TSB
screen printing => Screen Making => Topic started by: mimosatexas on October 02, 2014, 02:20:44 PM
-
So I finally bought and received a more powerful loupe because I want to start improving some of the "little" things in my process and decided to look at my halftones on my film...and they SUCK. I'm honestly shocked I am even holding the dots on my screens. There has to be HUGE room for improvement though based on what I am seeing. A standard dot looks a lot like the attached image, complete with gaps and translucent areas. More ink lay down seems to bleed or make the problem around the edges worse, less means more gaps and translucency.
I'll definitely admit to being a relative newbie when it comes to optimizing my film output. I use stock epson inks and whatever version of WP film I am sourcing locally. I am not using a rip. Obviously this isn't ideal.
What are some QUICK steps I can use to improve my output on this particular printer. Are there specific inks and films (I know the other recent thread goes over some film options, but seems geared toward higher end printer models) that will immediately improve my output? What are the more advanced tips or tricks to really getting a higher quality dot? Do I just need to look into a better printer?
A lot of why I'm asking is I was about to buy a CIS for the 1400, but don't necessarily want to invest in that if it will still produce poor halftones.
-
Next time you come by we'll look at ours under the loupe so you can compare it with what you've got. We have the 1400 with Accurip.
-
I'm guessing it's because you don't have a rip. I'm assuming you are filling your halftones in illustrator?
I'm thinking if you get accurip you'd be fine.
*I only say Accurip because I think it's the most economical one. I could be wrong. But any rip software should give you the control you need.
Now Pierre and Dot Tone can get on technical in here.
-
That would be awesome. I thought I had it at least somewhat dialed in, but holy crap was this eye opening...
I make my halftones in photoshop using the bitmap method with a 600 or 1200 dpi file to prevent the jaggies (depends on LPI).
-
I've never liked Photoshop's halftones, always found them rough.
I'll second that it's likely not using a RIP.
GhostScript is free and if your printer is supported pretty easy to get setup.
-
You're right to use 1200 ppi as your Output Resolution when you convert to bitmap in photoshop, that's the cleanest those dots will come out, from the program. However, if you aren't using a RIP, then you probably don't have the control you need; spend the money, it'll come back in no time, you want to be printing shirts, not films over and over. Also, if you want cleaner dots once you get a RIP, just print from one channel, not all six, AccuRIP lets you control that. Also, print out a test film and have someone read the results with a densitometer so you can plug the readings into the RIP and it will control your shades better. Also, you can buy much less expensive ink from Cobra or MIS, don't buy it from Epson with their 10,000% markup...
Steve
-
All great points. I was going to buy the Cobra CIS with pigment inks so I can do inkjet transfers as well. Will that mess with my ability to get great films? Will a rip fix the issues inside the dot AND the jagged edges, or at least improve them? I was under the impression the rip primarily automated the creation of the halftones from the grayscale image, while the issue seems to be with the way to ink and film are actually "creating" the dot. The digital photoshop file with manually created halftones has very nice edges at that resolution.
Also, with a rip, will it affect the clean edges of text? Part of why I never got a rip was people complaining that accurip made text edges jagged while they were clean in the digital file. The same issue happens with manually created halftones in photoshop, but you can omit the text and halftone only what you need, them merge the layers cleanly in photoshop. I hope that makes sense...
-
RIPs also control how the ink goes down on the film. And a single channel will print once, 6 channels will pile up, and your edges lose something. My last 1400 was giving us trouble, but I kept switching channels, but eventually it was time for a new printer, so I picked up a 3800 for a great price that was hardly used, and AccuRIP had it in it's list of printers, so it was an easy change. As for the 1400, don't use it for both films and color; you can of course, but it would be better to have both dedicated to their particular jobs. And lastly, we didn't bother with a CISS, we just top of the ink every morning with a syringe. I would recommend detaching the waste tube from the waste pads and putting it into a waste bottle. Eventually, your printer will think it needs service, but you can get the Epson reset utility to reset it when that happens.
Steve
-
I use an Epson WF1100 with Cobra pigment inks, Inkjet Film film, and Ghost, and while I am far from highend, when I look through a loupe, my dots are pretty clean, my type is sharp, and my films are opaque.
-
Here's a phone shot through a loupe, but as it enlarges, the lack of focus becomes apparent.
And with a better RIP, it could only improve
-
We have the 1400 with Accurip.
I used to use an Epson 1800 with Accurip and i was able to hold 3% dots on a 270 mesh exposing with flo black light bulbs so yeah, it's not the printer.
If you can get one, a densitometer will allow Linearization of your output so you can be sure that 35% halftone you print, is a 35% dot.
-
I'm sure if you ask nicely, a certain moderator might take those readings for you...
Steve
-
Also, with a rip, will it affect the clean edges of text? Part of why I never got a rip was people complaining that accurip made text edges jagged while they were clean in the digital file. The same issue happens with manually created halftones in photoshop, but you can omit the text and halftone only what you need, them merge the layers cleanly in photoshop. I hope that makes sense...
It's all about knowing what the right input and output will produce... proper sampling technique before the RIP.
For example, in photoshop if I have a clean-text solid-pixel edge, no anti-aliasing already.... and it is only 300 DPI, with fades and gradients elsewhere... what I do is image-size upsample to 1200 DPI -- choosing the NEAREST NEIGHBOR (retain hard edges) setting, this does no anti-aliasing/blurring on edges that are already perfect.... this is how to prevent the "RIP" from creating halftone edges on what were clean edges. Also if you don't upsample to 1200 or 1440 etc then you'll have halftones that are not high-enough resolution to hold the shapes especially in the lighter percentages. Just clarifying that you don't have to run it as two separate pieces, it is just the way upsampling is performed from 300 DPI - or rasterizing artwork if it is high-resolution like vector already you would rasterize at 1200 DPI or 1440 but NO ANTI-ALIAS to retain the clean-edges so they translate just as they would going from vector postscript to a separate RIP.
I use an Epson 1430 without any modification, converting the black and white halftone or stochastic image to CMYK and threshold 100% for each channel, choosing print settings to give it the best result with the film. Personally for me I just don't use the "blind RIP" method, I always want to work with the halftones and preview them in color, trying to get the best press-simulation of the separations and films/ink etc before ever making films and screens. Working in dot-gain compensation that is different for different colors and meshes etc.
Was just testing some 45 LPI halftones on 86 mesh the other day, created a dot-response curve profile very easily by working with basic tools like using test patterns and scanning them, just using proper color-management and calibration logic in the workflow to make a compensated or linearized film output to test for the press-curve. The same settings for everything but switching one factor - like black ink or white ink, through an 86 mesh and it is going to be a difference on press... hitting white twice, another difference, another curve or type of curve. But it won't be that hard to build up the right calibration and print tests to figure out the information flow from computer to film to screens to press and have scenarios with the separation and RIP methods that will get me working as automated as possible while retaining the accuracy I need to be able to trust what I do on the computer will translate to press the way I need it to. Some of the response-curve has to do with the film and density, resolution / RIP method etc, some has to do with exposure type and timing, emulsion type and EOM, mesh count to a degree, etc... lots of variables but they are your friends and not your enemy, once you have a logical workflow and easy methods and tools to measure and sample, you can get very close to a streamlined WYSIWYG model (What you see is what you get) - and this is the purpose behind using computers to aid in the design and manufacture of products.
-
As I already mentioned, I create my halftones and upsample artwork properly to print using 1200 dpi and my digital file is fine...I also already use CMYK 100% and have optimized my settings within the default printer settings. The issue I mentioned occurs regardless of how clean your edges art when outputting halftones manually in photoshop. It WILL create jagged edges on text if you do not halftone separately from your clean hard lines, even if it is a small change.
The issue is not with my digital file, my workflow during artwork and output file creation, or anything to do with compensating for gain on press or during exposure, EOM, mesh count, etc. It is specifically with the dot as it is being created by the printer on my film.
-
Got ya. I didn't know there was a way around the cruddy dots photoshop makes for you.
Give Ghostscript a whirl and see if it's actually your printer or lack of RIP. Then you'll know which way to go.
FWIW FilmMaker is pretty badass.
-
I guess this is why I don't do my text in Photoshop; I like to do the image in Photoshop, save the channels, export as DCS, import into Illustrator and create the text there, always clean edges. The RIP converts the grayscale to the halftones and we can simply print the film. I have to read over Full Spectrum's post though when I have more time and check it out, knowledge is power after all. Also, I notice your mention of 86 mesh; we haven't used it in years because I discovered this 81 SDE from SAATI that will put down more white ink than you could ever need. It's the "S" thread of low mesh counts, check it out...
Steve
-
If your Rez is at 600-1200, that's all you need when doing manual halftones. Keep in mind tho, that the halftones comiing from Photoshop are not 100% true line screens to what you choose. A 55lpi will not be an exact 55lpi. That function is entirely based on resolution settings. If you choose 55lpi from a 100 rez file, it will be different from a 55lpi using 300 rez.
When I say "different" it might have 56-57 lines in it and not really 55.
It's true that like SBrem said, images in photoshop (because they are raster) will be less sharp and can have jagged edges. Still, at 600-1200, those little jaggies make do difference (or very very little). It would ahve ot be something pretty low (in resolution) in order for it to be visible or make a difference. Many choices we make in file prep pertaining to detail are overkill and also that goes to say we also underkill something. Like many don't use the right mesh to halftone ratio and live with (and feel comfortable with) using a 55lpi on a 110 mesh (for example). I see it a good bit of the time.
Here is an example of why the little jaggies are not so important.
You print small type on high mesh to hold detail, but you coat a high mesh thinner than the others physically.
Many coat it 1:1 with a shapr edge with a fast pull. Some a little heavier but with a sharp edge. That difference alone, contributes to wether you kick out a jaggie edge curve on type. So in print, you end up brining back some (mesh thread) jaggie edges from your clean vector type anyways. So the option and time used to save as dcs2 provides minimal benefits visually. Then, add to that the wet ink on the tee, smashed around a little in the stroke pull, is never an exact true shape of the small letters on top of the fabric threads. All of that combined, contributes to making the need for clean vector type null. AN extreme example of this is if you were to use a 110 mesh coated 1:1 and print fine type with it. Very jaggie. You can see mesh tracks in the print at times.
I love the cool factor of adding in vector with raster but at high resolutions only, it's benefits are still minimal at best. For those that use low Rez psd files, I can see using the dcs method as the lower the rez, the larger the jaggie.
For the issue of the digital film dots, it's like the original poster had indicated. You can bring in high quality files but if the setting and Rez of the printer is low, output will be low. A 300 dpi printer is not going to provide very good dots. You probably need 600 min these days to begin to compare to others.
Digital printers all depends on the process used and the settings options. For example, some have multiple passes and at different resolutions. A single pass is know for being able to put down finer detail. In theory this is correct since the more lay down, the more fill in you have. THink of shadow tones. THe more passes, the more fill in of those negative lines like a negative .5 line and also your negative shadow tone dots. Everything around it is being compounded. Your 95% dot starts to fill in and become more of a 98%.
I've seen those dots put down before like you describe in the art file image above. They are oblong shapes that seem to be placed randomly in one spot to create a dot (size) representation. That in and of itself (I say) is not a bad thing. The question should be, How do you get them to be more opaque? You should look at your settings and determine if you have the option of building in a few more passes and/or a different resolution. THis process is how Pierre was able to achieve his better dots.
-
Sorry for the messy posting. Typing fast on phone at the Dr. Office.
I've found over the years that (shape) of a dot is not at all as important as the size that actually gets put on the shirt. Similar to wether or not you have small jaggies on the edges, is the question of shape.
I would be the first to search out the cleanest, roundest, smoothest dot possible in the past. Pierre's current dots would not pass for me at one time.
By the time a beautiful round dot gets put onto a shirt, under a loupe, it's nothing but a blob with an undefined shape resulting from its process to get there.
An ugly, undefined rough edges squiggle shape (to start with) can be used with equal print results. It's not the shape of the dot that is important, but the % that it is to represent once on the garment. It's all in the compensation and calibration settings of any device.
The only concern should be ink lay down to make that dot. Is it opaque? If not, use a setting where you put down more ink, that might be by using more passes or more heads.
-
My main concern is with holding the smallest dots and holding dots in a consistent way. I think the issues I see with banding and loss when exposing are mostly due to the film having shitty dots. The exposure unit I am using is holding fantastic detail, but I am having weird issues with dots next to eachother not exposing the same despite being identical dots. Under a loupe it isn't an issue of mesh interference, so I assume it is light coming through the gaps in my messy films. Honestly when I get some more free time I am going to test more, and I am going to grab a few rips to test some stuff as well.
I appreciate all help and the info as always!
-
What kind of inks are you using?
some are more opaque than others. the more opaque or UV block is has, the more light you can expose with to compensate for any film cloudiness.
I was using this stuff, it's dye so may not work for you.
Jet Black Dye Based Ink in the liter bottle or you can scroll up and buy the pre-filled cart for the 1400, or just about any epson 'film' printer.
http://www.filmoutputsolutions.com/dyeink#bulk_1 (http://www.filmoutputsolutions.com/dyeink#bulk_1)
-
If your Rez is at 600-1200, that's all you need when doing manual halftones. Keep in mind tho, that the halftones comiing from Photoshop are not 100% true line screens to what you choose. A 55lpi will not be an exact 55lpi. That function is entirely based on resolution settings. If you choose 55lpi from a 100 rez file, it will be different from a 55lpi using 300 rez.
When I say "different" it might have 56-57 lines in it and not really 55.
It's true that like SBrem said, images in photoshop (because they are raster) will be less sharp and can have jagged edges. Still, at 600-1200, those little jaggies make do difference (or very very little). It would ahve ot be something pretty low (in resolution) in order for it to be visible or make a difference.
Let's clear something up, it seems like things are going into the territory of opinion or misguided understanding.
I'm trying to wrap my brain around what the missing piece is...
When you say "because they are raster" -- and I think you've said in another post before "clean vector"... Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
-
Raster is the term for an image made from dots.. ie photoshop.
This will print your images as dots and cause thinks like text to lose their sharp edges.
Vector is a mathematical line from point a to b. Ie illustrator.
Your images will print with clean lines or edges
-
Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I wonder if FSS is categorizing most of our output devices as printing with dots, and therefore printing raster, even if the file is vector.
-
If your Rez is at 600-1200, that's all you need when doing manual halftones. Keep in mind tho, that the halftones comiing from Photoshop are not 100% true line screens to what you choose. A 55lpi will not be an exact 55lpi. That function is entirely based on resolution settings. If you choose 55lpi from a 100 rez file, it will be different from a 55lpi using 300 rez.
When I say "different" it might have 56-57 lines in it and not really 55.
It's true that like SBrem said, images in photoshop (because they are raster) will be less sharp and can have jagged edges. Still, at 600-1200, those little jaggies make do difference (or very very little). It would ahve ot be something pretty low (in resolution) in order for it to be visible or make a difference.
Let's clear something up, it seems like things are going into the territory of opinion or misguided understanding.
I'm trying to wrap my brain around what the missing piece is...
When you say "because they are raster" -- and I think you've said in another post before "clean vector"... Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I will define/clarify my statements.
Files that are created in vector, are of course achieved by being plotted/drawn using math rather than physical pixels and resolution as Jsheridan had indicated.
The quality of output (from vector) is based on your quality of the imagining device. A high dpi/resolution printer will produce cleaner images. Cleaner being (edge jaggies) on curved type.
The quality of raster can be affected by the resolution of your file at output, no matter the output device.
You can have a 3600dpi wet film imagesetter to produce your film...and use a 200dpi halftones in photoshop...and they come out much jaggier via the manual halftones in photoshop (because of the resolution used at your starting point) when creating the manual halftones.
There are plenty of imaging devices that convert both vector and any good/high ppi raster file to it's own dpi "raster" quality at output as well.
Most any time I refer to vector, I refer to is as "clean" as it does present the best chance of having clean edges at output versus raster images, although, I myself have no issues using photoshop with clean edges even if it does have a tad of the jaggies. That is not to imply that raster is non clean. Here, I am specifically speaking of an isolated topic of edge cleanliness such as on the curve of the letter S.
Below, I gave my personal opinion (and past experiences) that are for some opinions, and for me as fact in the topic of using DCS2 files for vector output on type (below). I tho, don't proclaim that anyone should NOT use something they are comfortable with based on my own experiences or even what I believe as facts. The differences are minimal and can be proven, but I do not claim that people should not use what they are familiar with or comfortable with. For me, I find it to be yet another step that I do not need.
-
Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I wonder if FSS is categorizing most of our output devices as printing with dots, and therefore printing raster, even if the file is vector.
Right, exactly. Am I wrong to make this categorization?
Can you please show me an example of an output device that prints VECTOR?? Perhaps a vinyl-cutter-plotter... that is the type of device that may use vector instructions sent to the motors and print-head. Just about everything else prints with DOTS. I am raising the question because it seems like somehow you guys think you're outputting VECTOR onto your films??
The printer rasterizes the data whether you sent it raster at the same resolution as the print-engine will do it, or vector, it will turn into the same exact thing on film -- the only difference is if the in-device rasterization has some other factors besides resolution that can be measured and shown as a different/better quality factor in the process.
But the conversion of a "vector" to a shape on the film is almost always a Raster Image Process - conversion to Raster, at a specific resolution.
-
Raster is the term for an image made from dots.. ie photoshop.
This will print your images as dots and cause thinks like text to lose their sharp edges.
Vector is a mathematical line from point a to b. Ie illustrator.
Your images will print with clean lines or edges
"Raster is the term for an image made from dots.. ie photoshop." --- First part is ok...
"This will print your images as dots and cause thinks like text to lose their sharp edges." --- Has nothing to do with printing. This is where you're confused.
"Vector is a mathematical line from point a to b. Ie illustrator. " --- Ok, your definitions of Raster and Vector are not the best... a little too narrow, but in relation to graphics fine..
"Your images will print with clean lines or edges" --- again, this has nothing to do with printing. Your "images" will print with jagged lines or edges that follow the resolution of the RIP of the print device, or the vector lines/edges themselves, whichever resolves at the DPI of the device.
You are mixing up digital graphics with physical print machines.
I can output high resolution raster from photoshop at the same resolution as the print device will rasterize, and nothing changes, it produces the same exact CLEAN SMOOTH LINE on the film as the file sent as vector, because they both get RASTERIZED to the same points/dots. Only if the RIP in the print device is creating a measurably better digital image than the one you rasterize yourself will it have any chance of being "cleaner" on the film.
-
If your Rez is at 600-1200, that's all you need when doing manual halftones. Keep in mind tho, that the halftones comiing from Photoshop are not 100% true line screens to what you choose. A 55lpi will not be an exact 55lpi. That function is entirely based on resolution settings. If you choose 55lpi from a 100 rez file, it will be different from a 55lpi using 300 rez.
When I say "different" it might have 56-57 lines in it and not really 55.
It's true that like SBrem said, images in photoshop (because they are raster) will be less sharp and can have jagged edges. Still, at 600-1200, those little jaggies make do difference (or very very little). It would ahve ot be something pretty low (in resolution) in order for it to be visible or make a difference.
Let's clear something up, it seems like things are going into the territory of opinion or misguided understanding.
I'm trying to wrap my brain around what the missing piece is...
When you say "because they are raster" -- and I think you've said in another post before "clean vector"... Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I will define/clarify my statements.
Files that are created in vector, are of course achieved by being plotted/drawn using math rather than physical pixels and resolution as Jsheridan had indicated.
The quality of output (from vector) is based on your quality of the imagining device. A high dpi/resolution printer will produce cleaner images. Cleaner being (edge jaggies) on curved type.
The quality of raster can be affected by the resolution of your file at output, no matter the output device.
You can have a 3600dpi wet film imagesetter to produce your film...and use a 200dpi halftones in photoshop...and they come out much jaggier via the manual halftones in photoshop (because of the resolution used at your starting point) when creating the manual halftones.
There are plenty of imaging devices that convert both vector and any good/high ppi raster file to it's own dpi "raster" quality at output as well.
Most any time I refer to vector, I refer to is as "clean" as it does present the best chance of having clean edges at output versus raster images, although, I myself have no issues using photoshop with clean edges even if it does have a tad of the jaggies. That is not to imply that raster is non clean. Here, I am specifically speaking of an isolated topic of edge cleanliness such as on the curve of the letter S.
Below, I gave my personal opinion (and past experiences) that are for some opinions, and for me as fact in the topic of using DCS2 files for vector output on type (below). I tho, don't proclaim that anyone should NOT use something they are comfortable with based on my own experiences or even what I believe as facts. The differences are minimal and can be proven, but I do not claim that people should not use what they are familiar with or comfortable with. For me, I find it to be yet another step that I do not need.
"Files that are created in vector, are of course achieved by being plotted/drawn using math rather than physical pixels and resolution as Jsheridan had indicated."
---No they are not, if you're talking about the actual PRINT OUTPUT. Perhaps only in cases like a vinyl cutter-plotter is this even close to true. They aren't even displayed on the computer monitor without being rasterized. The math exists in the digital information... the image or physical-world thing exists when there is digital-to-analog conversion. On a computer monitor or a laser printer or inkjet printer or a cutter-plotter. Almost always there needs to be sampling, the creation of points that become signals to the devices... it is not "vector-printed".
"The quality of output (from vector) is based on your quality of the imagining device. A high dpi/resolution printer will produce cleaner images. Cleaner being (edge jaggies) on curved type."
---- The quality of output (from vector) is based on the RESOLUTION of the Raster Image Process - and other factors.. but the actual "quality" of the image itself that goes through the rest of the steps is determined by the RIP -- if you perform the RIP at the same resolution and the pixels are all in the same places, there is no difference, nothing "cleaner" or "edge jaggies" --- you still get edge-jaggies always at the resolution of the RIP - then modified by the resolutions of the droplets or other factors of the printing process. The "effective" resolutions are not always the same as the digital version - but if you rasterize the same image that the in-device raster-engine creates, then absolutely there is no difference whatsoever - the same image gets sent through the rest of the process.
So am I to take it you really do think that there is vector output on film for screenprinting? Please show me the device.. I have worked with many over the years, even a thermal imagesetter, one of the densest and best positive-devices, creates 1200 dpi at best, it appears rounded under the microscope because just like emulsion etc, the film/heat can't create an exact reproduction in the physical world of the digital pixels that are sent to it. But a vector text vs. a raster text at 1200 dpi - no difference, the vector still gets rasterized at 1200 in the RIP of the imagesetter.
Until someone shows me a device that supposedly outputs vector without performing a RIP in order to do it, then there seems to be no clear evidence that there is any distinction between sending vector to a print device or rasterizing first at the SAME RESOLUTION as the device does to the vector data IN-RIP. But let's look for a printer that doesn't RIP.
-
Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I wonder if FSS is categorizing most of our output devices as printing with dots, and therefore printing raster, even if the file is vector.
Are you printing films with a plotter?
http://en.wikipedia.org/wiki/Plotter (http://en.wikipedia.org/wiki/Plotter)
"Pen plotters print by moving a pen or other instrument across the surface of a piece of paper. This means that plotters are vector graphics devices, rather than raster graphics as with other printers. Pen plotters can draw complex line art, including text, but do so slowly because of the mechanical movement of the pens. They are often incapable of efficiently creating a solid region of color, but can hatch an area by drawing a number of close, regular lines.
Plotters offered the fastest way to efficiently produce very large drawings or color high-resolution vector-based artwork when computer memory was very expensive and processor power was very limited, and other types of printers had limited graphic output capabilities.
Pen plotters have essentially become obsolete, and have been replaced by large-format inkjet printers and LED toner based printers. Such devices may still understand vector languages originally designed for plotter use, because in many uses, they offer a more efficient alternative to raster data."
It is about the only case in which real vector information is going to the print-head and not being raster-image-processed first.
The LED toner and inkjet printers of today all perform high-resolution 600/1200/2400 etc RASTERIZING of the vector. They don't print vector.
-
Are you under the impression that somehow there is "non-raster" or "vector" that ends up printed on your film?
I wonder if FSS is categorizing most of our output devices as printing with dots, and therefore printing raster, even if the file is vector.
Right, exactly. Am I wrong to make this categorization?
Can you please show me an example of an output device that prints VECTOR?? Perhaps a vinyl-cutter-plotter... that is the type of device that may use vector instructions sent to the motors and print-head. Just about everything else prints with DOTS. I am raising the question because it seems like somehow you guys think you're outputting VECTOR onto your films??
The printer rasterizes the data whether you sent it raster at the same resolution as the print-engine will do it, or vector, it will turn into the same exact thing on film -- the only difference is if the in-device rasterization has some other factors besides resolution that can be measured and shown as a different/better quality factor in the process.
But the conversion of a "vector" to a shape on the film is almost always a Raster Image Process - conversion to Raster, at a specific resolution.
Hey, don't pick on me. I'm the one who answered your quasi riddle ;)
-
This is getting a little off topic honestly. I think or assume that almost everyone here understands that printers print using dots and are essentially rasterizing vector input as a result. That is the main reason I manually create my halftones vs using a rip in the first place, because I do not see the benefit of a rip in this specific area. The dots I am creating digitally mirror what a rip would create specifically when it comes to the interpretation of the grayscale information. I am using a high resolution digital file, and outputting 1 to 1 (or close to it) with my dots, but they are not as clean as they are on my screen due to the splatter of the ink as it is laid onto the film.
As for RIPs, controlling other factors like ink density etc are benefits which I am excited to explore as a result of the conversation in this thread and are unrelated to the raster/vector tangent.
Off topic, but I feel like it needs to be addressed...FSS: you need to work on your communication skills. I often see you respond in threads where you repeat what others have already said, but you do so in a way that comes across as condescending and it often seems like you simply haven't read or understood the rest of the posts. Posting the wiki definition of a plotter for example comes across as condescending on a forum full of seasoned professionals who very obviously know what a plotter is... No one in this thread has said anything about printers outputting as vector, yet you seem to be arguing about it due to some misunderstanding or hurried reading of other's points. In fact the majority of the thread mentions printers outputting in dpi and how the digital dpi correlates to the quality of the dots. Vector was never mentioned in regards to the actual film output beyond how it would relate to outputting at the default dpi of the printers...
-
Full,
I suspect you have inadvertently supported my initial post regarding the limited differences of using DCS2 and raster type. It's all in the resolution of the imaging device and the starting point of your art file.
The topic was more about edge definition and I think you took a general term of "clean vector" too literal to imply that "we think" or "I think" our output is actually vector cut/output/plotted onto film or something, I don't know. Not sure what that means for sure, but no. Printing vector art comes out more clean (because it used the optimum quality of your imaging device. No matter wether it's vector, digital "raster" or something else.
Typically, now days, psd art is at 300ppi. So any "type" used straight in photoshop can have various levels of edge artifacts based on your individual type setting preferences or how you addresses that issue. When you have anti-aliasing turned off, it's hard edged, yet (resolution dependent). So if the file is 200ppi, it will be more jaggie than a 300ppi file. Far less (and basically vector clean" if it were 1200ppi. Most people don't have their whole art file at 1200ppi tho, so only do that if you are at the halftone conversion point.
To go a step further, and (here is where I will contradict myself), For my own separations, for my own sales and use, I don't even turn anti-aliasing off. I like the softer edges for the simple reason that at a high res of 300ppi, with somewhat of a fuzzy edge on my type, (on press), it helps smooth out the jaggies. I mean, who sees a 1%-2% dot on the edge of my 14pt type (at 55lpi on a 305 mesh (that is mixed in with 2-8 other colors creating a photo real image anyhow? The exposure process alone takes care of that for me and rounds out my edges. Having a 1-2% dot at the edge of the type seems to smooth out or obliterate and hard jaggedness of the curves. It's the same thought process of why it's not so important to have your type done in vector and your art done in psd and combining them in a vector for a DCS2 file. It can be redundant. Thats a bit maverick I know, but it looks better to me when I have soft curved edges on letters and art and I get that from not using hard edges. That falls into the same idea that a great dot must be a perfectly round, smooth vector smooth and clean dot. Thoughts like that seems to go against the grain and on most occasions, I don't mind.
Back to jaggies. If you were to create 55lpi halftone art that has type in it and your file resolution is 200 or even 300 and you manually create halftones as the discussion was about, then you get visible jaggies. Lets say that is even if your imagine device is archaic and 300dpi. A 600 dpi printer would be better, but still somewhat visible. Higher, and you start to lose that...but it's still there. At the 600dpi and 1200 dpi printers the edges are indeed much smoother coming onto film or screen via vector program. It's provable. I see it every day. You can see it in the preview (before) going to the substrate of film or screen. Vector is cleaner.
What I had been mentioning, (as it pertains to edge definition) are two things.
1, The shape of the dots is not as important as the solidity of that dot on film or screen (and the intended size of the dot it is to represent. SIZE matter more here, then the shape does.
2, Is that the use of DCS2 files (on tee shirts) at high resolutions with the art done in either vector or raster at high rez, (makes no difference) or very little visible difference. Mostly due to all of the other factors involved. You touched on that a little as well. Stuff like RZ value on the surface of the screen, EOM, ink (types etc. Is it a think or a thin ink) All of the little things compounded on the print, make very little difference for using DCS2. Sure, initially it looks great at print preview and maybe even on the screen, but once exposed and printed, the results of the two images will be almost identical. For that reason, I forego the use of DCS2 (as a means of getting crisp clean type) on a tee. It's an added step for me that I find not as beneficial as others do.
It seems your stance here is to make sure we know we are not printing vector physically (onto) film. I think we understand that. It has to be laid out/plotted, mathed out, however you want to call it.... on the film square by square of the devices resolution (and higher resolutions devices) are what create that smoother looking letter S edge. It's all the same, but thats a bit literal and off topic really. For example, when we talk about "blending halftones on press" we don't really mean that we take a spatula and physically mix the dots on press till they blend smoothly. Thats literal. When someone refers to putting vector on films, or raster art on films, they are more so referring to what program they used to do so.
The difference here (at this point) is that we were talking about manually creating the original posters halftones in photoshop (to film) with no rip and since photoshop files are typically at a much lower rez than the imaging device, we needed to address the differences and discuss resolutions. It's just the 99x out of 100, When people use a resolution in photoshop, it's (typically) or most often, NOT at the same resolution as their imaging device at the time of creation or arrival from the customer. At the point of bitmapping, it should be, but not at the stage of creation. Not many artist have a 30 layer rgb psd file at 14x17" at 1200ppi as a standard. HERE, we are telling them, at the point of sending your separations to the printer, and you want clean edges (the least amount of jaggies), and you are manually creating your own halftones in photoshop, you should do so (at a very high resolution). Something that matches the imaging device for example. 1200dpi would be great.
-
HA! sorry Mimosatexas,
I continued the off topic part, (but i did try to clarify my intentions). :) I posted right about the same time you did.
-
Mimosatexas,, After all of that, I think you should look at (the process) of how your imaging device) delivers the dots. It's got to be a different setting in your options. Again, I've seen those dots you illustrated before (in photo) from someone else and I believe it might have even been Pierre. He did two things.
Don't quote me on this, but from memory (often bad), he was getting similar dot shapes and opacities as you and this person did this. (Maybe it wasn't even Pierre) but someone on this forum posted about this a year or so back.
1, He purchased another RIP program for the same device....and got much more of a different shape, more opaque printed dot results. This has to do with how the ink is delivered. He then made some drastic resolution/settings/print pass changes and is getting great dots now.
2, He later purchased another printer with much better results using the newer rip.
I haven't seen Pierre chime in here, so I hope he sees this soon and can elaborate.
-
No worries Dan, you addressed basically what I was trying to address. I think everyone is on the same page when it comes to the raster/vector and output results of either, and the tangent seemed to derail a bit based on miscommunication.
I am definitely going to get a rip and start to experiment much more with the actual printer output settings. I will honestly probably continue to create my halftones manually at high rez in PS as I like being able to see them on the screen, and I often clean up actual dots in areas that may not be obvious from looking at just the grayscale image. Exposure often takes care of these low percentage "artifacts", but not always and I like being able to see them before printing.
I really appreciate the discussion here and I will update with my progress as I make it. I am starting to get to a point where I am trying to incorporate sim process-like aspects into designs whenever possible simply to push myself as a printer and keep things interesting. As a result, the issues with my films became obvious and I can't wait to fix/improve them.
-
That's when you know your making progress, when these little things get unearthed. Aka, the awe haw! moment.
-
Off topic, but I feel like it needs to be addressed...FSS: you need to work on your communication skills.
passionate people tend to take it a little personally when a topic they really know something about gets a little tangled on the web, look at how I lecture people on here about screens.
-
Off topic, but I feel like it needs to be addressed...FSS: you need to work on your communication skills.
passionate people tend to take it a little personally when a topic they really know something about gets a little tangled on the web, look at how I lecture people on here about screens.
But those of us who know you have learned what to filter out. With those who don't post very much. the filters have not been developed, and the baby may tend to get thrown out with the bathwater.
-
Having a pretty good understanding of this discussion, I see everyone's methods and points. If you want to rasterize the text in PS, then you need to be at the highest resolution you can handle. What I like about the RIP though, is that I can leave the raster image at a lower resolution, typically the "double the line count", and files that don't clog up your machine or network, then doing the text in Illustrator. Before I had a RIP though, I would convert to halftone with the bitmap method, and 1200 ppi produced a smoother edged dot than 300, so that was what I used. The physical evidence proved that. So I might work at 300 creating my channels, but my output resolution in the Bitmap dialog would be 1200. I just find it much faster to have a RIP set up and let it do it's job, no manual converting to halftones. Send all colors to print and get on with something else. Thanks to all for sharing
Steve
-
there are many ways to skin a cat, in the end, we use a RIP. It is easy and fast. When on budget you can use the workarounds, but in the end some tools work better than the others (for a variety of reasons).
pierre
p.s. filmmaker is actually cheaper than the AccuRip.
-
Filmmaker and accurip are both $495 I think, no?
-
Filmmaker and accurip are both $495 I think, no?
aaaah, interesting. AR used to be $100 more!
pierre
-
Fresners T-Rip is at $399, I could be wrong but I think it is the same engine as Filmmaker?
-
Fresners T-Rip is at $399, I could be wrong but I think it is the same engine as Filmmaker?
It is!
Pierre