>From Creatve Pro:
>http://www.creativepro.com/story/howto/25527.html?cprose=daily
As inaccurate as it is poorly written (and a sad commentary on
iStockphoto.com):
The Truth About Resolution
Whenever an article proclaims "truth", baloney is sure to follow!
When I became a pixel pusher years ago,
2? 3?
one of the toughest concepts
to wrap my brain around was that of image resolution:
And now she's an expert. :)
I know better now, but many professional designers don't. They buy a
high-quality photo from a stock agency, open it in Photoshop, and
black out with rage when they see "resolution = 72" in the Image Size
dialog. An irate call to customer service usually ensues, followed by
an embarrassing explanation of Resolution 101 by someone half their
age. ...
Now we know how old she is.
The Dark Art of Upsampling.
Be afraid, be very afraid! :)
The resolution measurement dictates how closely the pixels are packed
together. Increasing an image's resolution means the pixels will be
packed together more tightly, resulting in a smaller physical size,
Other way around, but who's keeping track.
but generating a smoother, higher quality print. Lowering an image's
resolution means loosening the pixels, resulting in a larger physical
image size, but generating a blocky, lower quality print.
Sounds like slutty pixels to me! :)
In fact any decent current printer driver does a good job of upsampling
before printing.
Think of the resolution measurement as density. For example, the
tighter a substance is packed, the denser it is and the less surface
area it takes up (like brown sugar). The more loosely a substance is
packed, the more surface area it consumes and it becomes less dense.
Are we talking crack or digital images? :)
Because our eyes can only process so much information, a 72 ppi image
onscreen looks identical to a 600 ppi image onscreen. ...
Only if you have a crappy monitor or are legally blind.
This gorgeous photo (courtesy of iStockphoto.com/Lisa Gagne) is
measured at 72 ppi. Does that mean it's a low-quality image?
Negatory, good buddy.
She's my buddy? Does that mean what I think it does? :)
Evil upsampling method #1: Open the Image Size dialog in Photoshop,
take a deep breath, and leave the Resample Image box checked. (I
realize I just told you not to do this, but bear with me.) Choose
Bicubic Smoother from the pop-up menu to its right and change the
document dimension pop-up menus to Percent. Enter 110% in the width
box and press OK (Figure 3). Repeat this process as many times as
necessary to enlarge the image. For some reason, adding data 10% at a
time doesn't cause a huge amount of quality loss. Do resist the urge
to increase the size more than 10% at a time, unless you want that
chunky look.
Stairstep interpolation is actually a bad idea. Bicubic Smoother works
best in a single step.
Evil upsampling method #2: Buy a plug-in, such as Genuine Fractals by
onOne Software, Blow Up by Alien Skin, or PhotoZoom Pro 2 by
BenVista. They pull off some serious voodoo and somehow manage to
increase pixel data without totally destroying the image.
As my own tests show ("Adobe Bicubic Smoother upsize vs Genuine
Fractals"), there is very little difference between Bicubic Smoother and
Genuine Fractals in large upsizing, and none at all in modest upsizing.
I Wouldn't Lie to You
When someone says "I Wouldn't Lie to You" you can be sure they're about
to lie to you. Think used car salesperson. :)
Resolution doesn't mean squat until that image is headed for a
printer. ...
And not even then, since the printer driver will upsample for you.
Lesa Snider King is the chief evangelist of iStockphoto.com and
founder of GraphicReporter.com.
Yet she doesn't know how printer drivers work, thinks stairstep
interpolation works, and doesn't know how well bicubic smoother works.
Yikes!

Signature
Best regards,
John Navas <http:/navasgroup.com>
"Usenet is like a herd of performing elephants with diarrhea - massive,
difficult to redirect, awe inspiring, entertaining, and a source of mind
boggling amounts of excrement when you least expect it." --Gene Spafford
> From Creatve Pro:http://www.creativepro.com/story/howto/25527.html?cprose=daily
I didn't think much of the article. When I first found rpd many years
ago I got on a soapbox about resolution and didn't change very many
minds.
Problem is, there are MANY definitions of resolution. There is no
final arbiter of English language.
However, to my mind the best definition of resolution is what film
photography used to use- an actual MEASURED value based on test
charts.
To me the number of pixels in the image is just that- a number!
Theoretically there is a ratio of the resolution in each direction to
the number of pixels in that direction. In a PERFECT camera with a
perfect lens, the actual measured resolution would approach that
number (about 0.7 times the number of pixels). But of course, no
camera and no lens is perfect.
Marvin - 28 Nov 2007 17:03 GMT
>> From Creatve Pro:http://www.creativepro.com/story/howto/25527.html?cprose=daily
>
[quoted text clipped - 15 lines]
> number (about 0.7 times the number of pixels). But of course, no
> camera and no lens is perfect.
I agree, but there is confusion. See
http://www.dpreview.com/learn/?/Glossary/Digital_Imaging/Resolution_01.htm/
One part says resolution is "defined" by the number of
pixels in the image. The next part brings up a chart used
in film photography to determine the amount of detail that
the camera can capture. The pixel count does control the
maximum capability of the sensor to capture detail, but the
capability of the camera also depends on the lens. Camera
makers and sellers emphasize the pixel count because that is
something most buyers think they understand, though it is
often just in the sense that "more is better". The game is
called "specsmanship".
Don Stauffer in Minnesota - 29 Nov 2007 15:07 GMT
> >> From Creatve Pro:http://www.creativepro.com/story/howto/25527.html?cprose=daily
>
[quoted text clipped - 27 lines]
> often just in the sense that "more is better". The game is
> called "specsmanship".
I agree. But the problem is that ANYONE can create a definition. The
part that reads "resolution is defined by" is really meaningless.
That can merely be HIS definition. Personally I find the science of
film imaging to be far more consistant than that of digital, because
there because the professional societies dealing with film photography
for many decades did a good job, and the photo industry in those days
was not as likely to attempt that specsmanship game, like they do now.
Also, in the early days photography was dominated by really serious
people. It wasn't until well after Eastman and his Kodaks that it
became a consumer product.
On the other hand, the drafts of the ISO spec on digital camera
resolution was very good. Unfortunately, getting a current version
requires spending bucks. I retired before the draft was finalized, and
don't want to spend my own money just to see the spec. But as I say,
the draft spec anyway was pretty good, but as far as I know, no
advertiser uses ISO resolution in ads.
Marvin - 29 Nov 2007 17:13 GMT
<snip>
> I agree. But the problem is that ANYONE can create a definition. The
> part that reads "resolution is defined by" is really meaningless.
[quoted text clipped - 7 lines]
> people. It wasn't until well after Eastman and his Kodaks that it
> became a consumer product.
Kodak dominated film commercially, and they set up a
research lab early on. That lab dominated the technical
aspects of photo films, including terminology. It is
similar to the way that Microsoft has dominated conventions
in computer programming. Everyone else's document creation
software has to be able to open Word files. All camera
makers had to make products that accepted Kodak's films.
> On the other hand, the drafts of the ISO spec on digital camera
> resolution was very good. Unfortunately, getting a current version
> requires spending bucks. I retired before the draft was finalized, and
> don't want to spend my own money just to see the spec. But as I say,
> the draft spec anyway was pretty good, but as far as I know, no
> advertiser uses ISO resolution in ads.
Few people seem to have any idea what ISO ratings mean in
photography, or - for that matter - what ISO is. Sex sells
cars; pixel counts sell digicams.
Don Stauffer in Minnesota - 30 Nov 2007 15:51 GMT
> Few people seem to have any idea what ISO ratings mean in
> photography, or - for that matter - what ISO is. Sex sells
> cars; pixel counts sell digicams.
I was referring to the resolution spec, not the speed spec.
I can remember early high fi days, people did sort of fall into line
on defining bandpass, even in absence of a trade or professional group
spec.
On the other hand, I don't really feel TOO upset about the modern
trend, since pixel count is certainly one of the main drivers of
measured resolution.