CompuServe Thread

James, I might be of some assitance here… after spending many hours with Pixel Pro producing 10 logos from BW scans… I've found some tricks that make the program very useful. First, import the monochrome bitmap into Pixel Pro with smoothing, beveling, etc OFF! You will then have an object with a point plotted in every corner of the outside pixels. It's from this thing that you will begin to hack away at the jaggies with ONE smoothing algorithm at a time. Next, pop up the smoothing requester and forget about the silly Small, Med and Large buttons. (I'm working from memory here so forgive the lack of detail.) The parameters in the middle of the requester are used for several smoothing algorithms at once. Try using only one algorithm at a time by setting the parms for the unused ones to zero or -1. To get the maximum quality – study carefully what each algorithm does and the parameters involved in the manual – the confusing part is trying to combine the effects! But we're not doing that here so things should be a little more streight forward. If the first method you chose doesn't work right, undo it and try different settings or a different algorithm – but never use more than one at once – if you don't, you won't know what the heck happened. Also (this is pretty important), only operate on selected points instead of all of the points (this means, of course, you'll have to select them). This comes in handy when you're using the angle difference algorithm. Deselect the points on corners – the run length will not be included here so sharp angles will not magically dissappear. Invariably some manual cleanup is neccessary – but I've yet to have to place individual points…. only select ones that shouldn't be there and delete them. Keep in mind you can use many of the smoothing algorithms to clean up small sections. For instance – resetting curves on letters. Key things to remember: Use only ONE algorithm at a time, rely heavily on the UNDO button, work in small sections until you are absolutely certain what each technique does. Oh, some more hints: If you move the object back and forth between LW and Pixel for editing – you're making it more difficult on yourself. Pixel seems to need pionts in a certain order. If you cut and paste points in LW modeler and go back to Pixel – you may find yourself in the twilight zone. I hope this helps. If you need further help… I'll try to repond within a few days or so… I don't get on CIS every day (I'm not rich). About beveling – now there's some real tricks! (but that's another story). I want to encourage you to persist in your attempt to understand the algorithms – The forthcoming upgrade will see Postscript importing so you can use another program (ProDraw maybe?) to trace your bitmaps and clean the B-splines and then import to LW or Pixel (but that upgrade won't be available for a few months according to Axiom). Study hard and have fun.

No, we don't handle EPS yet. And I suspect you'll be disappointed with any program that claims to be able to do that. To do that, you need to implement a full-blown PostScript interpreter. This is quite tough to do right.