Print resolution question

On Wed, 4 Feb 2004 08:45:35 -0800, "Norman & Nancy Perry" wrote:

I noticed something that I can not explain. I scanned a 35mm slide using a Nikon Coolscan 8000 film scanner. The scan was done at 4000 ppi. I made some corrections (levels, contrast, brightness and unsharp mask) with PhotoShop 7. I then printed the image on my HP Deskjet 970 printer at the full 4000 ppi. I then changed the size of the image (re-sampled) with PhotoShop to 300 ppi and printed the image again. Both images were printed the same size (4 X 6). When reviewing the two prints I noticed that the printed detail looked identical, but the 300 ppi print was much warmer in color. It had much less blue and more yellow / red overall. I would appreciate any comments or information that explains why this occurs. Thanks Norm.

Just fyi, printing at 4000 ppi is overkill, printing at about 300 ppi is probably the most you need, and sometimes as low as 150 is okay for inkjet printing. But that aside…

And assuming you didn’t inadvertently change a color mode or embed a color profile and that the images are identical except for the resolution, then…

The print engine processes the digital image in order to translate the image’s resolution AND color mode (greyscale, rgb, or cmyk, ect) to the printer’s output resolution and ink colors, I’d assume that when you downsampled the ratio of colors in the sampling process changed slightly and therefore the output colors changed slightly. (i.e, a 4000 ppi image printed at 1440 dpi has a different ratio than a 300 ppi image printed at 1440 dpi, and therefore the sampling of colors is different).

The jump from 4000 ppi to 300 ppi is pretty severe. I doubt you’d see much difference in most prints when going from say a 300 ppi image to a 150 ppi image.

That’s a rough explanation, but is generally the rationale for saving an original scan, then working on a copy of the file in order to color correct and set resolution for the particular output device it will be printed on.

Norman & Nancy Perry wrote:

I noticed something that I can not explain. I scanned a 35mm slide using a Nikon Coolscan 8000 film scanner. The scan was done at 4000 ppi. I made some corrections (levels, contrast, brightness and unsharp mask) with PhotoShop 7. I then printed the image on my HP Deskjet 970 printer at the full 4000 ppi. I then changed the size of the image (re-sampled) with PhotoShop to 300 ppi and printed the image again. Both images were printed the same size (4 X 6). When reviewing the two prints I noticed that the printed detail looked identical, but the 300 ppi print was much warmer in color. It had much less blue and more yellow / red overall. I would appreciate any comments or information that explains why this occurs.

It could be the following: Inkjet printers cannot print millions of colors. They can only print a drop of ink or not, so they can only print those colors they have. In order to print the other colors, they use a method where each pixel is represented by a number of dots (ink drops). That is why you should set the image at 300 dpi maximum, even if the printer is 1200 dpi or even 2400 dpi. That way a 1200 dpi printer can use 4×4 drops to represent one pixel (and 6×6 drops in case of a 2400 dpi printer).

If you send an image to the printer at 4000 dpi, the printer may try to print more pixels, and consequently use less drops per pixel. That may alter the colors of the print.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

-xiray- wrote:

On Wed, 4 Feb 2004 08:45:35 -0800, "Norman & Nancy Perry" wrote:

I noticed something that I can not explain. I scanned a 35mm slide using a Nikon Coolscan 8000 film scanner. The scan was done at 4000 ppi. I made some corrections (levels, contrast, brightness and unsharp mask) with PhotoShop 7. I then printed the image on my HP Deskjet 970 printer at the full 4000 ppi. I then changed the size of the image (re-sampled) with PhotoShop to 300 ppi and printed the image again. Both images were printed the same size (4 X 6). When reviewing the two prints I noticed that the printed detail looked identical, but the 300 ppi print was much warmer in color. It had much less blue and more yellow / red overall. I would appreciate any comments or information that explains why this occurs. Thanks Norm.

Just fyi, printing at 4000 ppi is overkill, printing at about 300 ppi is probably the most you need, and sometimes as low as 150 is okay for inkjet printing. But that aside…

i’d point out that a scan of a 35mm frame @ 4000ppi will print @ ~1000ppi at 4×6".

And assuming you didn’t inadvertently change a color mode or embed a color profile and that the images are identical except for the resolution, then…

The print engine processes the digital image in order to translate the image’s resolution AND color mode (greyscale, rgb, or cmyk, ect) to the printer’s output resolution and ink colors, I’d assume that when you downsampled the ratio of colors in the sampling process changed slightly and therefore the output colors changed slightly. (i.e, a 4000 ppi image printed at 1440 dpi has a different ratio than a 300 ppi image printed at 1440 dpi, and therefore the sampling of colors is different).

The jump from 4000 ppi to 300 ppi is pretty severe. I doubt you’d see much difference in most prints when going from say a 300 ppi image to a 150 ppi image.

That’s a rough explanation, but is generally the rationale for saving an original scan, then working on a copy of the file in order to color correct and set resolution for the particular output device it will be printed on.

Norman & Nancy Perry wrote:

I noticed something that I can not explain. I scanned a 35mm slide using a Nikon Coolscan 8000 film scanner. The scan was done at 4000 ppi. I made some corrections (levels, contrast, brightness and unsharp mask) with PhotoShop 7. I then printed the image on my HP Deskjet 970 printer at the full 4000 ppi. I then changed the size of the image (re-sampled) with PhotoShop to 300 ppi and printed the image again. Both images were printed the same size (4 X 6). When reviewing the two prints I noticed that the printed detail looked identical, but the 300 ppi print was much warmer in color. It had much less blue and more yellow / red overall. I would appreciate any comments or information that explains why this occurs. Thanks Norm.

I have never encountered a description in the literature of this happening before, or a theoretical discussion of why resampling would cause a color shift.

Pilot error aside, here are several guesses:

1) The printed image is so large that is is exposing a bug in the HP printer driver. For example, a failed memory allocation may cause the profile conversion to fail, causing the printer driver to revert to its default color handling – typically manufacturers err a bit on the warm side, for obvious reasons – people like healthy warm faces more than pale cold ones.

2) If the high rez image has a lot of noise, then resampling the image in gamma space, as normally occurs for a printer driver, could cause an overall color shift. If this is the case, you would see the color shift right away in Photoshop, rather than after printing.

3) You are running afoul of some built-in sharpening or image enhancement "feature" in the printer driver.

As far as tracking this down – I’d suggest that you do the conversion again in Photoshop and see if you see a shift in color at that point. —

Mike Russell
www.curvemeister.com
www.geigy.2y.net

On Wed, 04 Feb 2004 13:46:50 -0900, James Connell
wrote:

i’d point out that a scan of a 35mm frame @ 4000ppi will print @ ~1000ppi at 4×6".

I’d point out that a 1000 ppi image is overkill for a 4 x 6 print.