Ad
Ad
Ad
Pages: [1] 2 3 »   Bottom of Page
Print
Author Topic: How Sharp Is Your Printer? How Sharp Are Your Eyes?  (Read 11196 times)
wolfnowl
Sr. Member
****
Offline Offline

Posts: 5807



WWW
« on: January 06, 2010, 01:42:47 AM »
ReplyReply

from Mike Johnston's 'The Online Photographer' page...

Interesting read...

Mike.
Logged

If your mind is attuned to beauty, you find beauty in everything.
~ Jean Cooke ~


My Flickr site / Random Thoughts and Other Meanderings at M&M's Musings
Ernst Dinkla
Sr. Member
****
Offline Offline

Posts: 2911


« Reply #1 on: January 06, 2010, 03:09:57 AM »
ReplyReply

Quote from: wolfnowl
from Mike Johnston's 'The Online Photographer' page...

Interesting read...

Mike.


While I do not think much will be changed in the conclusion that 450 PPI input is the optimal input for best image quality possible in a print, it might be a better method to keep up- and downsampling limited to one step in the process. The method as described uses two steps and possibly two different algorithms to get to native printer input resolution. There's no mention of smart print sharpening which could make a difference too.



met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
Logged
Deepsouth
Jr. Member
**
Offline Offline

Posts: 81


« Reply #2 on: January 06, 2010, 12:52:50 PM »
ReplyReply

This is an interesting approach, but I question how much will be visible in thumbnails. I recommend 8 x 10 prints of unfamilar subjects and compare no more than three steps (in octaves, e.g. 180, 360, 720)  at a time. In fact, I'd skip the lowest octave and go to the next two higher. Then repeat for each media you use.

As I've mentioned before, manufactuer claims about print resolution are fairly meaningless above 600 dpi. If the Epsons and Canons would be forthcoming about how their upsampling algorithims work, then we could all form educated opinions about upsampling to the "printer resolution".

There are a lot of variables in play here.  My simpler advice: find a resolution that works with your printer, your media, your finished print size and your subject matter. Don't worry about it being too high-we are long past the era of worrying about computer memory resources. Stick with it until you have a compelling need to change, whether that reason is ink set, printer, media, or print size.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #3 on: January 06, 2010, 03:36:07 PM »
ReplyReply

Quote from: wolfnowl
Interesting read...


"Interesting" perhaps...but filled with some "wrongess" for sure...

Resampling anything anywhere will have an impact down the road when it comes time to sharpen...which by the way the author failed to address at all. (the failure to even address resampling for the contact sheets and the use of output sharpening for the printing makes this whole test nothing other than mildly "interesting").

The other main part of the equation that the author seems to have glossed over (since he may not understand it) is the issue of print SIZE and the human eye's visual acuity (loosely called "resolution").

It's not easy to compare the "PPI/DPI" of human vision to printed output but you can. Human visual acuity is generally defined as being capable of resolving a spacial design whose features are separated by one minute of arc (or 1/60th of a degree). What that one minute of arc represents is entirely dependent on viewing distance. The eye can see more resolvable detail closer than further away (assuming you don't bee reading glasses–all of this is based upon 20/20 vision).

Given that 1/60th is 0.00029089 radian, you can calculate the threshold of visual acuity for a given distance. Calculate the limit L of visual acuity at distance D with this formula  L=D*TAN(0.00029089).

What this will give you is the following viewing distance and resolution required table:

Viewing Distance (inches)____ Limit (inches)______Resolution (DPI)

8 _______________________ 0.00232 ______________ 428
12 ______________________ 0.00349 ______________ 286
15 ______________________ 0.00436 ______________ 229
18 ______________________ 0.00524 ______________ 191
20 ______________________ 0.00582 ______________ 172
24 ______________________ 0.00698 ______________ 143


A couple of things related to the above...these are not my numbers but Bruce Fraser's numbers from his Real World Image Sharpening book )both the original and the one I updated). But since I'm coauthor of the current edition I stand by them (I was also the guy who asked Bruce the original question of "how many DPI can our eyes see Bruce?" which he referred to as my sending him down yet another rabbit hole).

The above assumes 20/20 vision...there are a lot of people whose vision is better or worse. The older one gets, the less close focus vision most people have (and the more people who need to wear "reading glasses"). So you milage may vary...

The above also assumes high contrast line pairs which while useful for measuring don't automatically translate directly to low contrast photographic textural detail. So, the odds are the practical limit may be a bit larger and hence the required resolution for continuous tone is prolly lower.

In general the "intended viewing distance" is said to be about the 1-1.2x the diagonal. So if you are holding an 8"x10" print in your hand, you'll prolly be holding it about 13-15 inches away from your face and be able to resolve about 286DPI. If you move it closer, you'll need more resolution as the eye can resolve more. If you move it away you need less. If your image is 30x40" the intended viewing distance is about 50" and you would need far less resolution for the print.

The downside about intended viewing distance is it assumes the viewer wants to see the entire image in their visual field...which of course is not always the case. Bruce used to say that the "intended viewing distance for a photographer was limited only by the length of their nose (or the magnifying power of their loupe)"

The next issue is the resolving capability of the medium. Even the smoothest matte papers can't resolve as much detail as glossy or semi-gloss surfaces. Beside the effect of dot "gain" you also have the issue of dot diffusion. So how much resolution you NEED for a continuous tone appearance in a print depends on the viewing distance and the nature of the media.

In general Bruce used to suggest 180PPI-480PPI depending on the viewing distance (either intended or likely) and the media.

The other question (which is of course close to my heart) is how the heck you sharpen an image for optimal printed output. That's a different subject that somebody could write a whole book on...oh, wait, we DID :~)

So,while the original author has some "interesting" ideas and suggests some "interesting" tests, I'm not at all sure his conclusions are as "interesting"...
« Last Edit: January 06, 2010, 06:13:03 PM by Schewe » Logged
Randy Carone
Sr. Member
****
Offline Offline

Posts: 545


« Reply #4 on: January 06, 2010, 04:10:55 PM »
ReplyReply

Thank you for that Jeff. My first thought was that sharpening was not addressed. When I first became "aware" of the subject of sharpening, it was similar to the first day I got glasses and played slow pitch softball. WOW. Move forward to the day I became aware of PK Sharpening. It was like getting tri-focals!
Logged

Randy Carone
Farmer
Sr. Member
****
Offline Offline

Posts: 1631


WWW
« Reply #5 on: January 06, 2010, 04:59:19 PM »
ReplyReply

It's odd that in one of the responses to a reader's comment, he says that in the real world you can't control the resolution because your file is fixed and your print size is whatever you want it to be and he doesn't recommend resampling.  If that's the case, then it makes no sense to recommend to people to "test" where the sharp point is on their printers because the same "real world" limitation will come into play.

I would agree that resampling is sub-optimal and that you're better providing, say, a 259ppi file to the driver than resampling to 300 or 360 etc in most cases (in some cases, with excellent algorithms, you might see an improvement in upressing yourself).  Using the likes of PK sharpener really is a better way of controlling the quality of the printed result.
Logged

Neuffy
Newbie
*
Offline Offline

Posts: 42


« Reply #6 on: January 06, 2010, 05:42:24 PM »
ReplyReply

Quote from: Schewe
The other question (which is of course close to my heart) is how the heck you sharpen an image for optimal printed output. That's a different subject that somebody could write a whole book on...oh, wait, we DID :~)

Thanks for the post. I just finished that book, actually.

It seems to me that even if one could print a very large print at 300 dpi, when printing that large the optimal sharpening for a standard viewing distance would necessarily preclude optimal sharpening for nose-to-the-print viewing. As an inevitable tradeoff, we simply can't have it both ways. My inclination is to compromise and sharpen to the point that viewing a 40x60" print at 2 feet it is still (to steal a word) non-crunchy. I know that I definitely move to about two feet when viewing large pieces in galleries or other peoples' homes, so I've found this to be a fairly reasonable compromise.
Logged

Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #7 on: January 06, 2010, 06:08:22 PM »
ReplyReply

Quote from: Neuffy
Thanks for the post. I just finished that book, actually.

It seems to me that even if one could print a very large print at 300 dpi, when printing that large the optimal sharpening for a standard viewing distance would necessarily preclude optimal sharpening for nose-to-the-print viewing.


Well, the viewing distance doesn't actually have anything to do with the optimal output sharpening...the only two critical factors are pixel density (PPI) and media. Viewing distance does have an indirect impact in that you choose the output resolution based on your desired output size. So, bigger printed image size would equal lower PPI which would require the sharpening for the lower pixel density.
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3907


« Reply #8 on: January 06, 2010, 06:30:17 PM »
ReplyReply

Quote from: Schewe
Well, the viewing distance doesn't actually have anything to do with the optimal output sharpening...

Hi Jeff,

I beg to differ on that. The human visual system does have a different contrast sensitivity at different spatial frequencies, and those depend on viewing distance.

One could quibble on how important it is, but stating that it has nothing to do with it is just not true.

Try this if you still don't agree ...

Kind regards,
Bart
Logged
Justan
Sr. Member
****
Offline Offline

Posts: 1886


WWW
« Reply #9 on: January 06, 2010, 07:10:56 PM »
ReplyReply

Quote from: Schewe
The other question (which is of course close to my heart) is how the heck you sharpen an image for optimal printed output. That's a different subject that somebody could write a whole book on...oh, wait, we DID :~)

What book?

Logged

Randy Carone
Sr. Member
****
Offline Offline

Posts: 545


« Reply #10 on: January 06, 2010, 07:18:07 PM »
ReplyReply

If I may, Real World Image Sharpening by Bruce Fraser, updated by Jeff Schewe.

And a quick note about viewing distance, Gregory Crewdson produces amazing images with stunning sharpness in a format that runs to 54" x 80" (printed on an Epson 11880). He encourages viewers to study details of the image from a few inches away. It seems to me that if an image, no matter how large, has excellent sharpness at 2", it'll look good from 5 feet away as well.
« Last Edit: January 06, 2010, 07:23:14 PM by Randy Carone » Logged

Randy Carone
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #11 on: January 06, 2010, 09:29:16 PM »
ReplyReply

Quote from: BartvanderWolf
One could quibble on how important it is, but stating that it has nothing to do with it is just not true.

As it relates to the needs of image output sharpening for digital printed output, no, viewing distance is NOT a factor in image sharpening needs...(the links you cite not withstanding)

How much resolution you NEED for what size print is related directly to viewing distance and that final resolution (pixel density) IS the determining factor, along with the media type, for determining proper image sharpening for printing.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #12 on: January 06, 2010, 09:37:08 PM »
ReplyReply

Quote from: Randy Carone
It seems to me that if an image, no matter how large, has excellent sharpness at 2", it'll look good from 5 feet away as well.

Maybe, maybe not. Ever see Seurat's Sunday Afternoon on the island of La Grande Jatte?

The relationship between looking real close and looking at the intended viewing distance are not necessarily related. I've seen wonderful images whose prints from a distance are great that look like crap from up close...just because something may look technically good up close doesn't mean anything definitive when judged from a distance.
Logged
bradleygibson
Sr. Member
****
Offline Offline

Posts: 829


WWW
« Reply #13 on: January 07, 2010, 12:07:55 AM »
ReplyReply

Quote from: Schewe
I've seen wonderful images whose prints from a distance are great that look like crap from up close...

I think this is the opposite of what Randy is saying...  

Quote from: Schewe

I'm curious what you've seen in this image that makes you bring it up?

I don't see how an image that is well-printed and holds up at close-range will somehow "fall apart" at a distance.  If the subject material is too busy or some such, that may impact the aesthetics, but I'm at a loss to see how printing it differently will save the day.  

I feel it's just the case that it's just *impractical* (not techically feasible)  to have enough information to make every image we print, at any size, look good from 2" away.  Assuming infinite data, computing/printing horsepower, and sufficiently low cost, is there any reason you wouldn't print at a very high resolution?
Logged

neil snape
Sr. Member
****
Offline Offline

Posts: 1433


WWW
« Reply #14 on: January 07, 2010, 02:13:27 AM »
ReplyReply

Quote from: Randy Carone
If I may, Real World Image Sharpening by Bruce Fraser, updated by Jeff Schewe.

And a quick note about viewing distance, Gregory Crewdson produces amazing images with stunning sharpness in a format that runs to 54" x 80" (printed on an Epson 11880). He encourages viewers to study details of the image from a few inches away. It seems to me that if an image, no matter how large, has excellent sharpness at 2", it'll look good from 5 feet away as well.



The accuracy of the inkjet drops is not a constant nor is the acutance on the substrate.
I like Jeff's definition further down which is a perfect way of saying pixel peeping on prints has no relation to the visual appearance to your eye when distance is considered.

I suppose what could be a very good addition to printer driver sharpening algorithms is the intended viewing distance changing the unsharp mask other than just up or down ressing. Does that exist already in some of the cited plug ins or scripts like PK?

Qimage has very good sharpening, based on output size but viewing distance nor output size can change the printer driver masking>

Logged
Ernst Dinkla
Sr. Member
****
Offline Offline

Posts: 2911


« Reply #15 on: January 07, 2010, 03:51:34 AM »
ReplyReply

Quote from: neil snape
The accuracy of the inkjet drops is not a constant nor is the acutance on the substrate.
I like Jeff's definition further down which is a perfect way of saying pixel peeping on prints has no relation to the visual appearance to your eye when distance is considered.

I suppose what could be a very good addition to printer driver sharpening algorithms is the intended viewing distance changing the unsharp mask other than just up or down ressing. Does that exist already in some of the cited plug ins or scripts like PK?

Qimage has very good sharpening, based on output size but viewing distance nor output size can change the printer driver masking>

Neil,

Qimage at least has a flexible and transparent tool to adjust the smart sharpening. In practice with print sharpening one usually hits the limitations set by upsampling artefacts first before the other aspects like viewing distance play a role. It takes some courage to enlarge a print that the artefacts show and still keep the smart sharpening active with the excuse that the viewing distance will dissolve all the enhanced artefacts to the eye. I rather tell my customers that there wasn't enough information in the file to get a sharp blow up and do no print sharpening then. It might make the print slightly worse at viewing distance but looks more natural nearby. This has nothing to do with Qimage's algorithms, they are excellent. If there's data enough for the large print then I do not see why one wouldn't use the smart print sharpening within the viewing distance if that doesn't compromise the image quality at viewing distance.

On the article, I didn't do enough close reading to check whether aliasing on downsampling could have influenced the outcome as well. At some point you consider what you read as not very scientific and then the brain isn't involved anymore.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/


Logged
neil snape
Sr. Member
****
Offline Offline

Posts: 1433


WWW
« Reply #16 on: January 07, 2010, 04:04:12 AM »
ReplyReply

Quote from: Ernst Dinkla
Neil,

Qimage at least has a flexible and transparent tool to adjust the smart sharpening. In practice with print sharpening one usually hits the limitations set by upsampling artefacts first before the other aspects like viewing distance play a role. It takes some courage to enlarge a print that the artefacts show and still keep the smart sharpening active with the excuse that the viewing distance will dissolve all the enhanced artefacts to the eye. I rather tell my customers that there wasn't enough information in the file to get a sharp blow up and do no print sharpening then. It might make the print slightly worse at viewing distance but looks more natural nearby. This has nothing to do with Qimage's algorithms, they are excellent. If there's data enough for the large print then I do not see why one wouldn't use the smart print sharpening within the viewing distance if that doesn't compromise the image quality at viewing distance.

On the article, I didn't do enough close reading to check whether aliasing on downsampling could have influenced the outcome as well. At some point you consider what you read as not very scientific and then the brain isn't involved anymore.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
Yes Ernst I know, I use it often. Yet this is adjusting the sharp level before the driver. Driver level masking is the solution that as far as I can see on inkjet printers isn't addressed. Cymbolic Science was doing something of this nature with Lightjets but masking technologies were at the time and probably are still way out of my knowledge.

Sharp handed to the printer could have toggles for variables, as I think they already with the recent Epson drivers. Yet they are not basing sharpening masks on relative viewing distances, rather assumptions at a line level.

For some reasoning behind driver screening have a look at this site:

http://www.iro.umontreal.ca/~ostrom/


A lot of the HP driver sharpening mask are coming from this>
Logged
Ernst Dinkla
Sr. Member
****
Offline Offline

Posts: 2911


« Reply #17 on: January 07, 2010, 05:25:56 AM »
ReplyReply

Quote from: neil snape
Driver level masking is the solution that as far as I can see on inkjet printers isn't addressed.

A lot of the HP driver sharpening mask are coming from this>

Alright, a misunderstanding.  But is there still interpretation of the image possible at that stage if more images of different sizes are nested on one print page like Qimage allows? And Qimage isn't the only application that does that.

A familiar site. 90% that I do not understand but always fascinating. Referred to it when I asked whether Qimage could get an extrapolation routine for vector origin pixel images.         I see that what I thought was not discovered yet for digital imaging "Cairo Tiles" is now also dicussed there. Ordered a custom made carpet with that pattern some years ago when I got fascinated with the possibilities of that pattern. I thought it had potential for CCD-CMOS sensor mosaïcs but became interested in the first place when Stork's galvano fabricated silkscreen mesh with the honeycomb structure showed the extra moiré of 3 axes symmetry. Something Cairo Tiles pattern doesn't have.  Deviating here .... I will read it.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/


Logged
NikoJorj
Sr. Member
****
Offline Offline

Posts: 1063


WWW
« Reply #18 on: January 07, 2010, 05:56:08 AM »
ReplyReply

Quote from: Ernst Dinkla
On the article, I didn't do enough close reading to check whether aliasing on downsampling could have influenced the outcome as well. At some point you consider what you read as not very scientific and then the brain isn't involved anymore.
As far as I understand it, the question is more adressed in the comments than in the article ; basically, Ctein just says that in mixed photographic subjects (that's what his experiment is about using a 'contact print'), aliasing is far from significant, which sound rather reasonable to me (unless of course artefacts are amplified by strong sharpening? I do also regret that output sharpening is not adressed at all in the article).
If you only shoot architectural subjets (brick walls and tiled roofs), your mileage may implode.
« Last Edit: January 07, 2010, 05:56:50 AM by NikoJorj » Logged

Nicolas from Grenoble
A small gallery
Ryan Grayley
Full Member
***
Offline Offline

Posts: 217



WWW
« Reply #19 on: January 07, 2010, 05:57:51 AM »
ReplyReply

Co-incidentally I have been running some A4 print tests at 360 dpi on Canson Photographique 310 gsm to compare speed and print quality between my Epson 9600, Epson 7900 and HP Z3200. Using a loupe the best dithering and dot pattern to my eyes seems to be the Epson 7900 at 2880 dpi (but only just) while the worst dithering and dot pattern to my eyes is the Epson 9600 at 720 dpi. However, without the use of a loupe I can't really tell the difference when comparing dithering and dot patterns.

Edit1: Specified the three printers I compared.
« Last Edit: January 07, 2010, 10:11:00 AM by Ionaca » Logged

Ryan Grayley BA IEng MIET ARPS
RGB Arts Ltd, London, UK
Pages: [1] 2 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad