Epson not showing posterization in proof

[zulbat]

I'm producing contract proofs for product catalogs with product shots and when I output Epson proofs they look great and gets approved. Once on press some images show strong posterization on press. I have to fix these using blur, noise after upping to 16 bit. Is there a way to catch these images at the proofing stage before plates are made? What could be the cause of this posterization in the customer supplied images. Out of hundreds it seems this occurs only in one or two images. The customer did tell me that one of these images was re-res-ed prior to being supplied to us.

 

[gordo]

I'm producing contract proofs for product catalogs with product shots and when I output Epson proofs they look great and gets approved. Once on press some images show strong posterization on press. I have to fix these using blur, noise after upping to 16 bit. Is there a way to catch these images at the proofing stage before plates are made? What could be the cause of this posterization in the customer supplied images. Out of hundreds it seems this occurs only in one or two images. The customer did tell me that one of these images was re-res-ed prior to being supplied to us.

If you can see the posterization/shade stepping in the original image viewed on screen then a better way to fix the problem is using Photoshop's "Spatter" filter rather than noise or blur. In images it's usually caused by extreme tone curve adjustments on 8 bit images - typically seen in large gradients - like blue skys/clear sky sunsets. Or just a long gradient or a gradient rasterized from an Illustrator image. Sometimes doing an "Equalize" image adjustment (and then reverting back) can exaggerate the visibility of the posterization making it easier to see and fix. To avoid the problem in images, the client should make such strong tonal adjustments with the image in 16 bit rather than 8. Then converting to 8 for final use. Or use the "Spatter" filter to fix.
If the posterization only appears on press but not in the original image then the problem might be the grey levels setting in the RIP or a poorly crafted dot gain compensation curve.

 

[Puch]

Proofs made on inkjet tends to mislead in case of gradients, as they're inherently noisy. You can't really judge the posterization of a fine gradient on an inkjet proof. If you can't see the steps on a calibrated monitor, then the error is in the CTP curves, as Gordo said.

 

[zulbat]

Gordo,
Thank you for the feedback. I have used the Spatter filter as well to fix banding and prefer Spatter to fix. This seems to be a much deeper degree of posterization. I did consider the dot gain characterization curves and ran without them to see if this changed the look. It did not. I will look at grey levels setting in the RIP and see what that yields. I'm just trying to get ahead of the posterization issue on press and try to see if there is a way the Epson could show this before plating. Although I think if anything the Epson hides this type of issue more then show it.

 

[zulbat]

I tend to think it's not press curves because there can be several images on the same sheet yet only one of the images is the issue.

 

[zulbat]

Tried re-ripping test files with GCR on and off with no change to ripped files. Normally we have it turned off.

 

[sejer_sdps]

What about embedded profiles in images being different or images being different color space? RIP may have trouble with certain images. Are there extra channels in the images that give you trouble? We have done inkjet proofs for 25 years, back to the Iris days, and I have never encountered a problem like that.

 

[Puch]

One more thing to try: select ZIP image compression when producing the PDF output. JPEG compression can increase (or create) posterization in images with fine gradients.