Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: keithcooper on July 20, 2015, 05:18:05 am
-
Having recently moved from a Canon 1Ds3 to a 5Ds as my main camera, I've been looking at how the new camera fits into my work.
Although not the biggest part of my business, architectural/landscape print sales do matter, so I was interested to see how I might need to change how I handle the bigger files for print.
Having made steps from 11MP to 21MP to 51MP over the years, I thought a simple print test would show the massive improvements since my 2002 vintage Canon 1Ds (effectively my first DSLR)
http://www.northlight-images.co.uk/article_pages/cameras/canon_5ds-print-comparison.html
Sure, the differences are there, but the real surprise came when I asked some non photographers to compare the prints.
If ever there was a reminder that what we sometimes think of as important doesn't matter one jot to (most) people who actually buy stuff, this was it ;-)
Of course, I could take some solace by viewing it as a testament to my upsizing and printing abilities, but that's not fooling anyone for long!
-
Thanks for this interesting, real world, article…
-
Hi,
I would say I had a similar experience when comparing 39 MP MF with my 24 MP Sony cameras. No really obvious difference in A2 size prints.
Best regards
Erik
Having recently moved from a Canon 1Ds3 to a 5Ds as my main camera, I've been looking at how the new camera fits into my work.
Although not the biggest part of my business, architectural/landscape print sales do matter, so I was interested to see how I might need to change how I handle the bigger files for print.
Having made steps from 11MP to 21MP to 51MP over the years, I thought a simple print test would show the massive improvements since my 2002 vintage Canon 1Ds (effectively my first DSLR)
http://www.northlight-images.co.uk/article_pages/cameras/canon_5ds-print-comparison.html
Sure, the differences are there, but the real surprise came when I asked some non photographers to compare the prints.
If ever there was a reminder that what we sometimes think of as important doesn't matter one jot to (most) people who actually buy stuff, this was it ;-)
Of course, I could take some solace by viewing it as a testament to my upsizing and printing abilities, but that's not fooling anyone for long!
-
Consider what print size would equal the PPI of a MacBook Pro 15" retina screen? The MBP 15 " retina screen is 220PPI. A print of 66,6x100cm is 221PPI. So viewing at 1:1 in Lightroom at distance of about 50cm from the MBP screen would be closer than you would view a 100cm wide print. So any detail that cannot be seen in 1:1 view on a retina screen is in fact irrelevant as long as printing is not exceeding that size. This also means (I would postulate :)) that any differences between lenses and RAW converters not seen in 1:1 are irrelevant. The differences between the 5RsR and the 645Z and the difference between the Otus and the Canon 85 f/1.8 are tiny seen in 1:1 view in my view. Yes, I see small differences but I doubt that would see them on a 100cm side print. Just as I did not see any noticeable difference between the Canon 5D III, the D800E and the IQ160 on an A2 print.
I'm no longer zooming in further than 1:1 in Lightroom on my MBP. I'm looking to get a 4K display for my office at a slighly longer distance the MBP typical viewing distance to see things more realistically than the 30" 2560x1600 display. In other words if a picture looks sharp in 1:1 one should be happy and move on :) I fully realize that there ar details smaller than that, but it would only be seen at a print size significantly larger than 100cm wide.
I guess a lot would disagree with my views here ;)
-
Yes Hans, I'm sure they would ;-)
There certainly are differences, but whether I truly care about them really does depend (to some extent) on the end client and how much I'm being paid ;-)
I've just had to send a crop of a 5Ds image to a client for their printer (big display print) and it was indeed nicer to work on the 5Ds image than one I took earlier in the year on the 1Ds3. My tests with the 9 prints have reminded me though that the image they selected was based on what it showed, not what camera I used...
-
Where I believe you will see the difference, is in a 30 x 40 at 360dpi (for an Epson), 50MP with a good lens will make a better overall print, with less uprez (interpolation by either software ((LR or Epson print driver)).
I have been reworking several images from 11MP, 16MP Canon's all single, attempting to get them to a 30 x 40 at 360 dpi. With all the tools I know of the images will not quite reach that size IMO. Sure the average person may not see the difference, but it's very clear. I also strongly believe in the view at 100% as still the best overall test. Even the 20MP from my 5D MKII has some trouble getting up to this size.
It's also very image subject matter dependent. Thus a macro or image without a lot of smaller details, trees, leaves, etc. will show better with less MP.
After reworking many of my old P45+ files 39MP, taken with various Mamiya lenses, I was pleasantly surprised by the output both at 30 x 40 and 40 x 60.
To me for a large print megapixels do make a difference.
40 x 60, 24 x 72, 36 x 72, 36 x 96 etc.
Paul Caldwell
-
Yes Paul - at those sizes I'd expect to see it quite clearly - don't get me wrong MP does make a difference ;-)
However I limited the size in the experiment to ~20"x30" since that's about as high as I go very often with single image prints. For larger ones they tend to be specially commissioned and quite often stitched, so effective MP can be very high. There are a few 1Ds images that work at 60"x40" but they rely on visual impact and quite a lot of selective processing during the up-rez process.
I would note though that if the average person can't see the difference then that is perhaps a big hint for me to step back and think what I am happy with for the price I'm getting...
I draw a distinction between very big prints that are really going to be looked at carefully and ones that are frankly there for decoration. I find my standards of 'acceptable' already exceed most large scale decorative images I see. That said I've long believed that old engineering adage that 'Perfection is the enemy of excellence' - particularly in a business context ;-)
-
Yes Paul - at those sizes I'd expect to see it quite clearly - don't get me wrong MP does make a difference ;-)
However I limited the size in the experiment to ~20"x30" since that's about as high as I go very often with single image prints. For larger ones they tend to be specially commissioned and quite often stitched, so effective MP can be very high. There are a few 1Ds images that work at 60"x40" but they rely on visual impact and quite a lot of selective processing during the up-rez process.
I would note though that if the average person can't see the difference then that is perhaps a big hint for me to step back and think what I am happy with for the price I'm getting...
I draw a distinction between very big prints that are really going to be looked at carefully and ones that are frankly there for decoration. I find my standards of 'acceptable' already exceed most large scale decorative images I see. That said I've long believed that old engineering adage that 'Perfection is the enemy of excellence' - particularly in a business context ;-)
Hello Keith,
Yes for sure in the 20 x 30, maybe even 24 x 36 size, which are my two most common sold sizes and commercial sizes, I totally agree. This job came out quick and they pulled many images from my older archives. I bid it thinking about the 20 x 30 size and how well I was getting those done. I was a bit surprised when I tried to get to 30 x 40 with some of the file. The raw conversion is much better using modern software like LR 6 or C1 8, but the trick is still taking a native 13 x 9 @ 300 dpi to 30 x 40. I had high hopes for size fixer, but they seem to gone away, as their website is not up anymore.
Nice article, and I am glad to see that the Canon 50MP cameras are getting a good reception as they seem like a major advance in all areas.
Paul
-
Where I believe you will see the difference, is in a 30 x 40 at 360dpi (for an Epson), 50MP with a good lens will make a better overall print, with less uprez (interpolation by either software ((LR or Epson print driver)).
I agree with Paul, that it will be visible, but the differences may be subtle if the output sharpening was less than optimal. The additional real image detail of the 50MP can/should be used for less interpolation guesswork/blur and better sharpening.
One should also use/upsample to 600 PPI on the Canon (720 PPI on Epson) for better large format quality, and even more pixels for sharpening. If known beforehand, one can even optimize for viewing distance. I use Topaz Labs Detail for creative sharpening and either Qimage DFS smartsharpening, or TL Detail, for halo free output sharpening after resampling to the final 600PPI printfile. A tool like Focusmagic can even restore some of the original sharpness from upsampling blur, although it gets slow of large files.
The upsampling quality itself can also make a difference, especially PhotoZoom Pro can create additional resolution on sharp edge detail while remaining realistic on non-edge detail. Lightroom does a decent job of upsampling, but it doesn't add resolution, and its output sharpening is limited to a few fixed settings. IMHO it lacks the amount of control needed to make a real difference.
It is a known phenomenon that we can print larger than we expect in many cases, but that has also to do with the viewing distance, and the fact that we often do not have comparison material at hand to show what we're missing.
Cheers,
Bart
-
I would note though that if the average person can't see the difference then that is perhaps a big hint for me to step back and think what I am happy with for the price I'm getting...
Hi Keith,
Maybe the difference becomes clearer when resampled and printed at, and output sharpened for, 600 PPI for the output size on your Canon printer. For 300PPI output there is just enough resolution for not seeing the differences, but no spare pixels to sharpen better.
I've met/spoken very few people who could not see the difference between 300 and 600 PPI output. Most people do see an improvement, but how much of an improvement depends on several factors (subject/workflow/tools/viewing conditions).
Cheers,
Bart
P.S. My cruel print resolution target (http://forum.luminous-landscape.com/index.php?topic=99905.msg818035#msg818035) can also reveal some alignment/media/viewing distance effects on resolution.
-
Hi Keith,
Maybe the difference becomes clearer when resampled and printed at, and output sharpened for, 600 PPI for the output size on your Canon printer. For 300PPI output there is just enough resolution for not seeing the differences, but no spare pixels to sharpen better.
I've met/spoken very few people who could not see the difference between 300 and 600 PPI output. Most people do see an improvement, but how much of an improvement depends on several factors (subject/workflow/tools).
Cheers,
Bart
P.S. My cruel print resolution target (http://forum.luminous-landscape.com/index.php?topic=99905.msg818035#msg818035) can also reveal some alignment/media/viewing distance effects on resolution.
Thanks - I do sometimes take architectural images with a lot of detail up to 600 on the 8300 - this with the 1Ds3. In general the improvement, whilst (sometimes) visible to me, is not always something anyone else would notice, so does depend to some extent on the effort I want to put in ;-)
I've recently added masked application of Piccure+ to my sharpening tools - Focus magic is still a big part though.
I've not had an image taken with the 5Ds yet that was going to printed for high end display, so this is something I'll be paying more attention to in future, where an image may already be at more than 300 for the size I'm printing.
The important thing for me is to experiment and know what you -can- do if you really want to, and then to decide what you will do.
-
I also strongly believe in the view at 100% as still the best overall test.
I assume this was a response to my post. What I suggested was the 1:1 or 100% view on a 220PPI screen like the MBP is sufficient and the super pixelpeeping is not needed with the 5Ds(R) up to 100cm wide prints. In other words if there are no real differences visible in 1:1 then all is fine. I'm just trying to put things a bit in perspective and suggesting that some very small differences does not really matter and I understood the article as exactly that.
-
I have an Epson 7900, and have come to a general paper size matrix that also agrees with your findings and Paul's:
Up to 13x19 Sheet: Image size limited to max dimensions of 11x17. All captures in my portfolio are printable up to that size, including Coolscan 5000-scanned 35 slides and images from a 20D (8mp).
Up to 24" Long Dimension: Image size limited to 22" on the long edge. Some 5D images (12.8mp) are successful, but many of the scanned slides fall apart to my eyes. 5DII images are fine (21mp).
Up to 24" Short Dimension: The MFD images (80+mp) are fine as big as I can make them. I've farmed out larger prints up to 40"x60", and the MFD images seem like they will hold up an any size given reasonably-aggressive viewing distances. 5DII images start falling apart once I rotate them on the roll paper to get a long edge larger than 22". I haven't printed enough of the a7R images to know where they fit in all this, but my guess is that camera matches my printer quite well, meaning I could print as big as I want assuming an un-cropped image (22x33).
This correlates with a 180-240 native ppi limit for "aggressive" viewing distances. All that being said, I have clients that are perfectly happy with 20x30 prints of a scanned slide, and 24x36 prints from a 5D. If someone is interested in a print that goes beyond my above matrix, I caution them that I will give them the option to preview the print before I do any mounting work where my time and variable costs start to rise.
Of course for long viewing distances and for a lot of other artistic styles, anything may work!
Dave
-
Hi,
My impression is that:
- 12 MP is OK up to A2 size, but more is beneficial
- 24 MP is absolute OK up to A2
- 24 MP is probably OK at A1, but weaknesses may show up
- 24 MP is pretty similar to Velvia 67 scanned at 3200PPI when printed at 30"x40" decent but not really good. Can be probably be pushed quite a bit with excellent processing.
- 39 MP on MFD looks like 24 MP in A2 prints. At A1 size the MP advantage starts to be noticable.
Best regards
Erik
I have an Epson 7900, and have come to a general paper size matrix that also agrees with your findings and Paul's:
Up to 13x19 Sheet: Image size limited to max dimensions of 11x17. All captures in my portfolio are printable up to that size, including Coolscan 5000-scanned 35 slides and images from a 20D (8mp).
Up to 24" Long Dimension: Image size limited to 22" on the long edge. Some 5D images (12.8mp) are successful, but many of the scanned slides fall apart to my eyes. 5DII images are fine (21mp).
Up to 24" Short Dimension: The MFD images (80+mp) are fine as big as I can make them. I've farmed out larger prints up to 40"x60", and the MFD images seem like they will hold up an any size given reasonably-aggressive viewing distances. 5DII images start falling apart once I rotate them on the roll paper to get a long edge larger than 22". I haven't printed enough of the a7R images to know where they fit in all this, but my guess is that camera matches my printer quite well, meaning I could print as big as I want assuming an un-cropped image (22x33).
This correlates with a 180-240 native ppi limit for "aggressive" viewing distances. All that being said, I have clients that are perfectly happy with 20x30 prints of a scanned slide, and 24x36 prints from a 5D. If someone is interested in a print that goes beyond my above matrix, I caution them that I will give them the option to preview the print before I do any mounting work where my time an variable costs start to rise.
Of course for long viewing distances and for a lot of other artistic styles, anything may work!
Dave
-
I assume this was a response to my post. What I suggested was the 1:1 or 100% view on a 220PPI screen like the MBP is sufficient and the super pixelpeeping is not needed with the 5Ds(R) up to 100cm wide prints. In other words if there are no real differences visible in 1:1 then all is fine. I'm just trying to put things a bit in perspective and suggesting that some very small differences does not really matter and I understood the article as exactly that.
Hello Hans,
I totally agree with your statement, I was just trying to put more around it, that's all. For my work the 1:1 view is very important, for both details and noise that may interfere with the final print.
A lot of photographers no longer seem to care about the "print' anymore. It seems all well and good to post on various websites, and be done with it. The thought that that same image may be out of gamut, or not able to stand up to a 300dpi print from Canon even at 13 x 19, no longer seems to matter very much. The web is a totally different space and much more forgiving.
Paul
-
I like to print large on my 9890. I shoot a D810. Am waiting on the new Sony A7rII 42mpx camera.
From first hand experimentation, the megapixels are no longer the limiting factor. It is the lenses, choice of aperture, and shake or vibration. Perhaps none of Canon's line of lenses could resolve 50 mpx. I don't trust it but DXO rates lenses, and I've seen some zooms rated at 8-10 mpx on a 24 mpx sensor. You won't really get a more detailed image @ 50 mpx using that same lens.
At reduced print sizes it may not matter, but larger prints will. Plus, nothing is more exciting than looking up close on a 40x60 print and seeing perfect detail in a cityscape, every window, person on the street.
But seriously, you must compare the highest resolution lens, at its optimal aperture on a tripod. Remember such tightly packed pixels will be effected by the tinniest vibration during an exposure.
Thankfully, the Sony will have 5 axis stabilization.
-
Whilst I agree that good lenses help, I've just been testing my 5Ds with an Olympus 50/1.2 Zuiko, via an adapter. The bits that are sharp at f/1.2 are sharper on the 5Ds than the 1Ds3, so the rising tide really does float all boats.
As to the pixel density and need for super stable tripods etc... Well, the pixel density of my 100D is only a bit less than the 5Ds, and I only put that on a studio stand (with Stackshot) for macro work.
The 5Ds is excellent but I still won't be using a tripod for my daytime landscape photography (I do for my architectural and much commercial work). Nor will I be rushing out to buy new lenses.
Where I want really high res, I'll still be stitching. The limits of the Canon iPF driver give me a maximum print length (at 300 ppi) of around 17 feet (~61k pixels - limit found by expensive experiment ) to print.
For a truly sharp 60x40 I'd want to have 300ppi of real resolution or 18k x 12k pixels - this is what I mainly have a gigapan for, not panoramas.
The 5Ds means I could easily take this to 600ppi of real resolution, which I know from the 9 test prints would make a difference - mind you, a difference that I'd likely need to leave a box of magnifying glasses near the print for anyone viewing to actually notice ;-)
-
I like to print large on my 9890. I shoot a D810. Am waiting on the new Sony A7rII 42mpx camera.
From first hand experimentation, the megapixels are no longer the limiting factor. It is the lenses, choice of aperture, and shake or vibration. Perhaps none of Canon's line of lenses could resolve 50 mpx. I don't trust it but DXO rates lenses, and I've seen some zooms rated at 8-10 mpx on a 24 mpx sensor. You won't really get a more detailed image @ 50 mpx using that same lens.
At reduced print sizes it may not matter, but larger prints will. Plus, nothing is more exciting than looking up close on a 40x60 print and seeing perfect detail in a cityscape, every window, person on the street.
But seriously, you must compare the highest resolution lens, at its optimal aperture on a tripod. Remember such tightly packed pixels will be effected by the tinniest vibration during an exposure.
Thankfully, the Sony will have 5 axis stabilization.
Lenses are improving too. Not just OEM but Zeiss and Sigma lenses as well. New companies competing at lower image quality market segments.
For reproduction work the sensors with image stabilization get the ability to shift the sensor one pixel 4x or 8x times. The Olympus OMD E-M5 II and a Pentax model have that already, the Sony FF versions likely candidates. It demands less of the lens resolution compared to sensors with 4x the resolution and dampen the color artefacts in neutral subjects to a high degree compared to Bayer sensor single exposures.
The increasing resolution route without revolutions in sensor or camera technology is still the strategy of Canon. Samsung and Sony show the real improvements with back illuminated sensors that have better low light and dynamic range properties. Not that interesting for reproduction work but as important as resolution in almost any photography job including work for large prints. The Sony A7R II is the one I aim for too, if pixelshift is introduced on that model it will be hard to beat that camera with any high resolution sensor on any photographic job. A lot of Canon lens owners change route now, Metabones adapter offers AF on this A7R II body for several Canon mount lenses.
I wonder whether Canon's own wafer stepper machines are still used to make Canon FF and APS-C sensors while Samsung and Sony probably have switched to the ASML equipment for their sensors, they have ASML wafer steppers for more than a decade now including the latest technology. Both Canon and Nikon lost their market share in wafer steppers to ASML that now has 80% of that market. Zeiss delivers the very special optics in the wafer steppers of ASML. Nikon has been more pragmatic in the use of third party made sensors than Canon. Next to innovating new camera/sensor designs, Canon may be forced to switch to third party suppliers for its FF and APS-C sensors like they already did for compact camera sensors.
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots
-
I understand how sensor shift works. What I don't understand is how "super sampling" say a 1980's era lens will get higher resolution, actual detail from the lens. To me the maximum resolution is still limited by the optics.
On a side note, I don't always care about sharpness. I've used some very cheap lenses for floral photography and they render beautiful soft gradations.
So, with sensor shift, can you really ernd up with a much sharpewr image from say an old Nikor 28mm f 2.8 lens?
-
Since system resolution is (essentially) a combination of sensor and optics, improve one and the total is improved. Of course, the fact that system resolution is improved doesn't actually mean it's a useful improvement ;-)
In another experiment this afternoon I put my 1958 Tamron 135/4.5 (the 'Duo') onto the 5Ds (with an AF confirm M42 adapter).
Yes indeed, there is more detail in the sharp bits, but to be honest - so what? I've tried out lenses like this in the past and whilst it's fun, I can't think what I'd actually do with them.
The only old lenses I do sometimes make use of are some M645 ones that I use with a shift adapter.
-
My view on MPs includes my experience that every geometric adjustment -- lens corrections, stitching, perspective control, etc. -- causes a loss of information/sharpness. So, the more information I start with, the more likely I'll end up with an image file that is, at printing time, still sharp.
I also think that whatever issues are introduced by the Bayer pattern will end being reduced as the MPs increase, whether or not the pure resolution in ideal circumstances is affected noticeably.
So, while balancing noise is critical, all else being equal, I want to see lots more MPs. I don't think we're at the end of the line there for those of us who do landscapes and large prints. (And the cell phones may take over most of the rest of the market.)
(Yes, my pre-order is in for the a7rii.)
Paul
www.PaulRoark.com
-
I'm all for more MPX. My Uncle is getting the A7rII in a week or so. It is more important to pick lenses, modern lenses intended for high MPX. Luckily, Zeiss is making the Sony FF lenses, and they are aware of the sensors they need to resolve to.
The saddest thing is finding a favorite photo you shot years ago at non optimal mpx or even only from a jpg. and know you won't be able to print it larger.
I have 8,000 photos I shot in Cuba @ 15.9 mpx. Because I know what I'm doing I can get very nice 24x36, even some 30x60s. I always shoot at low ISO, well exposed raws, and a good noise reduction/ uprezzing workflow.
So 42 mpx will be very nice. Like I said, I have a new roll of 44" glossy to play with:-)
-
I understand how sensor shift works. What I don't understand is how "super sampling" say a 1980's era lens will get higher resolution, actual detail from the lens. To me the maximum resolution is still limited by the optics.
On a side note, I don't always care about sharpness. I've used some very cheap lenses for floral photography and they render beautiful soft gradations.
So, with sensor shift, can you really ernd up with a much sharpewr image from say an old Nikor 28mm f 2.8 lens?
I wouldn't describe The R>G>G>B pixelshift as super-sampling or over-sampling, the Bayer sensor/filter concept is more a kind of subsampling on color. You still need the lens resolution to address the individual pixels but the color information is limited (not considering the anti-aliasing sensor lenses). If the sensor resolution would be described per cell of RGGB pixels (or AA lens) we would read way lower sensor resolution figures. The dilemma Fovean had to deal with to get realistic figures for comparing both systems. R>G>G>B pixel shift is a 1:1 sampling on Bayer sensors, like a monochrome sensor + R>G>B filtering in time is a 1:1 sampling for RGB color. Alright, there is possibly some Green oversampling done in the first example.
Olympus adds oversampling on top of the R>G>G>B pixelshift with 4 extra steps halfway the pixel pitch, Pentax doesn't. Olympus does not deliver more than 64 MP in the RAW file (sensor 16MP) and concentrates the data even more in the JPEG output. Like with scanners that oversample, there is a gain on the signal/noise ratio which translates in better information. A solid base for upsampling and sharpening if this output is not directly expressed in sharper images.
I do not advocate the use of low quality lenses but at a point either the sensor or the lens out-resolve one another. With all sensor and lens tests we see the warning that they should be compared with one lens or one camera system. Enough lenses of the analogue past out-resolved the first digital sensors, that is hardly the case with today's sensors. Imaging Resource used the Sigma 70mm macro a long time for camera tests and it out-resolved a lot of sensors in the past. If the lens is optimal for a given Bayer sensor, both not out-resolving one another, then pixelshift R>G>G>B with that combination will still deliver more information than a single exposure with that system. Olympus' over-sampling can add information on top of that. Take the same lens and put it on a Bayer sensor with half the pixel pitch so 3x or 4x the resolution and you will not gain the same information with a single exposure, a lens with more resolution is needed. That is what I had in mind.
We have seen the pixel shift used in MF digital backs to compete with analogue large format and digital scanback resolution figures. It is introduced in Micro 4/3 as the sensor size limits higher resolution numbers, not to mention few lenses in the Micro 4/3 systems can actually keep up. Panasonic announced a 20MP M4/3 sensor recently so I wonder what it will improve. The first APS camera has pixel shift and I expect some FF cameras will follow. Back illuminated sensors will improve with a wider dynamic range, better low light specs or more resolution for these sensor formats too but there is an end to it. Either sensors hitting physical laws on visible light or lenses becoming too expensive to keep up with the sensor resolution. Decades later than the wafer steppers hit on similar conditions and they go beyond the visible spectrum, the cameras we use can not. Yes, this is speculative. On the other hand some years ago I did wonder whether mechanical aspects of DSLRs or mirror-less would not limit further pixel pitch shrinking as the vibration goes beyond pixel pitch. Canon and Olympus had to deal with that aspect if you read in between the lines of recent DPreview camera reports. Sony encountered the same issue.
On the use of (old) soft portrait lenses on high resolution sensors. There is no law forbidding it, there are tastes, there are conventions and any print that satisfies the viewer makes this whole thread pointless anyway. Upsampling from lower sensor resolutions with algorithms that deliver soft images has some analogy with the use of said portrait lenses though.
Going back to the subject line. I estimate that in best case; optimal inkjet paper coating, smallest ink droplet, best droplet addressing, the required PPI at printer output stage will not exceed 450. Bart van der Wolf's print target could add the right numbers. That said, I expect that if one starts from 225 PPI quality pixels and use best upsampling routines and sharpening it will be damned difficult to make the distinction between the same size prints made from either starting point. Upsampling on the 450 PPI data not translating in a better print, the information then simply out-resolving the printer/paper combination. The numbers falling back to 300 PPI and 150 PPI with heavy textured, matte papers. The best camera + lenses today will exceed A2 format data requirements with a good workflow on a 3.5 picoliter droplet printer, no printer model that can print A2 has smaller droplets than 3.5 picoliter. Starting from 50 MP quality pixels, with 150PPI input, sizes of 2 M2, 22 square feet on heavy textured matte paper will still be acceptable.
Few people start from the print size they use in practice and decide then what camera quality is needed.
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots
-
A slightly tidied up version of my original article has been added to this site.
In another related test, I looked at the 5Ds with a couple of old lenses, confirming that more MP can help with almost any lens, even if spotting the difference needs big prints or tight crops. Pretty much what I'd expected ;-)
http://www.northlight-images.co.uk/article_pages/cameras/canon_5ds-old_lens.html
The Tamron Twin-Tele 135/4.5 (from 1958)
-
The Tamron Twin-Tele 135/4.5 (from 1958)
So cute ;)
Indeed, a denser sampling of the same lens projection will allow to extract more detail from an image, although it's a tandem of lens and sensor so the worst of the two will set a ceiling. By the time we reach 1 micron sensel pitch values, then we'll reach the practical limit of what a lens has to offer.
Cheers,
Bart
-
Obviously an improvement from previous Canon offerings, but not enough to make me regret switching to Nikon and this 3 year old camera:
http://www.dxomark.com/Cameras/Compare/Side-by-side/Canon-EOS-5DS-versus-Nikon-D800E___1008_814
But if you have a stock of Canon glass you don't want to give up, it's a no-brainer. Still, I have to say it's the dynamic range of Nikon's Sony sensors that really made the trade-up worth while to me. And I'm happy with the 40"x60" prints (largest I can print.)
Maybe Canon will be a leader again some day.
-
For my own work, it's still primarily about the lenses. I've just added the Canon EF11-24 to my EF8-15, TS-E17 and TS-E24mm
The DxO figures mean somewhat less for me if a lens type I use isn't available ;-)
-
And that's obviously the best way to approach gear...it's just frustrating that Canon seems to be so half-hearted about their upgrades, not only for this series ever since the 5DII, but also for the video capabilities of their DSLRs, what with offerings by Panasonic and Sony that show those manufacturers apparently more interested in listening to their customers.
For my own work, it's still primarily about the lenses. I've just added the Canon EF11-24 to my EF8-15, TS-E17 and TS-E24mm
The DxO figures mean somewhat less for me if a lens type I use isn't available ;-)
-
A good thorough article as usual Keith.
-
..., Zeiss is making the Sony FF lenses, and they are aware of the sensors they need to resolve to.
Is that actually true, or are Sony having 'Zeiss' lenses made to Zeiss specs?
Brian A
-
Incredibly precise job this article.
But, I read the sentence: "If you don’t print over A3+ size (13″x19″) then any advantage of the 5Ds over the 1Ds3 (or 5D mk3) is going to be minimal."
OK I bought a 5D Mk III 2 years ago.
OK… I keep it and am happy
-
Obviously an improvement from previous Canon offerings, but not enough to make me regret switching to Nikon and this 3 year old camera:
http://www.dxomark.com/Cameras/Compare/Side-by-side/Canon-EOS-5DS-versus-Nikon-D800E___1008_814
But if you have a stock of Canon glass you don't want to give up, it's a no-brainer. Still, I have to say it's the dynamic range of Nikon's Sony sensors that really made the trade-up worth while to me. And I'm happy with the 40"x60" prints (largest I can print.)
Maybe Canon will be a leader again some day.
Canon glass can be used on the A7R II. Add image stabilisation too for Canon lenses without that feature.
No DXO tests of that camera/sensor yet but it would surprise me if it would not equal or surpass the D800E.
An interesting image quality test of pixel shift cameras versus the D810:
http://www.imaging-resource.com/PRODS/pentax-k3-ii/pentax-k3-iiTECH2.HTM
Both the Pentax K3-II and Nikon with the Sigma 70mm macro lens.
Any reason why Sony would not use pixel shift on the replacement of the A7 II or A7R II ?
Or prolong the A7R II's place in the market with a firmware upgrade like that ?
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots
-
Hi Keith, Thanks for a great review.
I advise a few garden photographers in the UK, and every so often 1 metre+ prints are required by their clients or print sellers for exhibition stands etc.
As you can guess general garden views with lots of plant detail would really benefit from a good many more quality pixels that this camera could potentially offer.
In the past I found that the deshake filter works wonders on 1ds Mk111 garden shots, even mounted on a tripod when the mirror is up and a separate release is used, guess that's why the block around the shutter on the 5dsr was made a little more robust.
-
Any reason why Sony would not use pixel shift on the replacement of the A7 II or A7R II ?
Or prolong the A7R II's place in the market with a firmware upgrade like that ?
Met vriendelijke groet, Ernst
It may be a question of licensing the pixel shift technology. I think it may belong to Olympus.
-
It may be a question of licensing the pixel shift technology. I think it may belong to Olympus.
Pixel shifting must be more than two decades old now, it has been used even before Sinar MF digital backs started to use it in 2000: http://www.jenoptik.com/en-progres-25-years-1989-1999
Maybe an Olympus patent on pixel shift combined with sensor image stabilisation but we see that Pentax can use that method too. Sensor image stabilisation has been used since 2003 by Konica-Minolta so before Olympus did and Sony must have the KM patents or at least a deal to use that technology. It has used the technology for other camera models since. So far I mentioned actual camera/sensor introductions and no patents, often enough patents are in the hands of companies that will not use the technology themselves.
No, I do not see patent issues preventing Sony to go that route even if Olympus has control, the market segments of FF and M4/3 differ enough to make a license deal possible.
EDIT: to put this thread in perspective http://www.dpreview.com/forums/post/56218279
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots
-
It may be a question of licensing the pixel shift technology. I think it may belong to Olympus.
Patents normally protect how something is done, not the concept itself which in this case is several shots combined into one.
You cannot patent washing windows for example, but you can patent a handheld fancy vacuum device's workings.
-
In the US Apple etc could today patent a round wheel and a square one IMHO. Nothing in patent law is impossible meanwhile.
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots
-
Having recently moved from a Canon 1Ds3 to a 5Ds as my main camera, I've been looking at how the new camera fits into my work.
Although not the biggest part of my business, architectural/landscape print sales do matter, so I was interested to see how I might need to change how I handle the bigger files for print.
Having made steps from 11MP to 21MP to 51MP over the years, I thought a simple print test would show the massive improvements since my 2002 vintage Canon 1Ds (effectively my first DSLR)
http://www.northlight-images.co.uk/article_pages/cameras/canon_5ds-print-comparison.html
Sure, the differences are there, but the real surprise came when I asked some non photographers to compare the prints.
Diminishing returns!
First, the limitation of the printing technology comes into play. Epson will tell you that 360ppi is the limitation of the technology so if the print size requires less than the original image, you have hit a printer limitation.
Also, printing dimensions doubling requires a quadrupling of the MP count (and print area) so 48MP allows double the size of 12MP, not 24MP, at the same output ppi.
There are more uses for those pixels as well. And more options. I shoot the D810 in 1.2X crop mode at 25MP for some sports situations.
-
Hi,
My impression is that:
- 12 MP is OK up to A2 size, but more is beneficial
- 24 MP is absolute OK up to A2
- 24 MP is probably OK at A1, but weaknesses may show up
- 24 MP is pretty similar to Velvia 67 scanned at 3200PPI when printed at 30"x40" decent but not really good. Can be probably be pushed quite a bit with excellent processing.
- 39 MP on MFD looks like 24 MP in A2 prints. At A1 size the MP advantage starts to be noticable.
Best regards
Erik
That's exactly what I have found. You'd think there would be more difference between FF 12 mp and 24 mp at A2 size, but in fact it's subtle at normal viewing distance.
However you can see the benefit of 24 against 12 when cutting out, and I guess 50 will be better again for this, assuming decent technique.
-
That was a great review. I was very impressed by the comments made by average people who didn't really notice the difference with all those extra pixels, that content and interest appears to be more important.
-
That was a great review. I was very impressed by the comments made by average people who didn't really notice the difference with all those extra pixels, that content and interest appears to be more important.
Hi Alan,
Do note that the prints IIRC were made at 300 PPI, and maybe output sharpening was (therefore) not as good as is possible when one does have that extra, 4x as much, detail at 600 PPI to sharpen with. Of course relatively casual viewers will look more at the image itself than the technical quality of it. They will subconsciously 'experience' a difference but will not be able to put that into words.
It's a burden I happily carry with me all my life, but technical image quality is important to me because, besides showing attention for detail in preparing the image (and thus taking the viewer serious by not holding back), it does work on the subconscious observation. Material texture adds 'life' to the subjects we image, and if done well, it's not readily noticeable but it's present, as if subjects become more tactile, but not harsh.
I remember when I did my exams for my Photographers license (was mandatory for settling as a professional photographer many moons ago), and one of the examiners who had to judge my assignments looked at my assignment 'glass'. I had shot many different subjects with glass as the main motif, but finally settled for one beautiful stained glass window. The examiner looked at the large print that I had submitted and asked me with which view camera I had taken the image. He then turned over the printed image with my negative attached to the back in a sleeve, looked at the tiny 35mm negative, and said: "Okay, next assignment ...".
Cheers,
Bart
-
https://luminous-landscape.com/canon...t-performance/
I still haven't yet pulled the trigger on that new body...I found that treatise interesting as he
compared it with the same body that I've used for years.
That fact that the "normal" client, for the most part, couldn't discern a difference in printed
images...he could and, IMHO, that matters!
I have made a habit of photomerging images, simply to create those large images which could
be routinely coughed out by that 5Ds.
My questions are...would my large, merged images, equal the print IQ as might be obtained
by that high MP body, assuming my skills were up to the task and equal pixel/dpi numbers?
-
I'd say a stitched image would be sharper. The percentage of enlargement of the optical image will be far less. But you are limited to stationary objects, and occasionally, I've messed up and missed a corner. I like both techniques.
-
Hello All,
I recently pulled the trigger on a 5Ds; I can say clearly that the difference in prints as small as 8.5X11 is noticeable; the difference at a print size of 11X17 is visible, and larger is more clearly visible. At the 8.5X11 size, the difference is subtle, but real. The print has more realism..like a contact print from chemical days of yore. At 11x17, the images is "touchable" compared to a similar image taken with a 5D3 or 20ishmpx camera. These are my observations...very happy with the camera. I have had more than a few images that were handheld, 1/100 of a second, 30 or 50mm (the 24-70LII) and the IQ was outstanding.
The handling is very refined - everything works and works together...EOS concept at a high level.
For those of us who still print and enjoy the printed image, it is a real treat in a small package...
Regards,
George
-
Hi,
What I think Keith has noted is that the higher resolution images actually have more detail, clearly observable with a loupe or using a macro lens. On the other hand it seems that the extra detail doesn't matter to viewers. On the other hand viewers can pick up thinks like differences in global contrast and small differences in composition.
This is pretty much what I have seen comparing prints in A2 size from my 24 MP and 39 MP cameras. More detail in print but little perceived difference.
My guess is that we see is that the human vision is most sensitive to medium frequency detail. The eye can resolve fine detail but medium frequencies will dominate perception, see linked figure from Wikipedia. So if we discuss detail visible at 360PPI at 25 cm, the contrast sensivity of vision is several magnitudes below maximum. To me that indicates that we should focus our sharpening efforts more on medium frequencies than actual pixel detail.
(https://upload.wikimedia.org/wikipedia/commons/f/fa/Contrast_Sensitivity_vs._Spacial_Frequency.png)
What I have observed that 12 MP was pretty good enough for A2 size prints. At my recent exhibition I had a mix of 12 MP and 24 MP images, reflecting the state of shooting irons in my possession at that time. On most images 12 or 24 MP mattered little. Could be that 24 MP or higher would make a noticable difference.
But, on one images there were a lot "dust specs". I realised that those were birds and not dust, but 12 MP was not enough to resolve them. Now, 24 MP may not have been able to resolve them either, that is just a 41% increase in resolution.
The impression I got was that the visitors placed that image #2 of all pictures at the exhibition.
(http://echophoto.dnsalias.net/ekr/Articles/Exhibitions/BergDalOchVatten_1/20080928-DSC00956.jpg) (http://echophoto.dnsalias.net/ekr/Articles/Exhibitions/BergDalOchVatten_1/20080928-DSC00956.jpg)
This was #1:
(http://echophoto.dnsalias.net/ekr/Articles/Exhibitions/BergDalOchVatten_1/20140604-_DSC4093.jpg) (http://echophoto.dnsalias.net/ekr/Articles/Exhibitions/BergDalOchVatten_1/20140604-_DSC4093.jpg)
Best regards
Erik
Hi Alan,
Do note that the prints IIRC were made at 300 PPI, and maybe output sharpening was (therefore) not as good as is possible when one does have that extra, 4x as much, detail at 600 PPI to sharpen with. Of course relatively casual viewers will look more at the image itself than the technical quality of it. They will subconsciously 'experience' a difference but will not be able to put that into words.
It's a burden I happily carry with me all my life, but technical image quality is important to me because, besides showing attention for detail in preparing the image (and thus taking the viewer serious by not holding back), it does work on the subconscious observation. Material texture adds 'life' to the subjects we image, and if done well, it's not readily noticeable but it's present, as if subjects become more tactile, but not harsh.
I remember when I did my exams for my Photographers license (was mandatory for settling as a professional photographer many moons ago), and one of the examiners who had to judge my assignments looked at my assignment 'glass'. I had shot many different subjects with glass as the main motif, but finally settled for one beautiful stained glass window. The examiner looked at the large print that I had submitted and asked me with which view camera I had taken the image. He then turned over the printed image with my negative attached to the back in a sleeve, looked at the tiny 35mm negative, and said: "Okay, next assignment ...".
Cheers,
Bart
-
My guess is that we see is that the human vision is most sensitive to medium frequency detail.
Hi Erik,
That's correct, but it is very much connected to things like output sharpening and printing at 600 or 720 PPI.
The eye can resolve fine detail but medium frequencies will dominate perception, see linked figure from Wikipedia. So if we discuss detail visible at 360PPI at 25 cm, the contrast sensivity of vision is several magnitudes below maximum.
Yes, unless we boost the contrast of the lowest spatial frequencies of the output image!!! Then we do not only see a bit more micro detail, but also the spatial frequencies from medium frequencies to those highest spatial frequencies will receive more signal/contrast. That would partly compensate for the dropping sensitivity of human vision, and the combined (boosted) input signal x CSF = output signal drop will be slower. The image subjects will look more 'tactile' as a result. Having the additional pixels at 600 or 720 PPI, helps in boosting the signal even more, without creating artifacts that would be more likely to become visible with only 1/4th of the pixels to calculate with.
That is separate from all the other cognitive processes involved in 'vision', but in a double blind test when specifically scoring on perceived detail, it should make a difference.
To me that indicates that we should focus our sharpening efforts more on medium frequencies than actual pixel detail.
The problem with that is, that we usually cannot control the viewing distance from the output. So what may be medium spatial frequencies at a proper viewing distance that allows us to take in the entire image/composition without turning our heads too much, will become low frequency detail, and high frequency will become medium frequency detail, when we get closer. That's why we need to also sharpen the higher spatial frequencies, and that will automatically also start boosting the medium frequencies. So that makes the image quality less sensitive to different viewing distances. It's like boosting the MTF curve near the Nyquist frequency, it will lift all spatial frequencies below it as well, but in a controlled gradual/monotonic way.
Figure 24 at this site (http://webvision.med.utah.edu/book/part-viii-gabac-receptors/visual-acuity/) shows that the contrast sensitivity for the spatial frequencies also depends on the average luminance level or viewing conditions. Spatial frequencies of 60 cycles/degree (= average 20/20 vision, younger people can do better than that) would be around 300 PPI at reading distances, and are worthwhile to boost if we want to exploit that. We can do it better and with fewer artifacts when we boost the levels with even finer detail, say 600 or 720 PPI, and the lower frequencies will smoothly follow as well.
BTW, as cited in the link above, the receptor array in the human visual system can resolve in the order of 6/1 (20/3) or ~150 cycles/degree, which one might add would roughly match the 600-720 PPI at reading distance range. However, the rest of the eye's structure (a.o. lens) will reduce that to closer to the earlier mentioned 60 cycles/degree.
The impression I got was that the visitors placed that image #2 of all pictures at the exhibition.
That may well be because they loved the scene (as they should), more than they scored it on perceived resolution, crispness, 'tactile' quality. But those qualities do help to enhance the immersive quality of being there, if done well.
At the risk of repeating myself too much, Topaz Detail does just that, extremely well (although it gets slower with all the calculations needed as sizes increase), and it not only allows to control the small details but also those 'medium' detail frequencies separately for even more control. Detail can even be locally brushed in (or out) with edge aware masking, to avoid sharpening the noise in e.g. smooth sky gradients. Every enhancement smoothly transitions between different spatial frequencies, and can be finely controlled for overall tonality, and/or additionally for shadows and/or highlights.
Cheers,
Bart
P.S. In the book, A System Engineering Approach to Imaging (https://books.google.nl/books?id=2V9dMucKgf0C&pg=PA392&dq=threshold+contrast+curve+human+visual+acuity&hl=en&sa=X&ved=0CCEQ6AEwAGoVChMIlordtY-SxwIV6WtyCh0X3Q5e), By Norman S. Kopeika, there is a chapter on the "Threshold Contrast Curve". I've attached a chart from that chapter. It basically shows, translated to our print resolution challenge, that we need to boost the contrast of the highest spatial frequencies of the input signal (towards an MTF=100%) more, to offset the loss of contrast sensitivity of the human visual system's MTF . It basically peters out at 60 cycles/degree, but Nyquist taught us that we need more than twice that sampling frequency to reliably resolve (and edit) those frequencies.