Help needed! one angle halftone printing Am i using dot on dot or flamenco?

[Hugomartin]

 I use the one angle method for all my halftones... According to ryonet thats the flamenco method... I googled and googled and nothing... But found a thesis on one angled vs multiangled screen printing... Pretty cool but it says there is two methos one angled: dot on dot one angle points printing on top of each other vs flamenco one angle staggered with slight offsets in the vertical and horizontal dimensions... ryonet doesnt say anything about no slight offset dimensions... So may quiestion is what are this dimensions... and what techinque am i using Dot on dot or flamenco if i only use one angle...? Are the ryonet guys missing something? and maybe their technique is not really flamenco?

[jvanick]

for those interested in the article...

http://scholarworks.rit.edu/cgi/viewcontent.cgi?article=4813&context=theses

[blue moon]

one angle with dots on top of dots. And if anybody tells you that it does not work, we won two Golden Image awards and three or four ISS awards with it. So, quality of print will be more dependent on your other factors. . .

If you are using a RIP it will produce the dots that work for you.

pierre

[ZooCity]

Single angle works fine for our Sim pro work here.  Balanced seps with even coverage is key to this method. 

Spot color gradients are another story.

Sent from my SM-G900V using Tapatalk

[AdvancedArtist]

Halftones are an interesting topic so is flamenco which I am not a fan of by any means as it is known in this industry. years ago standing over the 10 color auto as an Art Director looking at prints from separations that we ordered from Mark Coudreys netseps which were just off. Why off? Was it netseps or my shop? Today I would say it was me because I did not understand at that time in my life.

Well...

Got dot gain?
Base peaking? More dot gain.
White highlight wiping out the highlights.
Wrong pigments or not exact pantone matches. Pantone is utterly insane by the way.
Dot on dot?
And so much more...

There are so many factors that can throw your prints off it is amazing but you have to look at all the factors organize them and deal with them.

I have no desire to start a pissing match I have dedicated to so much time to understanding all this it would blow you mind. Not months years in fact perhaps thousands of hours. Why because as an artist I used to do everything I could to avoid problems on press for the sake of production or profit.

I have the utmost respect for Pierre and his awards but even more for his openness. But do they judge awards based on prints or art? Based on original VS print? I do not think they do. All the shows I ever went to they never posted the original next to the print.

I have posted one of Blue Moons award winning prints close up next to the original. This is not to put down or insult Blue Moon or start a pissing match.

Question did the art win the award or did the print win the award?

Granted the photo is camera based on that the camera might push magenta to red and we can only take so much from what comes from a digital camera but.. well the evidence is there.

[Maxie]

This is a subject that can be argued about forever.
I am happy with results I get using 22.5 on all the screens.
Dot, no eclipse with 45-55 dpi.
I need to work on my screen system, (lamp, vacuum, emulsion) to improve the 55 results.
There is a good article In the blog section of the Seperation Studio website by Charlie Fancini.

[jvanick]

what a about the terminology? (stupid thought over coffee this morning)

where does the term 'flamenco' originate? 

[AdvancedArtist]

I wanted to get into all this last night but did not have the time. Looking at the attached image I set up a red to yellow gradient. The halftone printing method has the objective of simulating the orange from the red and yellow. Two types methods of halftone printing are displayed.

Simple dot on simple dot

Simple dot against simple inverted dot

Last I knew most RIPs had inverted dot options for their RIP output.

Dot next to inverted dot as we have analyzed it and used it has a couple of benefits.

The effect of one color stepping on the other is reduced in some cases significantly producing more accurate color simulation. Sub-straight or back ground color is more effectively covered or blocked out by the ink coverage produced by the inverted dot which reduces color contamination from the background or sub straight color. Ink build up can be reduced as colors/Inks are not printed on top of each other but next to each other. Especially in printing areas where more than two colors or inks are being printed in the same area. For example, a print area with, white base, red, yellow, black and white highlight which could be used to simulate a brown or a tan color on a dark garment. Here you would have 5 layers of ink. If the whites or black are inverted dot and the red or yellow are inverted dot the ink build up would be reduced closer to 3 layers of ink as the inks are printed next to each other as opposed to on top of each other.

The way we have approached this is by looking at the grayscale gradients in the color separations and determining where the gradients are blending and using inverted dots on the colors that are blending. For example, if you have a print a with red, orange and yellow and orange is blending into the red and yellow then I would use the inverted simple dot on the orange as it is blending into the other colors.  As with the attached image the yellow is an inverted simple dot.

I often think of this method as a sort of hybrid mix of index printing and AM halftone printing.

I have a close friend at the PHD level of color education who holds patents in half toning processes regarding dot gain etc. Interestingly he told me that inverted dot was at one time widely used in textile printing on various fabrics.

Off course dot on dot can work we have seen the results of it for years. But is it always the best option? At least for me I use inverted simple dot any time I am simulating color or have color blends. I also use it on the white base, white highlight and black when doing simulated process printing. I have been very happy with the results.

I am not saying this is the only way but it might be worth exploring if you are planning on printing color simulation or photographically. 

You can do this without RIP software in PS, Corel and other applications if you want to look at this option and experiment with it and half toning in general.

You can for example, with PhotoShop take something like the gradient in the attached image and separate that to grayscales. Then change your resolution to say 1440 DPI. Then convert the red separation with mode bitmap in PS and select the method of halftone screen. Set that to 45 LPI and an angle of 22 convert the red to a halftone screen. Then for the yellow go to adjustments and invert the greyscale separation and run the same PS bitmap halftone screen conversion on the yellow. Once that is completed go to adjustments and invert the yellow halftone screen you would then have your yellow as an inverted simple dot. Then copy and paste the results on top of each other in layers. After which you can convert the image back to RGB and you could change the color on the layers and have a more detailed look at how this process would work.

As a side note you always want to use high resolutions when half toning in graphics applications. Most inkjets can output 1440 DPI and this resolution provides sufficient greyscale data and resolution for good quality halftone conversions.

If you have the desire to do so once you are familiar with these conversions, you can look at and analyze half toning for yourself and even do some test prints and compare results.
If you do decide to do some test prints I cannot emphasize enough that you should do some dot gain testing and make yourself a custom dot gain curve. Dot gain really is an issue in some aspects of printing especially when there is high contrast in the print, such as white bases and white highlights against black or dark colors.

Finally, I am not trying to rock the boat just presenting some of the things I have come to understand in looking at color, halftones and screen printing.

[ZooCity]

AdvanceArtist do you know of any RIPs specifcally or for that matter pre-press software that allows you to draft your gradients and separations as normal but will then output interlocking dots to your films or cts? 

The manual steps one has to take in adobe suite previewing and outputting to any common RIP are many to get the interlocking dots and, for me at least as the sole separation artist here, the time needed can be prohibitive to the client's budget and the time constraints placed upon me to complete pre press work. As a printer/designer/artist/whatever I prefer the results I get from choosing a dot shape for each channel- circle, square or otherwise -and size that is best fit to the ink and substrate that is then interlocked for even coverage and controllable to some degree to compensate for excessive gain.  i.e., if I had time to manually bitmap everything at high res, using inverted areas of color as desired and punch everything out of each other then manually choke the dots back I would do it every time. 

I think all three common methods have their place- single angle, multi angle and interlocking.  But I think the approach of channel seps to stacked or partially stacked dots is used primarily because that's what our tools allow, it's what's expedient and it produces acceptable results.   

[AdvancedArtist]

I don't use any of the commercial RIPs for inkjects but inverted dot is basically the same thing as interlocking. I am not sure what the RIPs these day are offering or if they have an inverted dot option I would hope they would as it is a standard postscript halftone type. Actually covered in Adobes PostScript documentation see page 303 in the document below.

http://www.adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/PDF32000_2008.pdf

The last place I worked we printed to a Xante laser jet and it did have inverted dot options. However at that time I was just clueless about halftones and color separation and never used it.

Personally I prefer having access to the dots before I print them out. That is just me and there a few reasons for that such as being able to choke the actual halftone dots which I have found to work well in some printing circumstances. For example when a color is fading into the garment color on say a black tshirt then I choke that halftone dots where that color is fading into the garment to eliminate the white base peaking out in the area of the color fade.

I am sure someone here the forums would know if there are inverted dot options with currently popular industry RIPs.

[blue moon]

few things . . .
As far as the image, the color was changed from the original to produce higher impact. I have a version with proper colors and it does not have as much punch. Same was done with our first award winner and the artist liked our changes and agreed that it looked better.

When it comes to reverse dots, I would have to do some testing, but there are several issues that I don't have answers to. Dot inside the dot would only look as represented in the image if the inks were 100% opaque. As it is, the simulated process (and 4CP) inks are very translucent and when the yellow is printed on top of the red they would blend and produce orange.* If the inks were opaque and we were looking for our eye to interpret the color and join the yellow and red into overall perception of orange, this would work. Problem is, even the opaque ink systems (which suffer from reduced color spectrum) are really translucent to a large degree when printed through the fine mesh. Yellow in particular has very limited coverage ability even when printed through a 110 mesh. So the ink opacity is the first issue.
Next one (if we managed to get the 100% opaque dots) would be holding the registration. Anything that is off (dots and inverted dots overlapping area varying throughout the print) would now create a variance in tonal range that would come in splotchy areas (don't ask how I know about this). So for this method to work, registration would have to be dead on with every single dot. In all our years of printing and trying, we've managed to pull this off on one print only! Funny thing is, it was on the Dragon print used as an example.
creating an underbase for reverse dot image would be a nightmare. Rather than creating the halftone dot based on the final color of the (blended) ink it would have to be based on the actual color of the individual dot on top of that piece (another potential registration issue). For example, with orange, the ubase is probably in the 80-90%. So the original image is converted to grayscale and the gray value is the value of the underbase. Looking at the example with inverted dot, we would need a 50% under the red and 100% under the yellow. So the red part of the underbase would have to be generated from the red screen and the yellow part would come from the yellow screen. And what happens when you add the third color to the mix and you are trying to create a brown? Do you now go and remove the underbase from the areas where black will print on top? Does the black go in the inverted area or in the regular dot space?

What happens when the orange you are creating is 20% red and 50% yellow? What goes in the remaining empty space (30%)? What if it's 5% red and 20% yellow? Now three quarters of the space are exposed to the substrate!

Beauty of the HSB seps is that the black garment is utilized as the B component in the HSB and is actually contributing to creating the proper color. With the interlocking dot I am not really sure how this would work. . . How do you adjust the underbase percentage under the yellow and red? It would mean going back to the already complicated solution of calculating the ubase under the red and yellow dot from their perspective screens and making another adjustment. And if there was a 25% gap left after the 60% yellow and 15% red, what do you put in there? As both white and black will change the color perception.

And if this was to be used in automated color separations, we would be printing on only 75% of the surface. When separated by hand, yellow would be laid down at 100% and red would be phased in on top. This would provide a gradual and seamless blend with at least 100% coverage on the garment. (prints without 100% deposit often look "airy" and like there wasn't enough ink on the shirt).

I had a couple more, but can't remember of the top of my head. . .

pierre

*in this case the two oranges do not look the same. one on the right is significantly darker than the middle image. Orange in general has less red than yellow, so in this case the larger dot would have to be yellow and the smaller center dot would be red to get correct tonal value.

[AdvancedArtist]

These all good questions I will take some time in the next few days to prepare some more graphics and a detailed response. So far I have had nothing but great feed back from shops using this method in fact some have said some of the prints have very noticeable improvements over the same prints with dot on dot. This feed back came from shops that had repeat orders that were previously printed dot on dot but when do again with inverted dots the color blends printed better.

Briefly and I will prepare a graphic for this also.. if your red was 15% and yellow was say 30% then you have a 45% ink coverage coming from the red and yellow with inverted dot which would leave the correct percentage from the sub-straight or the remaining 55% which would usually be your saturation or white component. If you were printing dot on dot then the 30% dot would be you maximum ink coverage from either the red or the yellow leaving an incorrect 70% sub-straight color component. Of course your registration is not going to be dead on but I think you still end up with much closer color to the original color components via inverted simple dot.

[jvanick]

These all good questions I will take some time in the next few days to prepare some more graphics and a detailed response. So far I have had nothing but great feed back from shops using this method in fact some have said some of the prints have very noticeable improvements over the same prints with dot on dot. This feed back came from shops that had repeat orders that were previously printed dot on dot but when do again with inverted dots the color blends printed better.

Briefly and I will prepare a graphic for this also.. if your red was 15% and yellow was say 30% then you have a 45% ink coverage coming from the red and yellow with inverted dot which would leave the correct percentage from the sub-straight or the remaining 55% which would usually be your saturation or white component. If you were printing dot on dot then the 30% dot would be you maximum ink coverage from either the red or the yellow leaving an incorrect 70% sub-straight color component. Of course your registration is not going to be dead on but I think you still end up with much closer color to the original color components via inverted simple dot.

This is intriguing. .. would you be willing to possibly process/make a photoshop file with say 45 and 55 lpi bitmap style halftones in both techniques so we could play along at home?

Would be interesting to compare on our presses/techniques.  I'd certainly be willing to set up a few screens to try, and then post the results.

[blue moon]

These all good questions I will take some time in the next few days to prepare some more graphics and a detailed response. So far I have had nothing but great feed back from shops using this method in fact some have said some of the prints have very noticeable improvements over the same prints with dot on dot. This feed back came from shops that had repeat orders that were previously printed dot on dot but when do again with inverted dots the color blends printed better.

Briefly and I will prepare a graphic for this also.. if your red was 15% and yellow was say 30% then you have a 45% ink coverage coming from the red and yellow with inverted dot which would leave the correct percentage from the sub-straight or the remaining 55% which would usually be your saturation or white component. If you were printing dot on dot then the 30% dot would be you maximum ink coverage from either the red or the yellow leaving an incorrect 70% sub-straight color component. Of course your registration is not going to be dead on but I think you still end up with much closer color to the original color components via inverted simple dot.

agreed, there will be an open area in both cases. My point was that with HSB dot on dot, the underbase will fill in the open area to the correct level of B(rightness). In this case (orange) is close to 100% so the open space would be white and we'd get a lighter shade of Orange. With inverse dot, the calculations for the correct underbase seem to be pretty complex. Can you explain how would you generate the ubase for the inverse dot?

pierre

[blue moon]

These all good questions I will take some time in the next few days to prepare some more graphics and a detailed response. So far I have had nothing but great feed back from shops using this method in fact some have said some of the prints have very noticeable improvements over the same prints with dot on dot. This feed back came from shops that had repeat orders that were previously printed dot on dot but when do again with inverted dots the color blends printed better.

Briefly and I will prepare a graphic for this also.. if your red was 15% and yellow was say 30% then you have a 45% ink coverage coming from the red and yellow with inverted dot which would leave the correct percentage from the sub-straight or the remaining 55% which would usually be your saturation or white component. If you were printing dot on dot then the 30% dot would be you maximum ink coverage from either the red or the yellow leaving an incorrect 70% sub-straight color component. Of course your registration is not going to be dead on but I think you still end up with much closer color to the original color components via inverted simple dot.

agreed, there will be an open area in both cases. My point was that with HSB dot on dot, the underbase will fill in the open area to the correct level of B(rightness). In this case (orange) is close to 100% so the open space would be white and we'd get a lighter shade of Orange. With inverse dot, the calculations for the correct underbase seem to be pretty complex. Can you explain how would you generate the ubase for the inverse dot?

pierre

as I reread your post it seems we are both claiming our method provides the correct B in HSB. It will take some thinking and math to figure this one out.
I would still like to know how do you calculate the ubase though. . .

pierre