Converting to bitmap in Photoshop

[kidink]

When converting a solid 1 color image to bitmap in photoshop using 50% threshold, do i need to up the output resolution to mimic the output device?

To clarify, if the original source image is at 300ppi, once converted to grayscale and then to bitmap should i change the output resolution to say 1200ppi or keep it at 300ppi

Will this create any differences with the positives?

[Sbrem]

For solid patches, the 300 should be OK, but I always increase the resolution to 600, then the printer will do what it does to that image. I proof to a laser printer @ 300 dpi, and it that looks good, I know the film will too. But, if I'm converting a gradient to halftones, then I make the output resolution 1200 ppi, as it makes for much smoother dots. Again, a better image for the printer to work with.

Steve

[Dottonedan]

Yes, sort of.  Let me explain.

When the original source image is say 300ppi, and you are re-sizing the resolution to 600. What that does, is it will produce or mimic your 300 resolution more accurately. I know you are referring to soldi art. Lets look at an extreme example to understand what’s happening to it.

Take some the letter S in Time regular capital letters at 72ppi and for our discussions sake, have anti-aliasing on. This is used for a softer edge or more photo like presentation. You can uncheck that and the type will be built using the pixels at the file resolution of 72.

Flatten that 72ppi with the anti-aliasing turned on and it will remain with soft edges. Flatten it with anti-aliasing turned off, and it will be flat with hard edges in noticeable pixels.  Now increase the resolution to 600ppi. The anti-Alisiaing will be blurry. The amount will depend on the re-smaplingoptions you choose. Neret neighbor is hard edges. It will re-produce the image as best it can with hard edges.  All that does is reproduce your 72ppi (more accurately with 600 pixels. The look will be the same. Very chunky and we would not want to use that in print.  If you re-sample with softer edges, the same applies. The soft 72pixles will not be soft at 600pixles

In the end with 600 pixels, (if I really want the benefit of the 600ppi). I would go back in and look at the details, (corners where color meet up) and re-sharpen areas to help maintain clean or hard edges.  Some files you don’t need to to much if anything, while others you may need to put in a lot of effort.

The very best results come from vector files, copied and pasted into a new file (as 600ppi) to being with so that there is not re-sampling going on.
(Copying to the clipboard) bringing the file in at size. When you go into photoshop,  Type Command or control N for NEW FILE. The very next thing I do in this window is change pixels to inch and enter 600 resolution. Then RGB. Hit enter.  (If you past in as pixels instead of inches, it changes the document size proportionately).

After you’ve hit enter, the document size shows, but no art yet. Hit PASTE or command/Control V. Then paste as PIXELS. At this point, you should see the box with an X through it. With the X still showing, If you want hard edges with no anti-aliasing, you must now UNCHECK Anti-Aliasing in the top menu bar while the placement still has that X in it.

This makes solid areas super easy to select using color range. Even for the subtle color changes of 5% to 8% for example. You can select those subtle differences in areas by adjusting the sliders and color range where it’s much harder at lower resolutions.

[kidink]

Hey Dan, yeah i get that in regards to the input resolution, obviously the higher the ppi the cleaner the edges will be as there are more pixels to work with but i'm thinking about the point at which we convert to bitmap in Photoshop which is essentially mimicking a RIP right?

So if the supplied artwork file is at 300ppi and we convert to bitmap, do i up the resolution to 600dpi or even 1200dpi which is similar to the output resolution of a printer? I'd always do this if I was converting to a greyscale image to halftones but i'm talking about a standard 50% threshold bitmap.

[Sbrem]

I've left out one of my steps to help smooth the edges; after going up to 600 ppi, I add a little blur with Gaussian Blur, then open either Curves or Levels and move the endpoints inward, which hardens the edge, but with much less discernible pixillation. This while still in grayscale, then I convert to Bitmap after that. The source image is key as Dan says.

Steve

[Dottonedan]

Hey Dan, yeah i get that in regards to the input resolution, obviously the higher the ppi the cleaner the edges will be as there are more pixels to work with but i'm thinking about the point at which we convert to bitmap in Photoshop which is essentially mimicking a RIP right?

So if the supplied artwork file is at 300ppi and we convert to bitmap, do i up the resolution to 600dpi or even 1200dpi which is similar to the output resolution of a printer? I'd always do this if I was converting to a greyscale image to halftones but i'm talking about a standard 50% threshold bitmap.

Yes, to mimicking a RIP.  When they say that a RIP is actually better, it is for a couple reasons.

1, the algorithm or math for doing so is better, more specifically designed for such a purpose...More accurate.

2, There are various other great benefits of using a RIP. But use this method up till you can afford a rip.


If I’m given a 150ppi file, and I need to create halftones for the customer yes, I convert to bitmaps and rip at (the end output devices resolution). Typically 600 or 1200 depending, but 600 at minimum. The reason is more to provide a more detailed and well formed dot shape. 600 is the bare minimum for that. It does not create goods small dots because at that resolution, the small dots in a 55lpi (1-7% range) are more or less a couple square pixels.  But yes, convert any resolution to create ehe halftones at 600 or 1200.  There is some resampling going on, and as a results, (if there are any blurred or fuzzy edges, you will get halftones on those edges but your dots will be well formed.  I sometimes clean up edges (after I convert to bitmap) using the pencil tool and paint in black or white.

[Dottonedan]

I've left out one of my steps to help smooth the edges; after going up to 600 ppi, I add a little blur with Gaussian Blur, then open either Curves or Levels and move the endpoints inward, which hardens the edge, but with much less discernible pixillation. This while still in grayscale, then I convert to Bitmap after that. The source image is key as Dan says.

Steve

Yes.  This is similar to deciding if you want anti-aliasing turner don or off before hand. Similar results.  I do a lot of adjusting either way.