Recently there have been several machines capable of applying ink, wax, or other media directly onto a screen's layer of emulsion to replace the positive. Basically everything we need from the positive without the film. More on this later.

Without waxing poetic about the past or pontificating too much lets look at the issue of positives, RIP software, and exposure.

Back many years ago when many of us were foolish enough to think that we could get away with "laser printer" positives, we tried to replace our far superior silver/chemical based film positives. Our camera shots, even the PMT shots had both darker "blacks" (a higher D-Max) and a clearer base film (D-Min) than vellum or the frosted film. We thought that we were saving time and money by shooting toner based positives quickly from a laser printer, we sprayed them with a repackaged artist-fixitive (something that always worked better than hair spray - it was the evaporative solvents that did the "melting" of the toner) and passing the quickly printed vellum to the screen room.

Of course this only cost the "screen room" time and effort - far more than we were saving in the "art room" we were soon forced to shoot paper positives with the old manual cameras to make reasonable positives. This did not last long, and as soon as we were able to justify to the ownership we needed an image-setter, life was good. We had more time for more artists to print more product. Production with excellent positives shot to the roof, and that was when a Harliquin RIP and a VLR image setter with chemical developer was a 48K investment.

In the industry there has always been the old argument over what was more important D-Min or D-Max. the REAL issue is CONTRAST it is not just the clear of the film or just the dark black but BOTH. We want the best of both, simply better is better, high quality always trumps mediocrity.

Discounting the need for a good D-max is showing a basic misunderstanding of how electromagnetic energy or "Light" works and how emulsion is exposed. Light waves never "slow down" - light is light and it either is blocked, bent, (refraction), or filtered. Light continues on - always the same speed - the speed of light. Reaction times change in emulsions based on the volume of light energy.

The lower the D-Max the less light is blocked (more accurately filtered) and the more electromagnetic energy passes through the dark areas. The dark areas of the positive control how much light is filtered, the lower the density/opacity the more light will pass through into the emulsion layer. Regardless of how powerful or fast the exposure the light always is the same speed and any light that passes into the dark areas of the positive and is not filtered continues on to the emulsion and starts the exposure reactions in the emulsion.

Here is an example of a poor black on a positive, even with a ridiculously thin coating of emulsion to compensate by shortening the exposure - it clearly did not work. A correct and full exposure has this and other negative results with a poor positive.

It is always in the facts, the microscope tells all, poor positive, poor results or you are forced to underexpose.

What does this have to do with a RIP?

Inkjet printers without a RIP in front will only deposit a particular pre-determined amount of ink on each dot from the piezo head, this is measured in picoliters (One millionth of a microlitre; 10-12 liters) a very small amount of ink, and that is always set at about 3 to 4 with the free software for the printer.

Epson printers have this setting infinitely variable via the information sent by the printer driver, in other words Epson left their inkjet printers open to performance modification by software.

To modify the inkjet ink deposit you have to have software that will change the ink deposit. The free driver supplied by the manufacturer is designed for ink deposit to give pleasing or accurate color and will be limited in ink deposit volume to about 4 picoliters per dot including the black inks.

With a RIP the printer is "told" to "squirt" enough ink to fulfill the demands of the user, and the RIP makes this much easier to adjust often with settings from 6 picoliters to 27 picoliters or more per machine dot printed - that is each machine printed dot not the formed "dots" you would adjust as "LPI/frequentcy" in your vector or raster program. Contrary to assumptions a RIP not only converts the image but tells the printer EXACTLY WHERE (so it holds registration) to put a dot and HOW MUCH INK to deposit.

With a thin, poor black as the dark areas (D-Max) the black becomes like the darker areas of an exposure calculator - not dark enough to block or filter all of the light energy and some of the emulsion is exposed behind the black, the more full and complete the exposure the more the open areas under the black are exposed leaving remnants after wash development and forcing the user to "blast" the open areas excessively leaving ragged edges.

With the poor covering of ink from the standard driver you do not create a black that will block sufficient light to keep the emulsion from forming a skin of slightly exposed emulsion over the face side UNDER THE BLACK. To then open the stencil we would have to put enough pressure on the open areas in wash development to "break" the edges of this "skin" off and that will result in an edge that is rough and will present a less than desirable printing "gasket edge" against the product.

Professionals use professional tools, using a RIP has other advantages, the biggest one in art production is using the graphic tools we have correctly. No real graphics professional wants to go back to breaking jobs up and fidgeting with photoshop to get dots or separations, directions for this procedure you can find on the internet for free, but the results are a poor D-Max and the resulting problems.

Postscript a beautiful thing, no serious artist is going to go back to the days before the introduction of Adobe separator (a program that is now imbedded inside of Illustrator and has been for over a decade).

Not using good quality film, quality inks, and a good adjustable RIP, a printer cannot get the real and required contrast from D-Min to D-Max to get positives that will allow professional results with full and complete exposure. The combination of special film for positives, ink designed for positives, and an adjustable RIP are the tripod of good positives with an inexpensive inkjet.

First, Stochastic is a bit of a play on words from the Greek into english - to aim, and to guess...

Basically to aim for a "general" result knowing that the end result will be slightly random and unpredictable.

1. Is a stochastic separation virtually the same as diffusion dither?

Almost, but even using first order stochastic dots with a good RIP there are better formations of dots.

Advanced RIP output often works by exchanging fixed grid square dots in a semi-random pattern for round dots semi-randomly placed in a similar fixed grid.

First order stochastic dots are considered primitive and problematic because of the secondary pattern.

The secondary pattern you see here in the red square...

Luckily for screen printers, using stochastic dots is FAR more forgiving of printing with mesh than with other transfer method printing procedures like offset printing.

2. What is the advantage of stochastic separations versus conventional elipitical dots?

On that, how about a bit on what is a conventional or stochastic halftone. Regardless of type "halftone dots" are an optical illusion, where the color is broken into small printable areas to create the illusion of "grey value" or lighter tints of the same color.

Conventional halftones are referred to as amplitude modulation or AM screening. This is an example...

Stochastic dots are known as frequency modulation or FM screening. This example is the more primitive "first order" stochastic screening, an example of the typical "diffusion dither" type.

The good and the bad...

Good--------------

a. Less chance of moire pattern

b. Less chance of frequency interference (moire type pattern on the screen caused by mesh pattern and the dots even before printing)

c. Less chance of moire type pattern forming from the weave of the garment and the dots printed

d. More forgiving in registration

e. More consistent in tone for longer into the print run

Bad--------------

a. Soft focus look (sometimes not a bad thing, like babies, women, flowers and nature prints)

b. Grain like visual effect to open and mid tone areas

c. Dot gain compensation is more difficult

d. Advanced Second Order, combination, and hybrid stochastic dots require expensive sophisticated RIP software

3. Some basic playing around appeared to have the stochastic seps spool/print much faster than the conventional. True?

If you mean that printing diffusion dither conversions from photoshop is faster - yes, the files are much smaller and take the RIP far less time to image because photoshop has already dictated the print placement. Stochastic screening generated from a RIP especially second order or hybrid takes longer to process because of the complex load on the RIP.

This is first and second order stochastic - first order is more primitive and second order is more random with random placement and random dots.

There are two forms of second order the first was random round dots outside of the typical grid and then, more advanced dots of random size AND random non-grid placement.

First order stochastic...

Second order advanced stochastic dots...

4. An article I found has me creating a layer for each chanel, converting to greyscale, then outputting as a bitmap at 150 with diffusion dither. Is this correct, or is there a better way?

You can adjust the dot formation DPI grid, but to form stochastic of the first order without a RIP programed to make stochastic you have to form them in photoshop as described above. You can also set photoshop to use patterns you pick, so you can play with the results, some will be horrid and some cool...

There is also a new hybrid screening technology where the dots change on the same positive from AM to FM screening and back, this is considered the most advanced technology. Hybrid technology uses each type of screening for its strongest features in the most advantageous positions.

Also many printers use a combination technique, where the yellow screen in four color process is stochastic and the other colors are AM screening - this allows the use of three of your best angles and discarding your worst performing angle of the four...

Other types of printing, like offset are working hard on this technology in a desperate attempt to gain an edge with image printing on flat goods like paper publishing. "Art publishing" where image of the print is the reason for the final consumer purchase, has been working with stochastic for many many years...

The first attempts at random pattern stochastic dots were with separations painstakingly sprayed by a skilled separation artist with an airbrush - long before computers!!!