Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-07-2014, 05:05 AM   #1
Site Supporter
Site Supporter
Stone G.'s Avatar

Join Date: Jun 2010
Location: North Zealand, Denmark
Photos: Gallery | Albums
Posts: 1,516
Diffraction Compensation - What and How?

I was studing the Pentax/Ricoh Pentax 645z product description and stumbeled over this line:
  • Automatic compensation of lens distortion, lateral chromatic aberration, brightness level at edges, and diffraction
Now, I know what 'diffraction' is and what 'diffraction limited' means. But can anyone explain to me what 'automated diffraction compensation' does and how it is done?

07-07-2014, 05:22 AM   #2
Veteran Member




Join Date: Jul 2010
Location: Lyngby, Copenhagen
Photos: Albums
Posts: 742
I'll guess that it's a form of sharpening tailored to counter exactly the diffraction blur at the aperture and focal length (end whatever else determines diffraction - perhaps aperture shape, determined by an in-camera lookup table).

Regards,
--Anders.
07-07-2014, 05:25 AM   #3
Pentaxian




Join Date: Nov 2011
Photos: Gallery | Albums
Posts: 3,764
QuoteOriginally posted by Stone G. Quote
But can anyone explain to me what 'automated diffraction compensation' does and how it is done?
I suspect that it's just like on the Olympus EM1:

"According to Olympus, the EM1 now applies additional sharpening to images with small aperture to reduce the effect."

Olympus OMD EM1 diffraction compensation | Cameralabs
07-07-2014, 05:26 AM   #4
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,895
Some interesting prior disussion of this here, and some links:
https://www.pentaxforums.com/forums/172-pentax-k-3/266419-rumoured-k-3-f-w-up...provement.html

Not only will diff.comp be in the 645Z, but also likely be in the next K-3 firmware update too.

07-07-2014, 05:55 AM   #5
Pentaxian
Na Horuk's Avatar

Join Date: Feb 2012
Location: Slovenia, probably
Photos: Gallery | Albums
Posts: 10,934
Interesting. But is this for jpeg-only, or does it affect raw files in any way?
Hope this makes it to the Q series, as well
07-07-2014, 06:20 AM   #6
Site Supporter
Site Supporter
Stone G.'s Avatar

Join Date: Jun 2010
Location: North Zealand, Denmark
Photos: Gallery | Albums
Posts: 1,516
Original Poster
QuoteOriginally posted by Na Horuk Quote
......does it affect raw files in any way?
If it does affect the raw files, I would hope that it is a matter of on-off selection - I still like the idea of having my raw files as "raw" as possible.
07-07-2014, 07:01 AM   #7
Veteran Member
Imageman's Avatar

Join Date: Dec 2013
Photos: Albums
Posts: 461
If there's any truth to this I wouldn't want it.


Diffraction effect as we know is a lens aperture effect that causes loss of sharpness or increased blur, however you want to call it.


There is no such thing as sharpening a digital image, there is only local edge contrast adjustment which has the appearance of sharpening.


This can only be a measured increase in local contrast dependent on the aperture used, or in other words a progressively more fierce sharpening of the image as the aperture reduces. I have a dislike of such automated effects.


Sharpening is supposed to be the last thing you do in a professional workflow and here it will be the very first thing that is done to the image.


This is totally contrary to good established practice and I want none of it.


I don't even see that it can give any benefit over lets say unsharp mask correctly applied at the end of the workflow.


Its a marketing exercise.
07-07-2014, 07:31 AM   #8
Pentaxian
Na Horuk's Avatar

Join Date: Feb 2012
Location: Slovenia, probably
Photos: Gallery | Albums
Posts: 10,934
QuoteOriginally posted by Imageman Quote
I don't even see that it can give any benefit over lets say unsharp mask correctly applied at the end of the workflow.
Well, hopefully the benefit is that the engineers take into account many things that we might not, as someone stated above (sensor and lens properties, aperture and focal length, etc.) and apply advanced algorithms and not merely "+2 to sharpening). Similarly to distortion, vignetting and CA correction - quite handy if you shoot jpeg, and nobody knows the lenses and cameras better than the Pentax engineers (at least, I hope so!)
But yes, of course, it should be an option that you can turn on or off.

07-07-2014, 07:50 AM   #9
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 26,256
QuoteOriginally posted by Imageman Quote
If there's any truth to this I wouldn't want it.


Diffraction effect as we know is a lens aperture effect that causes loss of sharpness or increased blur, however you want to call it.


There is no such thing as sharpening a digital image, there is only local edge contrast adjustment which has the appearance of sharpening.


This can only be a measured increase in local contrast dependent on the aperture used, or in other words a progressively more fierce sharpening of the image as the aperture reduces. I have a dislike of such automated effects.


Sharpening is supposed to be the last thing you do in a professional workflow and here it will be the very first thing that is done to the image.


This is totally contrary to good established practice and I want none of it.


I don't even see that it can give any benefit over lets say unsharp mask correctly applied at the end of the workflow.


Its a marketing exercise.
If you think a bit about what diffraction is....light bending as it passes a solid edge...( or even different temperatures in the air for that matter) you know different wavelengths diffract differently...it might be possible to shift the light by wave length, in such a matter that the light that was at one place together at the edges that caused the diffraction, is back together after a digital shift. Thus it is theoretically possible that the altered image would be better than an unaltered image, both sharper and a more accurate rendition. Whether in fact that has happened, let's not assume that things cannot be done to correct both sharpness due to diffraction and correct CA, that don't actually improve the image.

The complexity of the algorithm would be mind boggling, at least to me, but lets not assume some smart guy hasn't done it. The proof will be in side by side comparison's of images, not theoretical postulations on the feasibility of the technology by the un-imaginative.
07-07-2014, 08:04 AM   #10
Site Supporter
Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 16,434
QuoteOriginally posted by Imageman Quote
If there's any truth to this I wouldn't want it.


Diffraction effect as we know is a lens aperture effect that causes loss of sharpness or increased blur, however you want to call it.


There is no such thing as sharpening a digital image, there is only local edge contrast adjustment which has the appearance of sharpening.


This can only be a measured increase in local contrast dependent on the aperture used, or in other words a progressively more fierce sharpening of the image as the aperture reduces. I have a dislike of such automated effects.


Sharpening is supposed to be the last thing you do in a professional workflow and here it will be the very first thing that is done to the image.


This is totally contrary to good established practice and I want none of it.


I don't even see that it can give any benefit over lets say unsharp mask correctly applied at the end of the workflow.


Its a marketing exercise.
I assume that what you describe is not what Pentax is doing. Adding sharpening to a RAW file makes no sense and adds nothing, of course. If there is some sort of algorithm that can compensate for softness due to being shot at f22 then it does make sense. I doubt, though, that there could be detail that could be added in to the image that wasn't there originally...
07-07-2014, 09:02 AM   #11
Veteran Member
Imageman's Avatar

Join Date: Dec 2013
Photos: Albums
Posts: 461
The true cost of manufacturing a k3 is likely to be 300 dollars. a hifi that sold for 80 dollars was usually manufactured for 20 dollars, a car that sells for 18,000 dollars usually costs less than 5,000 dollars to build. This wont be some sophisticated and costly wizardry, it will be a cheap attempt to make fuzzy images sharper without compromising the sale price of what is already a high priced camera.


The 645z is even more expensive so to hike the sale price significantly to introduce another new technology would be folly.


I have never seen a convincing explanation of truly sharpening up an image that is already out of focus, what's lost is lost and unrecoverable. Taking a blurred image and applying a software fix that lets you see details that were smudges is the province of Hollywood dream makers.


As for suggesting that anyone who doubts the wizardry of technology is unimaginative, that is somewhat short sighted and slightly unfair. most of this type of claim turns out to be simplistic in execution, im simply being realistic here.


Are we going to see a real breakthrough in sharpening already blurred images, I seriously doubt it. If that were possible, they wouldn't have had to spend countless millions changing the optics in the hubble when they were giving blurred images. They would just have used magical sharpening technology on those blurred out of focus images it produced. The reason they chose the 50 million dollar option was because it was the only option.


The only way I know of, to have a sharp image is by keeping it sharp, not by taking something already out of focus and making details that have already been lost reappear, I do know a little about optics so I think my position is a reasonable one.


I agree speculation has its place and theres no substitute for seeing the actual results, but I thought at this early stage speculation was welcome, that seems to be what everyone else is doing.


Im simply stating the other side of the case and my opinion, for what its worth.

---------- Post added 07-07-14 at 05:20 PM ----------

For you guys who thought I was wrong to suggest that diffraction compensation is likely to be simply additional sharpening at small apertures, I have found the Olympus method of diffraction compensation explained by Camera Labs and I quote it here:-


"The Olympus OMD EM1 becomes the first Olympus camera to offer compensation for the effect of optical diffraction, where choosing small apertures can result in the overall image becoming soft and losing detail. According to Olympus, the EM1 now applies additional sharpening to images with small aperture to reduce the effect"


This explanation sourced from Olympus according to the article, is exactly as I speculated, additional in camera sharpening when small apertures are used.
07-07-2014, 09:50 AM   #12
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 26,256
QuoteQuote:
I agree speculation has its place and theres no substitute for seeing the actual results, but I thought at this early stage speculation was welcome, that seems to be what everyone else is doing.
QuoteQuote:
As for suggesting that anyone who doubts the wizardry of technology is unimaginative, that is somewhat short sighted and slightly unfair. most of this type of claim turns out to be simplistic in execution, im simply being realistic here.
Just to play devils advocate... so which is it, what's realistic, or speculation? Those are two different things... or speculation on what's realistic, which is a really small subset of speculation? I find it hard to believe that a company is going to sharpen my raw files and call that limiting the effects of diffraction. And I wouldn't buy a camera based on that. For one thing, if you get CA or purple fringing, you need to turn your sharpening off...sharpening tends to make it worse. So if you make up for diffraction with sharpening, you've just blown your CA and fringin out of the water... so the only thing I can think here is that this is necessary to try and make up for the lower diffraction limits of smaller sensors, which is made worse by the even higher pixel densities, since diffraction and CA are measured most effectively compared to the width of the pixel. , but I'm not sure that would apply to the 645z, it's hardly a "smaller sensor", and doesn't have a particularly small pixel site. In fact it's low light performance would suggest large pixel backlit construction maximizing photon capture.

Just saying,

Last edited by normhead; 07-07-2014 at 10:29 AM.
07-07-2014, 01:31 PM   #13
Veteran Member
Imageman's Avatar

Join Date: Dec 2013
Photos: Albums
Posts: 461
Lol,


Come on now, its speculation because we haven't seen the hardware and what theyr doing so were speculating about it, its realistic because ive seen it all before, ive worked in manufacturing and companies make trivial improvements and call them fabulous and marvelous when they'r not. Its all double talk and smoke and mirrors.


Im trying to present the voice of reason suggesting that what you fear, is exactly whats going to happen.


You nailed it when you said:-


"I find it hard to believe that a company is going to sharpen my raw files and call that limiting the effects of diffraction."


And I suggest that's exactly whats going to be done and presented as the solution. I refer you to the Olympus quote:-


"The Olympus OMD EM1 becomes the first Olympus camera to offer compensation for the effect of optical diffraction, where choosing small apertures can result in the overall image becoming soft and losing detail. According to Olympus, the EM1 now applies additional sharpening to images with small aperture to reduce the effect"


Notice they said "applies additional sharpening" to "reduce the effect", that means in manufacturing speak, massaging the output file to resemble a sharper image.


All this means is, beyond f11 they automatically sharpen and make it progressively more aggressive as the aperture reduces. If there had been anything novel or new other than the simple sharpening presented in the text, they would have lost no time advertising the fact. Consequently theres nothing other than simple sharpening.


You have far too much faith in these companies. They are introducing simplistic sharpening and calling it diffraction compensation.


We have to assume that Pentax will follow the lead of Olympus given that we now know how Olympus are dealing with it. Pentax will im sure do exactly the same. No other approach than this simplistic and almost worthless sharpening makes any sense given that the cat is out of the bag and they've said how they do the compensation.


But I will await the arrival of the Pentax hardware before making the final judgement.
07-07-2014, 02:42 PM   #14
Site Supporter
Site Supporter
Stone G.'s Avatar

Join Date: Jun 2010
Location: North Zealand, Denmark
Photos: Gallery | Albums
Posts: 1,516
Original Poster
QuoteOriginally posted by Imageman Quote
We have to assume that Pentax will follow the lead of Olympus given that we now know how Olympus are dealing with it. Pentax will im sure do exactly the same. No other approach than this simplistic and almost worthless sharpening makes any sense given that the cat is out of the bag and they've said how they do the compensation.
I agree on all of your comments about contrast enhancement along edges. Knowing nothing and being a layman, I am just speculating - perhaps way out in the blue:

Could THIS 'sharpening' perhaps have something to do with the way the signal is being sampled at higher f-numbers? I.e.: when you reach 'empty spatial magnification' you could sample over larger blocks of individual pixels????? That wouldn't add anything in terms of information contents, (but wouldn't detract any information either, for as long as you are in the region of empty magnification).

That just might result in "nicer-looking pictures" (at least if you are a pixel-peeper)???
07-07-2014, 03:06 PM   #15
Pentaxian




Join Date: Oct 2011
Location: Savannah, GA
Posts: 5,005
QuoteOriginally posted by Imageman Quote
The true cost of manufacturing a k3 is likely to be 300 dollars. a hifi that sold for 80 dollars was usually manufactured for 20 dollars, a car that sells for 18,000 dollars usually costs less than 5,000 dollars to build. This wont be some sophisticated and costly wizardry, it will be a cheap attempt to make fuzzy images sharper without compromising the sale price of what is already a high priced camera.


The 645z is even more expensive so to hike the sale price significantly to introduce another new technology would be folly.


I have never seen a convincing explanation of truly sharpening up an image that is already out of focus, what's lost is lost and unrecoverable. Taking a blurred image and applying a software fix that lets you see details that were smudges is the province of Hollywood dream makers.


As for suggesting that anyone who doubts the wizardry of technology is unimaginative, that is somewhat short sighted and slightly unfair. most of this type of claim turns out to be simplistic in execution, im simply being realistic here.
The day is coming, as they have already accomplished this in research on a small scale...I've seen it. There is a paper describing it with examples posted around here somewhere. What it was...was they used extremely simple optics -- one or two element lenses that produced crappy blurry images but then applied the knowledge of the way the light was spreading etc (point spread functions and things I don't understand) to "fix it" in the computer and end up with images that look they were shot with much higher-end lenses. I don't see any reason why such ideas could not be applied to diffraction. I've actually been dabbling with something similar myself, but with a totally different approach (since I don't understand point spread functions, etc). I've seen similar "fixing" of motion-blurred images by use of motion sensors at capture time to record the vibrations. The idea (put simply) in either case is that the light is supposed to be here, but it is somewhere else instead (but still captured), and so if you can model or have a profile of the way it went wrong, you can undo it and put the light back where it is supposed to be. Seems like making a diffraction profile of a particular lens at a particular aperture would actually be one of the easier and most obvious applications of such tech, and so it starts...
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
645z, apertures, automated, camera, compensation, diffraction, diffraction compensation, dollars, dslr, focus, image, images, olympus, pentax, photography, technology
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Rumoured K-3 f/w update: diffraction correction and mirror flap improvement JPT Pentax K-3 42 07-11-2014 10:34 AM
Flash Compensation, Overall Compensation and Ambient Lights Rashedul Flashes, Lighting, and Studio 3 01-11-2013 04:19 PM
Diffraction and our Q hnikesch Pentax Q 17 12-05-2012 10:59 PM
Hyperfocus and diffraction help. ladybug Troubleshooting and Beginner Help 14 05-13-2012 05:05 PM
Exposure compensation..umm what? Raptorman Pentax DSLR Discussion 14 01-18-2010 12:33 PM



All times are GMT -7. The time now is 11:02 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top