here.....
 sandy b's gear list:sandy b's gear list
Nikon Coolpix A Nikon 1 J1 Nikon D750 Nikon D7500 Nikon AF-S Teleconverter TC-14E II +10 more
Muntz Contributing Member • Posts: 967
Re: Sample variation

Joe0Bloggs wrote:

There were reports of a banding problem with some early D200s. This
could be put down to a mismatch between channels. (which were
apparently interlaced) That this didn't happen for D80s probably
means that there's a greater quality variation among sensors for
the D200s. AndreasE claims to see a difference even in RAW, and you
don't. You probably have a better D200 or worse D80 than he does.

But the banding problem was only in the first issues, which I believe were all fixed (if users returned their cameras). So I don't think there's a varaition amont sensors any longer.

Then too, RAW is not RAW--whatever converter you're using, it has
probably done calculations to level out differences between the
readout channels, otherwise I'd think you'd be practically certain
to see banding in the RAW files.

Woah, I never saw banding in my raws (no more than from any othere dlsr at any rate). I'll even get wicked banding with my D2Hs ... at ISO6400

But seriously, I guess anything's possible (as you said, good D200, bad D80, who knows), but as you can see from this thread (and what I've read) the raws are pretty similar (there are some that even give the D200 the edge ... which I find improbable) but in the jpgs I'd be willing to wager major cash that it's NR. Just my opinion, but when I see such a disparity between the differences between raw and jpg rendering I really find it hard to believe it's anything but NR.

Greg Henry Veteran Member • Posts: 3,841
Sensitive Swedes?

LilKnytt wrote:

Erik,

as a Swede I have to tell you. Careful with such statements.

I could do this in a far more direct manner in Swedish for you. But
as a D200 shooter you've "ticked me off" & I would be anything but
kind to you. So therefore I'm choosing my words carefully.

Wow - it's certainly easy to tick off a Swede, isn't it?

He's asking a simple question about NOISE LEVELS between two cameras, folks.. he's not slapping your first born children! GEEZUS.

Emile Vermeulen New Member • Posts: 4
Re: Nice comparisom here.....

To be very honest...

...from these results I would conclude that the difference in resolution between the D1-series and the later models is much more important than the actual noise levels...

Owning a D200 by the way I am satisfied with the noise levels.

It's about making pictures people!!!!!

drip01
drip01 Senior Member • Posts: 2,905
Re: Analog circuit

Muntz wrote:

Nikon engineer:
No, the NR of D80 and D200 is almost same.
But the path from CCD to the analog front end is different.
It could happen that 4ch CCD picks much noise than 2ch CCD.

Not that I have anything less than complete faith in the engineer
in question, but how convincing does "IT COULD HAPPEN that ...".

Very typical engineering talk from engineers, not architects. I woudn't worry about that.

I realize stuff might get lost in translation, but if it IS in fact
the analog circuit that has extra noise, I'd love to hear a figure
like "the D200's analog circuit is 1.3% noisier than the D80's. I
believe Nikon engineers probably would know this (not necessarily
the guy in question, but someone must).

I say that was pure speculation just to get out of the interview.

My opionion is that, while I'm sure a 4ch pathway may be noisier
than a 2ch I think the difference would be negligibly minimal.
Whereas I think it's obvious that the answer lies in NR. I've
compared pictures from the two, and while the jpgs are a tad less
noisy in the D80 I fiind the raws pretty identical. If it were the
channel it would affect raws as well.

BUT, I could be wrong.

I think you're probably right. The NR on D80's JPEG (even when it claims NR is turned off) is apparent.

NR is like faked boobs. If done well, they are easy to fool horny people (those obessed with noise) ... 8-)

-- hide signature --
 drip01's gear list:drip01's gear list
Sony a6400 Canon EF 85mm f/1.8 USM Canon EF 200mm f/2.8L II USM Sigma 18-300 F3.5-6.3 DC Macro OS | C Sigma 150-600mm F5-6.3 | C +2 more
RudivanS Veteran Member • Posts: 4,480
Re: Why is D200 that much noiser than D80?

Ilkka Nissilä wrote:

Other people (one very knowledgeable pro in particular) have
reported that D80 images at iso 100 are much noisier than D200
images shot under similar conditions.

Ilkka

Firstly, good to see you posting here again Ilkka.

Is it possible to verify the base 100 iso difference between the 2 cam's in question?

Am sure it would be a minimal difference .. if any. Purely as a matter of interest of course ;^)
--
Rudi -
15 years in Blade Runner Tokyo - back in Sunny Sydney now

 RudivanS's gear list:RudivanS's gear list
Nikon D4 Nikon D810 Nikon D5 Olympus PEN-F Canon Pixma Pro-10
Muntz Contributing Member • Posts: 967
Re: Analog circuit

drip01 wrote:

Muntz wrote:

Nikon engineer:
No, the NR of D80 and D200 is almost same.
But the path from CCD to the analog front end is different.
It could happen that 4ch CCD picks much noise than 2ch CCD.

Not that I have anything less than complete faith in the engineer
in question, but how convincing does "IT COULD HAPPEN that ...".

Very typical engineering talk from engineers, not architects. I
woudn't worry about that.

LOL. I suppose "it could happen that a certain support won't hold" would be a little more worrisome

I realize stuff might get lost in translation, but if it IS in fact
the analog circuit that has extra noise, I'd love to hear a figure
like "the D200's analog circuit is 1.3% noisier than the D80's. I
believe Nikon engineers probably would know this (not necessarily
the guy in question, but someone must).

I say that was pure speculation just to get out of the interview.

That would be my guess.

My opionion is that, while I'm sure a 4ch pathway may be noisier
than a 2ch I think the difference would be negligibly minimal.
Whereas I think it's obvious that the answer lies in NR. I've
compared pictures from the two, and while the jpgs are a tad less
noisy in the D80 I fiind the raws pretty identical. If it were the
channel it would affect raws as well.

BUT, I could be wrong.

I think you're probably right. The NR on D80's JPEG (even when it
claims NR is turned off) is apparent.
NR is like faked boobs. If done well, they are easy to fool horny
people (those obessed with noise) ... 8-)

You kill me

arachnophilia Veteran Member • Posts: 3,362
my mirror makes just as much noise at iso 100! (nt)

Different mirror construction, the dampers in the d200 don't damp
down the noise at ISO 1600 nearly as well as in the d80. If you
got rid of the mirror, there'd be no noise.

 arachnophilia's gear list:arachnophilia's gear list
Nikon D700 Nikon D300S Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 50mm f/1.8G +3 more
Toermalijn
Toermalijn Forum Pro • Posts: 15,993
Re: My point

No, i don't. people buy a camera, they can't test it as they can when they actually owna camera. So, before your testing and getting familiar period, they have to decide wich way to go. So in a sence they are allready bashing a brand by not picking them.

Kluso Senior Member • Posts: 1,904
Chain oil

I use Castrol Chain Oil in a pressure can, silky smooth, no noise, do remember to shake the can for 30 seconds before applying to mirror hinges, a quick squirt on the shutter also helps.

Don't overdo it as that can cause a yellow cast and we don't want that !

Ednaz wrote:

Different mirror construction, the dampers in the d200 don't damp
down the noise at ISO 1600 nearly as well as in the d80. If you
got rid of the mirror, there'd be no noise.

-- hide signature --

Inspector Kluso

OP egrc Junior Member • Posts: 32
Re: Well written yet again....... Re: Here we go again.

Muntz wrote:

You know something, (this came to me this morning I think even
if it were possible AND simple (to port new algorithms into the new
hardware) I think companies wouldn't do it because it would be
tantamount to a product recall. Basically anyone with a D200
(especially one on warranty) would go back to where they bought it
(or straight to Nikon) and say "Hey, I got a defective model with
old algorithms). I really think that's the main reason a model is
never upgraded without at least a small change in name (usually the
extra "s" ;).

I agree, I thought the same thing when I had the D70, and the
cheaper D50 came out with "apparently" better IQ.

Your so right, that can be yet another reason.

erik

OP egrc Junior Member • Posts: 32
Re: Well written yet again....... Re: Here we go again.

Matt Sergeant wrote:

egrc wrote:

This is exactly my point. If the jpegs in D80 is better and it is
due to software why isn't Nikon just porting it to D200 in a
firmware update? Could there be some other difference that makes
this impossible, such as pre AD signal processing, the sensors, the
higher speed of D200 in continous mode etc.

Because the JPEG engine isn't in software. It's in hardware. And
not in flashable hardware either. It's hard coded. You can't just
upgrade the firmware and improve it, you have to put a new chip in.

Maybe you'll get these changes in a D200s (though I'd suggest
you're making more of an issue of it than is necessary). Or maybe

Yes, I probably do make more of an issue of it than necessary. One part of it is curiosity however.

erik

OP egrc Junior Member • Posts: 32
Re: It's not

Kim Letkeman wrote:

Toermalijn wrote:

otherwise, it would be a limitation of the hardware.

You seem to be a bit confused. This was his whole point. The D80 is
not all that much superior in jpeg mode to the D200 ... very
slightly less noisy on equivalent exposures. Read a few reviews by
competent reviewers and you will see.

Hence ... the user who started this thread is probably
underexposing and thus driving noise levels up sharply. That sits
firmly in the "user error" camp.

Well, I haven't done any exposures at all, I've just examined pcitures, for example those in the reviews here at dpreview, taken with the cameras.

erik

Keyboard shortcuts:
FForum MMy threads