|
|
Post by Greg on Apr 7, 2013 17:14:57 GMT -8
Looking around the internet for these pics or how to do it makes me think it must be some sort of black magic because the info is too technical or not available at all so I'm going to try to share what I have learned.
The easiest way to make one of these pics that I have learned so far is to open your pic in a program like Gimp, Gimp is what I am using in this example you should be able to cross it to other programs.
After I open a file the first thing I do is make several transparent layers on top of the picture.
Next use the selection tool (the wand) and pick a spot any spot this is just to show how it will work, after you pick the spot select your first transparent layer you will see the selection showing through now color it white or some other color OR pull a gradient with the colors selected for your gradient.
Now go to the next transparent layer and put a white color over the same thing you just did on the first layer but your going to adjust the opacity so you can see through and let the graident show up on the third layer.
Thats about it the rest is adjustments or blur and you need to do this until you get the whole picture done.
Just doing this fast step will show you how it's done the easy way and you'll see after a couple of times what you can actually do by usine this method.
Greg
|
|
|
|
Post by gullyfoyle on Apr 7, 2013 18:25:48 GMT -8
Looking at the supplied images and various other depth maps, it seems as if one of the layers is a sketch or copy of the lines. This layer looks to be rejoined after any color manipulation. The lines themselves are not blurred as you would get with Gaussian, they are definite and crisp. It makes me wonder if something like a 3d conversion using grayscale instead of red/blue would work. It's also easier to remove the color from an image and work with that than start with b&w. Screwing around with Photoshop filter Difference Cloud has done some nice grayscale stuff. Emboss works wonders if you are mainly looking at text, and extrude has a bevel function which will round out text. I think drop shadow does something similar. Then again any decent word processor has a design function allowing for the creation of special effects for flyers. If you can't save to an image format, you can do a screen cap and crop out what you don't need. From playing around with the color calibration charts it looks as if Yellow has a better contrast than White. I-Picture only reads light to dark, and often leaves untouched surface with the brighter white. There are a couple of programs which will turn a 2d photo into 3d mesh. The mesh can be made grayscale and the depth is already included. Haven't played with one yet. I think one is included with Artcam, turns a photo face into 3d. Really though all those grayscale depth maps are hand created and not converted pictures. More for the artist. As long as I'm rambling check out the Digital Juice fonts, www.digitaljuice.com/products/product_lines.asp?plid=10www.digitaljuice.com/products/products.asp?pid=944www.digitaljuice.com/products/products.asp?pid=1343www.digitaljuice.com/products/products.asp?pid=1343Some really cool stuff that looks to be on sale for $10. Make some truly interesting signs and placards. |
|
|
|
Post by Greg on Apr 8, 2013 1:42:34 GMT -8
You can put a gradient inside any text actually text is the easiest ones to do.
The process I outlined above will let you get those lines you talk about because it show through the upper layers.
When you put a gradient on your image and come back to a transparent layer and paint he layer white you can r3educe the opacity till you see what you want to see plus it smothes the gradient out. this is how you get the see through look.
|
|
|
|
Post by gullyfoyle on Apr 8, 2013 3:39:16 GMT -8
|
|
|
|
Post by Greg on Apr 8, 2013 9:37:58 GMT -8
All you have to do with text is to pull a gradient and use super sampling tickbox if if you have the black in the center just invert the color then add blur and you'll have nice slopped text .
|
|
|
|
Post by Greg on Apr 8, 2013 10:02:52 GMT -8
|
|
|
|
Post by gullyfoyle on Apr 8, 2013 11:02:05 GMT -8
Like I said we use image processing software for other than CNC. It's better to know too much than too little when you wish to apply to other kinds or processing, such as 3d printing or even just design for things like flyers or art works. But yes I do get involved when screwing around with images. Some pretty cool and fun things can be created. But they don't always translate well, or they look different but more interesting than what was expected. I can waste hours just applying and undoing filters and manipulations. I cut this the other day www.nipic.com/show/2/25/7779670k13677979.htmlIt was the first cut that caused me to realize the greater potential of the machine. Much more complex and detailed than most of the other cuts. Third best cut I've made after a Doctor Who pic and a railway car full of lumber. If I were interested in signage or placards the machine met the requirements out of the box with very little creativity necessary. I read that when the Romans switched from hand craving to a more automated type of carving the productivity increased but the quality decreased. Statues became more for the masses and less about art. With the correct amount of tweaking these can create art. Things which are original and unique and not just stuff translated to wood. Each different image, style, experiment takes me a step closer to the artistic nature that may be achieved with these machines. Sure we are limited by I-Picture much more than someone using Artcam or any of the Vectric software. We have no idea how something will look until it is cut, unlike the clear creation of the other design software. Even tossing a model on Makerware gives a clear idea of what should turn out. If it doesn't print correctly it is far simpler to narrow down the problem. That same clarity can't be achieved here. Just too many variables that run blind. How can you tell if a flawed cut is due to a bug in I-Picture or a a code error written in the firmware. Or even fluctuation in power during either the processing or the cutting. I suppose we could learn to read the generated code and work through it line by line to check. But that is pretty tedious and time consuming. Essentially we are beta testers. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 8, 2013 12:02:35 GMT -8
Here is simply an "A" created in MS Paint on the left, the one on the right modified in GIMP by only using gaussian blur. You can see that the letter has a rounded appearing surface and that the greys merge toward the background. Attachments:
|
|
|
|
Post by Greg on Apr 8, 2013 12:49:30 GMT -8
Gully if you notice that pic you link to is only 180K you get a better outcome from the full size file.
|
|
|
|
Post by gullyfoyle on Apr 8, 2013 13:10:39 GMT -8
|
|
|
|
Post by gullyfoyle on Apr 8, 2013 13:41:15 GMT -8
|
|
|
|
Post by Greg on Apr 8, 2013 17:53:20 GMT -8
|
|
|
|
Post by gullyfoyle on Apr 9, 2013 5:36:28 GMT -8
The idea is to make any cut as interesting, eye grabbing, attention getting, as possible.
Making the work "pop" as they say.
The competition is every ten year old who has started creating their own art from the time the parents allowed them to touch the home pc. How many times have you heard someone say " I had this computer problem, but the neighbors kid fixed it"?
To them it's like breathing.
|
|
|
|
Post by Greg on Apr 10, 2013 1:14:26 GMT -8
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 11, 2013 19:18:34 GMT -8
Greg, great find.... also a fellow I know that has a laser machine told me that there is a free program called draft sight, which is supposed to work to convert images into plain line drawings, it's what he uses on all of his work (which looks pretty good).I really have not tried it yet. The guy gets lots of business out of that laser, trophies and plaques for organizations which one order was for 140 of them. And I'm a selling one at a time.
|
|
