Printing Gradients - Fountain Steps

Hi, Im a Graphic Designer and I recently sent a Bar runner job to a printer.
The Grey linear Gradient background came back a bad Stripey and a greeny colour.
The printers reason for this was that as Im using a Mac the “Fountain steps” are only set to 256 and I need to increase them to 512 (which is apparently what they are on a PC).
I’ve never heard of this, Does anyone know how I can do this, or if this is even a valid reason.

Thank you

I half expect the printer is making some excuses. They’re not totally made up — gradient banding is always a real thing as are color density variations in those gradients.

However, I’ve never heard of gradient banding referred to as “fountain steps,” but, hey, terminology differs from place to place. Gradients are not smooth, continuous transitions. They’re a series of steps that, under the right conditions, can be seen. I’ve also never heard anyone cite a difference between the steps in a gradient between a Mac and a Windows machine. There are those here who know more about the technical end of them than I do, so I’ll let them explain. @PrintDriver, where are you

As for the greys heading off into green, any neutral grey composed of the four process colors is apt to skew off in one direction or another. A grey composed of, say, 10c, 10y, 10m, 10k should produce a nice (but lifeless) gray. However, all it takes is one of those inks printing slightly darker or lighter to skew the color off visibly in that direction.

In a 4-color process grey gradient, at each point in the gradient — especially at the very light color densities — you’re apt to get one of the colors stepping down a notch or getting darker or lighter at slightly different rates than the others throughout the gradient. This will end up causing an the unwanted rainbow effect that you described. For example, if a two percent magenta suddenly drops out between two and zero percent in the gradient, that portion of the gradient will appear greenish. in addition, the printer RIP might not get the gradient perfect across all four colors or the dots on the litho plate might disappear at light densities or the droplets of ink sprayed onto the paper in digital printing might not be exactly calibrated perfectly at all densities. There are a half dozen minor errors that can affect neutral 4-color process greys.

You really should avoid neutral greys in 4-color printing — not only in gradients, but in general. It’s much better to create a grey composed of different percentages of the process colors. That way, those one- or two-percent differences in ink density won’t make as big of a difference and won’t be nearly as noticeable. And asking for all the colors in 4-color process gradient to evenly thin themselves down to white, um, you’re asking for problems.

As for minimizing the banding that occurs in just about any gradient, there are ways to do that, like dithering and adding noise, but like I said, others here likely have better techniques and know more about it.

Thank you very much for your detailed response, I appreciate it.

What application are you using — a vector program, like Illustrator or InDesign, or a raster application, like Photoshop? Which versions?

Without going into detail, if you create gradients in a vector program and subsequently rasterize them without making the right adjustments to your application settings, it could accentuate the problem you experienced. For example, rasterizing transparent vector objects might be done somewhere between 72ppi and 300ppi depending on your settings. Again, though, there are others here with more technical knowledge than mine.

Hi file is created in Indesign & then output as a High res PDF (300dpi).
The PDF is then opened in photoshop and saved as a tiff.
Thank you

Why are you doing that? Sounds unnecessary just use the pdf

Yeah, that’s just what I was going to ask. There might be reasons for doing it that way, but I can’t think of very many. For type, especially, you don’t want it rasterized at 300ppi. You want it output at the maximum resolution of the output device, which might be ten times that.

They want me to open the PDF in photoshop and change to mode to RGB then when I save it as a tiff they want me to uncheck the RGB Colour profile box.

Like I mentioned, there might be situation-specific reasons for this, but it’s out of the ordinary. I have a feeling there’s more to this than meets the eye. I can understand why they don’t want a color profile associated with the art, but why they want the vector art converted to a 300ppi bitmap, doesn’t make a whole lot of sense without further information.

Even so, this doesn’t have much to do with the neutral grey color rainbowing or the banding.

By the way, it’s Pixels Per Inch (PPI), not Dots Per Inch (DPI). DPI another measurement for things like halftone dots or output resolutions, although lots of people use the term DPI when they really mean PPI.

Most RIPs strip out colour profiles anyway.

Vector gradients can be problematic - but haven’t seen anything like that in years.
It’s a grey linear bar - it should technically be in grayscale - not RGB.

The Tiff method was a favourite many years ago - so I can only hazard a guess they are using antiquated machines and technology that cannot handle modern workflows.

That being said - I wouldn’t do it 300 ppi - I’d do it 1200ppi.
And you can add a bit of gaussian blur to the gradient to distort the banding so it’s not noticed.

Stick to what they’re asking for - and if you’re not getting anywhere see if you can get a printed sample from a printers nearby you.

Need to know how it’s printed. I only know wide format digital and what happens there may be different from conventional print.

When using the gradient tool in Indesign, I cannot find in the documentation any reference to the number of steps or any way to change that. I suspect it’s the same as Illustrator and is limited to 256 steps, no matter the length. See this chart to see why wide format printers hate gradients in Illustrator:
If your gradient is over 8" long you could potentially have issues. Not always. Depends on the color difference and how the stops are built. We often suggest creating the gradient in Photoshop at the correct size and resolution as a linked image (adjustable on our end) and if you are working in scale, be darn sure your Raster Effects Settings are set appropriately.

Your printer may be referring to a color blend, which is different from the gradient tool. You can set the number of steps in a blend to well above 500.
Depending on the color mix, the printer rip may not have the capacity to interpret all of those 500 steps.

Neutral gray is an issue for all the reasons Just B stated. There is nothing worse than that rainbow striping as you get toward the low end of the numbers. That bright pink stripe you sometimes get, yeah…LOL.

We use “neutral gray” all the time. And it always takes a test print and it almost always takes some adjustment by eye. Where you are doing file gymnastics and ending up RGB and tif with an unknown print rip profile, I would say you get what you get. I can’t fix that without globally adjusting your tif image. I might get that gray correct, but the rest of it may be off by quite a bit. That’s why wide format asks for native files. So we can adjust important elements such as this.

So, what is your final print goal. What are you making? And why would a .tif with rastered text be acceptable to you? I mean, I print from tifs all the time and sometimes quite large (40’ x 16’ is about the largest we’ve done) but anything hand held? I’ll do it, but…wellllll…maybe you can’t see the quality difference, but I can.

1 Like

Is that a piece of carpet?

Oh. Didn’t see that. (actually I did and it didn’t register at first.)
That’s usually a temporary adhesive vinyl thing, or a nonskid back thing, these days with an antimicrobial overlam, that goes on the top of a bar where drinks are served.

And if they are using an RGB tif, they’re using a high volume gang printer to do it.
You get what you get.
Good luck.

I’m not 100% certain, but that seems to ring a faint bell from the old days when I was using CorelDRAW.

Weird term in any case.

1 Like

That’s exactly what it is. Fountain Fill is a Coreldraw term where the steps can indeed be specified.

It sort of rang a bell of a related sort with me too — not with CorelDraw but the technique of placing two different color inks in the same printing ink tray (or fountain). A different color on each end, so that they blend together in the middle on the printed paper. It’s a way of getting two colors out of one pass on a one-color press.

We used to do that occasionally when I had a summer job as a screen printer. Place a glob of ink on one side of the screen and another color on the other side, then run the squeegee across so they blend together as the ink is pressed through the screen stencil. That was called a “split fountain.” Sometimes we’d try it with three or more different inks, but after four or five passes, all the inks would all start to blend together into one big, uneven, ugly color.

I didn’t know if there were a set number of steps or whether the number was determined by the application, an inherent limitation in PostScript or by the output RIP. That’s useful information.

Assuming there are always a set number of steps in, say, an Illustrator gradient, that would mean banding would be most apparent when going from black to white as opposed to, say, going from a 60% to a 30% screen tint where the dot or ink density between the steps would be less.

See the link I posted. It explains the whole adobe gradient algorithm, with math to calculate the number of steps you’ll get.
Bearing in mind the max length in the chart is 7.7" step 4 of the calculation says this:
4. Using the number of steps calculated in step 3, see if the length of the gradient is larger than the relevant maximum length indicated in the next chart. If it is, reduce the length of the gradient or change the colors. (my italics.)
Makes me laugh and shake my head. LOL.

1 Like

InDesign definitely has no vector gradient as such. InD and Illy use smooth shading, which is PS Level 3 or higher. The rendering is done on output and is device-dependent.

On-screen or with non-PS devices, these are rendered with Adobes renderer.

But when you use a PS Level 3 or 1.4 and above, the shading is optimal for the device, which is device-dependent on the resolution and technology of the output device.

Sometimes some devices and drivers are geared as RGB devices. So you end up getting a CMYK/Grayscale image shoehorned through an RGB colour profile.

That’s probably where the green is coming from.

1 Like

Hmmm… but if you output to PDF using a best-guess profile then rasterize it to RGB tif? Good luck with that on any output device. That gradient is flat locked in there now with no way to adjust except globally.

I see Dov Isaac’s 2008 post on the Adobe forum regarding smooth shading. Adobe never seriously considers wide format in a lot of their thinking. We still have issues with banding, optimized or not, usually more so with certain light blue-to-white fades. We don’t have as many issues on the gray ones as previous but that is more due to the Caldera rip and better profiling control - and the availability of 16-bit processing - than it is to any “smooth shading” through postscript.

©2021 Graphic Design Forum | Contact | Legal | Twitter | Facebook