Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-09-2015, 11:38 AM - 4 Likes   #1
Forum Member




Join Date: Sep 2008
Location: throughout the US
Posts: 79
Is diffraction killing your 645Z images?

Much is made on the internet about the problem of sharpness-robbing diffraction which increases as a lens is stopped down to small apertures. From what you read by some testers you are virtually wasting your money buying a 645Z if you stop down much beyond f/8. This becomes very concerning for landscape photographers especially, where your composition in the field may dictate f/11, f/16, and even f/22 to realize your vision.

Yes, diffraction is real. But it is not the big boogeyman many believe it to be when considering the entire photographic chain from camera to lens to aperture setting to processing software to printing. For me, processing workflow can make diffraction a non-issue on the 645Z. The main element is proper sharpening, along with global and micro contrast settings.

What I've found most helpful is deconvolution sharpening (not typical edge sharpening) for the capture sharpening step. This is available in various software (Iridient Raw Developer is one I'm aware of), but I am especially fond of using it in Lightroom since that is my preferred workflow. In the sharpening panel, when you move the Detail slider all the way to the right (+100) it turns the algorithm into essentially all deconvolution (according to Jeff Schewe who helped design this part of LR). Start with amount between 40 to 50, radius 0.8, and masking at 0 (and raise the masking amount depending on the image content, such as if there is sky or smooth tones you don't want sharpened). The 645D, and even more the Z, will take large amounts of sharpening extremely well without visible artifacts. In addition to losing sharpness to diffraction, it also robs your image of micro and global contrast, but adding a few points of Contrast and Clarity in Lightroom nicely restores that.

Processing in this way allows my 645D and 645Z files to look truly excellent even at 100% pixel view, and a recent 54" wide print for a customer from the 645D showed how well this works for real-world saleable photographic art. The 645Z is even more diffraction-resistant when using this workflow, and I have not ceased to be amazed at the excellent sharp nuanced detail and contrast it's rendering when shot at f/16 and even f/22.

Below is a link to a full resolution 38MB test image I recently did showing the beautiful file quality obtainable at f/22 with the Pentax 67 55mm f/4 lens after Lightroom processing. I saved the jpeg in sRGB at 100% so it should come close to showing you the massive amount of extremely fine textural details while retaining a very natural look. I encourage you to check it out at 100% view.

Hope this is helpful info for someone, and that it gives landscapers the confidence to enjoy the 645D/Z fully at whatever aperture you need for a given scene, knowing you can largely restore your diffraction-affected files and make gallery quality prints of enormous size!

Here is the file:

https://www.dropbox.com/s/p2ixj1mkosv46wc/_IMG1021.jpg?dl=0

Cheers,
Ross


Last edited by SeattleDucks; 07-10-2015 at 11:44 AM. Reason: clarity
07-09-2015, 11:48 AM   #2
CDW
Veteran Member




Join Date: May 2011
Location: Big Island, Hawaii & Utah
Posts: 378
Lens reviews of P645 lenses have revealed many of them to resolve their highest resolutions at F8-F13. The sweet spots vary with format and with medium format many Pentax lenses are not their sharpess at F8. Otherwise, your workflow is logical and is similar to mine. I've found in practice the Z files require little sharpening. I regularly shoot at F16, sometimes higher, with no real world issues. My clients aren't pixel peepers looking at 100-200% views on an LCD display.
07-09-2015, 12:10 PM - 1 Like   #3
Veteran Member
MJSfoto1956's Avatar

Join Date: Dec 2013
Location: Boston
Posts: 1,287
"Back in the day" we generally shot f/22-f/16 for 4x5, f/16-f/11 for medium format and f/11-f/8 for 35mm film. Frankly not much has changed...

Michael
07-09-2015, 01:22 PM   #4
Veteran Member
Kolor-Pikker's Avatar

Join Date: Dec 2014
Posts: 341
Even without special sharpening settings, I get excellent results up to f/16 no problem.

While the 645Z should technically become diffraction limited at around f/9~10, in reality you're never going to have image content that pushes the limits of the sensor and optics short of a brightly-lit test chart accidentally ending up in your images.

Now going from f/16 to f22 is where things start to get visibly soft, but not by too much, especially if the added DoF ends up creating a greater impression of sharpness than the presence of actual micro-detail.

I like to think about digital sensors more in terms of pixel pitch rather than sensor size, since nowadays the size of a sensor has little to do with its actual performance; in this case the 645Z has 30MP if cropped down to 35mm, or 12MP if cropped down to APS-C, which aren't nearly the highest values photographers have had to deal with. The D800 cameras and Sony A7R are 36MP at 35mm size, with the coming A7RII being 42MP, while many APS-C cameras today range from 16~22MP of resolution, which is a good chunk more than the Z's equivalent.

Are all the photographers using these cameras more concerned about diffraction than the 645Z users?

07-09-2015, 01:33 PM   #5
Forum Member




Join Date: Jun 2014
Posts: 73
Much bigger problems with pentax glass to steer around, like VERY bad CA and flare on some. Diffraction is the least of your worries.
07-09-2015, 01:41 PM   #6
Forum Member




Join Date: Sep 2008
Location: throughout the US
Posts: 79
Original Poster
QuoteOriginally posted by MJSfoto1956 Quote
"Back in the day" we generally shot f/22-f/16 for 4x5, f/16-f/11 for medium format and f/11-f/8 for 35mm film. Frankly not much has changed...

Michael
Yep. In my 4x5 days f/32 was what I most needed even after proper tilt, and certain comps could not be done without f/45 (even one very difficult scene needed f/64 to give the depth required). The 40x30 prints are still stunning.

---------- Post added 07-09-15 at 01:43 PM ----------

QuoteOriginally posted by Kolor-Pikker Quote
Even without special sharpening settings, I get excellent results up to f/16 no problem.

While the 645Z should technically become diffraction limited at around f/9~10, in reality you're never going to have image content that pushes the limits of the sensor and optics short of a brightly-lit test chart accidentally ending up in your images.

Now going from f/16 to f22 is where things start to get visibly soft, but not by too much, especially if the added DoF ends up creating a greater impression of sharpness than the presence of actual micro-detail.

I like to think about digital sensors more in terms of pixel pitch rather than sensor size, since nowadays the size of a sensor has little to do with its actual performance; in this case the 645Z has 30MP if cropped down to 35mm, or 12MP if cropped down to APS-C, which aren't nearly the highest values photographers have had to deal with. The D800 cameras and Sony A7R are 36MP at 35mm size, with the coming A7RII being 42MP, while many APS-C cameras today range from 16~22MP of resolution, which is a good chunk more than the Z's equivalent.

Are all the photographers using these cameras more concerned about diffraction than the 645Z users?
Check out the full size image I linked to - at f/22 the micro detail present is simply astounding. This is largely due to the effectiveness of the deconvolution algorithm.

---------- Post added 07-09-15 at 01:48 PM ----------

QuoteOriginally posted by CDW Quote
Lens reviews of P645 lenses have revealed many of them to resolve their highest resolutions at F8-F13. The sweet spots vary with format and with medium format many Pentax lenses are not their sharpess at F8. Otherwise, your workflow is logical and is similar to mine. I've found in practice the Z files require little sharpening. I regularly shoot at F16, sometimes higher, with no real world issues. My clients aren't pixel peepers looking at 100-200% views on an LCD display.
There is one particular well-known reviewer on the internet who has made a big deal out of diffraction. And every now & then I see forum discussions arising from that where folks assume you lose the benefit of a high-res sensor such as in the 645Z if you are shooting in the diffraction range. The problem is, this reviewer fails to consider the power of deconvolution-type sharpening to recover much of the detail. In writing this up and sharing the test image file I hoped to help perhaps some some newer folks who are unaware of how to activate the deconvolution within Lightroom, and to show just how effective that routine is. I knew it worked well, but I was still amazed viewing at 100% the test image shot at f/22.

---------- Post added 07-09-15 at 01:52 PM ----------

QuoteOriginally posted by rob8888 Quote
Much bigger problems with pentax glass to steer around, like VERY bad CA and flare on some. Diffraction is the least of your worries.
Thankfully, neither of those issues have surfaced for me. I'm using a mix of manual A 645 lenses (35, 75, 150, 200), FA 300 & 400, and 67 55mm & 105mm, and all are performing superbly for me with Lightroom processing on the 645D and 645Z. But I did have to carefully test & return several copies to arrive at near-perfect specimens.

Last edited by SeattleDucks; 07-09-2015 at 02:34 PM.
07-09-2015, 01:59 PM   #7
Veteran Member
MJSfoto1956's Avatar

Join Date: Dec 2013
Location: Boston
Posts: 1,287
QuoteOriginally posted by SeattleDucks Quote
YBut I did have to vigorously test & return several copies to arrive at near-perfect specimens.
For the benefit of those who are new to testing their camera gear, you might want to describe your testing methodology.

Michael
07-09-2015, 02:30 PM   #8
Forum Member




Join Date: Sep 2008
Location: throughout the US
Posts: 79
Original Poster
QuoteOriginally posted by MJSfoto1956 Quote
For the benefit of those who are new to testing their camera gear, you might want to describe your testing methodology.

Michael
Hey Michael. I definitely am not a test chart shooter. But I just use common-sense techniques to eliminate variables, and replicate typical landscape distances. So I'm on a sturdy Gitzo CF tripod, mirror lockup, remote release, manual exposure, low ISO, focusing careful at 16x magnification in live view with a Hoodman loupe. The most important test for me is subject outdoors near infinity, as that is where the image circle is smallest and most likely to reveal decentering or field curvature issues. I use neighborhood homes and foliage that are roughly 200 yards away, placing them in a corner of the frame and repeating shots for all four corners and the center, aperture range from f/8 through f/16. This lets me evaluate sharpness, contrast, and color across the frame and chromatic aberration. I also have nearby foliage in the frame which lets me evaluate depth of field and rendering of elements that are outside the focus zone (and find a lot of variance between lenses here).

The lenses that look best near infinity I then test indoors at about 25 feet which is similar to shooting distance of intimate landscapes such as a group of autumn trees. The scene is my living room so it's a ready reference for my eye.

Image evaluations are on a Retina level resolution monitor after my usual Lightroom processing has been applied, at 100% and 200% pixel levels. When making prints such as the 54" wide monster I did recently I am interpolating up quite a bit on a 645D file to match the 300ppi native resolution of my Canon IPF8300 printer, so evaluating at 100/200% onscreen when testing lenses is worthwhile for me.

I don't presume to advise others on what methodology they should use and I respect the lens chart testers as well as the folks who just go make pictures - but the above provides me what I need to know in finding a good lens copy that performs well for my use. It has taken me a lot of work and many months of purchases to settle on my current set of Pentax glass, and I'm extremely pleased with the results.

Ross


Last edited by SeattleDucks; 07-09-2015 at 02:36 PM.
07-09-2015, 03:55 PM   #9
Pentaxian




Join Date: May 2011
Location: All over the place
Posts: 2,580
I have as much interest in shooting/looking at test charts as I do in watching paint dry....no, I'd rather wait for it to peel, in the words of Garfield the cat! I'd rather see what the equipment does in the real world with real light.
07-09-2015, 04:29 PM - 1 Like   #10
Veteran Member
MJSfoto1956's Avatar

Join Date: Dec 2013
Location: Boston
Posts: 1,287
QuoteOriginally posted by itshimitis Quote
I'd rather see what the equipment does in the real world with real light.
I think many of us (me included) do our tests "in the real world" rather than with charts. Personally I think it important to test your equipment before your return period end. The most common flaw I have seen is "decentering" -- super easy to test: put your lens+camera on a tripod, open the lens to its widest aperture, and turn on live view. Zoom in to focus on a center star and then manually adjust the focus until the star is a clear, distinct point. Take the picture using mirror lockup and a remote then examine the stars in the corners: if all four corners look the same then your lens is fine. If one of the corners is out of focus, then you have a decentered lens.

CAVEAT: since pretty much all the big retailers now allow returns within 30 days, the chances of you receiving a lens that is less than perfect have increased. So IMHO it is now pretty much de rigeur to test your gear. My quick test above is just one way to evaluate a lens. It is important to find a methodology that works for you.

YMMV

Michael
07-09-2015, 05:36 PM   #11
Forum Member




Join Date: Sep 2008
Location: throughout the US
Posts: 79
Original Poster
QuoteOriginally posted by MJSfoto1956 Quote
I think many of us (me included) do our tests "in the real world" rather than with charts. Personally I think it important to test your equipment before your return period end. The most common flaw I have seen is "decentering" -- super easy to test: put your lens+camera on a tripod, open the lens to its widest aperture, and turn on live view. Zoom in to focus on a center star and then manually adjust the focus until the star is a clear, distinct point. Take the picture using mirror lockup and a remote then examine the stars in the corners: if all four corners look the same then your lens is fine. If one of the corners is out of focus, then you have a decentered lens.

CAVEAT: since pretty much all the big retailers now allow returns within 30 days, the chances of you receiving a lens that is less than perfect have increased. So IMHO it is now pretty much de rigeur to test your gear. My quick test above is just one way to evaluate a lens. It is important to find a methodology that works for you.

YMMV

Michael
I agree with you Michael on the importance of testing your lenses before the return period is up. This has been even more urgent with the Pentax lenses because the vast majority of used copies are in Japan via eBay where the return window is usually only 7 days.

I made the mistake once of taking a used Nikon lens on a photo trip without thorough testing first. All the images were ruined from strong decentering on one entire side. Never again will I make that mistake.
07-09-2015, 05:43 PM   #12
Site Supporter




Join Date: Dec 2012
Location: Pekin, IL
Posts: 237
QuoteOriginally posted by MJSfoto1956 Quote
. . . The most common flaw I have seen is "decentering" . . . CAVEAT: since pretty much all the big retailers now allow returns within 30 days, the chances of you receiving a lens that is less than perfect have increased. So IMHO it is now pretty much de rigeur to test your gear. My quick test above is just one way to evaluate a lens. It is important to find a methodology that works for you. . .
Do you think this applies also to the lowly APS-C-format lenses, or is it primarily an MF problem?
07-09-2015, 05:56 PM - 1 Like   #13
Site Supporter
gavincato's Avatar

Join Date: Apr 2015
Location: NSW, Australia
Posts: 247
fred, happens to all lenses. I have a big canon setup as well and have often been through 2-3 lenses to find a good copy.

My current 24-70mm is an absolute cracker and for a zoom is incredible. It's my 4th copy.
07-09-2015, 08:07 PM   #14
Pentaxian




Join Date: Sep 2009
Location: Southern California
Photos: Gallery | Albums
Posts: 3,181
I'm not a lens technician, but sometimes a "bad" copy of a lens can quickly be turned into a good one with a simple service center visit. Perhaps this is all many of these returned lenses actually need.

This can be especially useful when buying a used lens. My K-mount FA20 I purchased used was this way. I couldn't even describe exactly what looked wrong with the photos from it. Sent it in and ~$50 later I had a "good" copy.

QuoteOriginally posted by MJSfoto1956 Quote
CAVEAT: since pretty much all the big retailers now allow returns within 30 days, the chances of you receiving a lens that is less than perfect have increased. So IMHO it is now pretty much de rigeur to test your gear. My quick test above is just one way to evaluate a lens. It is important to find a methodology that works for you.
Michael, I realize you wouldn't want to spend either service time (waiting) or extra money to adjust a brand new lens, but do you think this is all many of these "bad" new copies need?
07-10-2015, 12:08 AM   #15
Veteran Member
Kolor-Pikker's Avatar

Join Date: Dec 2014
Posts: 341
In theory, the smaller the lens and its intended format, the greater the tolerances needed for it to function properly and vice versa... But again, digital throws a wrench into this concept because now we're all about pixel density, and that's what tolerances must be measured against.

For instance, my 5Dmk2 has 21MP of resolution and at the time it was the highest resolution camera, one on which many Canon lenses have shown their weakness. However, the 645Z not only has 50% greater pixel density, but a larger surface area as well, meaning that an optimized lens must be considerably over-engineered in every measure, we see this with the 28-45mm that's nearly the same size and weight as the camera itself.

On the whole, smaller formats do have the tendency to require higher tolerances, but I think the practical implications don't really kick in until you've gone smaller than APS-C. Now when you're down to sensors that are only a few mm in size, even things like changes in temperature are enough to require re-calibration of the lens system; I used to work with video cameras that had a back-focus adjustment that had to be readjusted if you so much as went from indoors out to a cold environment.
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!
35mm, 645d, 645z, 67 55mm, amount, aps-c, camera, cameras, contrast, deconvolution, diffraction, f/16, f/22, format, image, lens, lenses, lightroom, medium format, photographers, sensor, size, test, workflow
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Some 645Z Images michaelz Pentax Medium Format 23 02-12-2015 08:03 AM
In-camera diffraction correction with 645Z? jwoodland Pentax Medium Format 6 09-26-2014 09:45 AM
645Z Sample Images interested_observer Pentax Medium Format 2 04-20-2014 11:41 PM
all this wait is just killing me oxidized Pentax Full Frame 16 07-24-2013 07:52 AM
K-x Killing Images? Shub Pentax DSLR Discussion 28 03-20-2010 05:29 PM



All times are GMT -7. The time now is 09:13 AM. | 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