[Ken Styles]

Always have been curious why the art needs to be vectorized?

I always thought you just print out the color for each screen on a regular transfer paper. Burn each color on their own screen. Then print

The only reason I would see that it needs to be a vector would be to resize the image.

What did people do before these computer programs came out??

 

[frank white]

vector artwork is great for resizing. it also produces lossless artwork that isnt pixel based. type is sharper, edges are smoother.

i usually do all my artwork in illustrator, but my screen printer prefers it in photoshop as different channels for each color.

i usually start it in illustrator and then move it into photoshop.

before all of this, they shot film and made negs from that.

 

[identityburn]

As Frank said, when artwork is in vector format you don't have to worry about resolution, you can resize it to any size and lose no detail.

Say you created a logo in photoshop at 72 dpi at around 3 x 5 inches. If you wanted to blow that up to put on the front of a shirt you'd be in trouble because you'd lose way too much detail.

Also it is easier to do color separations in vector, and change them on the fly.

 

[Fluid]

Frank and Jason are correct. Vector is easier and is resolution independant. One bad thing with vector is there are a lot of effects that are easier to do in raster images.

Ex:
Creating realistic beveled chrome is easier in a raster program than vector. it can be dne in vector, yet will take 3-4 times the amount of time to render

 

[dennymclain]

So the deal would be to create, say, a three color design in Illustrator (or import it from a program like Streamline which creates vectorwork from scans). Add registration marks, then duplicate the file three times. For each duplicate, use only one of the layers, removing the other two.

Then open each file in whatever image editing software you might have, and drag the other two into one file, lining up the reg marks. Once all images are registered, dump the other two files, leaving one file with all three images perfectly aligned, ready to be burned to a CD.

I'm not sure about chrome beveling or effects. That might work better on heat transfers or Direct to garment printers.

 

[Fluid]

Your confused.

If your design was simple text with 2-clr outlines. you can easily create your seps from that.
Chrome beveling, like your mentioning, is more than likely going to be separated in Photoshop as channel separations. - this would be a simulated process separation/print.

Here is a visual for the manual vector program separating.
Add your crop marks.
duplicate the image for each color separation
change the plate color to balc and everything else to white.
Do this for all plates/colors in the design.
Print to film.
http://www.yourimagehost.com/is.php?i=98574&img=Separate.jpg

 

[Fluid]

Here are simulated/chrome separations. Top- color comp image and seps below. Seps were created in photoshop as channel separations.
http://www.yourimagehost.com/is.php?i=98577&img=SimulatedSEPS.jpg
The channels were printed to an imagesetter from Photoshop with necessary angles and line screen info inserted.

 

[jshade]

Does Adobe Photoshop CS2 have any file extensions that would make it vector? Or is the only way to get a screenprint done and everything is if its a vector?

 

[smutek]

Photoshop can produce spot color line art to rival anything produced by a vector program, but pathetically few photoshop uses understand the terms, let alone know how to implement them.

fred

I agree 100% that vector is not a pre-req for screen printing but, assuming the artist knows what they are doing (ie. doesn't rely on livetrace for everything), how exactly does a vector conversion hurt an image?

Also, maybe I misunderstand you, but if you are talking about solid spot color line art how and when exactly is a pixel going to rival a clean vector line?

It's not, ever.

Not if we are talking about things like line art and text, this is precisely the type of thing illustrator excels at. Why would I want to convert a nice, clean, crisp, scalable vector image into a pixel based raster image when I can just assign spot colors and output directly from illustrator?

Photoshop is a painting program, and I know how to sep from it, but I use it for effects that are not achievable or maybe are not practical to do in illustrator; an example would be chrome, as mentioned above, or an automobile, a person, or any other type of photorealistic or complex painting style. Then I will output a dcs .eps and add my text in illustrator.

Also, when a client submits a CMYK file to me I will re-draw it, 9 times out of 10, in illustrator. And 9 out of 10 it can be sep'd in illustrator as well. If it calls for photoshop I will color it in photoshop.

Requirements for my shop are either a vector image with all text converted to outlines, or a clean raster image sent to size, 300 dpi. In a perfect world I would get that all of the time, but most of the time I don't. Now, i don't know how other shops run, but I know my boss isn't going to turn down work or send any customer to another shop because someone sent over a low res jpeg or something with halftones, he's going to tell me to re-draw it and make it work. And that is exactly what I will do. And no image goes to press w/o customer approval.

They key, for me, is understanding and knowing when to use which tool.

All due respect, and again I may have misunderstood, but your statements don't make a lot of sense to me. Would you mind elaborating a bit?

 

[out da box]

Under normal circumstances, raster images created in photoshop at 300dpi wll be ALMOST equal to a vector image when PRINTED on a t-shirt, especially to the eye of a non-printer.
I tend to seperate everything in photoshop to channels and output to film from there. As long as the finished product looks good, whatever method works best for your shop- go with it.

 

[Sheepsalt]

when exactly is a pixel going to rival a clean vector line?

Jimmy, first off I agree with everything you said - vectors are easier to manipulate, especially when scaling. I'll take an illustrator file any day.

Still, if someone brings me a PS file I'm not going to redraw it in Illustrator unless it's some 72 dpi 300X300 little thing. I'll sep the colors & print the films right from PS. For spot colors it's easy to change them on the fly in PS by using color overlay - if the art is done right, each spot color should be on its own layer with no background, and color overlay works great. Still, that's only for preview or proofing because no matter what the spot color is the only thing that file is directly outputting to the film is black.

To answer that above question, when you've got a raster scaled to final print size at 300 dpi and you're going to burn that image into a 156-mesh screen for screen printing, a vector line and the raster are going to give you exactly the same result. Even a 305-mesh screen will have one pixel per mesh opening at that resolution. I'm sure that's what he meant by "rival" in the overall context, not that you could do everything with a raster that can be done with a vector.

 

[smutek]

Totally understand where you are coming from. Maybe it was the tone of the post that threw me off a bit, Fred's post seemed very absolute, and even slightly..... jaded? Perhaps it was the use of the terms
"to rival anything" and "pathetically few" that had me wondering.

 

[smutek]

Still, if someone brings me a PS file I'm not going to redraw it in Illustrator unless it's some 72 dpi 300X300 little thing.

Totally agree. Whatever is going to be most time efficient while yielding the proper results is what I am going to do.

 

[smutek]

Me too, I'd take about 10 or so minutes to trace it in illustrator and print it across the side of a blimp if you wanted, with no loss in quality and a file size weighing in under 1 meg.



Sorry man, but that graphic you posted isn't going to be superior to a vector graphic, especially not at 72 dpi.

 

[Sheepsalt]

I'll take 72 DPI, any day of the week!

A 300x300 dpi graphic at 72 DPI would print at 4.2"x4.2"

Are you saying that you couldn't print this 300x300 image good at 4.2"x4.2"??? Heck, i think i could take this to tabloid size and still make it look good!

Well, if it's 72dpi 300x300 & they want a 12" x 12" print on the front of a shirt, I'm probably going to trace it and re-do the type if I have the font. Even still, sometimes resampling it in PS with "bicubic smoother" up to 300dpi & scaling it up to print size works well enough on its own, depending on the image.

But, Fred, you are the image manipulation guru as far as I am concerned.

 

[midwaste]

[/quote]

Perfect example of an image that is well suited to livetrace.
Why does everybody knock livetrace?
The only reason that I can see is people who are adept at tracing(which I am not), and are pissed that livetrace made the process incredibly faster and easier. Just by playing with the settings, you can get a great, much better than expected result.

 

[smutek]

Perfect example of an image that is well suited to livetrace.
Why does everybody knock livetrace?

Because there are people, and I'm not necessarily saying you, who consider this (see attached file) to be an acceptable result, and the shirts (or business cards, or brochures, or whatever) go out the door like that because they didn't take the time to learn how to trace it themselves.

I knock live trace quite often, but I'm personally not pissed about anything. I think live trace is a great tool, both as a time saver and as a creative asset, but I also think there are many times it does not produce acceptable results, such as the image I've attached, and that it pays to just buck up and learn how to do it yourself.

 

[smutek]

What happened to his forehead just above the glasses??

What happened to the stars?

Why do the letters look like their melting???

When you said you could blow it up to fit on a blimp, i didn't realize that you would be needing the blimp to get the graphic far enough away that the tracing mistakes wouldn't be visible!

Yes, and what happened to reading a persons post..... sheesh...



The image I posted is in response to the post directly above mine and was posted to illustrate why it is not a good idea to rely live tracing....



Had you taken a second to read and less time thinking of witty Dali remarks you'd have seen that.

Didn't think I could have made that much more clear.



I'm out of here. Have fun with your game.

 

[midwaste]

OK, I should have said a **** 3x3 72dpi image off the web wasnt necessarily what I meant by good prospect for LT, but a low-color image on a white background.

I took the image and resized to 300, then adjusted the levels, per fred. (BTW, levels is a fun tool for other pictures, also..) Then traced the image and attached is what I came out with. Needs a little work still, a few points to be manipulated, and all the stars need to be replaced with actual stars, but that is to be expected. It took all of 8 minutes to play with the settings.

I guess the point I was making is, LT is a very powerful tool, and learning to manipulate LT means also learning simple point manipulation.

Cool image to mess with too, Colbert 08, indeed.

 

[midwaste]

At the point you had the file resized to 300 DPI, what specifically did the trace do to "improve" the image???

fred

Made it resizeable to "blimp" levels.