A Photography forum. PhotoBanter.com

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

Go Back   Home » PhotoBanter.com forum » Digital Photography » Digital Photography
Site Map Home Register Authors List Search Today's Posts Mark Forums Read Web Partners

Why pixel peeping (usually) makes no sense



 
 
Thread Tools Display Modes
  #1  
Old November 16th 07, 03:47 PM posted to rec.photo.digital
John Navas[_2_]
external usenet poster
 
Posts: 3,956
Default Why pixel peeping (usually) makes no sense

There's a great difference between screen resolution (72-96 PPI) and
good printing, where rules of thumb for normal viewing distances are at
least 130 DPI for barely acceptable results, and up to 230 DPI for
excellent results. With current technology, anything more than 300 DPI
is pretty much wasted. This translates to:

4x6 5x7 8x10 11x14 16x20 20x30
------ ------ ------ ------ ------ ------
Acceptable: 0.4MP 0.6MP 1.4MP 2.6MP 5 MP 10 MP
Very good: 0.8MP 1.1MP 2.6MP 5 MP 10 MP 19 MP
Excellent: 1.3MP 1.9MP 4 MP 8 MP 16 MP 32 MP
Best: 2.2MP 3.2MP 7 MP 14 MP 28 MP 54 MP

For my 7.2MP Panasonic FZ8, 11x14 is "excellent" at about 216 DPI.
The same effect on my high-quality 96 PPI display is 96 / 216 = 44%
zoom, more realistically 50% zoom. On a standard 72 PPI display it's
72 / 216 = 33%.

Even these reduced zooms exaggerate the issue, since display pixels are
so much bigger and more distinct than high-resolution print dots. For
this reason display zoom needs to be reduced even further for meaningful
print judging, to about 25-33% (depending on display) for a high-quality
(8MP) 11x14 print.

Because screen pixels are so much larger and coarser than good printing
dots (effectively a magnifier), zooming in more than this doesn't make
sense (unless the ultimate objective is screen display of 100% crop
rather than normal screen display or printing). If you can't see an
issue at this reduced display zoom, then you're not going to see it in
an excellent print either.

Thus my normal practice is to assess images on screen at no more than
33% zoom, zooming in farther only to examine the effectiveness of
sharpening and/or noise reduction.

This relationship does of course change for larger print sizes, but then
print degradation due to pixelation becomes an offsetting issue -- more
pixels are needed to make larger high-quality prints. When I anticipate
printing larger than 11x14 with my 7.2MP Panasonic FZ8, I shoot multiple
overlapping images and then stitch them together, multiple-frame
super-resolution. 4 image stitching is sufficient for excellent 16x20
and very good 20x30 prints, and even larger sizes can be produced by
stitching more images. In which case the same reduced zoom is
appropriate.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
  #2  
Old November 16th 07, 05:07 PM posted to rec.photo.digital
David Arnstein
external usenet poster
 
Posts: 5
Default Why pixel peeping (usually) makes no sense

In article ,
John Navas wrote:

This is a nice paper, and you put a lot of work into it. You did not
have much to say about cropping, however.

One of the attractions of digital photography is the ease of post-
processing (compared to using a darkroom!). When I take pictures, my
first priority is to get the shutter button pushed quickly. I can take
care of composition in Photoshop or Thumbs Plus. I admit, I have no idea
how other folks frame their shots. But in my case, higher pixel counts
are useful.

By the way, what is this "print" that you speak of? Is it related to
darkroom techniques and stiff paper? I don't mess with that stuff any
more. Do you?
--
David Arnstein (00)
{{ }}
^^
  #3  
Old November 16th 07, 07:13 PM posted to rec.photo.digital
Ali[_3_]
external usenet poster
 
Posts: 216
Default Why pixel peeping (usually) makes no sense

Two points you didn't take into consideration is what about cropping images
post process and also it's very difficult stitch photos with moving
subjects. In addition to this, stitching adds to post process time and if
you want a certain perspective in the photo (very wide angle/fisheye),
stitching won't work.

I don't like cropping images myself and do my cropping in the view finder,
however there are times when it is not possible to get the crop you want in
camera (for example if you want a square image, you can't physically get
closer to the subject, you make a mistake with composition, etc), so more
pixels are always better as long as the pixel pitch isn't too small, as it
can increase noise.

Of course, I know where you're coming from and I'm happy with my 20D, which
is 8MP.

PS: From a commercial point of view, certain clients require photo's to be
a minimum of XYZ MB TIFF.


"John Navas" wrote in message
...
There's a great difference between screen resolution (72-96 PPI) and
good printing, where rules of thumb for normal viewing distances are at
least 130 DPI for barely acceptable results, and up to 230 DPI for
excellent results. With current technology, anything more than 300 DPI
is pretty much wasted. This translates to:

4x6 5x7 8x10 11x14 16x20 20x30
------ ------ ------ ------ ------ ------
Acceptable: 0.4MP 0.6MP 1.4MP 2.6MP 5 MP 10 MP
Very good: 0.8MP 1.1MP 2.6MP 5 MP 10 MP 19 MP
Excellent: 1.3MP 1.9MP 4 MP 8 MP 16 MP 32 MP
Best: 2.2MP 3.2MP 7 MP 14 MP 28 MP 54 MP

For my 7.2MP Panasonic FZ8, 11x14 is "excellent" at about 216 DPI.
The same effect on my high-quality 96 PPI display is 96 / 216 = 44%
zoom, more realistically 50% zoom. On a standard 72 PPI display it's
72 / 216 = 33%.

Even these reduced zooms exaggerate the issue, since display pixels are
so much bigger and more distinct than high-resolution print dots. For
this reason display zoom needs to be reduced even further for meaningful
print judging, to about 25-33% (depending on display) for a high-quality
(8MP) 11x14 print.

Because screen pixels are so much larger and coarser than good printing
dots (effectively a magnifier), zooming in more than this doesn't make
sense (unless the ultimate objective is screen display of 100% crop
rather than normal screen display or printing). If you can't see an
issue at this reduced display zoom, then you're not going to see it in
an excellent print either.

Thus my normal practice is to assess images on screen at no more than
33% zoom, zooming in farther only to examine the effectiveness of
sharpening and/or noise reduction.

This relationship does of course change for larger print sizes, but then
print degradation due to pixelation becomes an offsetting issue -- more
pixels are needed to make larger high-quality prints. When I anticipate
printing larger than 11x14 with my 7.2MP Panasonic FZ8, I shoot multiple
overlapping images and then stitch them together, multiple-frame
super-resolution. 4 image stitching is sufficient for excellent 16x20
and very good 20x30 prints, and even larger sizes can be produced by
stitching more images. In which case the same reduced zoom is
appropriate.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)


  #4  
Old November 16th 07, 08:22 PM posted to rec.photo.digital
frederick
external usenet poster
 
Posts: 1,525
Default Why pixel peeping (usually) makes no sense

Scott W wrote:

Shoot, hit the wrong button and send a blank replay, sorry about that.

Once your message appears in Thunderbird, select it, right
click, select cancel message, and if you've been quick
enough you'll get the message deleted from the news server
before it propagates.
OTOH if you're not quick enough, you can end up deleting it
from your news server after it's propagated, and end up with
replies to messages that you thought you didn't send.
  #6  
Old November 16th 07, 08:35 PM posted to rec.photo.digital
John Navas[_2_]
external usenet poster
 
Posts: 3,956
Default Why pixel peeping (usually) makes no sense

On Fri, 16 Nov 2007 08:49:35 -1000, Scott W wrote
in :

John Navas wrote:
[SNIP]


Well I don't see how looking at the screen at %25-33% can tell you how
the print is going to look, unless you are making a very small print.
Say you start with 8MP and view at 25%, you effectively are throwing out
15 out of 16 pixels in your viewing, reducing your 8MP image down to 0.5 MP.


Viewing at 100% on screen would only be comparable to printing if your
screen had the same high resolution as the printer, but it's not even
close, as I tried to explain.

I also would take some issue with your table of needed pixels, this is a
personal thing. For an 8 x 10 print I would say 4MP is acceptable, but
just and very good would be 8MP. And with a good printer you can see a
noticeable improvement going from 8 to 16 MP for a 8 x 10 print, so I
don't see how even 8MP could be call best for this size print.

A lot depends on the printer used, when I use Costco they can't seem to
use detail much past 200 ppi, maybe a bit past but not close to 300ppi.
With my cheap inkjet printer I can see detail to 400ppi.


Assuming your Costco uses the Fuji Frontier on Crystal Archive paper
(or something comparable), and assuming it's properly adjusted, then
print quality at 300 DPI should be comparable to your inkjet printer
at any DPI.

The reason cheap inkjets need more DPI is that they don't do as good
a job of reproducing a given DPI.

(PPI pertains to screen, not printing.)

The other thing that I have noted is over time I have found uses for
higher resolution photos, part of this is making bigger prints and part
of this is doing slide shows that have pan and zooms in them. I look
forward to the day that I can have a 4 x 6 foot wall mounted display
that has 200 ppi, that is going to take close to 140MP, and for it to
really look clean you need to down size a bit, so you might want to
start out at 200MP.


As I wrote, I stitch multiple images together when I need higher
resolution.

The point is you can never have too much resolution and your images can
never look too clean. Maybe not for how you are using today, but what
about 20 years from now?


You can likewise never have too much money and be too skinny.

I'm talking what's real and practical.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
  #7  
Old November 16th 07, 08:41 PM posted to rec.photo.digital
John Navas[_2_]
external usenet poster
 
Posts: 3,956
Default Why pixel peeping (usually) makes no sense

On Fri, 16 Nov 2007 19:13:37 GMT, "Ali" wrote in
:

Two points you didn't take into consideration is what about cropping images
post process and also it's very difficult stitch photos with moving
subjects.


True, although the best current stitching software may be better than
you think at removing ghosts. (I commented on cropping earlier in the
thread.)

In addition to this, stitching adds to post process time and if
you want a certain perspective in the photo (very wide angle/fisheye),
stitching won't work.


Actually it will -- the best stitching software easily produces multiple
different projections.

PS: From a commercial point of view, certain clients require photo's to be
a minimum of XYZ MB TIFF.


True, but I've not (yet) seen any that wouldn't accept what I can get
right out of camera for other than mural prints.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
  #8  
Old November 16th 07, 08:56 PM posted to rec.photo.digital
cmyk
external usenet poster
 
Posts: 115
Default Why pixel peeping (usually) makes no sense

Hi John,

"John Navas" wrote in message ...
There's a great difference between screen resolution (72-96 PPI) and
good printing, where rules of thumb for normal viewing distances are at
least 130 DPI for barely acceptable results, and up to 230 DPI for
excellent results.


I don't know what distance you're viewing the prints from, but 230dpi doesn't cut it for a anything smaller than a 22.5x15 print
viewed at *normal* viewing distances (as established by Kodak's reasearch - see "Handbook of Image Quality", ISBN-10: 0824707702,
ISBN-13: 9780824707705). And you're not going to get that kind of pixel density from a 7.2Mp camera. The best such a camera can
give, without degrading the image by interpolation, is about a 7x10.5 print, for which you need around 288dpi.

With current technology, anything more than 300 DPI
is pretty much wasted. This translates to:

4x6 5x7 8x10 11x14 16x20 20x30
------ ------ ------ ------ ------ ------
Acceptable: 0.4MP 0.6MP 1.4MP 2.6MP 5 MP 10 MP
Very good: 0.8MP 1.1MP 2.6MP 5 MP 10 MP 19 MP
Excellent: 1.3MP 1.9MP 4 MP 8 MP 16 MP 32 MP
Best: 2.2MP 3.2MP 7 MP 14 MP 28 MP 54 MP

Your assumption that print densities would remain the same regardless of print size is wrong unless you're one of those sods who
insists that everything has to be viewed from some arbitrary fixed distance. For a:
.. 4x6 print, you'd want about 360ppi
.. 5x7 print, you'd want 300-360ppi
.. 8x10 or 11x14 print, you'd want about 288ppi
.. 16x20 print, you'd want about 240ppi
.. 20x30 print, you'd want about 200ppi
prefereably uninterpolated, for view at *normal* viewing distances and assuming the CoC for the required visual resolution matches
the human eye's diffraction limits (ie about 90 arc-seconds for a 3mm pupil) under normally-lit indoor conditions. By what standard
do you define 'acceptable', 'very good', excellent', 'best'?

For my 7.2MP Panasonic FZ8, 11x14 is "excellent" at about 216 DPI.
The same effect on my high-quality 96 PPI display is 96 / 216 = 44%
zoom, more realistically 50% zoom. On a standard 72 PPI display it's
72 / 216 = 33%.


Hmm, for viewing at *normal* viewing distances (about 18 inches for am 11x14 print), that implies a CoC with an angular resolution
of about 105-110 arc-seconds. That seems to imply that the 90 arc-seconds I used for the above calculations is better than that -
your 'best', perhaps.

Even these reduced zooms exaggerate the issue, since display pixels are
so much bigger and more distinct than high-resolution print dots. For
this reason display zoom needs to be reduced even further for meaningful
print judging, to about 25-33% (depending on display) for a high-quality
(8MP) 11x14 print.

Because screen pixels are so much larger and coarser than good printing
dots (effectively a magnifier), zooming in more than this doesn't make
sense (unless the ultimate objective is screen display of 100% crop
rather than normal screen display or printing). If you can't see an
issue at this reduced display zoom, then you're not going to see it in
an excellent print either.

Thus my normal practice is to assess images on screen at no more than
33% zoom, zooming in farther only to examine the effectiveness of
sharpening and/or noise reduction.

This relationship does of course change for larger print sizes, but then
print degradation due to pixelation becomes an offsetting issue -- more
pixels are needed to make larger high-quality prints. When I anticipate
printing larger than 11x14 with my 7.2MP Panasonic FZ8, I shoot multiple
overlapping images and then stitch them together, multiple-frame
super-resolution. 4 image stitching is sufficient for excellent 16x20
and very good 20x30 prints, and even larger sizes can be produced by
stitching more images. In which case the same reduced zoom is
appropriate.


'Pixel peeping', as you put it, is useful if you want to be able to edit the image to maximise its qualities. Even with your FZ8, a
1024x768 screen is going to downsample by a factor of 10, which really limits your ability to work with fine detail unless you zoom
in. Of course, if you're limited to a camera that only outputs jpeg files, you don't have much as scope there as you get a decent
DSLR anyway (even one with less pixels).

Cheers
--
cmyk

  #9  
Old November 16th 07, 09:13 PM posted to rec.photo.digital
cmyk
external usenet poster
 
Posts: 115
Default Why pixel peeping (usually) makes no sense

For my 7.2MP Panasonic FZ8, 11x14 is "excellent" at about 216 DPI.

To produce an 11x14 print without interpolation from your FZ8, you can only print at 200ppi. Printing at 216dpi is only going to
degrade the image through interpolation. You can see what's optimum by dividing the print dimensions by the available number of
pixels. That gives an angular resolution of around 115 arc-seconds.

Cheers
--
cmyk

  #10  
Old November 16th 07, 09:23 PM posted to rec.photo.digital
Ali[_3_]
external usenet poster
 
Posts: 216
Default Why pixel peeping (usually) makes no sense

Mmm. I'm not talking about ghosts. I am talking about shooting moving
subjects, whether it maybe a F1 car or people.

I also can't see how stitching software is able to create the same effect as
a fisheye lens with multiple images, without a lot of time spent in post
production.


"John Navas" wrote in message
...

Two points you didn't take into consideration is what about cropping
images
post process and also it's very difficult stitch photos with moving
subjects.


True, although the best current stitching software may be better than
you think at removing ghosts. (I commented on cropping earlier in the
thread.)



In addition to this, stitching adds to post process time and if

you want a certain perspective in the photo (very wide angle/fisheye),
stitching won't work.


Actually it will -- the best stitching software easily produces multiple
different projections.


 




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
For some people film makes more sense Graham Fountain 35mm Photo Equipment 60 December 26th 06 10:02 AM
Olympus E-400 grip makes way more sense RichA Digital SLR Cameras 0 December 3rd 06 01:01 AM
Wide Angle IS makes sense... MarkČ Digital Photography 84 September 9th 06 01:42 PM
Wide angle IS makes sense... MarkČ Digital SLR Cameras 9 September 3rd 06 01:26 PM
Hopefully this makes sense HeritageMom Digital Photography 20 February 9th 06 07:08 PM


All times are GMT +1. The time now is 11:36 PM.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 PhotoBanter.com.
The comments are property of their posters.