Staples doing 3D Printing In Store

[C855B]

Gosh - deja vu all over again. We were having these exact debates back in the early days of high(er)-resolution 2D imaging. 30 years ago. We (well, some members of the community) couldn't see beyond the limitations of the technology of the day.

The problem everyone is wringing their hands over is called "aliasing". Yes, Virginia, there is a point where increases in resolution overcome the perceptions of aliasing. No, you don't eliminate aliasing, but you can eventually get the minimum imaging unit - "pixel" or "voxel" or whatever - small enough to where you don't notice it in combination with any texture influence from the output medium. We aren't there yet with stereolithography, but it's getting closer. There will eventually be an equivalent of 1200 dpi technology, roughly the point where plain-paper laser printers started to replace laser + emulsion film methods in "everyday" pre-press.

For the couple of us here who know the old tech, current 3D technology is roughly where we were with the Linotype 300. That was new in 1984. But even that exalted pioneering machine's output wouldn't stand-up under a loupe. It was only when the technology advanced to variable laser "dot" sizes that image rendering approached "analog" methods. That took nearly 25 years before becoming affordable for smaller operators.

We are apparently waiting for the 3D tech to advance to the equivalent of variable dots. I see another 5-10 years before we'll be satisfied with the output.

BTW, the extrusion printers such as Makerbot are the 3D equivalent of dot matrix printers - affordable but coarse output not useful for serious prototyping. They're OK toys to practice on, but the limitations of the medium are insurmountable and they'll soon be relics in our personal technology museums.

[bbussey]

Just because a flat made arc can look like a curve it will never be a true curve . It will always be a faceted curve like shape which will be proven when thin enough lines are used and viewed close enough under magnification . It is a way to mathematically create parabolic curves to fit tangent lines ...

But if the faceted curve is fine enough that you can't see the chords, then what is the difference?

Even the cutting of injection-molding cavity molds with ballpoint milling bits leaves visible stepping if the intervals are not fine enough.  Just look at the roof of the MTL 12-1 heavyweight and the underside of any Kato passenger roof.

The RP technology does keep improving, both in fineness and pricing.  It won't be much longer before it becomes a practical affordable modeling method.

[peteski]

Although I agree with Lemosteam (after all he does know this stuff), I think that the outlook is not as grim as being portrayed regarding the printing resolution (or the hardware part of 3D printing).  Learning and using the 3D modeling software is sill another subject.

Going back to the resolution. As most of the digitized representation of the world, computers round (or square) thing off.  Take digital photography for instance. My first digital camera (CASIO QV-100) took photos at a native resolution of 320 x 480 pixels and 640 x 480 fine resolution mode!  That was 20 years ago. Nowadays, we have digital cameras with image sensors that exceed the grain resolution of the emulsion on a 35mm film!

But the bottom line is, that if you magnify the image high enough you'll see the square pixels that make up the image.  But  at normal viewing magnification, a 12 megapixel 8 x 10" photos for example look just as good as equivalent 35 mm film images.

Even the standard CYMK color printing process uses dots (pixels) in the printed images. Magnified enough you will see them.  However at the current printing resolutions (2400 dpi or higher) we happily accept the printouts and they look just as well as conventional photographs produced the old fashion way.

Same applies to 3D printing. Sometime in the future, the voxel resolution will become fine enough that the printed objects under normal viewing conditions will look as good as injection molded ones.

If the resolution is fine enough, the visible layering will be so fine that a coat of primer might make the surface perfectly smooth. Maybe the printers themselves will have some method of automatically coating the printed object with some sort of coating that will smooth them out. But currently, we are not there yet.  The RP parts with glossy smooth surfaces (right out of the printer) are still not achievable.

I know of at least one model kit manufacturer (Tameo) that designs some of the model's parts on a computer (which results in perfect fit), has them RP printed, then they smooth out the visible surfaces of the RP part which then they use as master patterns for making resin cast parts.  This type of manufacturing process greatly reduces the labor cost and time and the parts fit much better than if they were hand-build from scratch.

[up1950s]

But if the faceted curve is fine enough that you can't see the chords, then what is the difference?

None to me , I don't care if they use a beaver to make a curve , if it doesn't show after painting I am happy . If I have to sand or scrape in confined corners or other entrapped areas I am less than happy to have to do it , but having something I can get no other way will make me more or less happy to be less than happy . 

[PAL_Houston]

Appropos of the current discussion, this add just appeared at the bottom of my screen while I was viewing this thread on the Railwire:
http://www.goengineer.com/site/pages/mojo_3d_printer.htm?gclid=CNm7jKfC97MCFQKRPAodNEQAXw.

is this targeted advertising or what?
 

[Robbman]

It will make you rich ,  famous , appreciated , and tie up all your modeling time .

No, no, no and yes

[Lemosteam]

Appropos of the current discussion, this add just appeared at the bottom of my screen while I was viewing this thread on the Railwire:
http://www.goengineer.com/site/pages/mojo_3d_printer.htm?gclid=CNm7jKfC97MCFQKRPAodNEQAXw.

is this targeted advertising or what?
 

LOL!  I get those all the time!!!

[John]

^^^ This is why I love Railwire...

rocket scientists, meteorologists, oceanographers, geophysicists, EEs, and a bunch of good modelers .. what a combination

[John]

+1

[John]

Appropos of the current discussion, this add just appeared at the bottom of my screen while I was viewing this thread on the Railwire:
http://www.goengineer.com/site/pages/mojo_3d_printer.htm?gclid=CNm7jKfC97MCFQKRPAodNEQAXw.

is this targeted advertising or what?
 

Click on it so Tom gets the revenue .. that way we won't have to ask for donations

[John]

No, no, no and yes

For Rob .. Infamous 

[conrail98]

Click on it so Tom gets the revenue .. that way we won't have to ask for donations

I got excited when I saw $185 then I saw the "per month" and it went away. I thought you'd like that pic, btw,

Phil

[DKS]

rocket scientists, meteorologists, oceanographers, geophysicists, EEs, and a bunch of good modelers .. what a combination

Exactly! Why go anywhere else?

[Mark5]

I'll get excited when someone post a photo of something in N scale scale that looks like it was injection molded...

+1! 

I am following this stuff but will be convinced when the output is convincing.  This will be in the year 20__.

I have a bunch of the Mark4 High Nose kits, so the potential is certainly there for good stuff.

Mark

[Bremner]

rocket scientists, meteorologists, oceanographers, geophysicists, EEs, and a bunch of good modelers .. what a combination

don't forget server nerds....