Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: In Camera Noise Reduction question  (Read 17932 times)
jvora
Jr. Member
**
Offline Offline

Posts: 90


« on: November 19, 2008, 08:43:12 PM »
ReplyReply

Hello !

I use the Nikon D200 and D300 and only shoot in the RAW format - Noting this, does the in built Noise Reduction feature have any effect on the image ?

I ask as I can see it having an effect on a jpeg or tiff, but is the RAW format not the most un-processed set of image data ??

Kindly confirm.


Jai
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5474


WWW
« Reply #1 on: November 19, 2008, 09:01:15 PM »
ReplyReply

Quote from: jvora
I use the Nikon D200 and D300 and only shoot in the RAW format - Noting this, does the in built Noise Reduction feature have any effect on the image ?

Not if you are using Camera Raw, yes if you use Nikon's software.
Logged
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #2 on: November 19, 2008, 09:02:39 PM »
ReplyReply

Quote from: jvora
I use the Nikon D200 and D300 and only shoot in the RAW format - Noting this, does the in built Noise Reduction feature have any effect on the image ?


The Noise Reduction does NOT affect the D300 raw data. I can not state state this re the D200, but I would be very surprized if that raw data were affected by the noise reduction selection.

I don't know of any other camera beside the Sony A700, which carries out noise reduction on the raw data. Perhaps the A900 too does that.

Note, that the Long Exposure Noise Reduction DOES affect the raw data; that is a completely different matter.
Logged

Gabor
mike.online
Full Member
***
Offline Offline

Posts: 207


WWW
« Reply #3 on: November 19, 2008, 09:15:13 PM »
ReplyReply

Quote from: Panopeeper
Note, that the Long Exposure Noise Reduction DOES affect the raw data

not to hijack the conversation, but how does it affect the raw information?
Logged

Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #4 on: November 19, 2008, 09:35:04 PM »
ReplyReply

Quote from: mike.online
not to hijack the conversation, but how does it affect the raw information?

The effect of Long Exposure Noise Reduction is, that a second shot will be made with the same shutter time but with closed shutter, after the first exposure is finished. The pixel-wise result of the second shot is then substracted from the first one; the raw data contains only the result of the substraction.

The usefulness of this feature is a different issue.
« Last Edit: November 19, 2008, 09:35:45 PM by Panopeeper » Logged

Gabor
mike.online
Full Member
***
Offline Offline

Posts: 207


WWW
« Reply #5 on: November 19, 2008, 10:04:15 PM »
ReplyReply

I sometimes get single red pixels (look like dead pixels almost...) sometimes in long exposures and dark images. would this negate that? (Canon 30D)
« Last Edit: November 19, 2008, 10:04:47 PM by mike.online » Logged

Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #6 on: November 19, 2008, 11:03:27 PM »
ReplyReply

Quote from: mike.online
I sometimes get single red pixels (look like dead pixels almost...) sometimes in long exposures and dark images. would this negate that? (Canon 30D)

My experience is, that the Canon in-camera substraction is brutal; the result of substracting such hot pixels can be a black hole (literally). One can achieve a better result by creating a black frame shot with the same shutter time (cover the lens AND the viewfinder), and substracting that in PS as a layer. That way one has control over the substraction.
Logged

Gabor
NikosR
Sr. Member
****
Offline Offline

Posts: 622


WWW
« Reply #7 on: November 19, 2008, 11:11:04 PM »
ReplyReply

Quote from: Schewe
Not if you are using Camera Raw, yes if you use Nikon's software.

Nikon's software can honour the in camera settings (not only NR) if asked to. I thought I should clarify since your answer leaves room for misinterpretation. So the raw data are not affected as they come out of the camera but the software can apply (on request) similar adjustments as the in-camera jpeg engine would to a jpeg file.
« Last Edit: November 19, 2008, 11:15:24 PM by NikosR » Logged

Nikos
jvora
Jr. Member
**
Offline Offline

Posts: 90


« Reply #8 on: November 20, 2008, 02:06:43 AM »
ReplyReply

Quote from: Panopeeper
Note, that the Long Exposure Noise Reduction DOES affect the raw data; that is a completely different matter.


Thanks Panopeeper -

Appreciate this important clarification !

Any verdict if one should keep NR On for Long Exposures ??


Thanks all for the replies  !!


Jai
Logged
edwinb
Full Member
***
Offline Offline

Posts: 143


WWW
« Reply #9 on: November 20, 2008, 03:11:35 AM »
ReplyReply

Quote from: jvora
Thanks Panopeeper -

Appreciate this important clarification !

Any verdict if one should keep NR On for Long Exposures ??


Thanks all for the replies  !!


Jai
as noise grows with time it becomes more significant with longer exposures and noise reduction techniques more likly to be beneficial
Edwin
« Last Edit: November 20, 2008, 03:12:16 AM by edwinb » Logged

Edwin Blenkinsopp
Technical Manager
image2output
m:  +44 (0) 7836 674749
e: edwin.blenkinsopp@image2output.com
w: Sinar | Foba | Inkjet
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #10 on: November 20, 2008, 10:51:16 AM »
ReplyReply

I forgot to post a demo for the effect of in-camera long exposure noise reduction.

The shot was only a few seconds long IIRC. The crop is from a turned off TV, thus the monotone appearance.

The first crop is w/o Long Exposure NR, the second one is with; look at the black hole. I have not saved the "manual" correction. I made a second shot and overlayed that as a layer in PS, as substraction. The result was pretty much the same as in-camera. Then I added a curve to the correction layer and adjusted it until the result was smooth.



Logged

Gabor
PierreVandevenne
Sr. Member
****
Offline Offline

Posts: 510


WWW
« Reply #11 on: November 20, 2008, 10:59:08 AM »
ReplyReply

Probably not that interesting for "normal" photography, but from an astronomical photography point of view, here is the analysis of C. Buil.

http://www.astrosurf.com/buil/nikon_test/test.htm
Logged
barryfitzgerald
Sr. Member
****
Offline Offline

Posts: 602


« Reply #12 on: November 26, 2008, 04:04:48 AM »
ReplyReply

ACR is doing noise reduction by default. No questions...

So shooting raw, you are stuck with Thomas Knoll's base level NR, which IMO is pretty terrible. See the samples below..ACR is hands down beaten with picasa.., you can see the effects of NR on ACR, sliders to 0 on luminance noise


ACR via Lightroom

[attachment=9876:LRACR.jpg]


Picasa


[attachment=9877:Picasa.jpg]

You have no way of avoiding this..sadly.

If you want the best quality high ISO work, ACR is not a good choice. Seems the folks at adobe are perfectly "ok" with being beaten with a freebie program..hmmmm


« Last Edit: November 26, 2008, 04:06:41 AM by barryfitzgerald » Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2108


« Reply #13 on: November 26, 2008, 09:15:33 AM »
ReplyReply

The real question is why you are using CR noise reduction of Luminance = 0 for high ISO work.
Logged

bjanes
Sr. Member
****
Offline Offline

Posts: 2794



« Reply #14 on: November 26, 2008, 09:52:37 AM »
ReplyReply

Quote from: barryfitzgerald
ACR is doing noise reduction by default. No questions...

So shooting raw, you are stuck with Thomas Knoll's base level NR, which IMO is pretty terrible. See the samples below..ACR is hands down beaten with picasa.., you can see the effects of NR on ACR, sliders to 0 on luminance noise

If you want the best quality high ISO work, ACR is not a good choice. Seems the folks at adobe are perfectly "ok" with being beaten with a freebie program..hmmmm

To me, your Picassa image looks soft and over filtered. Personally, I prefer the ACR image.

I've had excellent results with ACR with high ISO files. It is generally agreed that a dedicated NR program is best for heavy duty noise reduction. Personally, I use NoiseWare and turn off all NR and sharpening in ACR. If you don't want to store intermediate images, the process can be done via a Photoshop action.

Here is your ACR image with NR via Noiseware applied, using the defaults:

[attachment=9885:LRACR_NW.jpg]
« Last Edit: November 26, 2008, 09:54:17 AM by bjanes » Logged
barryfitzgerald
Sr. Member
****
Offline Offline

Posts: 602


« Reply #15 on: November 26, 2008, 10:43:51 AM »
ReplyReply

The point I made was that ACR leaves artifacts, simply because it does NR itself. This is a user choice, and should not be left to Adobe.

The picasa one, has a much  finer grain pattern, which is a better starting point. You can use noise reduction programs, but you get a better result by not using ACR in the ist place. ACR applies by default more contrast and sharpening than picasa does. Hence the different look.

Hence the softer look from picasa.

But what amazes me, is how complacent users can be, when you can clearly see the mess ACR makes of high ISO images. The blame goes firmly to Mr Knoll, who should not even be applying any kind of noise filtering or reduction anyway. The problem ist came to light with LR 1.1, it has been toned down since then..but it has not been fixed.

I am amazed anyone could prefer the ACR one...
There are plenty of other programs out there, for example raw therapee, that show you just how much adobe are playing about with the noise reduction, when they should be doing NONE at all.

The reason for not doing any luminance NR, was to show how different they look. And because I don't dislike a fine film like grain pattern for high ISO images.


And in case anyone thinks it's "just my camera" have a look here.


http://forums.dpreview.com/forums/read.asp...essage=30145534

I am going to keep moaning until Adobe get a grip on the problem, join me, and let's get something done!
« Last Edit: November 26, 2008, 01:25:56 PM by barryfitzgerald » Logged
Samotano
Jr. Member
**
Offline Offline

Posts: 88



WWW
« Reply #16 on: November 26, 2008, 11:21:03 PM »
ReplyReply

Quote from: barryfitzgerald
The point I made was that ACR leaves artifacts, simply because it does NR itself. This is a user choice, and should not be left to Adobe.
You may well have a point, but either your methodology is nonsensical or your language is unclear (or both).  I did not do any tests (formal or informal), but it would take more showing on your side to convince me that noise reduction (or lack of) in acr creates such artifacts.  Besides, are we really talking about noise reduction here or perhaps different demosaiking algorithms?
« Last Edit: November 26, 2008, 11:22:03 PM by Samotano » Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5474


WWW
« Reply #17 on: November 26, 2008, 11:37:11 PM »
ReplyReply

Quote from: barryfitzgerald
I am going to keep moaning until Adobe get a grip on the problem, join me, and let's get something done!

Well, Pissin&Moaning™ ain't gonna move much, ya know? It's useful to prove your case and the fact is, that thread did nothing (other than illuminate the fact you don't like censorship).

Noise ain't usable micro-detail...if you are evaluating image detail at 1:1 or larger, you aren't seeing "reality" you're seeing reality magnified x2 up to x4 (or more) and all the stuff people "think" they are seeing isn't usable data...not on prints and not online (unless you routinely crop to a tiny portion of an image or post full rez files on a web site).

I find it humorous when people work themselves into a tizzy about things they think they are seeing when what they are seeing ain't what they're getting. A modern LCD has about a 100 PPI resolution...if you are viewing an image at 1:1 (one screen pixel for one image pixel) you are putting a 2x-4x loupe on all that stuff (even worse if you zoom into 2:1 or higher). And while it's useful to evaluate sharpening settings and noise reduction at those zooms, they aren't showing you real image detail but enlarged detail. Detail at those zooms will NEVER print on paper even if you are outputting at 480PPI on glossy paper. At 480PPI, the image on a display at 1:1 is 4.8x bigger (and if you reduce the size, even to 25% the display is still a low resolution devise).

The only way to evaluate images intended for printing is to print them...and in tests, Camera Raw with "proper" capture sharpening and noise reduction performs really well. So all these posting of 1:1 or higher crops on web sites is really much ado about nothing.

If you think Camera Raw is deficient in dealing with noisy images, prove it. Say something is so ain;t the same as being so, ya know?

:~)
« Last Edit: November 26, 2008, 11:39:17 PM by Schewe » Logged
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #18 on: November 26, 2008, 11:44:53 PM »
ReplyReply

Quote from: barryfitzgerald
I am going to keep moaning until Adobe get a grip on the problem, join me, and let's get something done!
Your complaint would be better placed on the respective Adobe forum. There you could learn from earlier related threads that ACR's "basic" noise reduction is part of the demosaicing. There were lots of complaints IIRC when version 4.1 came out;  the next version reduced the inherent (i.e. not switchable) noise reduction, but some is still there. The effect is, that very fine details are eliminated.
Logged

Gabor
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #19 on: November 26, 2008, 11:46:42 PM »
ReplyReply

Quote from: Schewe
If you think Camera Raw is deficient in dealing with noisy images, prove it.
This has been done numerous times on the Adobe ACR user forum.
Logged

Gabor
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad