Lenticular Graphics

Does anyone have any experience in Lenticular graphics industry? (You know, the 3D, flip and morph software)

I do, and I need some help regarding resolution and vertical banding in my prints.

Anybody who can talk the talk? please let me know ASAP.

thanksabunch

I don't know much about 3D, I worked a LONG time ago on a CAD system but thats about it.


Resolution/printing I know....
I don't know if it is something I can help you with, but I have spent the last 11 years studying the intricacies of resolution and how it relates to printing.

Ok, thanks - give this a shot. I'll try to explain as best as I can.

I'm working with sliced up TIFF files. 3D software cuts up multiple images to be seen & mounted underneath a lenticular lens. These lenses are supposed to be at certain intervals (15, 20, 30, 40 lpi) Unfortunately, they aren't ever die cut exactly from the factory. they typically come out of LPI testing at 20.06 or 40.085, etc. I need to set the software to adjust to that, (printing at intervals of 1440 dpi). What this means, unfortunately, is that I need to resample the output from the lenticular software, thus "moving" the lines. - this kills the accuracy.

basically, due to the resampling of the sliced up tiffs, I am seeing banding and ghosting in the images, since it is impossible to A) print at 400.8 dpi. 360.27 dpi. etc, and B) get an accurate 20, 30,40 line lens.

:confused: thanks for your time.

Doing advertising are you?


It really needs to be that precise to avoid latency?



I hope torn can help you.

What kind of RIP is processing these line screens?

I'm assuming it is running thru a RIP software.

I'm using Wasatch SoftRIP. 99% of what we do is advertising, this week, though - it's Christmas presents. Looooong story....

However, I can't RIP it. I have to build the file completely, and only use the RIP to send it to the printer, not to tweak sizing at all, like typical with RIP. Same reason I'm having trouble with resampling in Photoshop - I just can't screw with the pixels after the 3D app completes it. I've been told even slight color adjustments can cause "stray pixels" that causes ghosting and banding...that's bad. ( I think that's mostly BS, personally, but that gives you an idea of how tightwad this process is!)

Sucks, since these files can be well over a Gig sometimes.

-this post may belong in the "other graphics apps" dept. - sorry if that's the case.

Thanks for checking - I'm going to play with some settings in the RIP now, too - I have a few ideas now.

How large are the lenses? I keep getting visions of telescopes and glasses..

But these things can be hand held can the not?

Ok, now that I have a little background. (and have read up on the "softrip" I haave to ask. Why exactly do you have to downsample?

The RIP takes your data (resolution) and discards what it doesn't need. ALL you need to print without pixelization is a minimum of 1.414 multiplied by the line screen at which you are printing. (Some rips use more but you get what I'm saying.)

If you are downsampleing for sizing, I might suggest using the crop tool's fixed target size, put in your physical size requirements and leave the resolution field blank.



Like I said I don't know a thing about 3D

the only time I have to downsample is when my 3D file's resolution is output higher than my target resolution. EX: lets just say these files (via a complicated formula) need to be at 360 dpi to print at a clear image flip, well, the lens (which is the technical name for the grooved plexiglass that sits on top of the art, sorry for any confusion there.) measures out to be a 20.08 lpi. - this means to establish a clean, perfect flip, it requires an output of 360.32, for example. thus, the resample...

I understand resampling in photoshop pretty well, so I tried to downsample from the 360.08 to 360. (less damage, pixel movement, etc.)The results were all generally crap when sampled down - yet - if I had a file output to 320.48, for example, and sampled up - the image looked much better.

basically - the opposite of traditional sampling rules was the case: up = good, down = bad. ? blew my freakin' mind, I tell ya...;) Unfortunately, neither were perfect. only the exact lens LPI (20, 30, 15, 40) and no decimal lens measurements work correctly.

This has been a 9 month trial and error process, so I'm really not expecting to solve it all tonight- though I do appreciate your time and comments. I was kind of shooting in the dark to see if anyone was running this stuff. I only know of a handful in US and Canada doing it.

thanks again

the only time I have to downsample is when my 3D file's resolution is output higher than my target resolution

Thats what I'm saying.... If the image has MORE than the resolution needed...."so be it" the RIP will take care of it, or at least the RIP I run will. (I print 175 LPI @ 2540 DPI = my image resolutions are 248 DPI, if I send images at 300 DPI, they are still going to print 175 LPI @ 2540....does that make any sense?)


I would NOT resample the resolution of the image. If that RIP is looking for a 308.xx DPI and you give it a 600.xx DPI it will discard the data it doesn't need.