Manual Cropping Workflow?

P
Posted By
patrick
Jul 12, 2005
Views
837
Replies
37
Status
Closed
I still prefer manual, step-by-step processes to automated ones. I’ve been retired for some 37 years now and have come to appreciate the process as much as the result in just about all my activities. That said, have I impaired the result with this method of cropping? (I seldom print above 8×10 and usually 5×7 is an enlargement for me.)

Cropping is my first step in the workflow, following initial processing in RAW.
Using the Marquee tool in Fixed Aspect, I select the desired area, regardless of size.
I crop to that selection.
Then I resample and resize to the actual final dimensions of the edit. (Bicubic: sharper for down; smoother for up)

Thanks for any comments. (Snide comments do not contribute substance, however!)
.. . . . patrick

How to Master Sharpening in Photoshop

Give your photos a professional finish with sharpening in Photoshop. Learn to enhance details, create contrast, and prepare your images for print, web, and social media.

BH
Bill Hilton
Jul 12, 2005
patrick asks ..

Using the Marquee tool in Fixed Aspect, I select the desired area, regardless of size. I crop to that selection.
Then I resample and resize to the actual final dimensions
have I impaired the result with this method of cropping?

No, not at all … this is the ‘normal’ way of doing it, with the caveat that you can’t upsample *too* much without losing image quality, which you probably already know.
P
patrick
Jul 12, 2005
"Bill Hilton" wrote in message
patrick asks ..

Using the Marquee tool in Fixed Aspect, I select the desired area, regardless of size. I crop to that selection.
Then I resample and resize to the actual final dimensions
have I impaired the result with this method of cropping?

No, not at all … this is the ‘normal’ way of doing it, with the caveat that you can’t upsample *too* much without losing image quality, which you probably already know.
Thanks for the reassurance, Bill!

Yes, I appreciate the caveat. I have Fractals but have not had occasion to use it.
I have read, and suspect it’s true, that as you enlarge to the limit, you can tolerate fewer dpi as the assumed viewing distance increases. On the Epson R300 printers I stay at 240dpi for anything up to 8×10. (720dpi being Epson’s native mode)

Surprisingly, when I had some 36 play plaques, shot at 6M and combined into a 48×52" composite, the graphic artist had me edit the final compo at 150 dpi for an assumed viewing distance of six feet. The results (to me) looked fantastic. So I’m willing to keep an open mind on that score.

Normally, my efforts simply seek to produce quality family album material. By "quality" I mean, that I’m trying to get beyond the recorded snapshot so that the result reflects how I remembered the occasion (or want to remember it) and that it suggests (and hopefully communicates) how I feel about it.

Always nice to hear from the "big boys," Bill. Thanks again! .. . . . patrick
O
Owamanga
Jul 12, 2005
On Tue, 12 Jul 2005 12:05:35 GMT, "patrick" wrote:

Then I resample and resize to the actual final dimensions of the edit. (Bicubic: sharper for down; smoother for up)

This bit sounds dangerous. Here you are damaging the data and *why* should be investigated.

Unless you are resizing for an email or inclusion into a webpage at a particular pixel dimension, I don’t see why this is necessary. Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it, but I wouldn’t be as convinced when it comes to a requirement to ever downscale the image.

The resample should certainly be avoided if you are saving certain crops to be possibly printed again in the future.

You also omitted the most important stage, just prior to printing: Intelligent use of unsharp-mask.


Owamanga!
http://www.pbase.com/owamanga
P
patrick
Jul 12, 2005
"Owamanga" wrote in message
On Tue, 12 Jul 2005 12:05:35 GMT, "patrick" wrote:
Then I resample and resize to the actual final dimensions of the edit. (Bicubic: sharper for down; smoother for up)

This bit sounds dangerous. Here you are damaging the data and *why* should be investigated.

Unless you are resizing for an email or inclusion into a webpage at a particular pixel dimension, I don’t see why this is necessary. Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it, but I wouldn’t be as convinced when it comes to a requirement to ever downscale the image.

The resample should certainly be avoided if you are saving certain crops to be possibly printed again in the future.

You also omitted the most important stage, just prior to printing: Intelligent use of unsharp-mask.


Owamanga!
http://www.pbase.com/owamanga

I don’t understand this, Owamangal!
My Nikon D70 yields a 6M image, 3000×2000 ppi. This loads from the RAW file at an arbitrary 300dpi which would yield a 10"X6.7" print. At 240 dpi it would print a 12.5" x 8.3" image..
I don’t see how I can get to any other print size without resizing and reampling as described above.
(Yes, I might luck out if the change from 300dpi to 240dpi resulted in exactly the print size I want. But that has never happened for me.)

I only resize and resample once for any print size. I archive all images in RAW and edited .psd formats as well as a flattened .tiff image for particular sized prints.. If I later want to print at a different size, I load both the RAW image and the edited .psd images. Then I resize and resample the new image to the desired print size (as described above) and drag the editing layers from the .psd image. Any print has seen but a single resize and resample from the RAW image.

Except for the copping theme, I omitted ALL editing steps in this thread, including the final unsharpen filtering.
.. . . . patrick
O
Owamanga
Jul 12, 2005
On Tue, 12 Jul 2005 19:53:41 GMT, "patrick"
wrote:

I don’t understand this, Owamangal!
My Nikon D70 yields a 6M image, 3000×2000 ppi. This loads from the RAW file at an arbitrary 300dpi which would yield a 10"X6.7" print. At 240 dpi it would print a 12.5" x 8.3" image..

Aggg! This is that damn dpi/ppi crap that confuses everybody again.

Here’s the key. 300dpi, is, as you state *arbitrary*. It means absolutely nothing. It’s just the number ‘300’ stored in the header of the image somewhere. Changing it to 900 for example can be achieved without *any* resampling of the image data (try it in photoshop, uncheck the ‘resample image’ box and change the dpi. This number is totally arbitrary and relatively meaningless in the context of what we need to do with the image.

I don’t see how I can get to any other print size without resizing and reampling as described above.

Just crop it and print. When you print, tell photoshop how big you want to print it. Photoshop will send *all* the data to the printer driver and it’ll internally resample (if needed), dumping anything it can’t use.

I mean, do you know for sure how many ppi your printer driver uses internally? Well, if it’s a Canon ink-jet, it’s 600ppi. If it’s an Epson ink-jet then it’s 720ppi. These numbers are fixed and don’t change with a change in paper quality or print resolution (dpi).

For you to resample to an arbitrary 300ppi or 240ppi ready for printing is pointless. If you are going to resample, at least make it match the printer driver’s internal resolution, so it doesn’t go and resample it again. Personally, I wouldn’t bother for a down-size.

I only resize and resample once for any print size. I archive all images in RAW and edited .psd formats as well as a flattened .tiff image for particular sized prints.. If I later want to print at a different size, I load both the RAW image and the edited .psd images. Then I resize and resample the new image to the desired print size (as described above) and drag the editing layers from the .psd image. Any print has seen but a single resize and resample from the RAW image.

Okay, this workflow is safe (only because you keep the RAW), except for the unnecessary resample prior to printing.

It is relatively pointless saving a resampled image that was made for a target printer at a target dimension because tomorrow’s printer drivers may well work at 900ppi, 1244ppi or any other number they invent. Technology is always improving and these numbers *will* go up.

Except for the copping theme, I omitted ALL editing steps in this thread, including the final unsharpen filtering.

Yep, okay, at least that bit makes sense now.


Owamanga!
http://www.pbase.com/owamanga
BH
Bill Hilton
Jul 12, 2005
Owamanga! writes …

Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it

Here’s a link to an article that discusses this for LightJet 5000, Chromira and Epson 9600 printers that just crossed my virtual desk today. The author runs one of the best custom labs in the western USA, favored by the likes of Robert Glenn Ketchum and Jack Dykinga … basically he says that while the native rez of the Chromira and LJ is 300 ppi (actually 304.8 or 12 lines/mm for the older LJ) these printers have exceptionally good built-in interpolation and 200 ppi looks great (something I’ve heard from two other Chromira or LightJet labs, Bill Nordstrom’s Laser Light and Calypso in Santa Clara).

On the other hand he says the Epson 9600, which has a native rez of 360 ppi, prints noticeably better if you resample to that size and sharpen correctly. I have an Epson 4000 and always try to hit 360 ppi, based on tests I’ve printed. Other printers will of course differ in their ability to resample well or poorly, so the best thing to do is test it yourself. Personally I think a lot depends on how well you can sharpen in Photoshop.

http://www.westcoastimaging.com/wci/page/info/photoshoptip/t ip25.html

Patrick wrote …

On the Epson R300 printers I stay at 240dpi for anything up to 8×10. (720dpi being Epson’s native mode)

I have two older Epsons with 720 native ppi rez, a 1280 and a 2200, and I agree with you that 240 seems to be a good number for these consumer printers. More is better if you have it native (ie without resampling) but for sure 240 – 300 ppi prints pretty well on these models.

Bill
C
Clyde
Jul 13, 2005
Bill Hilton wrote:
Owamanga! writes …

Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it

Here’s a link to an article that discusses this for LightJet 5000, Chromira and Epson 9600 printers that just crossed my virtual desk today. The author runs one of the best custom labs in the western USA, favored by the likes of Robert Glenn Ketchum and Jack Dykinga … basically he says that while the native rez of the Chromira and LJ is 300 ppi (actually 304.8 or 12 lines/mm for the older LJ) these printers have exceptionally good built-in interpolation and 200 ppi looks great (something I’ve heard from two other Chromira or LightJet labs, Bill Nordstrom’s Laser Light and Calypso in Santa Clara).

On the other hand he says the Epson 9600, which has a native rez of 360 ppi, prints noticeably better if you resample to that size and sharpen correctly. I have an Epson 4000 and always try to hit 360 ppi, based on tests I’ve printed. Other printers will of course differ in their ability to resample well or poorly, so the best thing to do is test it yourself. Personally I think a lot depends on how well you can sharpen in Photoshop.

http://www.westcoastimaging.com/wci/page/info/photoshoptip/t ip25.html

Patrick wrote …

On the Epson R300 printers I stay at 240dpi for anything up to 8×10. (720dpi being Epson’s native mode)

I have two older Epsons with 720 native ppi rez, a 1280 and a 2200, and I agree with you that 240 seems to be a good number for these consumer printers. More is better if you have it native (ie without resampling) but for sure 240 – 300 ppi prints pretty well on these models.
Bill

How do you know what the "native resolution" is? Epson and the others don’t publish that information. I know I’ve seen this debated for years, but I’ve seen no hard numbers from good evidence.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it. My view is that if I can’t see any difference, it doesn’t matter.

I also don’t resample down. I just let the dpi resolution adjust up to fit the size I want. I very rarely resample up either. I just lower the dpi to get the size. I have printed at 180 dpi and the quality has been surprisingly good. The average customer wouldn’t know the difference.

Clyde
O
Owamanga
Jul 13, 2005
On Wed, 13 Jul 2005 09:45:17 -0500, Clyde wrote:

How do you know what the "native resolution" is? Epson and the others don’t publish that information. I know I’ve seen this debated for years, but I’ve seen no hard numbers from good evidence.

Usenet legend I guess. Nobody has proven or even attempted to prove otherwise.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it. My view is that if I can’t see any difference, it doesn’t matter.

While that is true, I wouldn’t feel bad sending 240ppi even to the printer. Hell, I’m a great supporter of lower resolution prints, for massive posters so long as the observer isn’t intending to get too close. (eg, I support the 6Mpix is enough for most people argument) But there *are* reasons to want to feed the printer at 720ppi or higher (if your image has that quantity of data).

Firstly, the guy re-sampling 800ppi down to 300ppi is throwing image data away and there is really no good reason to be doing that. It costs nothing to send the printer more data, so send it all the data you’ve got.

Secondly, certain specific instances will look a little better when sent at the native resolution compared to 300ppi. *Most* stuff, you can’t see it. But I’ve printed a test page of fine black and white lines (found it online) which demonstrates that the printer/ink/paper combination *is* capable of much higher than 300ppi. This means that any high contrast, and possibly monochromatic detail in the image (which most photos don’t contain) will fare better at the native resolution.

Now this bit is pure conjecture: The process of un-sharp mask does exactly that to our image, adds fine-detail high-contrast ‘outlining’ to parts of the image. And this might be exactly the type of detail that would be damaged by lowering its resolution to 300ppi.

I also don’t resample down. I just let the dpi resolution adjust up to fit the size I want. I very rarely resample up either. I just lower the dpi to get the size. I have printed at 180 dpi and the quality has been surprisingly good. The average customer wouldn’t know the difference.

Right, but the important difference here is you are adjusting dpi as the tool for ending up with the right size print – *without re-sampling*, which is good. You could just as easily have described the print size in inches, cm, or pixels. DPI is just one part of the triangle.

None of this resolution voodoo is as important as getting the sharpening right.


Owamanga!
http://www.pbase.com/owamanga
P
patrick
Jul 13, 2005
"Clyde" wrote in message
Bill Hilton wrote:
Owamanga! writes …

Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it

Here’s a link to an article that discusses this for LightJet 5000, Chromira and Epson 9600 printers that just crossed my virtual desk today. The author runs one of the best custom labs in the western USA, favored by the likes of Robert Glenn Ketchum and Jack Dykinga … basically he says that while the native rez of the Chromira and LJ is 300 ppi (actually 304.8 or 12 lines/mm for the older LJ) these printers have exceptionally good built-in interpolation and 200 ppi looks great (something I’ve heard from two other Chromira or LightJet labs, Bill Nordstrom’s Laser Light and Calypso in Santa Clara).

On the other hand he says the Epson 9600, which has a native rez of 360 ppi, prints noticeably better if you resample to that size and sharpen correctly. I have an Epson 4000 and always try to hit 360 ppi, based on tests I’ve printed. Other printers will of course differ in their ability to resample well or poorly, so the best thing to do is test it yourself. Personally I think a lot depends on how well you can sharpen in Photoshop.

http://www.westcoastimaging.com/wci/page/info/photoshoptip/t ip25.html

Patrick wrote …

On the Epson R300 printers I stay at 240dpi for anything up to 8×10. (720dpi being Epson’s native mode)

I have two older Epsons with 720 native ppi rez, a 1280 and a 2200, and I agree with you that 240 seems to be a good number for these consumer printers. More is better if you have it native (ie without resampling) but for sure 240 – 300 ppi prints pretty well on these models.
Bill

How do you know what the "native resolution" is? Epson and the others don’t publish that information. I know I’ve seen this debated for years, but I’ve seen no hard numbers from good evidence.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it. My view is that if I can’t see any difference, it doesn’t matter.

I also don’t resample down. I just let the dpi resolution adjust up to fit the size I want. I very rarely resample up either. I just lower the dpi to get the size. I have printed at 180 dpi and the quality has been surprisingly good. The average customer wouldn’t know the difference.
Clyde
Howdo, Clyde! Nice to hear from you again.

Yes, the Epson native mode is definitely 720dpi. They advertise 1440 but that is by double printing.

You’re right that our Epson printers will show no improvement in anything greater than 300dpi. In fact, I can see no improvement between 240dpi and 300dpi.
Since 240 is an integral submultiple of 720, the Big boys recommend that number as an optimum density for print quality and ink conservation. Anything over 240dpi is simply a waste of ink.

So, if I start with a 3000x2000ppi image and want a 4×6 print without resampling, I have to print at 500 dpi — a colossal waste of ink with the risk of saturating the paper.
If I want to print that 3000x2000ppi image at 240dpi, I *must* resize and resample to.1440x960ppi.
And I much prefer to let Photoshop do the resampling than the printer. The same principle obtains when dealing with a cropped image. The only difference is that the starting ppi numbers are now different.

I may simply have a mental block on this issue, but that’s the way I understand it.
Thanks again for the input. . . . patrick
O
Owamanga
Jul 13, 2005
On Wed, 13 Jul 2005 17:04:56 GMT, "patrick"
wrote:

Howdo, Clyde! Nice to hear from you again.

Yes, the Epson native mode is definitely 720dpi. They advertise 1440 but that is by double printing.

Erm, dpi and ppi are not related. The native mode is considered to be 720ppi.

You’re right that our Epson printers will show no improvement in anything greater than 300dpi. In fact, I can see no improvement between 240dpi and 300dpi.

If you just change the ‘d’s to ‘p’s and restate it, then this is consistent with my observations, for *most* material.

Since 240 is an integral submultiple of 720, the Big boys recommend that number as an optimum density for print quality and ink conservation. Anything over 240dpi is simply a waste of ink.

Na. It doesn’t work like this. A single pixel from our (say 240ppi) image needs to be represented by more than a single dot from our printer if we want to see more than 16 different colors. The printer has to dither, spray a few hundred dots in a small area just to give me one dark red pixel. Variable dot size brings in a new dimension, but it’s not enough to change the rules of printer physics. Inkjets *must* dither.

So, if I start with a 3000x2000ppi image and want a 4×6 print without resampling, I have to print at 500 dpi — a colossal waste of ink with the risk of saturating the paper.

500dpi shouldn’t be enough to saturate standard paper. I print at 720dpi, 1440 dpi without saturation (following the printer driver’s built-in limitations based on paper type).

If I want to print that 3000x2000ppi image at 240dpi, I *must* resize and resample to.1440x960ppi.

Aggh!, no. The ppi and dpi bear no relationship to each other.

And I much prefer to let Photoshop do the resampling than the printer. The same principle obtains when dealing with a cropped image. The only difference is that the starting ppi numbers are now different.
I may simply have a mental block on this issue, but that’s the way I understand it.

Sorry, but you *don’t* understand it.


Owamanga!
http://www.pbase.com/owamanga
P
patrick
Jul 13, 2005
"Owamanga" wrote in message
On Wed, 13 Jul 2005 17:04:56 GMT, "patrick"
wrote:

Howdo, Clyde! Nice to hear from you again.

Yes, the Epson native mode is definitely 720dpi. They advertise 1440 but that is by double printing.

Erm, dpi and ppi are not related. The native mode is considered to be 720ppi.

You’re right that our Epson printers will show no improvement in anything greater than 300dpi. In fact, I can see no improvement between 240dpi and 300dpi.

If you just change the ‘d’s to ‘p’s and restate it, then this is consistent with my observations, for *most* material.

Since 240 is an integral submultiple of 720, the Big boys recommend that number as an optimum density for print quality and ink conservation. Anything over 240dpi is simply a waste of ink.

Na. It doesn’t work like this. A single pixel from our (say 240ppi) image needs to be represented by more than a single dot from our printer if we want to see more than 16 different colors. The printer has to dither, spray a few hundred dots in a small area just to give me one dark red pixel. Variable dot size brings in a new dimension, but it’s not enough to change the rules of printer physics. Inkjets *must* dither.

So, if I start with a 3000x2000ppi image and want a 4×6 print without resampling, I have to print at 500 dpi — a colossal waste of ink with the risk of saturating the paper.

500dpi shouldn’t be enough to saturate standard paper. I print at 720dpi, 1440 dpi without saturation (following the printer driver’s built-in limitations based on paper type).

If I want to print that 3000x2000ppi image at 240dpi, I *must* resize and resample to.1440x960ppi.

Aggh!, no. The ppi and dpi bear no relationship to each other.
And I much prefer to let Photoshop do the resampling than the printer. The same principle obtains when dealing with a cropped image. The only difference is that the starting ppi numbers are now different.
I may simply have a mental block on this issue, but that’s the way I understand it.

Sorry, but you *don’t* understand it.


Owamanga!
http://www.pbase.com/owamanga

OK, I’ll keep an open mind, Owamanga. It is still my understanding that ppi refers to the density of the pixels on the monitor and that dpi refers to the density of the pixels at the printer. "Usage of the Trade" A pixel is a pixel. You can spread those pixels on the monitor at 72ppi or 96 ppi, or whatever. I’m sure that is the terminology I’ve read in the literature. Same with the printer, except that trade usiage is dpi ain the printing business. They both refer to the density at which the pixels are displayed or printed. And, in the photo editing community on the Internet, they are being used interchangeably (even if wrongly to the traditional purist).

At any rate, thanks for your patience in trying to get through to me. Some night I may wake up with a great white light experience and see it all. Not today, though. Sorry.

Always enjoy your inputs, Owamanga.
cu . . . . patrick
H
Hecate
Jul 13, 2005
On Wed, 13 Jul 2005 09:45:17 -0500, Clyde wrote:

How do you know what the "native resolution" is? Epson and the others don’t publish that information. I know I’ve seen this debated for years, but I’ve seen no hard numbers from good evidence.
The Epson native resolution is 720. It’s been known for quite a while. And there have been several tests run showing that the best resolutions for printing with an Epson are multiples or divisors of that number. Before you ask, if you want further info ask Kennedy McEwen in comp.periphs.printers or comp.periphs.scanners. He’s usually around those 2 groups.



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
H
Hecate
Jul 13, 2005
On Wed, 13 Jul 2005 15:42:05 GMT, Owamanga
wrote:

On Wed, 13 Jul 2005 09:45:17 -0500, Clyde wrote:

How do you know what the "native resolution" is? Epson and the others don’t publish that information. I know I’ve seen this debated for years, but I’ve seen no hard numbers from good evidence.

Usenet legend I guess. Nobody has proven or even attempted to prove otherwise.

Not true..See my post to Clyde.



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
H
Hecate
Jul 13, 2005
On Wed, 13 Jul 2005 20:40:30 GMT, "patrick"
wrote:

OK, I’ll keep an open mind, Owamanga. It is still my understanding that ppi refers to the density of the pixels on the monitor and that dpi refers to the density of the pixels at the printer. "Usage of the Trade" A pixel is a pixel. You can spread those pixels on the monitor at 72ppi or 96 ppi, or whatever. I’m sure that is the terminology I’ve read in the literature. Same with the printer, except that trade usiage is dpi ain the printing business. They both refer to the density at which the pixels are displayed or printed. And, in the photo editing community on the Internet, they are being used interchangeably (even if wrongly to the traditional purist).
No, they do not refer to same thing at all. And they are wrongly used as the same thing. ppi refers to the image size – pixels per inch. dpi refers to the number of dots per inch when printing. You can print a 360ppi image at 1440 dpi, for example. That is, you8 can have more printed dots per inch than the equivalent pixels per inch of the image. If something is printed at 360 dpi it means the printer is laying down 360 dots per inch of paper. That is not the same as the image size which you can see quite easily by looking at image and changing the pixels per inch in PS, and seeing how the printable area changes. In terms of printing, ppi describes the printable area, not the number of dots laid down by the printer.



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
P
patrick
Jul 14, 2005
"Hecate" wrote in message
On Wed, 13 Jul 2005 20:40:30 GMT, "patrick"
wrote:

OK, I’ll keep an open mind, Owamanga. It is still my understanding that ppi
refers to the density of the pixels on the monitor and that dpi refers to the density of the pixels at the printer. "Usage of the Trade" A pixel is a pixel. You can spread those pixels on the monitor at 72ppi or 96 ppi, or whatever. I’m sure that is the terminology I’ve read in the literature. Same with the printer, except that trade usiage is dpi ain the printing business. They both refer to the density at which the pixels are displayed or printed. And, in the photo editing community on the Internet, they are being used interchangeably (even if wrongly to the traditional purist).
No, they do not refer to same thing at all. And they are wrongly used as the same thing. ppi refers to the image size – pixels per inch. dpi refers to the number of dots per inch when printing. You can print a 360ppi image at 1440 dpi, for example. That is, you8 can have more printed dots per inch than the equivalent pixels per inch of the image. If something is printed at 360 dpi it means the printer is laying down 360 dots per inch of paper. That is not the same as the image size which you can see quite easily by looking at image and changing the pixels per inch in PS, and seeing how the printable area changes. In terms of printing, ppi describes the printable area, not the number of dots laid down by the printer.



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
Sometimes it’s very hard to edit your own work!
I thought that was exactly what I said above: "ppi
refers to the density of the pixels on the monitor and dpi refers to the density of the pixels at the printer…."
I never equated the two.

I’m out of this thread. I have NO idea how "ppi describes the printable area, not
the number of dots laid down by the printer." In fact, I’ve never encountered the ppi term in the printer world. So I’m obviously not adequately informed to continue the discussion without further study. For now, however, I’m comfortable with using ppi to describe the number of pixels per inch on the monitor and dpi to describe the number of pixels per inch on the printer.

And I DO appreciate the frustration in attempting to explain something to someone who "just doesn’t get it!"

cu ….patrick
BH
Bill Hilton
Jul 14, 2005
Hecate writes …

The Epson native resolution is 720.

This is true for the consumer grade printers like the 2200, 1280 etc but for the Professional printers with wider carriages like the 9600, 7600, 4000 etc the native input rez is 360 ppi.

Bill
BH
Bill Hilton
Jul 14, 2005
Patrick writes …

It is still my understanding that ppi refers to the
density of the pixels on the monitor and that dpi refers to the density of the pixels at the printer.

In reference to inkjet printers, think of it this way … the input file resolution is (or should be, but often isn’t) expressed in pixels per inch or ppi. This is the correct terminology for an *input* file. The output of the printer is dots (or small droplets of ink) and is properly expressed as ‘dots per inch’ or dpi. The two are not directly related since you typically choose the output dpi based on the type of paper used, that is, a glossy paper for say the 1280 can handle up to 2,880 x 1,440 dpi while plain uncoated paper can handle only 180 dpi. So the same 300 ppi input file might be printed 720 x 720 dpi on matte paper or up to 2880 x 1440 dpi on PGPP with good results on either.
BH
Bill Hilton
Jul 14, 2005
Clyde asks …

How do you know what the "native resolution" is? Epson and the others don’t publish that information.

Epson *does* publish it for the Professional printers. The documentation for the 4000 says the native rez is 360 ppi and suggests upsampling to get there.

Epson is more coy about publishing the numbers for the consumer models but, IIRC, programs like Qimage can read this data off the Epson driver and supposedly it’s 720 ppi.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it.

I feel the same with my Epson 1280 and 2200 consumer models, but I’ve seen test patterns printed at higher rez that seem to back up the 720 ppi claims.

My view is that if I can’t see any
difference, it doesn’t matter.

It’s a visual art so that makes sense 🙂

Bill
AM
Andrew Morton
Jul 14, 2005
patrick wrote:
I’m out of this thread. I have NO idea how "ppi describes the printable area, not
the number of dots laid down by the printer."

Regarding inket/bubblejet printers, a /printed/ pixel (picture element) is made up of lots of tiny, tiny dots. To vary the colour, different numbers of different coloured dots are used.

If you print a one pixel image at the native ppi resolution of the printer (discussed elsewhere in this thread) and look at it with some magnification, you will see the individual dots.

It is like Pointillism (e.g. http://www.childs.mccsc.edu/art99/art2p6.htm) in that the dots are next to each other instead of being mixed before being applied to the paper.

Andrew
O
Owamanga
Jul 14, 2005
On Wed, 13 Jul 2005 20:40:30 GMT, "patrick"
wrote:
OK, I’ll keep an open mind, Owamanga. It is still my understanding that ppi refers to the density of the pixels on the monitor and that dpi refers to the density of the pixels at the printer.

Okay, yes, you can look at it that way.

A pixel is a pixel. You can spread those pixels on the monitor at 72ppi or 96 ppi, or whatever. I’m sure that is the terminology I’ve read in the literature. Same with the printer, except that trade usiage is dpi ain the printing business.

And her lies the problem. There is no 1:1 correlation between a printed dot and an image pixel. A printed dot can be at most, one of 16 different colors. Your image pixel can be one of 16 million different colors. How can the printer possibly represent 16 million different possible pixel colors when in a single dot, it can only spray:

Yellow or no Yellow,
Magenta or no Magenta,
Cyan or no Cyan,
Black or no Black

(16 different combinations). I’m simplifying here to demonstrate the point, there are of course a number of other advancements such as variable dot size, more color ink tanks and the affect of clusters of pixels of similar color. On the other hand, one could argue that if you spray black onto any mixture of CYM, you’ll get black, so the number of possible colors a single dot can be is only 9.

PPI = DPI in one situation only. When your source image is a 2-bit monochromatic image, ie it contains only 2 colors (say, black and white for example) and no shades of gray. Then, the printer can truly represent an image pixel with a single dot. For full-color prints, the printer needs 16, 32 or more dots per pixel, that when viewed together, our eyes merge to represent the ‘true’ color of the image pixel.

That having been said, this only applies to dithering devices (such as ink-jets and laser-printers). Continuous tone printers (some of the little 4×6 photo dye-sub printers) are not limited to 9 or 16 colors per dot, it can mix ink, just as your monitor mixes light and lay down a single ‘dot’ that can be one of several million different colors.

Same is true of digital wet printers, they can fire lasers at different intensities onto photo-sensitive paper so that each ‘dot’ (or printed pixel) can truly be one of millions of colors.

So, an ink jet dpi resolution of 300 dpi is no where near as good as a digital wet system or dye-sub printer of the same resolution.

(Here’s a dye-sub if you’ve never seen one)
http://www.steves-digicams.com/2005_reviews/cp600.html

Don’t run out and buy one though – From my own experience, a properly functioning ink-jet still wins. i.e. Resolution *is not that important* compared to a decent gamut. Today I have access to both but generally use neither, preferring to do digital wet prints from www.mpix.com instead.

They both refer to the density at which the pixels are
displayed or printed. And, in the photo editing community on the Internet, they are being used interchangeably (even if wrongly to the traditional purist).

This *isn’t* a purist thing. They are simply *not* comparable. A group of three RGB CRT dots that together, when illuminated provide one pixel capable of being any of 16 million colors can’t be compared to a grungy little inkjet dot that can be one of 16 (half of those being black).

At any rate, thanks for your patience in trying to get through to me. Some night I may wake up with a great white light experience and see it all. Not today, though. Sorry.

Don’t worry at all. It’s not you. It’s an *extremely* common misconception. It’s been drummed into the population from all angles – magazines, websites, the dude who’s still picking his spots at circuit city, and of course even this NG.

Unlike other places where marketing ‘cons’ people, (How many bytes in a megabyte – a 21" screen that isn’t 21" in any dimension even if you wrap a tape measure round it etc – processor speeds described in clock cycles rather than instructions per second – disk space described in it’s unformatted capacity etc), this one is a biggie.

Always enjoy your inputs, Owamanga.

Well, that *is* strange.

😉


Owamanga!
http://www.pbase.com/owamanga
O
Owamanga
Jul 14, 2005
On Thu, 14 Jul 2005 09:55:51 +0100, "Andrew Morton" wrote:

It is like Pointillism (e.g. http://www.childs.mccsc.edu/art99/art2p6.htm) in that the dots are next to each other instead of being mixed before being applied to the paper.

Hey, now I know what a pointillism is…

One in particular, especially the fill of sun in this image, looks just like a magnified ink jet print:
http://www.childs.mccsc.edu/art99/art2p6a2.htm


Owamanga!
http://www.pbase.com/owamanga
C
Clyde
Jul 14, 2005
Bill Hilton wrote:
Clyde asks …

How do you know what the "native resolution" is? Epson and the others don’t publish that information.

Epson *does* publish it for the Professional printers. The documentation for the 4000 says the native rez is 360 ppi and suggests upsampling to get there.

Epson is more coy about publishing the numbers for the consumer models but, IIRC, programs like Qimage can read this data off the Epson driver and supposedly it’s 720 ppi.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it.

I feel the same with my Epson 1280 and 2200 consumer models, but I’ve seen test patterns printed at higher rez that seem to back up the 720 ppi claims.

My view is that if I can’t see any
difference, it doesn’t matter.

It’s a visual art so that makes sense 🙂

Bill

I would like to see some documentation of that Qimage claim. I’ve been seeing all sorts of claims for about a decade and haven’t seen anything but hearsay. (Yes, that includes your reply Hecate.) Just once I’d like to seen something solid.

Not that it will matter for more than my curiosity. I’ll still keep printing at many resolutions from 240 and up. To me resampling is way more dangerous to my file than unseen printer resolution fine tuning.

Clyde
BH
Bill Hilton
Jul 14, 2005
Clyde writes …

I would like to see some documentation of that Qimage claim (that the input rez of consumer grade Epson printers is 720 ppi)

Go to Google and type in "qimage epson 720 ppi" and read away …
H
Hecate
Jul 14, 2005
On Thu, 14 Jul 2005 10:04:58 -0500, Clyde wrote:

I would like to see some documentation of that Qimage claim. I’ve been seeing all sorts of claims for about a decade and haven’t seen anything but hearsay. (Yes, that includes your reply Hecate.) Just once I’d like to seen something solid.

No, not hearsay. I referred you to Kennedy because he and a number of others were involved in a project to find the optimum printing resolution for those printers.



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
B
Bob
Jul 15, 2005
Hi, Bill.
Thank you for the link to the West Coast Imaging Photoshop tip. I found the article very helpful, and I’m enjoying the other Photoshop tips on the site.

Bob

"Bill Hilton" wrote in message
Owamanga! writes …

Possibly you can argue that upscaling in Photoshop is better than letting the printer driver do it

Here’s a link to an article that discusses this for LightJet 5000, Chromira and Epson 9600 printers that just crossed my virtual desk today. The author runs one of the best custom labs in the western USA, favored by the likes of Robert Glenn Ketchum and Jack Dykinga … basically he says that while the native rez of the Chromira and LJ is 300 ppi (actually 304.8 or 12 lines/mm for the older LJ) these printers have exceptionally good built-in interpolation and 200 ppi looks great (something I’ve heard from two other Chromira or LightJet labs, Bill Nordstrom’s Laser Light and Calypso in Santa Clara).

On the other hand he says the Epson 9600, which has a native rez of 360 ppi, prints noticeably better if you resample to that size and sharpen correctly. I have an Epson 4000 and always try to hit 360 ppi, based on tests I’ve printed. Other printers will of course differ in their ability to resample well or poorly, so the best thing to do is test it yourself. Personally I think a lot depends on how well you can sharpen in Photoshop.

http://www.westcoastimaging.com/wci/page/info/photoshoptip/t ip25.html
Patrick wrote …

On the Epson R300 printers I stay at 240dpi for anything up to 8×10. (720dpi being Epson’s native mode)

I have two older Epsons with 720 native ppi rez, a 1280 and a 2200, and I agree with you that 240 seems to be a good number for these consumer printers. More is better if you have it native (ie without resampling) but for sure 240 – 300 ppi prints pretty well on these models.
Bill
T
tinman334
Jul 15, 2005
Clyde wrote:
Bill Hilton wrote:
Clyde asks …

How do you know what the "native resolution" is? Epson and the others don’t publish that information.

Epson *does* publish it for the Professional printers. The documentation for the 4000 says the native rez is 360 ppi and suggests upsampling to get there.

Epson is more coy about publishing the numbers for the consumer models but, IIRC, programs like Qimage can read this data off the Epson driver and supposedly it’s 720 ppi.

In my experience, I can’t see any difference on my Epson R800 for anything over 300 dpi sent to it.

I feel the same with my Epson 1280 and 2200 consumer models, but I’ve seen test patterns printed at higher rez that seem to back up the 720 ppi claims.

My view is that if I can’t see any
difference, it doesn’t matter.

It’s a visual art so that makes sense 🙂

Bill

I would like to see some documentation of that Qimage claim. I’ve been seeing all sorts of claims for about a decade and haven’t seen anything but hearsay. (Yes, that includes your reply Hecate.) Just once I’d like to seen something solid.

Not that it will matter for more than my curiosity. I’ll still keep printing at many resolutions from 240 and up. To me resampling is way more dangerous to my file than unseen printer resolution fine tuning.

I’m with you on this one, Clyde. I can print my scans at ~300ppi on my 2200 without resampling. I can’t see any difference in the prints that are resampled to 360ppi in PS, so I say why bother. On the other hand, the 2200 is probably resampling to 360ppi if the file sent to it is not exactly that. So the choice is to either resample at PS or at Epson.
C
Clyde
Jul 15, 2005
Bill Hilton wrote:
Clyde writes …

I would like to see some documentation of that Qimage claim (that the input rez of consumer grade Epson printers is 720 ppi)

Go to Google and type in "qimage epson 720 ppi" and read away …

OK, OK. You made me look.

Very interesting. I guess there is some evidence for it. I’ll keep that in mind. However, I did notice that some of the samples that were shown as quality differences weren’t significant for me. i.e. The quality difference in their tests we barely noticeable in their own enlargements. (Several sites.)

So, in theory and technically 720 ppi is what the Epson drivers/printers work best at. In practice, I’m going to keep printing at 240+. OK, I’m may lean more toward 360 now rather than 300, but we’ll see.

Thanks for the ideas, Bill & Hecate.

Clyde
O
Owamanga
Jul 15, 2005
On Fri, 15 Jul 2005 11:13:19 -0500, Clyde wrote:

So, in theory and technically 720 ppi is what the Epson drivers/printers work best at. In practice, I’m going to keep printing at 240+. OK, I’m may lean more toward 360 now rather than 300, but we’ll see.

Whilst we are pulling numbers out of the air, for artistic work, I have determined the best image resolution destined for an Epson would be 445.71556 ppi, choose 446dpi as the next best thing.

….it was calculated using the Fibonacci sequence and some beer. Your prints should now have that ‘comfortable’ feel to them.

For the mathematically challenged, the equation used is quite complex, as follows:

720dpi / Phi = 445.71556


Owamanga!
http://www.pbase.com/owamanga
P
patrick
Jul 15, 2005
"Owamanga" wrote in message
On Fri, 15 Jul 2005 11:13:19 -0500, Clyde wrote:

So, in theory and technically 720 ppi is what the Epson drivers/printers work best at. In practice, I’m going to keep printing at 240+. OK, I’m may lean more toward 360 now rather than 300, but we’ll see.

Whilst we are pulling numbers out of the air, for artistic work, I have determined the best image resolution destined for an Epson would be 445.71556 ppi, choose 446dpi as the next best thing.

…it was calculated using the Fibonacci sequence and some beer. Your prints should now have that ‘comfortable’ feel to them.

For the mathematically challenged, the equation used is quite complex, as follows:

720dpi / Phi = 445.71556


Owamanga!
http://www.pbase.com/owamanga

I would be interested to see if you were to use a lupe on prints at 320 vs 240 if you would detect *any* difference.
("Significant" difference?)
[Printing black lines is not a predictor of RGB resolution.]

I still have not recovered from the shock that some actually print to a consumer inkjet at 1440dpi! That is simply beyond my ken.

However, that explains all of the confusion in the early parts of our thread. I presumed that 240dpi would be the first criterion for the printout.
If you’re going to print at *any* dpi, then the arithmetic to avoid resampling is trivial and there is no discussion — even if you end up at 445.71556dpi ;>).
I did not consider that option in the early rounds.

"All is forgiven. Please come home!"

cu …. patrick

PS to Owamanga: I noted earlier that I always "enjoy" your notes. Can I amend that to "respect"? ;>)
H
Hecate
Jul 15, 2005
On Fri, 15 Jul 2005 11:13:19 -0500, Clyde wrote:

Thanks for the ideas, Bill & Hecate.

No problem. You’re a hard one to convince 😉



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
H
Hecate
Jul 15, 2005
On Fri, 15 Jul 2005 20:17:23 GMT, "patrick" wrote:

I would be interested to see if you were to use a lupe on prints at 320 vs 240 if you would detect *any* difference.
("Significant" difference?)
[Printing black lines is not a predictor of RGB resolution.]

That depends on whether you’re talking ppi or dpi 🙂

I still have not recovered from the shock that some actually print to a consumer inkjet at 1440dpi! That is simply beyond my ken.
Usually a 360ppi image printed at 1440 dpi. (For smaller images)



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
P
patrick
Jul 16, 2005
"Hecate" wrote in message
On Fri, 15 Jul 2005 20:17:23 GMT, "patrick" wrote:

I would be interested to see if you were to use a lupe on prints at 320 vs
240 if you would detect *any* difference.
("Significant" difference?)
[Printing black lines is not a predictor of RGB resolution.]

That depends on whether you’re talking ppi or dpi 🙂

I still have not recovered from the shock that some actually print to a consumer inkjet at 1440dpi! That is simply beyond my ken.
Usually a 360ppi image printed at 1440 dpi. (For smaller images)


Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…

Oh year?!! How about lpi, then?

You sure know how to hurt a guy!
.. . . . patrick
C
Clyde
Jul 16, 2005
Hecate wrote:
On Fri, 15 Jul 2005 11:13:19 -0500, Clyde wrote:

Thanks for the ideas, Bill & Hecate.

No problem. You’re a hard one to convince 😉



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…

The older I get the more skeptical I get. I want evidence! Hum, maybe I’ve been using Google too much. 😉

Clyde
H
Hecate
Jul 16, 2005
On Sat, 16 Jul 2005 00:34:56 GMT, "patrick" wrote:

Oh year?!! How about lpi, then?

You sure know how to hurt a guy!
. . . . patrick
Different thing altogether 🙂



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
H
Hecate
Jul 16, 2005
On Sat, 16 Jul 2005 09:45:54 -0500, Clyde wrote:

No problem. You’re a hard one to convince 😉

The older I get the more skeptical I get. I want evidence! Hum, maybe I’ve been using Google too much. 😉
You and me both 🙂



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…
P
patrick
Jul 18, 2005
"Hecate" wrote in message
On Sat, 16 Jul 2005 00:34:56 GMT, "patrick" wrote:

Oh year?!! How about lpi, then?

You sure know how to hurt a guy!
. . . . patrick
Different thing altogether 🙂



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…

"It’s a joke, Son!"
.. . . . patrick
H
Hecate
Jul 19, 2005
On Mon, 18 Jul 2005 18:03:27 GMT, "patrick"
wrote:

"Hecate" wrote in message
On Sat, 16 Jul 2005 00:34:56 GMT, "patrick" wrote:

Oh year?!! How about lpi, then?

You sure know how to hurt a guy!
. . . . patrick
Different thing altogether 🙂



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…

"It’s a joke, Son!"
. . . . patrick
I know that’s why I put a smiley after it -and I ain’t no ones son 😉



Hecate – The Real One

Fashion: Buying things you don’t need, with money
you don’t have, to impress people you don’t like…

How to Improve Photoshop Performance

Learn how to optimize Photoshop for maximum speed, troubleshoot common issues, and keep your projects organized so that you can work faster than ever before!

Related Discussion Topics

Nice and short text about related topics in discussion sections