S9000 Microbanding example scans

Frank Nichols · Jul 9, 2002

I posted in another thread that I could see a difference between my MacOSX driver quality and my Windows XP drivers output to my S9000 printer. One of the issues is what I call micro-banding.

Below is a scan of printouts from the two systems to the same printer on the same paper (Epson Photo Paper - Costco) with the same settings (Glossy Photo Paper) using WeInk CRS system (the results are the same with OEM carts.)

The orginal image is from my Canon G2 and was taken in JPEG mode Large Superfine. NO editing has been done to the image.

The image was printed at 8x10 inches in both cases. The prints were scanned at 1200DPI to avoid any chance (hopefully) of artifacts from scannign at close to the printer resolution. A sample of about 600x400 pixels was then copied from each at approximately the same location, my wifes shoulder. On the 8x10 inch printout the sample area would measure 1/2 inch x 1/3 inch (1200 dpi / 600 pixels = 1/2 inch).

The banding, to me, is VERY obvious in the Windows example and just visible in the Mac output (it looks worse in the scan than under a loupe). Remember this is the equivalent of looking at the picture with an 8X loupe.

Comments???

Example 1 - Full picture of my wife and Jake my Lab pup. (My wife is the one with her head cut off..)

Example 2 - Windows XP Output (latest driver from canon Website)

Example 3 - MacOS X output

Example 4 Sample from Original jpg from camera.

--
I plan on living forever - so far so good!

pdqgp · Jul 9, 2002

Nice photo...nice print and nice scan....but I just don't see it. Not in your image or the many that I've printed on my S900. Maybe it's just me, but I'm sure if I pulled out a loupe and looked at photos printed at the camera store, they'd look bad to me. Perhaps I'm just not as picky.

tim

Frank Nichols said:
I posted in another thread that I could see a difference between my
MacOSX driver quality and my Windows XP drivers output to my S9000
printer. One of the issues is what I call micro-banding.

Below is a scan of printouts from the two systems to the same
printer on the same paper (Epson Photo Paper - Costco) with the
same settings (Glossy Photo Paper) using WeInk CRS system (the
results are the same with OEM carts.)

The orginal image is from my Canon G2 and was taken in JPEG mode
Large Superfine. NO editing has been done to the image.

The image was printed at 8x10 inches in both cases. The prints were
scanned at 1200DPI to avoid any chance (hopefully) of artifacts
from scannign at close to the printer resolution. A sample of about
600x400 pixels was then copied from each at approximately the same
location, my wifes shoulder. On the 8x10 inch printout the sample
area would measure 1/2 inch x 1/3 inch (1200 dpi / 600 pixels = 1/2
inch).

The banding, to me, is VERY obvious in the Windows example and just
visible in the Mac output (it looks worse in the scan than under a
loupe). Remember this is the equivalent of looking at the picture
with an 8X loupe.

Comments???

Example 1 - Full picture of my wife and Jake my Lab pup. (My wife
is the one with her head cut off..)

Example 2 - Windows XP Output (latest driver from canon Website)

Example 3 - MacOS X output

Example 4 Sample from Original jpg from camera.

--
I plan on living forever - so far so good!

WillieB38717 · Jul 9, 2002

Interesting comparison, this. A few questions, if you have time. Is the banding in the prints vertical? Can you see them with the naked eye? Also, I'm sure you have noticed the difference in color between the prints - the "Mac print" being darker. Have you tried darkening the "Windows print" before printing to see what effect it has on banding?
--
WillieB

Adam2 · Jul 9, 2002

Yup, its there under 8x magnification and therein lies the point. Try looking it at 1x magnification. Do you see it? I don't. If you enlarge prints from the 2100/2200 you will see it too, but in the horizontal direction. As far as I am aware every inkjet produces banding to some degree.

A comparison of the printers (s9000 and the 2100) documents this: http://www.photo-i.co.uk (See Page 11)

Frank Nichols said:
I posted in another thread that I could see a difference between my
MacOSX driver quality and my Windows XP drivers output to my S9000
printer. One of the issues is what I call micro-banding.

Below is a scan of printouts from the two systems to the same
printer on the same paper (Epson Photo Paper - Costco) with the
same settings (Glossy Photo Paper) using WeInk CRS system (the
results are the same with OEM carts.)

The orginal image is from my Canon G2 and was taken in JPEG mode
Large Superfine. NO editing has been done to the image.

The image was printed at 8x10 inches in both cases. The prints were
scanned at 1200DPI to avoid any chance (hopefully) of artifacts
from scannign at close to the printer resolution. A sample of about
600x400 pixels was then copied from each at approximately the same
location, my wifes shoulder. On the 8x10 inch printout the sample
area would measure 1/2 inch x 1/3 inch (1200 dpi / 600 pixels = 1/2
inch).

The banding, to me, is VERY obvious in the Windows example and just
visible in the Mac output (it looks worse in the scan than under a
loupe). Remember this is the equivalent of looking at the picture
with an 8X loupe.

Comments???

Example 1 - Full picture of my wife and Jake my Lab pup. (My wife
is the one with her head cut off..)

Example 2 - Windows XP Output (latest driver from canon Website)

Example 3 - MacOS X output

Example 4 Sample from Original jpg from camera.

--
I plan on living forever - so far so good!

stucole · Jul 9, 2002

The banding you mention looks more like noise.
I doubt that there is a problem with your printer.

It's either the original file or the way your computer processes the image and sends it to the printer.

WillieB38717 said:
Interesting comparison, this. A few questions, if you have time.
Is the banding in the prints vertical? Can you see them with the
naked eye? Also, I'm sure you have noticed the difference in color
between the prints - the "Mac print" being darker. Have you tried
darkening the "Windows print" before printing to see what effect it
has on banding?
--
WillieB

dgrogers · Jul 9, 2002

The first thing that's obvious to me is the difference in color, are you using any profiles? Different colors and shades may make any banding more obvious, however, since you have to blow up the print so large or view it through an 8x loupe to see the banding, I don't understand what the problem is. Another thing the "banding" may be is the natural ruts in the paper -- which may give the appearance of banding and grain. All coated paper has some kind of valleys, ruts, or pattern -- is the banding you see going along those imperfections in the paper (hold the print at an angle under a light source to see these imperfections, they are there).

--
dgrogers

http://www.pbase.com/drog

Frank Nichols · Jul 9, 2002

1. The banding is in the direction of the print head movement across the paper.

2. Well, at first I couldn't - my 52 year old eyes need some help. My friend can see them and pointed them out. Now if I catch the light just right I can see the banding.

3. The difference in color is my sloppy adjustment of the scan brightness. My scanner scans "dark" and I lighten them in photoshop. The prints themselves are virturally identical in color and brightness.

However, I think I will try changing the brightness alittle each way and printing some tests - it would be interesting if the banding has a very narrow zone of brightness and you can "save" a picture by a slight change in brightness.

fn

WillieB said:
Interesting comparison, this. A few questions, if you have time.
Is the banding in the prints vertical? Can you see them with the
naked eye? Also, I'm sure you have noticed the difference in color
between the prints - the "Mac print" being darker. Have you tried
darkening the "Windows print" before printing to see what effect it
has on banding?
--
WillieB

--
I plan on living forever - so far so good!

Frank Nichols · Jul 9, 2002

Unfortunately I can see this without magnification. Agreed at arms length the two prints are virtually identical. However, as soon as anyone hears that I printed the picture myself the first thing they do is stick it to the end of their nose and examine it!

fn

Adam2 said:
Yup, its there under 8x magnification and therein lies the point.
Try looking it at 1x magnification. Do you see it? I don't. If
you enlarge prints from the 2100/2200 you will see it too, but in
the horizontal direction. As far as I am aware every inkjet
produces banding to some degree.

A comparison of the printers (s9000 and the 2100) documents this:
http://www.photo-i.co.uk (See Page 11)

--
I plan on living forever - so far so good!

Frank Nichols · Jul 9, 2002

The color difference is my sloppy adjustment to the brightness after scanning, my scanner output is always "dark". The prints are virtually Identical.

No, I used NO special profiles, only the system level profiles provided by the print driver, and I made no adjustments on either system - QImage on Windows and IViewMeida Pro on Mac.

On the Epson paper I see more of a moteled or faint orange peel surface, not lines.

And I can see these bands without magnification. The point is not to be critical of the Windows printout, but to point out that the issue maybe a difference in the software/implemetation of the dither pattern used on the two machnies. This could explain a lot of differing results people are posting on here. It also gives encouragement that possibly Cannon could make changes to their drivers to reduce the effect some people are seeing.

fn

dgrogers said:
The first thing that's obvious to me is the difference in color,
are you using any profiles? Different colors and shades may make
any banding more obvious, however, since you have to blow up the
print so large or view it through an 8x loupe to see the banding, I
don't understand what the problem is. Another thing the "banding"
may be is the natural ruts in the paper -- which may give the
appearance of banding and grain. All coated paper has some kind of
valleys, ruts, or pattern -- is the banding you see going along
those imperfections in the paper (hold the print at an angle under
a light source to see these imperfections, they are there).

--
dgrogers

http://www.pbase.com/drog

--
I plan on living forever - so far so good!

Kiran P · Jul 9, 2002

I completely agree with all of this. I have noticed that I get different results when printing with different profiles if you believe that! I find that on Office depot paper, when I use the standard canon profiles and PPPG paper setting, I get more banding then when I use the OD profiles from Joe I think using the same PPPG paper setting. I was stunned to discover this but I see it over and over again. Its amazing how when printing on Canon PPP, all of this completely dissapears though. I too am very very picky and I have excellent vision and therefore can see the banding. I therefore do think that this is a software issue and Canon should be able to correct it with another driver update. Lets cross our fingers!

Kiran

ps. I tried to do some scans but only having a max of 600dpi, they just didn't turn out at all. You had to try hard to see it in the scan. It was almost easier to see it with just my own eyes!

Donald Cooper · Jul 9, 2002

Although inkjets use multi-levels, this doesn't give them enough greyscales, so they have to use a dither. The dither consists of rectangular tiles, where each block in the tile is divided into several sections.

Since the inkjet uses mutiple levels, the head has to put the ink down seveal times on the same tile. But the head is not perfectly positioned for each pass, tending to displace a small amout in the same direction. This displacement results in the tiles becoming elliptical, running into their neighbor in one direction before the other.

As a result you get artifacts where you have a regular pattern of something that looks like lines. Is this what you're talking about?

You can cut this down if the head only prints in one direction. But it's more or less endemic, until the tiles go bye-bye, which may happen with the new Epson inkjets that do not use variable drops at the highest resolution. I'm not sure.

Personally they don't bother me. I can't see them.

As far as the difference in the drivers, different drivers will give you different dither patterns. So it's not totally surprising that you get small differences in the output between OSs.

Frank Nichols said:
The banding, to me, is VERY obvious in the Windows example and just
visible in the Mac output (it looks worse in the scan than under a
loupe). Remember this is the equivalent of looking at the picture
with an 8X loupe.

Comments???

Frank Eersels · Jul 10, 2002

WillieB said:
Interesting comparison, this. A few questions, if you have time.
Is the banding in the prints vertical? Can you see them with the
naked eye? Also, I'm sure you have noticed the difference in color
between the prints - the "Mac print" being darker. Have you tried
darkening the "Windows print" before printing to see what effect it
has on banding?
--
WillieB

Mine went back because, in my case, the banding was very bad in light area's like sky's.

I got a call last week to pick it back up. The banding was gone (they replaced the printhead) but something went wrong during transport (I think). The head touched the first few cm's of the paper, leaving traces of blank ink on the paper. So, no banding, instead I got black ink marks on my prints.

Called up Canon again and I brought it back for service last week... still haven't heard a word from Canon.

Frank :-(

WillieB38717 · Jul 10, 2002

Frank Nichols said:
1. The banding is in the direction of the print head movement
across the paper.

2. Well, at first I couldn't - my 52 year old eyes need some help.
My friend can see them and pointed them out. Now if I catch the
light just right I can see the banding.

3. The difference in color is my sloppy adjustment of the scan
brightness. My scanner scans "dark" and I lighten them in
photoshop. The prints themselves are virturally identical in color
and brightness.

However, I think I will try changing the brightness alittle each
way and printing some tests - it would be interesting if the
banding has a very narrow zone of brightness and you can "save" a
picture by a slight change in brightness.

fn

Heck, my 52 year old eyes need help as well! Must be a naturally occuring phenomenon.

In your original post, you mention that driver settings are the same (paper choice). Does that include the quality slider bar? Also, have you run a nozzle test pattern and printhead alignment with each driver, on good paper? Are all nozzles present and accounted for with each driver, and are the alignment patterns similar? Using good paper for these tests helps, I think.

One observation here: when I get banding on my Canon 8200, the bands are separated by 16/600th's of an inch. This is exactly equal to the distance the paper advances with each pass of the (my) printhead. Your bands appear to be 32/600th's of an inch apart, and I'm wondering if this the "stroke distance" on your printer? Why the difference between drivers, though....don't know.
--
WillieB

Adam2 · Jul 10, 2002

Interesting. On my prints, no banding is detectable to the naked eye with PPP, Epson Glossy, HWM, Archival HWM, and Epson Premium Glossy. Granted the surfaces have different textures and thus yield different visual effects but the banding just doesn't exist. I have given sample prints to two professional photographers (one a commercial and the other a U professor) for comments. Neither detected any banding even after prompting.

If I loupe the prints some banding is there but so what? Out of curiousity are you using the latest drivers? If it is not a a print head problem, paper handling issue, paper or ink compatibility problem, then perhaps it is your driver settings.

Frank Nichols said:
fn

Adam2 said:

Yup, its there under 8x magnification and therein lies the point.
Try looking it at 1x magnification. Do you see it? I don't. If
you enlarge prints from the 2100/2200 you will see it too, but in
the horizontal direction. As far as I am aware every inkjet
produces banding to some degree.

A comparison of the printers (s9000 and the 2100) documents this:
http://www.photo-i.co.uk (See Page 11)

Click to expand...

--
I plan on living forever - so far so good!

Frank Nichols · Jul 10, 2002

up, like i said in the original post, all drivers are the latest from Canons website. And the fact that one computer produces different results from the other using the same printer, seems to say ithas to be the driver.

fn

Adam2 said:
If I loupe the prints some banding is there but so what? Out of
curiousity are you using the latest drivers? If it is not a a
print head problem, paper handling issue, paper or ink
compatibility problem, then perhaps it is your driver settings.

Frank Nichols said:

fn

Adam2 said:

Yup, its there under 8x magnification and therein lies the point.
Try looking it at 1x magnification. Do you see it? I don't. If
you enlarge prints from the 2100/2200 you will see it too, but in
the horizontal direction. As far as I am aware every inkjet
produces banding to some degree.

A comparison of the printers (s9000 and the 2100) documents this:
http://www.photo-i.co.uk (See Page 11)

Click to expand...

--
I plan on living forever - so far so good!

Click to expand...

--
I plan on living forever - so far so good!

Search

S9000 Microbanding example scans

Frank Nichols

Senior Member

pdqgp

Community Leader

WillieB38717

Leading Member

Adam2

New member

stucole

Well-known member

dgrogers

Veteran Member

Frank Nichols

Senior Member

Frank Nichols

Senior Member

Frank Nichols

Senior Member

Kiran P

Senior Member

Donald Cooper

Guest

Frank Eersels

New member

WillieB38717

Leading Member

Adam2

New member

Frank Nichols

Senior Member

Keyboard shortcuts