I just follow this tutorial on how to take a .pdf and convert it into a vector image so I could export it into SU.
I took the cover of a LG washing machine service manual that I had in .pdf as it had a nice LG logo on it. After doing a few quick steps as shown in the tutorial I soon was looking at the LG logo in SU.
At that point I could easily work with it and one of the first things I did was use the push/pull tool to make it 3d.
This seems to be so much better then importing a image and then tracing around things in the image.
I find this amazing and it seems like this would open up a lot of possibilities.
Being I had to do this using a .dxf file is this something I will only be able to do with SU Pro or are there other ways to go about doing it with SU Make?
I only have 7 days left of my 30 day trial of SU Pro, not that I’m counting or anything.
The ability to export pdf files to a CAD file format is useful in my workflow so I have adopted this for just about every project I undertake. I also rely heavily on the ability to export video files of the developed models. These features are only available in the Pro versions of SU which is a significant factor in my adherence to and appreciation for the software.
Pro Features vs. Make Features
Good to know.
So why when I take a pdf like this one and do that same as I did above the first time with the cover of a lg service manual I only get a blank page in SU.
22-172 Sears DDW18 Dishwasher.pdf (903.7 KB)
maybe this is why…
unable to map [c2] glyph name Euro encoded as 128 - using the PostScript code valu
**** This file had errors that were repaired or ignored.
**** The file was produced by:
**** >>>> <<<<
**** Please notify the author of the software that produced this
**** file that it does not conform to Adobe's published PDF
EDIT: and then I just opened it and it’s not a vector PDF, it’s just a raster images and text…
all you can do with raster PDF’s is use them as images in SU…
Ok, to get that pdf I used my Chrome browser to print the page as a pdf. Guess it don’t work like that?
if you open in Preview on your mac and zoom in, vector pdf’s will remain crisp…
I doubt this one was ever vectors…
If I understand what you did, no, that won’t work. The content displayed in a browser is rendered as an image by the browser, and save or print outputs a raster image of that screen regardless of what format you choose. If the original was actually a vector pdf, you would have to download it as such, not view and print it in the browser.
So when I use inkscape to convert the .pdf to .dxf the pdf has to already be a vector?
All inkscape is doing by me opening a .pdf and then “save as” .dxf is turning a vector pdf into a vector dxf?
Yes, that’s all it is doing.
I don’t know for sure since I’ve never used that particular scanner or its software, but I doubt it. A scanner is a raster device. It captures an image by repeatedly moving a sensor across the source image. So, it started with an image (not a vector model) and scanned it to a raster (also an image, not a vector model). If you think about it, such a device can not inherently capture vectors because it looks at only one tiny spot on the image at a time.
Software exists that examines a raster image, analyzes the pixels to look for patterns that seem to be edges, and outputs a vector representation of what it finds. But this is a sophisticated processing step usually seen in machine vision applications, and has significant limitations. One is that its precision is limited by the original raster image. In trying to make a single vector out of a series of “jaggies” it must seek a best approximation since the original is not really a line. It must make qualitative decisions about whether a “jag” is a raster effect or the vertex of two separate edges.
Hmm, guess I was dreaming that I would be able to trace the outline of a curvy object onto a piece of paper. Then run it through my scanner and import it into SU as a vector.
Not possible or I’m I going about it the wrong way?
for line art in a raster format you can use ‘potrace’ to extract vectors, I think Inkscape uses that under the hood…
it’s all a bit of a ‘black’ art and often using bezier tools in SU to manually trace will be quicker…
when scanning, I run VueScan rather than any scanners built in software and find using a black and white setting gives best results…
Sounds good, I’m going to check out VeuScan, thanks!
Indeed, Inkscape uses potrace beneath the covers for its Path->Trace Bitmap command. Here is their tutorial.
If you try this you will find that it requires some human guidance to choose the conversion method and set parameters to get best results. Depending on the source image, you may have to adjust things carefully to get the converter to generate single vectors corresponding to each line vs multiple vectors that trace the border of the line. Not exactly a fully-automated batch process, but maybe it will work for you.