Home/Support/

Forums

19 Replies Last post: Feb 28, 2009 6:31 PM by (joe_row)  
User 2 posts since
Jan 30, 2009
Currently Being Moderated

Jan 30, 2009 2:30 AM

Gradient banding issues in CS3

CS3 has a gradient banding problem that no-one seems to be openly acknowledging or addressing.

If i make a graduated background in either Illustrator or Photoshop there is obvious banding in the gradient fill.

I have researched the problem on the net and it seems that many people are experiencing this problem but that there is no known solution or patch form Adobe to fix this. Only useless suggestions about work arounds for this simple exercise. Interestingly several people report never experiencing this problem until they upgraded from CS to CS3. I also fit this category.

I run a mixed Mac/PC network and CS3 produces fairly large background gradients perfectly in PS on my PC, though not in Il. My macs fail to produce gradients of any useful size without significant banding in both PS and Il.

I know that many other people have this issue - when is Adode going to publish something useful to address it?
Wade S Zimmerman User 3,166 posts since
Apr 30, 2006
Currently Being Moderated
1. Jan 30, 2009 5:32 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
In illustrator did you check I two places the Document Raster Effects Settings and in the Document Set Up the Transparency Settings are they high enough.

In CS 4 I do not see this and do not recall if it happens in CS 3.
Wade S Zimmerman User 3,166 posts since
Apr 30, 2006
Currently Being Moderated
2. Jan 30, 2009 6:16 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
I took another look and I would say yes there is a problem though AI is not intend to make such a large gradient there is banding on the Mac and in both CS 3 and less so in CS 4 but still a small amount.

You would be better making the gradient in PS if it is a background which does a much better job and using that as a background.

You can try an give it a gaussian blur as well.
John Danek User 433 posts since
Oct 21, 2002
Currently Being Moderated
3. Jan 30, 2009 8:06 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
I wonder if the "View" setting is an issue ( i.e., InDesign's low vs. high quality ), too. And, maybe "Anti-Aliasing" is set to "OFF"? Document and raster res should be 2540dpi/300ppi minimum, but not sure that could be affecting the view. How does the file print? Is the banding showing up in output ( to a fairly reasonable Postscript 3 RIP or high quality "Press" PDF )?
storm371 User 128 posts since
Aug 11, 2006
Currently Being Moderated
4. Jan 30, 2009 8:13 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
"Document and raster res should be 2540dpi/300ppi minimum" ?? You're not talking about the Document Raster Effects setting are you? As in "Other: 2540" in the dialog (versus High (300ppi), etc.)
steve fairbairn User 975 posts since
Jan 7, 2008
Currently Being Moderated
5. Jan 30, 2009 8:45 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
One thing that will cause banding is jpg compression, even though you have got good resolution. This happens both in Photoshop and Acrobat if you use jpg somewhere along the line. For seamless gradients use another either LZW tiff or no compression. Your files will be considerably heavier, but that's what you pay for smooth gradients.
John Danek User 433 posts since
Oct 21, 2002
Currently Being Moderated
6. Jan 30, 2009 8:08 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
Jim, yes, raster effects setting should match document resolution ( I prefer 5080 dpi, but usually set it at 2540dpi minimum ). "300" is way too low ( if you are referring to document resolution, not image resolution ) and will result in banding, similar to what is seen in non-Postscript desktop printer output.
JETalmage User 1,938 posts since
Aug 13, 2005
Currently Being Moderated
7. Jan 30, 2009 9:25 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
John,

Where do you get the idea that Document Raster Effects settings should match a "document resolution" of 5080 or 2540?

> "300" is way too low ( if you are referring to document resolution, not image resolution )

Document Raster Effects is a raster image resolution setting: it's the resolution of the raster images that are created by Illustrator's raster effect features. Specifying a raster resolution that is 1:1 with an imagesetter's dpi is ridiculous.

Further, where exactly is it that you "usually" set a "document resolution" in Illustrator CS3? The document resolution setting in earlier versions was simply the flatness setting, and defaulted to "800 dpi" (which is a flatness of 3, assuming a 2400 dpi imaging device). That setting is no longer even present in Illustrator, since (as I recall) AI 10. Flatness had nothing to do with grad fills; it affects curve accuracy.

You guys are talking about a problem with banding--the OP implying that he has been investigating the problem for some time--responders even offering suggestions to "fix" it. Yet no one in this thread has yet stated any of the specifics that are necessary to even talk about banding:

Where is the problematic banding occurring? Paddy Rogers does not even say whether he is talking about banding on screen or in output.

Assuming the banding is in printed output, what kind of device is it? If it's an imagesetter, is it using PostScript level 3?

What are the CMYK color values between stops of the of the page-size grad fill?

What is the size of the page (distance the grad spans)?

What is the halftone ruling being used?

These suggested "fixes" are myths:

Rasterizing the artwork
This is nonsense. It doesn't matter if the color values across a uniform grad are from vector commands or actual colors of pixel bands. Raster grads yield banded results just as easily as vector grads do. Banding is a function of the dot size and halftone ruling of a printing device. Rasterizing doesn't magically increase the resolution of a printing device.

Setting Raster Effects Resolution to a higher value
This has nothing to do with grad fills in Illustrator. A grad fill is not a raster effect.

Transparency Flattener Settings
Again, has nothing to do with banding in grad fills. A low setting may result in pixelation, not banding.

Antialiasing
Has nothing to do with banding in graduated fills.

Paddy, you claim that you are experiencing increased banding in both Photoshop CS3 and Illustrator CS3, as compared to the CS versions of those apps. Have you actually compared the same files being sent to the same printing device? If so, provide the specifics of a page-size grad that reliably produces the problematic banding. Be sure to state exactly what device you are printing to.

JET
Wade S Zimmerman User 3,166 posts since
Apr 30, 2006
Currently Being Moderated
8. Jan 30, 2009 9:59 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
I think they may be trying to print this as a proof and perhaps to an inkjet.

You could at one time print it through Acrobat and get a fairly good print on an inkjet. You have a much harder time now.

And that is the point. And the screen display of this has taken a bad turn somewhere along the line.

And the point again is that it was not always this way.
blue_vision User 9 posts since
Nov 14, 2008
Currently Being Moderated
9. Feb 26, 2009 8:13 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
any other words on a solution to this topic?

If we create a white/gray radial gradient in illustrator, save it as a pdf, open it in flexi sign software to send to a Roland Wide Format Injet Printer... we are getting major banding. Plus, in illustrator on screen I can see major banding if I look closely.

What should be our trouble shooting steps to look at if all of what James states there is correct? James? Anyone?

Thanks!
Wade S Zimmerman User 3,166 posts since
Apr 30, 2006
Currently Being Moderated
10. Feb 26, 2009 9:09 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
Regardless of what James wrote I think that you would need a rip type software but what I see is that what you we on screen is what you get when printing and in CS this was not present, I am told that the display and the printing to an inkjet from OS 10.4.11 then from 10.5 and higher are an issue I do not get the display quality or the prints from Illustrator that other users who have 10.4.11 are getting.

If you are using Leopard that might be the problem I am thinking of installing 10.4.11 on an external disk and boot from there just to print.

This is the best solution I can offer.

But I do not know your configuration I can only assume it is Leopard you are working with on perhaps a MacPro?
storm371 User 128 posts since
Aug 11, 2006
Currently Being Moderated
12. Feb 27, 2009 9:22 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
storm371 User 128 posts since
Aug 11, 2006
Currently Being Moderated
13. Feb 27, 2009 9:33 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
Wow! I'm a month late on returning to this thread, and I hate having to concur with JET 8^P but this is patently incorrect. Document Raster Effects Resolution should never be the imagesetter resolution.

If you follow the same logic then any raster document you create in Photoshop should also be imagesetter resolution. How about a process color Business Card file? 172MB for 8-Bit color (344MB for 16-Bit for due diligence).

Document Raster Res should be 300 dpi maximum up to 175 lpi output. Industry standard is 400dpi for 200 lpi though 300 will suffice in most instances as well.

People have to be very careful about the answers they give in public forums like this.

Stupid double post on a forum hang! 8^(
(jean_p) User 465 posts since
Aug 17, 2004
Currently Being Moderated
14. Feb 27, 2009 5:17 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
>The problems i am having are all concerned with on screen gradient banding.

This is probably due to your monitor (hardware), not to your software.
Wade S Zimmerman User 3,166 posts since
Apr 30, 2006
Currently Being Moderated
15. Feb 27, 2009 5:22 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
Not necessary jean p I have an old version of Illustrator (CS) loaded on an old G3 and that can display perfect gradients and lines
but CS 3 and 4 do not on a Intel Mac. But the same screen
John Danek User 433 posts since
Oct 21, 2002
Currently Being Moderated
16. Feb 27, 2009 7:51 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
Jim Neumann, I'm still in v10 where I can set my output res @ 2540 "dpi" and my gradient mesh res @ 300 "ppi". Raster effects, in the case of the document res ( 2540dpi )should match the gradient mesh res of 300 "ppi". Sorry for the confusion. I should have been more clear. I wasn't referring to the raster res of 2540dpi. When I said 300dpi was too low, I was referring to document resolution, not raster res. I believe CS versions have made it a little easier to specify raster effects as low, medium, and high ( low = 72ppi, medium = 150ppi, and high = 300ppi ), if I'm not mistaken.

Anyway, my apologies to you and JET for the confusion. It appears the issue has been narrowed down to the monitor and/or some sort of view setting.
storm371 User 128 posts since
Aug 11, 2006
Currently Being Moderated
17. Feb 27, 2009 9:29 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
I was wondering what kind of supercomputer you were running, John! We had someone accidentally create and distribute a series of templates with DocRasterRes at 2540 dpi and it was making our new Intel Macs craaaaawwwwlllllll It's all good!
JETalmage User 1,938 posts since
Aug 13, 2005
Currently Being Moderated
18. Feb 28, 2009 7:07 AM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
> Raster effects, in the case of the document res ( 2540dpi )should match the gradient mesh res of 300 "ppi".

I don't understand that statement.

The Docuement "resolution" you are setting to 2540 in older versions is nothing more than the Flatness setting. 2540 would be a flatness allowance of 1 printer spot on a 2540 spi (spots per inch) imaging device. It has to do with the rendering (faceting) of vector curves, not of raster images.

The setting you use for Raster Effects should simply be 1-to-2 times the halftone ruling. (And since the purpose of most Raster Effects is to make "soft",--i.e.; blurred-- effects, like drop shadows, it should usually be fine to lean toward the low end of that range--setting it to match the halftone ruling will suffice to avoid pixelation. You can't see the square shape of a pixel rendered by one round halftone dot.)

For example: A raster image rendered as a 120lpi halftone does not need a resolution of 300ppi to avoid pixelation, regardless of whether that halftone is printed by a 1200spi device or a 3000spi device.

JET
User 3 posts since
Jul 14, 2008
Currently Being Moderated
19. Feb 28, 2009 6:31 PM in response to: (Paddy_Rogers)
Re: Gradient banding issues in CS3
This happens both in Photoshop and Acrobat if you use jpg somewhere along the line.

More Like This

  • Retrieving data ...