How do i get a smooth gradient?

[_Nelson_Maldonado_]

Hi all,

I have a vertical cymk background gradient: 0,0,0,0 -> 15,11,10,5.

I used the gradient feather on the object fill. It looks better on screen than it did with just a gradient although there is still some banding on screen - but it looks worse on print.

1. Is it just a function of the printer?
2. Am i using a gradient thats too narrow?
3. Should i use another app to create my gradients like Illustrator?

I run Vista 64 and I use a Epson 1400 inkjet printer.

[Jongware]

The banding may pop up even if you use Illustrator.

An alternative is to create an image in Photoshop. Draw the gradient and apply a blur until you don't see banding anymore. Save, place the image in ID. The banding disappears because blurring hides the continuous lines of a same colour, which is what your eyes perceive as discrete lines -- it's just perception, because calculated gradients are *exact*.

[_Nelson_Maldonado_]

Hello Jongware,

So i'm guessing that the "gradient feather" function in InDesign does not perform a "blur" in the same way that Photoshop's blur functions do?

If bluring produces a more uniformly sloped gradient instead of the more stepped gradient i see, why doesn't adobe just add the blur function to InDesign's fx?
>Quote: The banding disappears because blurring hides the continuous lines of a same color, which is what your eyes perceive as discrete lines -- it's just perception, because calculated gradients are *exact*.

Are you saying that banding that i see from calculated gradients created by InDesign and Illustrator produce are an illusion and that they are in fact smooth or a continuously linear slope gradients as apposed to a stepped slope?

Also, what do you mean by "exact" calculated gradients?

[_Nelson_Maldonado_]

Update...

Jongware, your idea of using Photoshop to create the gradient image was the answer. However, i did not need to apply a blur. The gradient image, after being placed into InDesign, looked remarkably smooth - no banding at all - on both the screen and print.

A note...
I first created a vertical gradient in Photoshop on a 1"x11.25" image strip that i placed into Indesign. It did produce a great vertical gradient until i stretched it horizontally to fill 17.25"x11.25". The horizontal stretching created a splochy stretch of color gradient.

So the only way to have a great gradient in InDesign, is to create it in Photoshop to the dimensions needed then place it into Indesign.

InDesign appears to do a great job of pulling elements together created from other applications to create a dynamic publishing documents, but it sucks at creating elements itself beyond solid color, transparent or semi transparent shapes.

[James Gifford—NitroPress]

There are different ways to create a gradient tone. Most gradient tools, in ID, Photoshop, Illustrator, etc. are very precise about it - every pixel's value and position are calculated by the gradient algorithm. The upside is that the gradient is very precisely created, but the downside is that most printers don't have the color and density gradation to create such a perfectly rendered color slope - so you get banding between the printer's levels.

It can be especially pronounced with a gradient of such subtle shift as yours.

By blurring the result, you break up the perfect, precise alignment of the color values and distort their lines, meaning that the same printer can create what is "mathematically" a less precise result but to the eye is much smoother.

There are a number of sophisticated techniques used when a very smooth, subtle gradient is needed (say, for product packaging or wrapping paper), but they often involve multicolor ink printing, much like duotones.

[_Nelson_Maldonado_]

Ok... I don't mean to beat a dead horse but...

but how do you explain the a non-blured gradient created and printed in Photoshop on my epson 1400 and even placed into InDesign and printed on the same printer looks perfectly smooth where as the same gradient created in InDesign looks and prints really bad?

In any case, you are saying that the visually bad gradient created in InDesign and printed on my Epson 1400 will look fine when printed by a printing service using better digital printers or the offset printing method?

[Peter Spier]

Two things:

InDesign's native gradients are vectors, like Illustrator's, so they don't use pixels, they use steps. I believe a Postscript 3 device should, in theory, be able to render them smoothly, but the results will probably vary with the printer.

The gradient feather tool is not really the correct tool for making a gradient background. It's a transparency tool designed to create a fade from solid to transparent across an object. You almost certainly don't need to do that with a background, and you introduce an incredible amount of complexity into the file unnecessarily. In the same way, unless you need to have another object show through, use a tint, rather than transparency to lighten a color.

Peter

[Dov Isaacs]

InDesign does not use vector gradients. InDesign as well as Illustrator use what is referred to as smooth shading, a feature of PostScript language level 3 and PDF 1.4 (and higher) to represent a gradient. What is stored in the InDesign document and subsequently in either PostScript or PDF is a numerical representation of how the gradient is supposed to look but not any specific rendering of same.

When displayed on a screen or printed to a non-PostScript or non-PDF device, these smooth-shaded gradients are rendered using Adobe's internal core technology graphics renderer. When printed to a device natively supporting PostScript language level 3 or minimally PDF 1.4, the smooth-shaded gradient by PostScript/PDF definition is rendered in a manner optimal for the particular device based on the device's resolution and available technology. Often, Adobe's PostScript and PDF OEMs go to great lengths to optimize this rendering for their devices.

As Peter noted, the gradient feather tool is effectively irrelevant to the smoothness of a gradient. It simply adds a dimension of variable transparency. However, taking a solid color from 100% opacity to 0% opacity using standard blending over a paper background could have a similar effect to that of having a 100% opaque gradient going from 100% tint to 0% tint, but not necessarily optimized for the device in question.

Gradients are always difficult in the sense that one is trying to translate a mathematical concept of a continuous range of color values into the real world of discrete color values, device resolutions, plate or toner drum characteristics, ink or toner characteristics, paper characteristics, etc.

One should always do printed, paper proofs of particular gradients to evaluate how they really will render prior to committing to same for layouts.

- Dov

[scottperezfox]

Thanks for clearing up all the whys and wherefores about gradients. But what remains, clearly, is the desire of users to have smooth gradients in InDesign. So let me phrase it in the form of a feature request.

Please make InDesign produce smooth gradients.

We don't care how you do it, but folks like me want to export PDFs, PNGs, or to create paper prints of designs which include gradients, and we desire those gradients to be smooth and continuous. Perhaps this can be a setting: "use vector gradients instead of Postscript" (warning, this will increase file size) ... or something. Maybe it's an export option only to allow fast rendering while working.

We love InDesign. We really do love it. But discovering a feature that doesn't quite work as expected is always disheartening. Finding out that "it's a feature not a bug" is a slap in the face.

[BobLevine]

If you create them properly they should be fine. Add a little noise to

help.

[shootsLX]

Useful response, thanks.

[_Nelson_Maldonado_]

thanks you all, you've been a big help on this issue. I believe i understand now.

One more question:
Other than postscript data producing a much smaller file than a pixel based data file, Does a vector based gradient produce a better gradient than a pixel based gradient on a high end digital or offset printer able to interperate the data correctly?

BTW, i did not use the gradient feather tool to "make" the gradient, i used it to try to make the gradient smoother but it didn't work. In any case, i realize that the feather tool is there to feather edges and create a smooth transparency. It was not a smart move on my part.

It seems that Indesign or the printer driver should be able to interperate (convert gradients from postscript to pixel) if the document is going to a non-postscript device.

[Peter Spier]

Pixel-based gradients are more likely to be smoother because they can have as many "steps" as there are pixels, plus you can add noise to break up the linearity of the borders, but keep in mind that the eye is unable to distinguish small differences in ink density, and there really isn't that level of precision in printing technology, so you can get by with fewer changes than you might think, as long as the steps are close together in value.

[Peter Spier]

Thanks for the clarification on the vector aspect. I stand corrected.

[_Nelson_Maldonado_]

Mr. Isaacs, Thank you for jumping in to clarify the gradient concept.

Actually, i assumed that Spier was referring to postscript data representing the gradient and that in his use of the term "vector gradients", he was really referring to concept of "vector graphics" since PostScript and PDF page description languages use a vector graphics model.
>One should always do printed, paper proofs of particular gradients to evaluate how they really will render prior to committing to same for layouts.

If by that you mean getting a proof from the "printing service " you'd be using as apposed to an in house print - on say- an Epson 1400 that does not support postscript then - yup - i agree completely. Getting a proof off of the machine that will be producing your production document or a compatible "print service" digital proofing printer is always wise.
>A quote from Wikipedia: There are instances when working with vector tools and formats is best practice, and instances when working with raster tools and formats is best practice. There are times when both formats come together. An understanding of the advantages and limitations of each technology and the relationship between them is most likely to result in efficient and effective use of tools.

So, obviously, one should know the characteristics of the output device and graphic element models before creating one's graphic elements. In my case, proofing the document containing "postscript based smooth shading" on a non postscript device like the Epson 1400 does not work since Adobe's internal core technology graphics renderer can't represent the gradient correctly for that device.

However, if where going to produce the final document on my Epson 1400 non postscript printer, then i should be using a raster graphics model (pixel image) for my gradients which Adobe's internal core technology graphics renderer handles quite well.

[Dov Isaacs]

Actually, PostScript and PDF use an imaging model that includes both vector and raster. In fact, smooth-shaded gradients cause the RIP to generate a custom raster image on-the-fly and not any vector data (such as a bunch of adjacent polygons of progressively changing colors).

The problem of going to a device such as the Epson inkjets is that although Adobe applications can present beautiful smooth-shaded gradients as raster to those devices, those devices and their drivers present themselves to applications as RGB devices. The RGB data from Adobe applications is converted by the drivers and printers into CMYK with their own secret sauce over which neither Adobe nor the user has any real control or insight into.

Bottom line is that you should see very little if any difference between a smooth-shaded gradient or a Photoshop-generated blend for this purpose unless you purposely add some noise into the blend to mask any banding.

- Dov

[_Nelson_Maldonado_]

Dang those rgb printers and their secret sauce! I wonder how much i'd have to spend to get a cmyk printer?

Thanks so much, Mr. Isaacs, for your insight!