Moiré Issue

[acp]

I would use other choice words to replace "Issue" but for fear of being reported I will keep it clean. I am looking for insight, help, confirmation, anything that can resolve the moiré issue I am having. I consider there to be two types of moiré, pixel caused by the camera and pattern caused by fabric or some sort of texture in the image. I am having trouble with the latter. On my monitor (LaCie 526) at 1-to-1-pixel ratio and on our proofs (GMG with Epson 9880, no dotproof) there is no pattern evident. Our vendor, however, sees moiré in both places and I am being told that they are printing on an Epson and not using any form of dot proof. The image in question was sent out as a test to two other vendors returning two separate results – one sees the moiré in both places and the other does not. So, how do I see moiré issues on screen? The vendors are using basic Apple or equivalent LCD monitors so I do not understand how moiré can be seen unless they are using software that simulates screened bitmapped images or they are previewing ripped files. If the proofs are being made on an Epson or any inkjet without dot proofing can moirés be picked up? What can be causing two separate results at four different places? What am I doing wrong that I am not seeing moiré either on screen or in a proof? Is there some trick out there to fix moiré - other than standing on your head, moving the camera back and throwing it out of focus, squinting, rotating the camera 45 degrees, changing the lighting or any other suggestions I have heard. I apologize in advance for rambling but this has been very tiring.

 

[Luc St-Pierre]

Moiré....

Moiré....

Moiré happens exactly in the situations you just described. But moiré is the combination of the original file's pattern and the pattern of dots on the printing plate. Therefore, no monitor or proof other than dot-to-dot proofing will show you the real moiré that will appear on the printed result. If you suspect an image will create moiré pattern, have a screened proof done for you. It is propably the only way to avoid these “issues”. And remember that some images will always keep the bad moirés unless you climb up in linescreen or fool around forever with the screen angles. I have seen a situation where new fabric had to be cloned into a necktie because nothing could be done with the file.
Good luck.

 

[gordo]

RE: "one sees the moiré in both places and the other does not. So, how do I see moiré issues on screen?"

You might want to check that there is no moiré already in the image. This is caused most often when the subject has more detail than the resolution of the the digital camera, this results in lower frequency harmonics that appear as “waves” in your image which might look like moiré on the monitor. This subject to sensor moiré can often be made worse by many Bayer Pattern demosiacing routines. (Bayer pattern is the arrangement of red, green, blue photosites on a digital camera's sensor.) Moiré can also be introduced if a transparency is scanned.

thx, gordo

 

[Tech]

Hi ACP,
See mine struggles with moires here... http://printplanet.com/forums/prepress-workflow-discussion/15153-moire-problem-dcs-files

Lessons I learned:
1) Don't trust your monitor

2) Don't trust you office printers b/w, color lasers and certainly NOT even an Epson with rip

3) Find out what you can regarding original sources, if you don't have this info...

4) Dotproof doesn't meaning much...at least not for what we went though...we also conducted two proofing with two different vendors...both test shown moires

5) If your product's quality is critical and you want to take the uncertainty out of the equation, I suggest spending the money and run a wet proof...in the end this is what work for us when everything else failed.

 

[Lukas Engqvist]

If you get moiree from some vendors and not others you will need a full documentation of the circumstanses. Did they use the same lpi? Same raster pattern? Do they have a server sharpening images? Is there an InkLimiting technology?
In practical life look at the image in an image editor (PS) there may be one channel that has too much noise that is causing the moiree, blur channels individually to maintain sharpness while reducing noise.
Avoid jpg comression as jpg looks for repetitive patterns to use as a base for compression.

 

[Jim Sewell]

Another item to look for is whether the image is scaled and rotated in your layout program, and how different rips handle that. I've seen it work better to set rips like Prinergy to not resample the image even if effective resolution is over 1,000 DPI. That being said, I've also seen it work to scale everthing in Photoshop to 300 DPI final size, and then soften the pattern as much as possible without killing too much detail. Staccato screen will usually take care of it, but I've even seen moires in Staccato with over sharpened images with very high resolution. You often can see the patter on the plate and don't have to wait to see ink on paper.

 

[Tech]

I would suggest against going Photoshop softening route. If you images are meant to be sharp, don't compromise the quality because it doesn't remove moires, but rather reduce the optical effect. People whom don't know what they are looking for in the first place won't care about moires, but for those whom can see it will only get annoy by the soften and degraded images.

 

[acp]

Thank you, I feel much better knowing I am not loosing my mind. You all have confirmed what I have been thinking and trying to relay to those involve with this project. I understand the only way to see moiré on output is to have a dot proof or press proof made but I guess I am still trying to figure out how the two vendors are seeing moiré on screen and one says they have moiré on an Epson output with no dot proofing. I would love to have documented specs from the vendors but they are not being very forthcoming. I can only tell you the variables I know and what I am told. The image was shot with a Phase One H25 and is being placed at 100 percent without rotation in ID. However, the issue begins before the placement when two vendors tell me they see the moiré in PS and have not processed the image in any way. In addition, the one vendor sees it in an Epson proof. Again the issue is from the pattern in a shirt not the camera sensors.

Gordo the sensor issue you mention - would this be the same as “camera noise or pixel burst” that cause the rainbow effect in an image? I want to make sure I am talking about the same thing.

 

[Lukas Engqvist]

When softening noise in Photoshop, without loosing sharpness to blur and then imediatly "fade" so that you change the softening to "colour" will remove colour noise without affecting sharpness in luminance.
(fade and changiing blend mode to colour can also work if using despecle, median or slight gausian blur)