Curiosity Round 2

[AdvancedArtist]

How do you look at color? How do you make decisions about separations? Perhaps some of you manual separation monsters can enlighten us. Lets compare notes 

[AdvancedArtist]

cricket sound

[nobrainsd]

Well, I'm partially colorblind and I look at an RGB screen and imagine what everything will look like when I print it discharged using my indexed seps... 

Provocative post attitude wins you attention, but that is about it. While I see your point regarding sepping into HSB color space you seem to want to stir it up. Hype or passion?

[starchild]

I think most take a mechanical approach to it... We just tilt our heads left and right to imagine how the color blends together on a shirt and use the tools provided in our program of choice to get the blend- the look.. We still got to focus on all the other variables after all.. (The customer, the emulsion, the screens, the registration, the print order, the shirts being delivered on time, the print looking good and on and so fought..)

So very few go that extra distance, nor have the time to go about understanding the scientific part of engineering the sep.. Maybe a graphic designer will. That's why our sep process gets better by coming on boards like these and getting the experience from the passionate few so we can add it to our repertoire.. It will still be a mechanical process in our list of todos.. We will just do it more efficiently..

[blue moon]

How do you look at color? How do you make decisions about separations? Perhaps some of you manual separation monsters can enlighten us. Lets compare notes 

don't do a lot of separating, but when I do, I am looking at the data and finding the best way to select the information that should be included in that particular channel. Understanding the color (ink) interaction guides the decision on how much of that data to include or exclude.

Coming from computer engineering background, I think like a programmer and use Photoshop as a tool that lets me manipulate the selection of the bits (or bytes) as necessary. Different options or selection tools are just different ways of filtering the recorder data.

pierre

[Sbrem]

I wouldn't call myself a monster separator, but I've done them since the '70's. Back then it was photo based film, shading film, rubylith. Then, we started designing by choosing a theme, shirt and ink colors, and would create a line drawing shot to film. Lay that on the light table, put tracing paper on top of that, and use pencil shading to create the textures, one color at a time, all the time considering how it should look. From there, it was halftone the shaded drawings with the camera and halftone screen to paper, touched up the paper, then shot those to film. Learning to use Photoshop later on was a lot less tedious. We would make duplicate files, one in RGB, and one in CMYK, which would give us 7 channels to start pulling pieces from and assigning colors in Multi-channel mode. Learned about Calculations and Apply image (still not sure of the actual difference) and it just got more intuitive with experience. Of course some images are flat graphics and don't need all that, and vectors are the way to go, separating while you build. I didn't watch the vid, sorry, no time right now, but later I will.

Steve

[Dottonedan]

How do you look at color? How do you make decisions about separations? Perhaps some of you manual separation monsters can enlighten us. Lets compare notes 

I use a hand-held wand ("Neutrino Wand" or particle thrower) connected to a backpack-sized particle accelerator. It fires a positively charged stream of protons that polarizes the negatively charged color and am able to them strip out the unwanted color by extracting the HBS obviously.

[Command-Z]

LOL @ Dot.


I invented a new color. I call it "Blenge." You can't see it unless you are a bird.

[Chadwick]

Dan's got some voodo.
 

I think in ink for seps, cause that's where they're headed at the end of the day.

[brandon]

Absolutely. I think of ink as well and proper mesh selection. Wow all the way

[Colin]

As most of the separators do that are here, we think as printers.

How is the ink going to react/print on the given shirt type and color.  Printing simprocess on a 6.1 standard t-shirt is much different than printing on a crappy promo weight shirt...  I have had to go back and over engineer base plates before because the shirt was so thin you couldn't create a smooth print with halftones.  5 shirts in and half your dots are still in the mesh...  Then you throw on a nicer shirt and it prints smooth as butter >.<

How opaque/transparent is the formula of the ink I will be using?  Does it shear well?  Do I have to adjust for dot gain more with these colors?  Less?  Will I NEED to throw in a flash on this design?  Will the shop I send it to need to throw in a few extra flashes because of their level of expertise?  Because of their screens and squeegee's?  Are we printing on 50/50?  Now we have to create our white plate different and prepare for dot gain a little different.

Those are some of the variables we think of.  Then we think about: is the design in gamut for plastisol/waterbase inks?  Can we recreate the full color spectrum in this image? Do I have to reduce colors for a 6-8 color press?  Will the color reduction be acceptable to the end customer?  And even more...

Last but not least:  WILL THE CUSTOMER BE HAPPY?

Also, I noticed in your tutorial with the superman logo, with your method you still have to make adjustments to your seps in order for it to print clean and smooth.  You may show that adjustment in a different tut, but I have not seen it.

I really like the idea you have of seping in different color modes.  But remember, in the end all seps need to be adjusted after the colors are generated.  I'm going to give your method a try on the next couple jobs I have coming up and see if it's smoother than what I have been doing.

Maybe I am talking out of my ass and my thoughts are backwards, but it's My 2 cents.

Dan, I want your proton pack... is it home made or can I get one on Amazon?

[ZooCity]

I look at it "old school".  I'm not old school at all, never had the necessity of separating with the old methods but I think the cumbersome nature of stat cameras, knife cut films, etc. really forces you to be methodical and hold the finished image, and how it will ultimately be achieved with your inks, screens, blades, machines, etc, on the unique substate, in your mind the entire time.  Yes, it's still costly now to make separation errors but it's all relative to our new tools.  I think in those basic terms when separating and try to act with conviction when working, even though I'm using PS to do it and a CMD+Z is all I need to undo it.  This is the part about separating I like most, it forces a certain state of mind.

Definitely into the HSB colorspace as it syncs with this mentality better than the others for textile printers- color + white level + black level -makes sense to me!  CMYK probably syncs better for those industries that use it predominately to print.  RGB is just a viewing colorspace, there as an adaptation to our primary visual when separating, the monitor.  The ultimate colorspace for us might be something like HSBO with the O relating to opacity of the ink on the particular garment or it's finished, printed, transparency.  Toss in an algorithm for the likelihood of a dot falling into the weave and being therefore lost in the final imprint and you'd have something pretty amazing there. 

At the end of the day, I want the method that lets me pull colors quickly and completely into channels where I can gauge what to do with the values I see there.  That's the key attraction to HSB.  Another guy has a youtube video on it that really demonstrates how many of the common methods used to pull color in PS (or Corel probably) don't get all the color data, they just can't.  That's been my #1 frustration with seps.  I'm always struggling to efficiently and completely pull color from the image.  #2 is the laborious nature of getting a somewhat accurate preview of the channels.

None of this should be so difficult but nobody has bothered to take one hot second of time and build some of these features into the software available.  It's like our industry is non-existent to outfits like adobe, epson, etc. yet I would be shocked if screen printers were not a significant portion of the market for all of these companies.   Maybe I'm wrong because, if we were significant, you'd think they would know and be catering, at least a little, to us.

[brandon]

That's been my #1 frustration with seps.  I'm always struggling to efficiently and completely pull color from the image.  #2 is the laborious nature of getting a somewhat accurate preview of the channels.

It's like our industry is non-existent to outfits like adobe, epson, etc. yet I would be shocked if screen printers were not a significant portion of the market for all of these companies.   Maybe I'm wrong because, if we were significant, you'd think they would know and be catering, at least a little, to us.

Same here, my number two problems with high end seps.

We did a simple one color discharge for Adobe a few months ago. They could not believe there was no hand to the print. One of the guys did not understand it. We tried to work a trade for software - that was out of the question!

[Rockers]

That's been my #1 frustration with seps.  I'm always struggling to efficiently and completely pull color from the image.  #2 is the laborious nature of getting a somewhat accurate preview of the channels.

It's like our industry is non-existent to outfits like adobe, epson, etc. yet I would be shocked if screen printers were not a significant portion of the market for all of these companies.   Maybe I'm wrong because, if we were significant, you'd think they would know and be catering, at least a little, to us.

Same here, my number two problems with high end seps.

We did a simple one color discharge for Adobe a few months ago. They could not believe there was no hand to the print. One of the guys did not understand it. We tried to work a trade for software - that was out of the question!

Astute Graphics Phantasm CS is a very good plug in for Illustrator with the print/screen print industry in mind. Check it out http://astutegraphics.com/products/phantasm/designer.html
As a matter of fact all their plugins are great.

[AdvancedArtist]

So far great replys and thanks... I will post some video in the next few days. In HSB there are 3 factors but as as been said ink.. density and well as for me I look at color in halftones and HSB pull back the S and B look at the Hue/Color to determine the dot density in my hue. Its all math in the end and the eye dropper. Oh the eye dropper what a tool it reveals everything in the end.