(2020-05-13, 19:16)Orion Pobursky Wrote: I had it planned but never started. You can go ahead. I have other parts I’m working on. Here’s my scan:

(2020-05-14, 0:35)N. W. Perry Wrote: Cool, thanks! Now I guess my next task is to decide how much to replicate the blotty-ness of the print…Wikimedia Commons has a very clean vector of the sign, but I'll have to thicken up those strokes and eliminate most of the negative spaces.

(It's going to be only a texmap version for now, so no help just yet for OMR purposes.)

(2020-05-14, 1:47)Orion Pobursky Wrote: This pattern is prolly perfect for using Rolf's img4dat.

(2020-05-14, 2:29)N. W. Perry Wrote: I will see whether I can get it working on my system; that's usually what holds me up. I can run most Windows programs under Wine, but sometimes I'm hindered by some fundamental but insurmountable difference.

This looks like it may work, though, especially once I have a clean vector image created. If not, of course anyone would be welcome to do the conversion if they have img4dat running smoothly.

(2020-05-14, 2:39)Orion Pobursky Wrote: I wouldn't bother. I just tried with my scan and results were unsatisfying.

(2020-05-14, 2:29)N. W. Perry Wrote: I will see whether I can get it working on my system; that's usually what holds me up. I can run most Windows programs under Wine, but sometimes I'm hindered by some fundamental but insurmountable difference.

This looks like it may work, though, especially once I have a clean vector image created. If not, of course anyone would be welcome to do the conversion if they have img4dat running smoothly.

(2020-05-14, 6:51)Vincent Messenet Wrote: I'm on Mac as well and run img4dat without any issue but I'm using Parallel. I never tried Wine so far.
I can also provide you the outcome if you send the image

(2020-05-14, 14:51)N. W. Perry Wrote: I am going to try it now, since I discovered/remembered that texmap does not always work with transparency, and this part comes in trans-clear. Just gotta whip up a quick template!

(2020-05-14, 15:49)Orion Pobursky Wrote: I might be able to get a better picture with my camera, tripod, and flash. The scan has weird shadows due to the trans clear nature of the part.

(2020-05-14, 17:26)N. W. Perry Wrote: I think I'm in good shape, since I was able to find a good source vector:
https://commons.wikimedia.org/wiki/File:Las_vegas_sign.svg

If anything, I could use a better sense of where the black printing is "supposed" to go (in my source photo, it's clearly offset from the matching gaps in the gold layer).

Here is my best guess so far:

(2020-05-14, 20:20)N. W. Perry Wrote: Clean-up was fairly quick with img4dat; it all depends on how fiddly you want to get, of course.

Texmapp'd and fallback'd version below; thoughts welcome. (Pattern code pending.)

(2020-05-14, 20:33)Orion Pobursky Wrote: I think that you should just get rid of the TEXMAP.

A couple of reasons:
- The gold ink doesn't show correctly with the TEXMAP image (primary reason)
- This pattern isn't complicated enough to warrant a TEXMAP

• It doesn't work in LDCad on the transparent tile
  
• It doesn't work in Studio either, so it's either use the fallback or generate a custom part just for rendering
  

(2020-05-14, 20:58)N. W. Perry Wrote: Also, should I try to optimize these triangles to quads? What's the simplest tool for that these days?

(2020-05-14, 20:58)N. W. Perry Wrote: I can think of two more reasons:

• It doesn't work in LDCad on the transparent tile
  
• It doesn't work in Studio either, so it's either use the fallback or generate a custom part just for rendering
  

Also, should I try to optimize these triangles to quads? What's the simplest tool for that these days?

(2020-05-15, 6:27)Vincent Messenet Wrote: Nice!
I agree with Orion the TEXMAP version is not necessary especially as your pattern is good enough.