Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
11-29-2010, 10:00 PM   #31
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,739
QuoteOriginally posted by Class A Quote
The idea that the notions of DOF and hyperfocal focusing change with digital photography is simply crazy.
It really isn't absurd as you think, while those rules and techniques are still applicable in regards to digital photography. At this level there is another issue to be considered: diffraction. On a 39Mp digital medium format back it begins to be visible at f/11. Some people will say " it won't be noticeable in a large print" while at print sizes below 16X20" that may be true, even a K5 can manage that print size with aplomb. The 645D is for prints significantly larger than that, it would be possible to get away with using f/16 if you're only going to print at 16X20 " but if you triple that print size, diffraction would be blindingly obvious at 48X60"

Sometimes hyperfocal focusing is not enough, and that is where T/S ( tilt and shift) lenses come in, they enable you to stay under the diffraction limit and provide extended DOF without any of the drawbacks of hyperfocal focusing - providing the lens is up to the job.


Last edited by Digitalis; 11-29-2010 at 10:06 PM.
11-30-2010, 02:09 AM   #32
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,175
QuoteOriginally posted by Digitalis Quote
At this level there is another issue to be considered: diffraction.
Diffraction can always be an issue. The larger the sensor (or film), the later it sets in, so I don't see the problem for MF. In particular the issue is independent from the image forming area being a digital sensor or analogue film.

QuoteOriginally posted by Digitalis Quote
Sometimes hyperfocal focusing is not enough, ...
Sure, that's why tilt & shift lenses are so useful, as you explain. But this isn't new. It's been the case in the analogue area already.

Nick's statements still make no sense to me.
11-30-2010, 03:49 AM   #33
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,739
QuoteOriginally posted by Class A Quote
Diffraction can always be an issue. The larger the sensor (or film), the later it sets in, so I don't see the problem for MF
As a Large format photographer I often use camera movements to stay under the diffraction limit, I rarely shoot with my lens stopped down to f32 on 8X10 format, I usually stop down to f/16(my lenses optimum aperture) and use tilt to get the results I want. But there is a problem with Digital MF is that with the higher resolution, Diffraction becomes much more obvious. This is one of the reasons I have a good chuckle when I hear some canon photographer hoping that canon will make a full frame camera with resolution upwards of 32Mp to compete with the 645D. With a sensor that small and resolution that high diffraction is going to be a significant hurdle, and there are only a handful of canon lenses that could handle that kind of resolution and none of them are wider than 50mm.

Nick's statements have been making perfect sense to me. He is simply pointing out a potential deficiency in the AF system Pentax has used in the 645D, and frankly I'm inclined to agree. In terms of focusing accuracy there are a lot of things that can go wrong at 40Mp that wouldn't be visible on a 16Mp APS-C sensor, and the faculty for manual focus is diminished due to the Focus screen's brightness.

Last edited by Digitalis; 11-30-2010 at 03:58 AM.
11-30-2010, 07:56 AM   #34
Senior Member




Join Date: Oct 2010
Location: Toronto, Canada
Posts: 132
QuoteOriginally posted by Class A Quote
Hmmh, perhaps you want to write to all the DOF calculator websites that they have to change their so

The idea that the notions of DOF and hyperfocal focusing change with digital photography is simply crazy.
The math doesn't change, but it's impact on the image changes in two ways. First, what looked "in focus" with film is different from what looks in focus with digital. The latter is simply more critical (this is not normative judgment). The accutance of digital is something many users love, and what die-hard film-users decry.

Second, there is a difference in outcome because of depth-of-focus at the sensor. Film has significant physical depth compared to a CCD. Producing a sharp image anywhere within the film's physical depth will render a sharp image. Withe a digital sensor, you are either right on the mark or not. This is why people are going to enormous lengths to "shim" digital backs to precise alignment with the cameras on which they are mounted.

Your tone of your comments suggest to me that you have not had much experience with large format digital work. If you doubt what I say about these issues, search out the articles by Joe Holmes and Mark Dubovoy.

The levels of precision required at every step of the process, both in the design and manufacture of the gear and the use of it, to obtain maximum IQ from larger-format digital gear, is enormous compared to what it was with film.

Take it or leave it.

- N.

ps. I gather the 645D's focusing screen is not replaceable. One reason for this likely is that, to maintain critical alignment with the sensor plane, it must be fixed. Allowing to it move about, at all, would erode this tolerance unacceptably.

pps. I tested all my 'long' glass on the 645D at near-infinity yesterday, under careful conditions. The 150-300 was best at f8 in the centre, just barely less good in the centre at f11, with a notable gain in IQ at the edges, and the whole thing dropping off very noticeably at f16. The 300 f4 was also best at f8, and also sub-par by f16. In fact, every lens I have tested suffers seriously noticeable degradation of the image due to difraction by f16. I have done a 13x19 print from a frame shot on the 33-55mm at f29, where I needed the DOF, such as it is, and the print looked fine even though on-screen one could tell there was a loss of accutance. By fine I don't means "wow!", I mean "fine".


Last edited by ndevlin; 11-30-2010 at 08:03 AM.
11-30-2010, 08:20 AM   #35
Veteran Member




Join Date: Mar 2007
Location: Boston, PRofMA
Photos: Albums
Posts: 3,053
QuoteOriginally posted by ndevlin Quote
pps. I tested all my 'long' glass on the 645D at near-infinity yesterday, under careful conditions. The 150-300 was best at f8 in the centre, just barely less good in the centre at f11, with a notable gain in IQ at the edges, and the whole thing dropping off very noticeably at f16. The 300 f4 was also best at f8, and also sub-par by f16. In fact, every lens I have tested suffers seriously noticeable degradation of the image due to difraction by f16.
Interesting....I had thought that a larger lens would allow you to get to an extra stop (closed) compared w/ APS-C before diffraction had an effect. Thanks for the testing...
11-30-2010, 09:40 AM   #36
New Member




Join Date: Nov 2010
Location: Bavaria
Posts: 7
@ nick:

Do you have really a look to your 645D? The Mattescreen IS replaceable. My Grid Screen is on the way from Taiwan :-)) If you look into the open Camera, you see a frame and a spring look around the mattescreen. Oh, on the Pentax webside they also list different screens. And the tolerances are easyier to manage as you think of. When the grid screen arrives, i bulid it in and post my experinces.

regards

Frank
11-30-2010, 02:03 PM   #37
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,175
QuoteOriginally posted by Digitalis Quote
But there is a problem with Digital MF is that with the higher resolution, Diffraction becomes much more obvious.
What higher resolution?
If you wanted the same pixel pitch of a 16MP APS-C sensor for the Pentax reduced MF sensor, it would have to have 62MP.

Enlarging an MF image works better because the "negative" is larger to begin with, not because you have more pixels per mm^2.

QuoteOriginally posted by Digitalis Quote
With a sensor that small and resolution that high diffraction is going to be a significant hurdle,
Diffraction is always present, no matter the sensor resolution and the aperture.There is always loss of actuance because of diffraction. Wide apertures just mean that lens aberrations are dominating but there's always diffraction too. A high resolution sensor just allows you to see diffraction earlier, just like it allows you to see a slight misfocus that would be harder or impossible to detect with a low-resolution sensor.

QuoteOriginally posted by Digitalis Quote
Nick's statements have been making perfect sense to me. He is simply pointing out a potential deficiency in the AF system Pentax has used in the 645D, and frankly I'm inclined to agree.
As with many other "issues", Nick doesn't know anything concrete yet, but feels it is OK to point out "issues". I take issue with that.

QuoteOriginally posted by kenyee Quote
Interesting....I had thought that a larger lens would allow you to get to an extra stop (closed) compared w/ APS-C before diffraction had an effect.
It is correct that larger sensors will shift the f-ratio value at which you are starting to see diffraction effects to higher numbers.
11-30-2010, 02:08 PM   #38
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,175
QuoteOriginally posted by ndevlin Quote
First, what looked "in focus" with film is different from what looks in focus with digital. The latter is simply more critical (this is not normative judgment).
You are confounding "digital" with "higher resolution". Say Fuji came out with an incredible high-resolution film emulsion that also provides excellent micro-contrast. You'd have the same "problems" you have now with high-resolution digital sensors. Conversely, take a low-resolution digital sensor and try to detect minute amounts of misfocus or misalignment to a wall. It won't work. Hence, "digital" is not the problem and "digital DOF" makes as much sense as "digital lens pouch".


QuoteOriginally posted by ndevlin Quote
Second, there is a difference in outcome because of depth-of-focus at the sensor. Film has significant physical depth compared to a CCD. Producing a sharp image anywhere within the film's physical depth will render a sharp image.
So you are saying that by not only capturing one plane of focus but multiple ones, and of course not only the planes but their out-of-focus neighbour planes, you get a sharper picture?

Then why don't you just emulate film thickness by not taking one shot, but say five ones, where each shot has a slightly different focus. By doing this, you are effectively moving the sensor plane and by overlaying all these shots, your final image will be as sharp as an image captured by film. I hope this makes it evident that your explanation doesn't make sense.

A point could be made that film has the different colour-sensitive layers stacked and that lenses might be optimised to project the respective wave-length into the correct layer. This would mean that a digital Bayer sensor would show a lens to have more lateral chromatic aberration than it would have on film. However, the effect is very small. IIRC, it is in the sub-pixel range. Also, it is a different argument than you made.

QuoteOriginally posted by ndevlin Quote
Your tone of your comments suggest to me that you have not had much experience with large format digital work.
The laws of physics don't change with large format digital work. I'm sure you have some experience that I'm lacking but that doesn't mean that everything you write makes sense.

QuoteOriginally posted by ndevlin Quote
The levels of precision required at every step of the process, both in the design and manufacture of the gear and the use of it, to obtain maximum IQ from larger-format digital gear, is enormous compared to what it was with film.
I suspect that is because you are now pixel-peeping into depths that a) people didn't used to do in film days and b) most film emulsions didn't support.

I think it is completely fine if one isn't on top of everything. There is no necessity to have intricate knowledge of underlying technology to make great images. The only thing I take issue with is that you write stuff you have no knowledge about in a professional review. Why not simply leave that stuff out? Your review would have been great without all the wrong statements, speculations, and (potentially unwarranted) homework assignments for Pentax.


Last edited by Class A; 11-30-2010 at 02:13 PM.
11-30-2010, 02:17 PM   #39
Inactive Account




Join Date: Nov 2010
Posts: 8
QuoteOriginally posted by Class A Quote
You are confounding "digital" with "higher resolution". Say Fuji came out with an incredible high-resolution film emulsion that also provides excellent micro-contrast. You'd have the same "problems" you have now with high-resolution digital sensors. Conversely, take a low-resolution digital sensor and try to detect minute amounts of misfocus or misalignment to a wall. It won't work. Hence, "digital" is not the problem and "digital DOF" makes as much sense as "digital lens pouch".



So you are saying that by not only capturing one plane of focus but multiple ones, and of course not only the planes but their out-of-focus neighbour planes, you get a sharper picture?

Then why don't you just emulate film thickness by not taking one shot, but say five ones, where each shot has a slightly different focus. By doing this, you are effectively moving the sensor plane and by overlaying all these shots, your final image will be as sharp as an image captured by film. I hope this makes it evident that your explanation doesn't make sense.

A point could be made that film has the different colour-sensitive layers stacked and that lenses might be optimised to project the respective wave-length into the correct layer. This would mean that a digital Bayer sensor would show a lens to have more lateral chromatic aberration than it would have on film. However, the effect is very small. IIRC, it is in the sub-pixel range. Also, it is a different argument than you made.


The laws of physics don't change with large format digital work. I'm sure you have some experience that I'm lacking but that doesn't mean that everything you write makes sense.


I suspect that is because you are now pixel-peeping into depths that a) people didn't used to do in film days and b) most film emulsions didn't support.

I think it is completely fine if one isn't on top of everything. There is no necessity to have intricate knowledge of underlying technology to make great images. The only thing I take issue with is that you write stuff you have no knowledge about in a professional review. Why not simply leave that stuff out? Your review would have been great without all the wrong statements, speculations, and (potentially unwarranted) homework assignments for Pentax.
I believe it has also to do with the thicker emulsion of film vs the thinner sensor.
11-30-2010, 04:10 PM   #40
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,739
QuoteOriginally posted by Class A Quote
What higher resolution? If you wanted the same pixel pitch of a 16MP APS-C sensor for the Pentax reduced MF sensor, it would have to have 62MP. Enlarging an MF image works better because the "negative" is larger to begin with, not because you have more pixels per mm^2.
Well the 645D is 40 Mp - it will make many faults in photographic technique and faults in lenses much more apparent. I don't know many MF digital photographers that print smaller than 13X19. And secondly, I work with 8X10 format. The reason why I use that format is because the larger negative gives a acceptable contact print size, frankly 4X5 is too small. I can also enlarge my 8X10 negatives, and they produce stunning detail. The resolution of the lens recorded on essentially grainless Kodak Tech-Pan gives a quality of tone and contrast rarely seen in an image from a Digital source, especially when printed on platinum.

From what I have seen the Pentax 645D essentially matches the quality of Velvia 50 on 4X5.
11-30-2010, 04:25 PM   #41
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,175
QuoteOriginally posted by paratom Quote
I believe it has also to do with the thicker emulsion of film vs the thinner sensor.
I responded to that argument by Nick.

The thinner sensor will allow better maximum sharpness while the thicker film may make it easier to make a slight mis-focus a little more acceptable.

I found the post by falconeye in which he speculates quantitatively about the impact of film thickness on longitudinal CA (aka purple fringing). But remember this is based on the assumption that film lenses have been optimised this way. We don't know this for sure.

QuoteOriginally posted by Digitalis Quote
Well the 645D is 40 Mp...
Yes, which means that, relative to sensor size, it has less resolution than a K-5. The sensor size of the 645D is part of its ability to provide stunning detail, not just the number of pixels. And if the next generation of sensors has even more pixels, this will not worsen things, on the contrary. While it will make it harder to achieve pixel-to-pixel sharpness, it will be easier to achieve overall image sharpness. A 2MP sensor will easily beat a 40MP sensor on pixel-sharpness, but of course the 40MP sensor produces the better image sharpness.

BTW, I'm not taking issue with anything you wrote. My only concern is that LL publishes a review in which Nick makes some very questionable remarks (including giving Pentax homework to do on the grounds of having a feeling). I very much appreciate the rest of the review and it is great that Nick found the 645D to be so nice to handle and to deliver up-to-par performance. That's why I think he just should have left the "digital DOF" stuff and the "homework assignment for Pentax" out.

BTW, Canon shooters are complaining hard about their camera having received a bad deal. There are issues with the lens chosen and whether or not true 100% crops have been shown. It seems like this part should have been done better or left out as well.

Last edited by Class A; 12-03-2010 at 10:01 PM.
12-02-2010, 01:40 PM   #42
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,175
Too bad that some of the posts in this thread have been lost due to the database issue.

I remember that Nick pointed me to these articles here:
Joseph Holmes - News: Medium Format Problems
Joseph Holmes - News: Medium Format Methods for Sharpness
and that I didn't know what to learn from them.

However, I quoted from them, e.g.,
"With lenses for film use, the diffraction was obscured by the inherent blurriness of the film until you had stopped down to roughly f/16 or f/22 or more, typically, but now you can actually see diffraction starting at f/8 or f/11 in captured images..."
My point was that we always had to suffer diffraction even in film days. Today we have better tools and they allow us to see slight imperfections such as slight misfocus or sharpness limitation due to diffraction much better. This is good news, not bad news as Nick wrote. Digital imaging didn't introduce these problems, it just reveals them earlier than film was ever able to do.

BTW, I asked how Nick aligned the camera to the wall (e.g., by using a mirror). In an LL forum discussion he explains that they just did it by eye. To conclude from this that the equipment might be at fault is... bold.
12-05-2010, 06:45 PM   #43
New Member




Join Date: Oct 2010
Location: Crowley Lake, CA
Posts: 22
Original Poster
I did not see my post regarding diffraction in larger prints so I'll redo my post.

In talking with a highly regarded photographer who's very familiar with large (40 inches on the short side) prints, this photographer finds little difference, between an image exposed at f 8 or f 22, with proper sharpening. "WITH PROPER SHARPENING"! Without the right sharpening a big difference. With, hardly noticeable.

Now before I get slain with look at this or that examples on websites xyz (diglloyd, cambridge in colour, etc., etc., etc.), I'm familiar with these sites. But the diffraction examples at these sites beg the obvious question; what did one do to process the image out of the camera? The answer of course is nothing. Post process is another question that needs a separate answer. Yes f 8 looks terrible compared to f 22 out of the camera. But the sharpening software that's out there now is impressive and effective.

In my mind this is great news! The tools to hone and perfect a big print are only getting better. And for me at least, big prints are the reason to buy a medium format camera. Plus not being stuck at f 5.6 to f 11 is really good news!! Additional field options are a good thing in my book.

Also, I'm not saying that given the chance I won't do what I can in the camera and avoid work, or a disappointing result at the back end.

The diffraction issue also depends on the standard of individual photographers. There are some who want prints as sharp as they can be from a viewing distance of three inches. Others are okay with prints that look okay from twenty feet away but look worse than news print at three feet.

To each his own.

Claude Fiddler

Wilderness Light
12-06-2010, 07:50 AM   #44
Senior Member




Join Date: Oct 2010
Location: Toronto, Canada
Posts: 132
Claude,

You are exactly right. An anecdote illustrates. Every year I have the chance to go shooting with a diverse and incredibly talented/knowledgable group of photographer friends. One of them tries to shoot at or near f8, since he knows from experience that his MF lenses perform best at this aperture, and lose quality to difraction rapidly thereafter. He will often import Helicon focus into his workflow to obtain DOF.

Another swears by f22, since he feels this gives his files (also MF) the appearance of DOF that he finds most visibly pleasing.

They are both photographers of the highest calibre, and technical experts who understand the underlying science better than some of the companies who make these products (and who thus hire them to consult). Both are right. It's all in how you relate to the medium and your demands/desires viz. output.

Of coruse, all of this assumes a good sharpening workflow. This, thankfully, is easier than ever. The sharpening tools in LR3 are now really very good (I believe they now incorporate the latest iteration of some of the proprietary sharpening software we used to buy). There is also Nik's excellent Sharpener plug-in. Both methods are pretty user-friendly.

I may have mentioned here that I printed a 13x19 of a fall scene with a carpet of fallen leaves stretching away out of the frame, which, at f29 on the 33-55mm, had very noticable difraction visible everywhere. As an experiment, I sharpened it to my acceptable limit (I hate 'crunchy' prints) and hit 'print'. The resulting image is very pleasing. At arms-length, no ordinary viewer would take issue with sharpness.

To be sure, it is rendering much less fine detail than the system is theoretically capable of, but, as Claude points out, and this thread has as its theme, theory and practice are rather different creatures in this realm.

Cheers,

- N.
12-06-2010, 04:17 PM   #45
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,739
QuoteOriginally posted by ndevlin Quote
at f29 on the 33-55mm, had very noticable difraction visible everywhere. As an experiment, I sharpened it to my acceptable limit (I hate 'crunchy' prints) and hit 'print'. The resulting image is very pleasing. At arms-length, no ordinary viewer would take issue with sharpness.
Until someone whips out a loupe, which is something I have seen people do when I have exhibited 16"X20" prints from 8X10 format.
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!
camera, medium format, nick, review
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Can't afford the 645D, what is the alternative for landscape. pcarfan Non-Pentax Cameras: Canon, Nikon, etc. 74 01-06-2011 03:25 AM
645D Review bikecoboss Pentax News and Rumors 7 11-16-2010 05:49 PM
The Luminous Landscape 645D review jerrymouse Pentax Medium Format 4 11-14-2010 07:16 AM
Luminous Landscape - 645D Review interested_observer Pentax Medium Format 15 09-08-2010 11:13 PM
http://www.luminous-landscape.com/ needs Pentax K-7 to prove it ogl Pentax News and Rumors 133 10-24-2009 07:50 AM



All times are GMT -7. The time now is 07:54 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