Enlarging images for printing

[apaflo]

Originally Posted by sbingham

You know you might very well be right. Granted the native resolution of Epson printers is either 360 0r 720 depending on the printer, however I have found through emperical print testing very little difference when using an even fraction such as 240. Eric Chan, who now works for Adobe as the number 1 guy behind Thomas Knoll in ACR, has pretty much stated the same thing. He is regarded by many as THE Epson printer expert. I also realize that Epson must resample to its native resolution before printing. I open all my Nikon raw files using ACR set at 240, whatever my camera is at the time.Using USM I usually view at either print size, 50%, or 100% (rarely). My final sharpening is always done for print size so I am not sure your theory would apply in my case.

"Print size" would be after the image is resampled to 360 PPI, rather than 240 PPI. And it only makes sense to do that while viewing a 100% crop, otherwise at least some resolution is lost when the image is resampled by the display system. Down sampling is literally a brick wall low pass filter, and a 50% crop removes virtually all of the high frequency spatial components affected by sharpening.

And whatever the PPI is when a file is opened for viewing on a computer is meaningless, as the value is totally ignored by the computer until the image is resampled for writing to a file.

[alanbutler57]

Check out Qimage. It's excellent for such things. If you choose to "print to file" be sure to disable printer color management though.

[ChicagoJohn]

Originally Posted by sbingham

After extensive testing over many years I have found 3rd party software never has equaled Photoshop resampling. There is also absolutely NO need to do this 110% at a time. In fact, this sometimes can be determental. This is pretty much an urban legent dating back to very early versions of Photoshop. Legends die hard!

Although I am now retired I taught photography and Photoshop for 33 years (high school and then college). Now I just shoot for FUN. No more commercial stuff.

You taught photography and Photoshop for 33 years, OK. I hadn't realized Photoshop was around 33 years ago. You know, I have actually been a scientist and technologist for 35 years. We didn't even have calculators when I started. We used slide rules. So did they have Photoshop on slide rules??

Here's the deal: in my corner of reality we have a slogan, "In God we trust; all others must bring data."

So I posted some data. Let's see your data showing that the iterative 110% Bicubic Smoother is not better than an single step method or than anything else.

That should be a slam dunk for anyone who has been teaching Photoshop for the past 33 years.

[apaflo]

Originally Posted by ChicagoJohn

Here's the deal: in my corner of reality we have a slogan, "In God we trust; all others must bring data."

So I posted some data. Let's see your data showing that the iterative 110% Bicubic Smoother is not better than an single step method or than anything else.

That should be a slam dunk for anyone who has been teaching Photoshop for the past 33 years.

Actually it's virtually impossible to adequately demonstrate that to be true for an upsized image if we are going to view the results on computer monitors that necessarily have to down size the image for viewing. But what we can easily do is show it is true for down sizing, and then suggest that while the results are not identical, the problem is just as significant for upsizing as it is for downsizing.

So lets first say that the idea that Adobe's Photoshop is as good as it gets is simply hilarious. But it is better than doing resampling in increments!

Want to see some results, here's one comparing ImageMagick using Lanczos to PhotoShop using Bicubic:
http://i605.photobucket.com/albums/t...75pct_sRGB.gif

A more complicated set of various comparisons can be found here:
Image Resizing for the Web and Email

These last two I've given previously, but lets show them again. First is a very good article on how to sharpen using ImageMagick:
Sharpening using Image Magick Red Skies at Night

Here is a tutorial from ImageMagick about resizing:
Resize or Scaling -- IM v6 Examples

[ChicagoJohn]

Thanks for the leads!

[blumesan]

Originally Posted by apaflo

Actually it's virtually impossible to adequately demonstrate that to be true for an upsized image if we are going to view the results on computer monitors that necessarily have to down size the image for viewing. But what we can easily do is show it is true for down sizing, and then suggest that while the results are not identical, the problem is just as significant for upsizing as it is for downsizing.

Two questions, just for my edification:

Could one not make some sort of comparison on a computer monitor if the images were not downsized for viewing (just display a crop at the higher resolution)?

What suggests that downsizing presents as significant (though not identical) a problem? How would you show this to be the case?

Thanks.

[apaflo]

Originally Posted by blumesan

Two questions, just for my edification:

Could one not make some sort of comparison on a computer monitor if the images were not downsized for viewing (just display a crop at the higher resolution)?

Technically viewing at 100% is fairly similar to viewing that much of a print (the effects of different paper surfaces is lost though), but...

Just do a web search on "pixel peeping" and see if you really want to get into that discussion! There's just too much baggage that comes with the territory to allow a reasonable chance of success.

Or, that's the way I look at it!

Another equally negative perspective is the example we saw here, where ChicagoJohn in all sincerity posted an example and said he could see a perceptible difference, which he said was due to aliasing. I have no idea what he's referring to, as I can't see any difference at all and have no idea what he means by aliasing in that context.

What suggests that downsizing presents as significant (though not identical) a problem? How would you show this to be the case?

Well the problems associated with down sizing have to do with artifacts created by what is called "ringing" when spatial detail is altered by a low pass filter. Something that should have a single sharp edged transition instead has an oscillation at the upper and lower edges of the transition. If just the right amount is allowed the effect is very much like an application of Unsharp Mask. That is the reason that Adobe uses Bicubic Sharper for down sizing.

The problems associated with up sizing are exactly the opposite though, because while the end result is more pixel resolution there isn't really any additional higher resolution detail to add. The existing low resolution detail is duplicated, hopefully in a way that doesn't add anything visibly different. Adobe uses Bicubic Smoother, which is configured to have less ringing than Bicubic Sharper by blurring it slightly. ImageMagick uses Mitchell, which also slightly blurs the detail too.

Because Unsharp Mask works on single edge transitions it has more effect on a down sized image than it does on an upsized image. And because a high pass sharpening algorithm works on multiple transitions it has more effect on upsized images than it does on downsized images.

So upsizing and downsizing each present significant, but different, challenges.

[sbingham]

Chicago John, as a supposed "scientist" you seem to have a reading problem. Please note the "and".

Actually, what I said was that I taught photography "and" Photoshop for 33 years - not that that I taught Photoshop for 33 years. Get the difference? I did, however, first use Photoshop version 1 on a Mac some 20 years ago. Been using it ever sense, except I switched to PC with version 2.5. I was also published in Popular Photography in November of 1994 ( a long time ago), along with Eric Meola, in a spread called "Digital Photography Comes of Age". I also have an MS from Brooks Institute of Photography. I have paid my dues. And you?? Perhaps google could help if I had your real name. Mine is Steve Bingham - one old crusty fart.

[shadowman]

Originally Posted by ChicagoJohn

This morning I downloaded the 30 day trial professional version of popular, patented plug-in enlargement software said to be based upon fractals and performed an experiment on the same file used for the comparison previously reported.

Comparing the results from this plug-in (which took 10 minutes to run on my computer) with the results from the 110% iterative method (which took under 3 minutes) at 400%, just at the borrderline of pix-elation, examining several areas of the image, I can see no difference in the two results except that in a few areas the plug-in shows very slightly more pixelation. However, with respect to resolution of detail and degree of artifacts, they are virtually identical.

Of course, you can do batch processing with the plug-in, but I may need to do this maybe once a month, if that, and it should be easy to automate the 110% iterative method as an Action in photoshop.

I read somewhere, and for the life of me can't seem to find the information anymore, that when enlarging a photo you should do it in 10% increments to lessen the pixelation and/or jagged edges. Obviously, this would be very time consuming and i haven't seen to much deterioration just be doubling my resolution. However, the imaes I practiced on had simple backgrounds and basic color schemes.

[Onslow]

Post deleted. New thread started.