In-camera processing of long-exposure RAW data

Started Jan 21, 2010 | Discussions
Iliah Borg Forum Pro • Posts: 26,501
Re: Correction to Number of neighbors used

Are you talking generically or nikon specific?

Is banding a general thing, or is it Nikon-specific?

-- hide signature --
Anthony Medici Veteran Member • Posts: 5,725
Re: The Sins of the Algorithm

eNo wrote:

That is truly disturbing. Would a fix be that the pixel should be declared "hot" only if it exceeds a certain threshold and it doesn't seem in line with its neighbors. If not, what fix would you propose to accomplish the intent (eliminate hot pixels) without unwanted/unnecessary averaging?

To me, a pixel is hot if it appears consistently over multiple images and even this is a problem if you are taking images to stack for Astrophotography.

Another test for a hot pixel would be a pixel above a certain value in a known dark frame exposure. If Nikon were to allow us to take a controlled dark frame (they don't open the shutter, record the values for the shutter speed and ISO) they could then use this control frame to removed hot pixels. I could even see that they could do this on startup for a known ISO and shutter speed and then use that same control as long as the camera stays on.

Pixel averaging simply helps to destroy detail that might have been captured in the image.

 Anthony Medici's gear list:Anthony Medici's gear list
Nikon D800E Nikon D7100 Nikon 1 AW1 Nikon D810 Nikon AF-S Nikkor 500mm f/4G ED VR +1 more
Anthony Medici Veteran Member • Posts: 5,725
Re: The Sins of the Algorithm

jim stirling wrote:

I would say that absolute critical image accuracy falls into a distant second place compared to delivering an aesthetically pleasing image . And in fact , images that are a more technically accurate when compared to the original scene are often much less aesthetically pleasing. And this can be evidenced by the extent of in camera or post processing adjustments that the vast majority of photographers do to ensure pleasing results as opposed to technical perfection.

Wouldn't that be something the photographer should decide on not the manufacturer?

And I would be genuinely interested in an a real life sample image where this issue shows significant degradation to the end result. If the problem is significant then it should be easily demonstratable . I do mainly event wedding shooting using D3s D700 and landscape work with the D3x and I am very happy with the results. And the D3x in particular seems to receive almost universal praise for its image quality { in direct comparison to the other 20mp+ models from the other makers} . I do wonder about the real life effects of this.

You mean beyond scientific work or images that include the night sky. It sounds like as long as there are little bright points in the sky you would be happy. It doesn't matter if they are the right brightness in comparison to each other or in the right places.

 Anthony Medici's gear list:Anthony Medici's gear list
Nikon D800E Nikon D7100 Nikon 1 AW1 Nikon D810 Nikon AF-S Nikkor 500mm f/4G ED VR +1 more
William Carson
William Carson Veteran Member • Posts: 6,205
Re: Hello William

Hi , this is in response to the aspect of the comments of a couple of posters in this thread that expressed concern about the treatment of hot pixels and Nikon neither informing nor giving an option to turn off the hot pixel surpression-adjusting algorithm that was being applied. The linked images were from a D3 and exposure times were 90 seconds at ISO of 2500 - 3200 and NR was turned off and it was RAW F/ was from 4 to F/8 . One was F/16 Read through linked thread to get more details. The Images show no reason for image quality concern even at lengthy exposure time.
Will

cluna Senior Member • Posts: 1,403
Re: In-camera processing of long-exposure RAW data

Thom Hogan wrote:

I've long written about trying to capture "optimal data" in the field. One thing I used to be able to do is set Long exp. NR to On, take a long exposure, then flip the camera off before NR had completed. This allowed me to bypass the "filter." Apparently, this is now "fixed" by Nikon's latest firmware and not possible.

May not work on the D3s but do you think the D300s with 14bit mode on might? Reasoning is since the A/D is off chip with the 300s and 14bit mode on...just a hackers guess of possibility.

-C

cluna Senior Member • Posts: 1,403
Re: Yes, that's a valid summary

Marianne Oelund wrote:

I try to avoid calling this NR, which I consider somewhat misleading. Its intended function is to identify and suppress hot pixels; the fact that it does somewhat more than that, is a side-effect of its inadequate design.

Marianne,

Is this possibly Nikons attempt to stabilize the sensor for the ambient temperature differences we might encounter in the field? If so it is a reasonable approach at zeroing the sensor. Id prefer that 'value' just be stored in the exif data rather than RAW being adjusted.

(Im approaching this from the perspective the nikon engineers arent dummies and didnt incur more processing overhead on EXPEED because they wanted to)

-C

OP Marianne Oelund Veteran Member • Posts: 7,779
Star trails do not invoke the algorithm

William Carson wrote:

Hi , this is in response to the aspect of the comments of a couple of posters in this thread that expressed concern about the treatment of hot pixels
.
.

The Images show no reason for image quality concern even at lengthy exposure time.

When star images are trailed, such as in these images taken with no clock-drive astro mount, the hot-pixel algorithm will not affect them.

The problem does become apparent in actual frames when star images are tightly confined, of low brightness, and well focused.

Thom Hogan Forum Pro • Posts: 13,659
Re: The Sins of the Algorithm

Anthony Medici wrote:

It sounds like as long as there are little bright points in the sky you would be happy. It doesn't matter if they are the right brightness in comparison to each other or in the right places.

Specular highlights can fall into this category if you're the right distance away from them. Indeed, that's one of the places where I first noticed the problem: in a pre-sunrise image with moonlight, I noticed that at some point the specular highlights in the water disappeared. Well, it was because they became a single pixel. This made for an artificial and strange looking transition. Those that think this is just an astronomy thing are mistaken.

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (21 and counting)
http://www.bythom.com

OP Marianne Oelund Veteran Member • Posts: 7,779
Example of detail loss and color shift

jim stirling wrote:

I would say that absolute critical image accuracy falls into a distant second place compared to delivering an aesthetically pleasing image . And in fact , images that are a more technically accurate when compared to the original scene are often much less aesthetically pleasing. And this can be evidenced by the extent of in camera or post processing adjustments that the vast majority of photographers do to ensure pleasing results as opposed to technical perfection.

I'm quite sure that no one is doing PP adjustments which corrupt their images in the manner that the hot-pixel suppression algorithm does.

And I would be genuinely interested in an a real life sample image where this issue shows significant degradation to the end result. If the problem is significant then it should be easily demonstrable . I do mainly event wedding shooting using D3s D700 and landscape work with the D3x and I am very happy with the results.

Here's an example in the vein of your work. These are 100% crops, but the differences are still readily visible at 50%. I shouldn't need to explain which is the 1/5 sec exposure, and which is 1/4 sec:

OP Marianne Oelund Veteran Member • Posts: 7,779
Demonstration of Modified Algorithm

I've created a modified version of the hot-pixel suppression algorithm, and used it to process the same image file which I presented in the original post. A few hot pixels were artificially added to the original file in order to test the modified algorithm's ability to suppress them. It isn't designed to remove hot pixels entirely; you will see that there are vestiges, but they're not visually distracting. Otherwise, the input file is largely undisturbed, in contrast to the camera's result.

Here is the result, alongside the camera's "After processing" image for comparison:

geo444
geo444 Contributing Member • Posts: 510
Bravo Marianne !!!

Marianne Oelund wrote:

I've created a modified version of the hot-pixel suppression algorithm, and used it to process the same image file which I presented in the original post. A few hot pixels were artificially added to the original file in order to test the modified algorithm's ability to suppress them. It isn't designed to remove hot pixels entirely; you will see that there are vestiges, but they're not visually distracting. Otherwise, the input file is largely undisturbed, in contrast to the camera's result.

Here is the result, alongside the camera's "After processing" image for comparison:

Bravo Marianne !!!

have still to convince Nikon....

Yamo Junior Member • Posts: 43
Re: The Sins of the Algorithm

Thom Hogan wrote:

Anthony Medici wrote:

It sounds like as long as there are little bright points in the sky you would be happy. It doesn't matter if they are the right brightness in comparison to each other or in the right places.

Specular highlights can fall into this category if you're the right distance away from them. Indeed, that's one of the places where I first noticed the problem: in a pre-sunrise image with moonlight, I noticed that at some point the specular highlights in the water disappeared. Well, it was because they became a single pixel. This made for an artificial and strange looking transition. Those that think this is just an astronomy thing are mistaken.

The way Marianne describes the algorithm it needn't even be bright points. Virtually any texture - sand, fabric, etc. is going to have many, many single pixel local maxima with the visual impact of overall reduced micro-contrast of the texture.

Imagine the image data as a height field (a surface with each pixel value represented as height)... Now, take a hammer to all the peaks and flatten them down to their nearest neighbor (height-wise).

Take any photo (taken on the shorter side of 1/4 second) and zoom in until you see individual pixels as squares 8 to 10 screen pixels to one image pixel. Any individual brighter pixel you see would have been hammered with this algorithm. Try individual channels

And... demosaicing will spread the impact over, what, an order of magnitude more pixels? It's just so bad...

-Yamo-

jim stirling
jim stirling Veteran Member • Posts: 7,356
Re: Example of detail loss and color shift

Marianne Oelund wrote:

And I would be genuinely interested in an a real life sample image where this issue shows significant degradation to the end result. If the problem is significant then it should be easily demonstrable . I do mainly event wedding shooting using D3s D700 and landscape work with the D3x and I am very happy with the results.

Here's an example in the vein of your work. These are 100% crops, but the differences are still readily visible at 50%. I shouldn't need to explain which is the 1/5 sec exposure, and which is 1/4 sec:

Marianne ,

Thank you for posting an image that demonstrates the effect , but I would still conclude that although the difference is visible on this static shot compared directly with another taken with the exact same framing and lighting, and with the only variable being shutter speed. In real life shot to shot of a highly reflective material, even the slightest of subject or photographer movement or any variation in light intensity or angle, would result in a completely different pattern of reflected highlights to the extent that the suppression of what seems on the whole lower magnitude highlights would never be noticed. I am not saying this could not be detrimental to certain specific photographic applications with astro photography being again a good example, but it is not an issue that I would be too concerned about.
Jim

 jim stirling's gear list:jim stirling's gear list
Panasonic FZ1000 Panasonic Lumix DMC-GF1 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GH4 Nikon D810 +12 more
geo444
geo444 Contributing Member • Posts: 510
and now Ladies Gentlemen & Moon Maids... The Next Question is ?...

.

is that bad median filter algorithm applied by NiKon BEFORE writing the NEFs :

  • 1 Hardware implemented ?... inside the low level DSP ??... Aie !!!

  • 2 Software ?... just a Firmware Correction ??

let's hope so ???

ejmartin Veteran Member • Posts: 6,274
Re: Example of detail loss and color shift

jim stirling wrote:

Marianne ,

Thank you for posting an image that demonstrates the effect , but I would still conclude that although the difference is visible on this static shot compared directly with another taken with the exact same framing and lighting, and with the only variable being shutter speed. In real life shot to shot of a highly reflective material, even the slightest of subject or photographer movement or any variation in light intensity or angle, would result in a completely different pattern of reflected highlights to the extent that the suppression of what seems on the whole lower magnitude highlights would never be noticed.

OK, I get it -- bad technique trumps bad firmware.

-- hide signature --
Thom Hogan Forum Pro • Posts: 13,659
Re: Example of detail loss and color shift

jim stirling wrote:

Thank you for posting an image that demonstrates the effect , but I would still conclude that although the difference is visible on this static shot compared directly with another taken with the exact same framing and lighting, and with the only variable being shutter speed.

The direct conclusion of your argument is that "lower quality cameras are okay." Taken to its extreme, your argument means that instead of saving up to buy that D3x you can safely get buy with a D3000. After all, the difference is only visible if you have taken a shot with a D3x and compared it to the D3000 shot.

Remember, we're paying US$5200 and up for these cameras. We do so because we expect them to produce the best possible results. Indeed, in some cases our clients demand it.

In real life shot to shot of a highly reflective material, even the slightest of subject or photographer movement or any variation in light intensity or angle, would result in a completely different pattern of reflected highlights

Two things. As others said, you're endorsing poor technique. But your assertion that "a different pattern of highlights" would "never be noticed" is way off the mark. In both cases, the highlights would not be recorded at shutter speeds greater than 1/4. As I noted before, it comes into play with landscape photography, and I've seen the problem directly in my images: long exposures mask detail. Might as well be applying heavy noise reduction to the image.

but it is not an issue that I would be too concerned about.

This says more about your level of photography than it does about the product being tested. I see a lot of this "I can't see it so it doesn't exist" type of argument. But it clearly does exist, and to those aspiring to the highest level of work, it's a clear problem (or perhaps I should say it's a problem of clarity ;~). Note that the competitors cameras don't do this.

Yes, I can just see Ansel Adams saying "you know, I really didn't need that highlight detail, I guess Tri-X is good enough developed normally."

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (21 and counting)
http://www.bythom.com

Thom Hogan Forum Pro • Posts: 13,659
Re: and now Ladies Gentlemen & Moon Maids... The Next Question is ?...

geo444 wrote:

is that bad median filter algorithm applied by NiKon BEFORE writing the NEFs :

  • 1 Hardware implemented ?... inside the low level DSP ??... Aie !!!

To my knowledge, it is applied by the ASIC after the initial data is put into the buffer as a "file" but before the file itself is stored on the card. That's why it used to be that you could invoke Long exp NR and interrupt that to get the original data intact: the file was there, but it hadn't been retouched by the ASIC yet.

Even if we can convince Nikon to change the algorithm, I don't think this is going to be a firmware update type issue.

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (21 and counting)
http://www.bythom.com

jim stirling
jim stirling Veteran Member • Posts: 7,356
Re: Example of detail loss and color shift

ejmartin wrote:

jim stirling wrote:

Marianne ,

Thank you for posting an image that demonstrates the effect , but I would still conclude that although the difference is visible on this static shot compared directly with another taken with the exact same framing and lighting, and with the only variable being shutter speed. In real life shot to shot of a highly reflective material, even the slightest of subject or photographer movement or any variation in light intensity or angle, would result in a completely different pattern of reflected highlights to the extent that the suppression of what seems on the whole lower magnitude highlights would never be noticed.

OK, I get it -- bad technique trumps bad firmware.

Whilst I am more than happy to bow to both your and Marianne 's superior knowledge of sensors , I do my work in real life scenarios. As with most event type photography I do a lot of my work handheld as the option of using tripods is just not viable. So with regards to my "bad technique" I would be delighted to take advice how exactly it would be possible to get identical back to back images , when the camera is hand held and my subject is alive especially in relation to reflective surfaces.

So in this instance and for the vast majority of my work reality triumphs over an "issue" that at worst has insignificant impact that would vary from shot to shot or a more likely scenario never be noticed at all. At no time have I contested the validity of the observation { I come from a scientific background } my only observation is in relation to the many typical shooting scenarios relevant to my shooting. And as i posted i am aware that for some niche areas it could indeed be an important issue.
Jim

 jim stirling's gear list:jim stirling's gear list
Panasonic FZ1000 Panasonic Lumix DMC-GF1 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GH4 Nikon D810 +12 more
jim stirling
jim stirling Veteran Member • Posts: 7,356
Re: Example of detail loss and color shift

Thom Hogan wrote:

jim stirling wrote:

Thank you for posting an image that demonstrates the effect , but I would still conclude that although the difference is visible on this static shot compared directly with another taken with the exact same framing and lighting, and with the only variable being shutter speed.

The direct conclusion of your argument is that "lower quality cameras are okay." Taken to its extreme, your argument means that instead of saving up to buy that D3x you can safely get buy with a D3000. After all, the difference is only visible if you have taken a shot with a D3x and compared it to the D3000 shot.

Remember, we're paying US$5200 and up for these cameras. We do so because we expect them to produce the best possible results. Indeed, in some cases our clients demand it.

In real life shot to shot of a highly reflective material, even the slightest of subject or photographer movement or any variation in light intensity or angle, would result in a completely different pattern of reflected highlights

Two things. As others said, you're endorsing poor technique. But your assertion that "a different pattern of highlights" would "never be noticed" is way off the mark. In both cases, the highlights would not be recorded at shutter speeds greater than 1/4. As I noted before, it comes into play with landscape photography, and I've seen the problem directly in my images: long exposures mask detail. Might as well be applying heavy noise reduction to the image.

but it is not an issue that I would be too concerned about.

This says more about your level of photography than it does about the product being tested. I see a lot of this "I can't see it so it doesn't exist" type of argument. But it clearly does exist, and to those aspiring to the highest level of work, it's a clear problem (or perhaps I should say it's a problem of clarity ;~). Note that the competitors cameras don't do this.

Well I have not had any experience of your photographic work only as a technical writer so I will not comment on this .As to my level of photography I have been a professional photographer for many years and have won several major awards including a number of national awards. And at the risk of giving the technical advisers a heart attack I would conclude that the aesthetics of the final image are paramount .And in fact if someone is looking at your work and counting the highlights reflected on a river I think you have failed. I have at no point contested the validity of the data , only made it clear that I see little evidence of it causing any significant impact on the quality of the results I can achieve.
Jim

 jim stirling's gear list:jim stirling's gear list
Panasonic FZ1000 Panasonic Lumix DMC-GF1 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GH4 Nikon D810 +12 more
bob elkind Veteran Member • Posts: 5,815
Re: Example of detail loss and color shift

jim stirling wrote:
...

... And at the risk of giving the technical advisers a heart attack I would conclude that the aesthetics of the final image are paramount. And in fact if someone is looking at your work and counting the highlights reflected on a river I think you have failed. I have at no point contested the validity of the data , only made it clear that I see little evidence of it causing any significant impact on the quality of the results I can achieve.
Jim

Jim,

The point I would like to make is that the aesthetic quality you've achieved has been accomplished in spite of this little Nikon image processing feature, not as a result of this feature. That's not a selling point that Nikon would be proud to make, nor one that will attract many serious photographers.

This is a 'feature' I would rather control on my own, and I see no compelling reason why this 'feature' should not be user selected rather than 'always on' .

-- hide signature --

Bob Elkind
Family,in/outdoor sports, landscape, wildlife
photo galleries at http://eteam.zenfolio.com
my relationship with my camera is strictly photonic

 bob elkind's gear list:bob elkind's gear list
Nikon D700 Canon EOS 500D Nikon D600 Canon EF-S 18-55mm f/3.5-5.6 Nikon AF Nikkor 50mm f/1.8D +5 more
Keyboard shortcuts:
FForum MMy threads