Pages: [1] 2 3   Go Down

Author Topic: Everything matters revisited  (Read 12955 times)

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Everything matters revisited
« on: January 25, 2012, 11:33:21 am »

I was posting the reply shown below as the thread was being locked. While some of Mark's claims about the per pixel quality of MFDBs as compared to smaller format sensors are not justified, the need for high resolution in making large prints is correct.

Another point I would make is that moder printers have very high native resolution. The Epsons are said to handle 720 PPI. A Phase One IQ 180 has 10380 pixels so it would print at 14.4" width for maximum resolution on an Epson. My Sony Alpha has only 6000 pixels so it would be able to print at 8.3" wide at maximal output quality.

I would call neither 14.4" nor 8.3" inch print very large, but I have no doubt Jeff Schewe could tell them apart in a 14.4" print using his loupe. Now, according to Norman Koren's writing 20/20 vision at 25 cm resolves about 180 ppi, so with normal vision we could essentially blow up the Sony to 33" and the IQ180 to 57" before an obvious difference would be seen. My experience is that my 12 MP APS-C is pretty good for A2 (about 23" wide) but the 24 MP full frame is marginally better.

Enough of this nonsense about high end stereo and fine wines and back to photography. Roger Clark has a good post on the acuity of the human eye and relates it to the print resolution necessary to represent that resolution. He did some blinded tests using multiple subjects and concluded that at least 600 ppi was necessary (he was using an HP printer). Jeff Schewe states in the Camera to Print and Screen tutorial that with Epson printers one can readily tell the difference between prints at 360 and 720 dpi. So, Roger and Jeff agree on the necessary resolution. Of course, this depends on viewing distance, but this assumes that the viewing distance was quite close. The subjects presumably viewed the prints at an optimum distance. Anyway, Roger concludes that only a large format camera is capable of this resolution when one is dealing with really large prints.

In another post he demonstrates various print resolutions of a 4*5 inch film image and also provides a chart to determine one's printer resolution. Of course, it makes no sense to send more data to the printer than it can print, so one should determine the resolution of his printer. I printed Roger's chart with my Epson 3880, setting the resolution to 1440 and using finest detail and examined the image under a microscope using a 2x PlanApochromat objective with the setup giving a magnification of 25x in the plane of the sensor. The results are shown. The individual ink dots are well shown. The 1 px lines are not resolved, but the 2 px are, giving a resolution of 720 ppi.



In another post entitled the MegaPixel Myth, Roger evaluates what resolution is necessary to literally knock your socks off with a large print. He concludes that 200 MP are necessary. In summary, Mark and the MFDB crowd are correct in that the high resolution of these backs is necessary to make really large prints. To reach the resolution that Roger suggests, one would have to stitch the IQ180, as Jeff demonstrates in the above mentioned tutorial. In summary, the main readily demonstrated advantage of the MFDB is resolution. The availability of high quality apochromatic primes is another factor. Some of the other claims are questionable.

Regards,

Bill
 

« Last Edit: January 25, 2012, 11:40:21 am by bjanes »
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: Everything matters revisited
« Reply #1 on: January 25, 2012, 11:55:57 am »

I had the same unfortunate experience re. posting...

But anyway, my post was absolutely respectfull, and it was to ask this: "eisegesis" means? I looked it up in the computer's brain, and the best it could reply was eisteddford or ejaculate. I suspect there's more to love than this; if anyone can illuminate this dark corner of my tiny mind, muchas gracias.

Rob C

Isaac

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3123
Re: Everything matters revisited
« Reply #2 on: January 25, 2012, 12:00:54 pm »

Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: Everything matters revisited
« Reply #3 on: January 25, 2012, 12:06:42 pm »

http://www.google.com/search?q=eisegesis


You deserve it: muchas gracias.

Trouble is, it brings up subjective/objective considerations and you know where they lead us!

Rob C

NikoJorj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1082
    • http://nikojorj.free.fr/
Re: Everything matters revisited
« Reply #4 on: January 25, 2012, 12:10:34 pm »

Rob : Wikipedia's entry seems of a different kind of romantism : shortly, it's a way to call each other names among the biblic scholars. Another analogy.

Oops, seems I've been preceded.
Logged
Nicolas from Grenoble
A small gallery

John Camp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2171
Re: Everything matters revisited
« Reply #5 on: January 25, 2012, 01:28:26 pm »

I think I commented three times on the original thread, and never really insulted anyone, and so I hope Michael will give me an extra inch here to summarize what I was thinking about this morning while I was walking the dog. The essential problem with Mark's post was that he created a line of reasoning which not only was demonstrably flawed, and that either intentionally or unintentionally insulted a lot of readers along the way, but that it led to an unpalatable, unreasonable and simply wrong conclusion: and he got there when he shifted from an objective assessment of MF qualities to a subjective assessment, and pronounced MF "the best." It is demonstrably the best by certain technical parameters -- but if you're trailing behind a Marine combat platoon in Afghanistan (to take an extreme example) would you really want to carry your IQ180, a tripod, a lightmeter, a five-pound ball head, etc.? Of course not, not unless you were also suicidal. For that kind of photography, MF is not the best.

The same thing holds in less extreme conditions. Not to put too fine a point on it, I can afford any kind of photographic equipment I want. I shoot M4/3, for my own good reasons. And they are good reasons. In my kind of shooting, MF is totally inappropriate.

If Mark had simply said, MF images can provide the most detailed large prints available from commercial photo equipment, I doubt that there would have been any comment at all, or perhaps a little controversy about how large you could push other formats, or how important difference between MF and m4/3 are when viewed on a monitor.

Again, the problem is in the shift between technical analysis and value assessment. I don't doubt for a moment that Mark is a very smart guy and a good photographer, but he has to be more careful about making that shift. Much of his post, and the subsequent thread, was a waste of time, but then, a lot of it was pretty interesting, too. I am actually quite interested in the concept of hyper-realism, and despite a certain amount of flak about Mark's use of the term, I understand what he's getting at and find it interesting.

Finally, I'd like to add that Carl Weese, who is basically a large and ultra-large format photographer, recently shot some m4/3 landscapes just to see how far he could push the format, and his subsequent article, on The Online Photographer, addresses many of the issues that Mark touched on, and does so in quite a congenial and interesting way, here:

http://theonlinephotographer.typepad.com/the_online_photographer/2012/01/raised-expectations.html   
Logged

fredjeang

  • Guest
Re: Everything matters revisited
« Reply #6 on: January 25, 2012, 02:15:37 pm »

I think john that what happens in internet goes way further than a Mark's semantic issue, if there was one.

I remember when bcooter did the Leica M essay and at one point was transmiting that he found certain unreliability or unfriendliness in the focussing accuracy, the forums were in flame and some posters didn't hesitate to discredit him as a photographer, pointing with arguments that if he wasn't able to focus properly with the M8 he shouldn't call himself a photographer...things like that. And we are talking about one experienced and prestigious awarded pro of this industry.

When Lagarfeld did the Pirelli calendar, tons of criticisms and jalousy postures all over the net by people who knew nothing about him, pointing the fact that Lagarfeld wasn't a photographer when he actually has quite a high level in photography (that a lot of people ignore), and knows and understands fashion way deeper than most photographers who call themselves fashion photographers when it's not high-end fashion photographers. Insults, disprestige of all kinds and we were talking about one of the most influent fashion designer of its time. And why? because he his not labeled "photographer" and because he his excentric.

More recently the Bloom affair, when mister Bloom had issues with a Red Epic on set and reported publically, the Red forum was absolutly in flame, directly insulting him and questionning his professionality just because he had the audacity to point wicknesses within the sacro-saint brand he owned. In part he was faulty, but in part he was true. The Red brand owner had to apologyse publicaly and warn his users that the insults and disprestige in his forum won't be tolerated any further and the poor Bloom had to cure his image the way he could.

Examples like those unfortunatly abund.

What happens to Mark appears IMO way over reacted and can ultimatly leads to a disprestige or the ridiculous of an active professional wich I find particularly 1) unfair, 2) unnecessary and ultimately, damaging.

Edit: and I have absolutly zero contact with Mark, we have never ever meet, but the form in some posts of the closed thread honestly shocked me.

No, there is something more in the internet than just a Mark's fault.

And maybe we should all think twice about the form when we want to disagree and opinate.  


 
« Last Edit: January 25, 2012, 02:46:09 pm by fredjeang »
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: Everything matters revisited
« Reply #7 on: January 25, 2012, 02:28:53 pm »

Rob : Wikipedia's entry seems of a different kind of romantism : shortly, it's a way to call each other names among the biblic scholars. Another analogy.

Oops, seems I've been preceded.


Thanks, Niko, I shall now hope to remember that word for future use.

However, I'm happy to see that in both cases of help, reference was made to an external answer rather than a direct piece of information; I hope that means I wasn't totally alone in my doubt ignorance!

;-)

Rob C

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Everything matters revisited
« Reply #8 on: January 25, 2012, 03:01:19 pm »

We know from the film days that collecting more information rather than less is important.  Thus Minox cameras (remember the ones the spies used?) don't do as good a job as 35mm which doesn't do as well as Hasselblad which doesn't do as well as 4x5 which doesn't do as well as 8x10.  Now a lot of Ansel Adams pictures were done with sheet film (4x5 or 8x10) but he used other formats as well.  I have an 8x10 of one of his Yosemite pictures hanging in my dining room which is a 2x enlargement of a 4x5 negative.  The detail is exquisite and and whatever hyper-reality might mean, I suspect that this picture is full of it (personally the only hyper-reality I have ever encountered was when the starship Enterprise moved into warp speed).  It's clear to me that with digital things are still the same.  Sure there are engineering tweaks that can be done and lens design is still an important factor but as was noted on the closed thread, the more photons that you can capture the better off you are in terms of not only having a good image but being able to print big.  With 35 mm Tri-X and my Nikon, I seldom enlarged beyond 7x10.  With a digital Nikon 300, I regularly print at 11x17 (on 13x19 paper) and have had a friend print some images on an Epson 7900 (22x34) with very good results.

While I understand where Michael is coming from with his decision to shut down the discussion on the previous thread, I think he was misreading a lot of us who were not taking issue with any of Mark's fundamental points (which I am largely in agreement with) but rather the tone of the post.

@John Camp - thanks for the link to Carl Weese's piece, I had missed that one.  Again it notes that mechanical issues can be a cause for problematic images (e.g., use a tripod!!!  which I do except when traveling to Europe and cannot lug it along with me).

@Bill - thanks for your printer test.  I did the real life test yesterday in printing a couple of images for my wife's new office.  I looked at the dpi as per Jeff Schewe's recommendation and saw that it was over 360 and so I set the dpi at 720 as per the recommendation.  LR gives a warning that it might not print out because of computer memory issues but I had no problem (and really didn't anticipate any since my computer has lots of memory).  I didn't use a Loupe to examine it but it did look a little better visually than the 360 dpi print (maybe a placebo effect).

While I would like to have a MF camera, I'm not sure that I can justify the expense or the added bulk that such a system would mean in terms of lugging it all around.  I'll take a close look when the new Nikon 800 comes out and see if this is an option.  The one thing I can say, is that my Nikon D300 has some pretty darn good pixels in it.

I'll be interested in Mark's next post on this topic.
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: Everything matters revisited
« Reply #9 on: January 25, 2012, 03:20:23 pm »

While I understand where Michael is coming from with his decision to shut down the discussion on the previous thread, I think he was misreading a lot of us who were not taking issue with any of Mark's fundamental points (which I am largely in agreement with) but rather the tone of the post.

Oh, I know exactly why Mike shut it down (and even predicted so in the thread). When posters start attacking people personally–such as calling into question somebody's honesty–stuff goes down hill very, very quickly. So, he shut it down...he didn't remove it, just shut it down.

This thread will live as long as personal attacks don't happen and people are civil–note I typed civil, you don't have to be polite, just debate amongst yourselves in a civil manner...cross any lines and this will get shut down as well. And, Mike has already warned that personal attacks will result in a ban on the perpetrator.
Logged

dreed

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1716
Re: Everything matters revisited
« Reply #10 on: January 25, 2012, 04:57:59 pm »

Following the shutdown of the previous thread, I went back to the story and skipped the sections that were not about photography...

Perhaps one of the more interesting conclusions is that photos that depict "hyper-reality" are better than those that are not, for some value of the word "better."

There's another side to this: the further a photograph goes down the "hyper-reality" path, the further it strays from being "natural." And if you venture down that path too far, does that then have a negative net impact on the reception of the photograph?

As an example of this, using the flower photographs in the article, photograph B doesn't look right to me because the shadows are too light. At least for me, the reaction is to reject photograph B, irrespective of whether or not it is considered to be technically "better".

If the shadow detail in B could be retained whilst darkening the shadows to the same level as in photograph A, then the result would be a lot more pleasing to at least my eye.
Logged

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Everything matters revisited
« Reply #11 on: January 25, 2012, 06:32:25 pm »

I forgot to add to my last post about the unseen.  Best example that I can remember in terms of capturing this was the photographer played by David Hemmings in Antonioni's 'Blow Up' who inadvertently captures the unseen which then gets revealed to him in the dark room!!! ;D
Logged

Fine_Art

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1172
Re: Everything matters revisited
« Reply #12 on: January 25, 2012, 10:29:32 pm »

I was posting the reply shown below as the thread was being locked. While some of Mark's claims about the per pixel quality of MFDBs as compared to smaller format sensors are not justified, the need for high resolution in making large prints is correct.

Enough of this nonsense about high end stereo and fine wines and back to photography. Roger Clark has a good post on the acuity of the human eye and relates it to the print resolution necessary to represent that resolution. He did some blinded tests using multiple subjects and concluded that at least 600 ppi was necessary (he was using an HP printer). Jeff Schewe states in the Camera to Print and Screen tutorial that with Epson printers one can readily tell the difference between prints at 360 and 720 dpi. So, Roger and Jeff agree on the necessary resolution. Of course, this depends on viewing distance, but this assumes that the viewing distance was quite close. The subjects presumably viewed the prints at an optimum distance. Anyway, Roger concludes that only a large format camera is capable of this resolution when one is dealing with really large prints.

In another post he demonstrates various print resolutions of a 4*5 inch film image and also provides a chart to determine one's printer resolution. Of course, it makes no sense to send more data to the printer than it can print, so one should determine the resolution of his printer. I printed Roger's chart with my Epson 3880, setting the resolution to 1440 and using finest detail and examined the image under a microscope using a 2x PlanApochromat objective with the setup giving a magnification of 25x in the plane of the sensor. The results are shown. The individual ink dots are well shown. The 1 px lines are not resolved, but the 2 px are, giving a resolution of 720 ppi.



In another post entitled the MegaPixel Myth, Roger evaluates what resolution is necessary to literally knock your socks off with a large print. He concludes that 200 MP are necessary. In summary, Mark and the MFDB crowd are correct in that the high resolution of these backs is necessary to make really large prints. To reach the resolution that Roger suggests, one would have to stitch the IQ180, as Jeff demonstrates in the above mentioned tutorial. In summary, the main readily demonstrated advantage of the MFDB is resolution. The availability of high quality apochromatic primes is another factor. Some of the other claims are questionable.

Regards,

Bill
 

I was about to reference the same articles by Clark. You last para is a different conclusion than his. Clark states he does his big landscapes now using large stitched arrays of 135 sized DSLR shots. Of course for wildlife or whatever you need the big resolution.

I appreciate the article. I think the message is fine but with poor examples. I know Mark could have shown us far more impressive shots. I do want to read part 2. If there are some faulty ideas they all get resolved in the discussion. Everybody end up with a better grasp of the subject with the discussion.
Logged

Fine_Art

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1172
Re: Everything matters revisited
« Reply #13 on: January 25, 2012, 11:27:21 pm »

A simple way for me to agree with Mark is to say until there is some way to view an image beside a window looking out over the image, and being unable to tell which is the window, which is the copy, cameras are lacking. No double blind is necessary, there is a cottage industry of HDR software companies trying to make money off filling in the shortcomings. Countless photographers play with sliders in their software to tease out more image quality.

Both screens and cameras have a long way to go.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Everything matters revisited
« Reply #14 on: January 26, 2012, 12:15:06 am »

Hi,

One aspect that is often ignored in this context is the effects of processing. For instance a raw image does actually not have a color, the color arises during raw conversion. There is a lot of discussion on different system having different color, but that color is added in processing and we have a healthy amount of control over the process.


Another factor is sharpening. Digital images need sharpening, but the amount needed differs, depending on OLP filtering but also on fill factor. So to compare two sensors we would need different sharpening. Look at the enclosed image. The three samples are scanned from prints corresponding to about A0 size at 480PPI (interpolated). Left image is from Nikon D3X, the right one from Pentax 645D and the center one is Nikon D3X with deconvolution sharpening. (From this article http://echophoto.dnsalias.net/ekr/index.php/photoarticles/51-a-closer-look-at-pentax-645d-image-quality ).

So to compare two images we would need to apply at least sharpening and color.

Next issue is scaling. Both upsizing and downsizing add artifacts to an image, but downscaling adds more artifacts. Any downsizing needs to be followed by sharpening. So if we compare two images of different size and downscale the larger one to smaller size we add a lot of aliasing artifacts and than sharpen.

The way I see it, the best way to compare is to scale the comparison images to "print size", like I maximally print 70x100 at 200 PPI so I would scale my image to 7874 pixels width before comparison. I would suggest that for comparison its basically more sound to upscale than downscale.

This article may also be of some interest: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/62-scaling-up-or-down?showall=1

Best regards
Erik




« Last Edit: January 26, 2012, 12:34:05 am by ErikKaffehr »
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Everything matters revisited
« Reply #15 on: January 26, 2012, 12:26:33 am »

Hi,

I'd say that there is much more info in an image than what can be reproduced in a print. The problem is in part that media (print) can only reproduce a DR of about seven steps. Would be compress ten or so EV into 1:128 range it would be very boring.

The sample here illustrates this: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/63-lot-of-info-in-a-digital-image

On the other hand a very similar effect can be achieved easily in Lightroom 4 (see attachement): http://www.luminous-landscape.com/forum/index.php?topic=61027.msg492127#msg492127

All the images here are made from a single raw file.

Best regards
Erik


A simple way for me to agree with Mark is to say until there is some way to view an image beside a window looking out over the image, and being unable to tell which is the window, which is the copy, cameras are lacking. No double blind is necessary, there is a cottage industry of HDR software companies trying to make money off filling in the shortcomings. Countless photographers play with sliders in their software to tease out more image quality.

Both screens and cameras have a long way to go.
« Last Edit: January 26, 2012, 12:31:35 am by ErikKaffehr »
Logged
Erik Kaffehr
 

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Everything matters revisited
« Reply #16 on: January 26, 2012, 03:11:30 am »

I was about to reference the same articles by Clark. You last para is a different conclusion than his. Clark states he does his big landscapes now using large stitched arrays of 135 sized DSLR shots. Of course for wildlife or whatever you need the big resolution.

Of course. Since everything matters, stitching has to be considered as an option everytime it has a chance to be relevant.

I don't understand how Mark can show examples shot with a Betterlight scanning back and not adopt stitching. This seems very incoherent to me.

Another thing I find surprising is the apparent assumption that all small cameras are identical in performance.

Cheers,
Bernard

NikoJorj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1082
    • http://nikojorj.free.fr/
Re: Everything matters revisited
« Reply #17 on: January 26, 2012, 04:17:07 am »

In another post he demonstrates various print resolutions of a 4*5 inch film image and also provides a chart to determine one's printer resolution. Of course, it makes no sense to send more data to the printer than it can print, so one should determine the resolution of his printer. I printed Roger's chart with my Epson 3880, setting the resolution to 1440 and using finest detail and examined the image under a microscope using a 2x PlanApochromat objective with the setup giving a magnification of 25x in the plane of the sensor. The results are shown. The individual ink dots are well shown. The 1 px lines are not resolved, but the 2 px are, giving a resolution of 720 ppi.


For me, there's a tad more than a simple resolution value in that...
What I see is that 6 and 4px lines are well rendered, while smaller ones are much mushier, and I'd think of it as a diminishing MTF - from what I've seen on my printer (R1800, older dither but 1pl drops) I'd say that from our metric 254dpi (100 pixels/cm) rule-of-thumb, MTF goes downward rather quickly (I even heard a custom french printer say that resolution beyond 200dpi is mostly wasted). Comparing two prints of the same image at 240 and 480dpi, the bigger one shows more information to my eye.
Alas, there is no way digital printing can equal the stunning clarity of details of a contact print... (see http://www.custom-digital.com/2008/09/bw-print-quality/ eg)
And my conclusion is that more resolution makes more impact if one prints big : for me, even after remplacing my clogged R1800 with a 17" printer, I may waste the resolution of a stitched 100MP file (I'm letting MFDB out of the debate for mundane reasons).

NB RNClark's post URL is http://www.clarkvision.com/articles/printer-ppi/index.html
« Last Edit: January 26, 2012, 04:19:12 am by NikoJorj »
Logged
Nicolas from Grenoble
A small gallery

NikoJorj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1082
    • http://nikojorj.free.fr/
Re: Everything matters revisited
« Reply #18 on: January 26, 2012, 04:27:30 am »

A simple way for me to agree with Mark is to say until there is some way to view an image beside a window looking out over the image, and being unable to tell which is the window, which is the copy, cameras are lacking.
We go deeper into philosophical territory here...
But I'd say that until now (and probably still for a little while), photography has always been a transcription of the reality : first B&W with the orthochromatic tonal rendition (white skies), then panchromatic, then color, but always 2-dimensionnal and within a limited output DR (and that is enough not to fool most of observers).
For me, the bigger part of a photographer's work is to deal with those constraints, and (if successful) to pass around them, adding some kind of transcendance to the image that makes it a symbol, rather than a depiction, of the reality.
That would mean, as Eleanor Brown said in the other topic, concentrate on content rather than form :
Quote from: Eleanor Brown
While I want high quality also, in the end the emotional impact of photographic images is what is most important to me personally.
Logged
Nicolas from Grenoble
A small gallery

viewfinder

  • Full Member
  • ***
  • Offline Offline
  • Posts: 124
Re: Everything matters revisited
« Reply #19 on: January 26, 2012, 05:17:01 am »

I've just come in to thank Micheal and 'LL' for this great article which I really enjoyed and learned much from!

My week has been quite good with two 20x16 inch prints now hung in my local Arts Centre.   They were made with my little Sony NEX 3, kit lens, ebay adaptor and 30 year old Canon FD lenses on salvaged and repaired tripod.   The quality of the prints has surprised me and they stand well among the 'full-frame' work of the other photographers.....

.......And now, this splendid essay which has indeed shown me that everything DOES matter!   Who could have thought that a camera phone could give a digital back such a good chase?   Further, I need never give any more thought to the slipstream from digital backs costing as much as decent cars and piloted by wealthy retirees posing as pro landscape photographers........
Logged
Pages: [1] 2 3   Go Up