24 Replies Latest reply on Jun 24, 2011 9:55 PM by Colin Flashman

# OT - Plating fine strokes lpi, degrees, & frequency

Whilst I ponder wasting some plates - I could use some techy help.

A pocket card, for sketching, is composed of a graph, off the press, the lines appear uneven. I had a 50/50 chance this would happen.

I set strokes, .5pt, 20% tint (black)

Stepped at .10487 inch (near as I can tell. To match a previous sample, 27 steps worked out to 3 inches. I stepped 28 strokes total, and used distribute)

Plated at 150 lpi/2400dpi, 45 degrees, there is unacceptable dot pattern of strokes two dots & single dot (due to the angles involved(?))

Is there some long math to handle thin strokes, at right angles?

Perhaps 0 degree angle...30 degree  change of frequency/lpi...

Hawkinson would probably love this one.

The imagesetter is not driven by any external RIP. The ppd allowed me AS, BAL, STD (Agfa - Accuarate Sceening, Balanced Screening, STD unknown)

Scan shown shows Sample, Konica 5500 proof and press sheet, in that order.

• ###### 1. Re: OT - Plating fine strokes lpi, degrees, & frequency

While I can not tell for sure from the scan, it appears as though the lines from sample are made of a very fine dashed rule as opposed to a solid line set to a 20% tint.

• ###### 2. Re: OT - Plating fine strokes lpi, degrees, & frequency

Could make your life a lot easier by doing .5 pt stroke and using Cool Gray 3 or 4.

• ###### 3. Re: OT - Plating fine strokes lpi, degrees, & frequency

To my old eyes, having done this for 20 + years, concentarting pretty hard with a 9x loupe,  they're definately traditional dot pattern, not stochastic or FM

• ###### 4. Re: OT - Plating fine strokes lpi, degrees, & frequency

Thanks MB, the other guy in the pressroom suggested that - he doesn't get lost looking at trees in the forest. That's gonna be my last resort after spoiling some film.

• ###### 5. Re: OT - Plating fine strokes lpi, degrees, & frequency

Throwing in a higher line screen I think would throw your pressmen for a loop. If the equipment can handle it without issue and the pressman are game I suppose that would reduce the visibility of the dot, but I think that's a lot more for the press guys to watch. Pretty sure a first or second or stochastic screen wouldn't help on straight line work. That's more just to benefit halftones and shadows and the like. More a benefit for gradients and sloppy registration than fine line work, at least the impression I have.

New line screen software is pretty expensive if i remember. I thought stochastic screens were around 5,000 and up. Last I priced them they were pretty ouchie stingy.

• ###### 6. Re: OT - Plating fine strokes lpi, degrees, & frequency

The sample appears that every line width is only one halftone dot, and why I suggest it could be an Illustrator stroke created with rounded-end dashes. If it were stochastic (which I am not suggesting) there would be no visible repeating pattern.

• ###### 8. Re: OT - Plating fine strokes lpi, degrees, & frequency

That screen shot definitely shows that it is an output halftone dot.

• ###### 9. Re: OT - Plating fine strokes lpi, degrees, & frequency

Ooh, I've done this heaps!  Whenever you're doing lines like this go for a 22.5 degree (half of 45) screen angle on the colour that has the lines, if you can't get that 23 will do fine.  This applies for when you only have horizontal lines as well.

Like someone else has already said though often it looks better if you go to a solid, even a .1pt stroke should still come out OK.

• ###### 10. Re: OT - Plating fine strokes lpi, degrees, & frequency

Looks like I can test this in PS. Opening the pdf in PS, at 1200 dpi, convert to bitmap, set output at 2400 dpi, Method > Use Halftone, Frequency 150, angle at 30 (or 22.5) . (Via a thread in Print Planet)

Tests at my original 0 & 45 angles gave a pretty exact representation of what the imagesetter did (unacceptable), so hopefully tomorrow a.m. I'll use 22 or 30 degrees.

I cannot yet get the near perfect near perpendicular result in Post 1, but this should image visually correct.

• ###### 11. Re: OT - Plating fine strokes lpi, degrees, & frequency

Wow, you disappear for an afternoon and someone calls out your name!

I don't have a lot of direct experience on this particular problem, so but the advice everyone else gave sounds good to me. I would definitely try to avoid using the gray at all, and make it a solid line that's slightly thinner. Or, if you can do it, a spot gray.

Another approach would be to rasterize it to a 2400dpi bitmap (no halftone) in Photoshop and see what you get. Hmm, that's maybe actually worth doing.

So starting with this:

Then Mode > Bitmap... with Halftone of 150/45° diamonds, I get:

Which is probably pretty crummy. Play with the settings until you find one you like? I think diffusion dither's not half bad:

You could even use a custom pattern dither...

of course this all looks much worse close-up, and maybe your halftone screen driver won't do exactly what photoshop does, but at least with photoshop you know exactly what's going to your imaging device...

• ###### 12. Re: OT - Plating fine strokes lpi, degrees, & frequency

That's good for simulation John but when you send that diffusion dithered exmple to the RIP it'll convert it to a half tone AM dot again.  And that will again depend on your angles...

Of course if you had a FM RIP you wouldn't have this problem anyway.

• ###### 13. Re: OT - Plating fine strokes lpi, degrees, & frequency

Wait, really? If you send a monochrome image to your RIP, it will halftone it? Are you sure? It sounds like such a RIP would be broken. What possible reason would it have to do so?

I'm pretty sure the last time I did this (years ago) it went through a Harlequin RIP and went fine.

Now, if you sent a grayscale image, I understand.

I'm certainly willing to believe it is broken, but it would be nice to know why. And surely it's not broken on all RIPs? Unless I'm missing something fundamental and/or I have fooled myself.

That said, I'm not sure the diffusion dither is the best choice, anyhow.

• ###### 14. Re: OT - Plating fine strokes lpi, degrees, & frequency

How do you make a monochrome image in Photoshop?

• ###### 15. Re: OT - Plating fine strokes lpi, degrees, & frequency

If this is sarcasm, I'm gonna look foolish...

Image > Mode > Bitmap = single channel

Although, Bitmap requires "Image > Mode > Show All Menu Items > Bitmap" since Adobe has started dumbing down things for us and I haven't had a reason to customize PS

• ###### 16. Re: OT - Plating fine strokes lpi, degrees, & frequency

Indeed, what Daniel said (and what we've been discussing in posts 10 and 11): A 1 bpp image, constructed in photoshop with Mode > Bitmap....

it maybe neccessary to go to Mode > Grayscale first though. And I suppose you have to save that in a format that understands 1bpp images. I'd use TIFF, though I don't know what formats actually break them, if any, and I'm sure PSD works fine too.

Daniel, I don't think I have any non-default photoshop menu settings (I went to Window > Workspace > Essentials (Default) and also the keyboard shortcuts menu...); but whatever. And I supose I used CS5 instead of CS5.1...

• ###### 17. Re: OT - Plating fine strokes lpi, degrees, & frequency

No that wasn't sarcasm!  I hardly use Photoshop and have never used bitmaps so wasn't aware that monochrome meant the same as bitmap, and realise I don't actually know how bitmaps work either

I thought that because bitmaps have a resolution a RIP would downsample them?  For example below are two screen caps of PDFs exported out of Indesign with a bitmap image, one set to 100ppi, the other to 400ppi.  Why wouldn't a RIP do the same thing?

• ###### 18. Re: OT - Plating fine strokes lpi, degrees, & frequency
No that wasn't sarcasm!  I hardly use Photoshop and have never used bitmaps so wasn't aware that monochrome meant the same as bitmap, and realise I don't actually know how bitmaps work either

Unfortunately the word "bitmap" does not mean the same thing to all people.

Monochrome pretty unambiguously means "one bit per pixel" (1 bpp). That's the kind of image that a RIP ultimately produces and what an imagesetter, CTP device, or even a laser printer uses to control the imaging apparatus.

Sometimes "bitmap" means that, but sometimes a bitmap means any raster image that may have greater "bit depth," such as an 8bpp grayscale image.

I thought that because bitmaps have a resolution a RIP would downsample them?  For example below are two screen caps of PDFs exported out of Indesign with a bitmap image, one set to 100ppi, the other to 400ppi.  Why wouldn't a RIP do the same thing?

What does it mean to export "with a bitmap image"? I don't believe ID has any such option. So i don't know what this paragraph means.

The whole purpose of a RIP is to generate monochrome images. It does so for grayscale and color images by halftoning them (usually; though there are other methods. I don't think "FM screening" is technically a halftone, not sure), for fonts by, err, following the font rules and implimenting the hinting and soforth, and for other objects by -- well, there's no need to be comprehensive here.

There is "nothing" to do for a monochrome image (at least, as long as the RIP has not been told to scale anything), so it should just pass it through unaltered.

• ###### 19. Re: OT - Plating fine strokes lpi, degrees, & frequency

OK we'll get to a common understanding soon!

Unfortunately the word "bitmap" does not mean the same thing to all people.

Sometimes "bitmap" means that, but sometimes a bitmap means any raster image that may have greater "bit depth," such as an 8bpp grayscale image.

Totally agree.  I was misinformed as to what a bitmap was early on (I thought it meant any raster image) and still find it hard to shake that misconception.

What does it mean to export "with a bitmap image"? I don't believe ID has any such option. So i don't know what this paragraph means.

By that I mean I placed a bitmap image with a resolution of 1000 ppi inside Indesign and exported it using the aforementioned ppi settings.  Indesign downsampled it.

The whole purpose of a RIP is to generate monochrome images. It does so for grayscale and color images by halftoning them (usually; though there are other methods. I don't think "FM screening" is technically a halftone, not sure), for fonts by, err, following the font rules and implimenting the hinting and soforth, and for other objects by -- well, there's no need to be comprehensive here.

Yup, agree on this.

There is "nothing" to do for a monochrome image (at least, as long as the RIP has not been told to scale anything), so it should just pass it through unaltered.

OK this is where I disagree.  If the bitmap has a resolution above what the RIP can output as a raster won't it have to downsample?  If not what can't you simulate FM screening on an AM RIP by giving it very high resolution suitably Frequency Modulated monochrome images?

Ah ha!  I've just realised I now know what monochrome images means in the Indesign export dialog!  Learn something new....

• ###### 20. Re: OT - Plating fine strokes lpi, degrees, & frequency
By that I mean I placed a bitmap image with a resolution of 1000 ppi inside Indesign and exported it using the aforementioned ppi settings.  Indesign downsampled it.

Err, so? Yes, if you let InDesign downsample the image then it will downsample the image. If you don't want that to happen, don't let it.

There's a reason that the default sampling threshold for monochrome images is very high -- "Bicubic Downsampling to 1200 ppi for images above 1800 ppi." That's with the expectation that your RIP rips to 1200ppi. Otherwise it can be increased.

OK this is where I disagree.  If the bitmap has a resolution above what the RIP can output as a raster won't it have to downsample?

OK, yeah, you got me! Of course I meant the RIP has nothing to do if it receives a monochrome image at the output resolution. All bets are off otherwise.

• ###### 21. Re: OT - Plating fine strokes lpi, degrees, & frequency

I've learnt something!  The Indesign example was just to prove a point visually.

• ###### 22. Re: OT - Plating fine strokes lpi, degrees, & frequency

John Hawkinson wrote:

OK this is where I disagree.  If the bitmap has a resolution above what the RIP can output as a raster won't it have to downsample?

OK, yeah, you got me! Of course I meant the RIP has nothing to do if it receives a monochrome image at the output resolution. All bets are off otherwise.

Monochrome bitmaps don't get translated into device space (the actual, tangible, physical laser-imaged dots) 1-to-1

For images like these, PostScript interpolates across device space and then samples the original. This will work correctly, even if

(a) the image-IN has a lower resolution than device space (and some input pixels may be duplicated)

(b) the image-IN is of exact the same resolution, but it does not align perfectly to the device space,

or

(c) the image-IN has a (much) higher resolution (in which case some input pixels won't ever be touched).

• ###### 23. Re: OT - Plating fine strokes lpi, degrees, & frequency

-- or

(d) if the monochrome bitmap is not perfectly aligned with the output device pixels.

• ###### 24. Re: OT - Plating fine strokes lpi, degrees, & frequency

@ John Hawkinson (post 11 specifically)

top tip using the diffusion... never thought of doing that!!! normally set our RIP's screen angle to 15 degrees to prevent the "saw-toothing" whenever we have these lines. in one respect it's a cheat's way of making a stochastic image.

as for the whole bitmap/monochrome/lineart thingy... totally agree. so long as the raster file is a 1-bit bitmap, it won't get rendered into halftone by the RIP creating fuzzy edges... at least with the AGFA Apogee RIP that I use at work.