Re: grayscale settings for offset printing
Re: grayscale settings for offset printing
- Subject: Re: grayscale settings for offset printing
- From: Steve Kale <email@hidden>
- Date: Sun, 18 Dec 2005 11:13:27 +0000
- Thread-topic: grayscale settings for offset printing
I think I follow this. With the first component, though, if I tell Colorlab
to create a patch for K=2 or K=4, presumably the relevant patch is created
with full 8 bit precision (I don't think the tiffs created by Colorlab can
be 16 bit - I'd have to go check). So while the reference file is expressed
in a 100 point scale the resultant colour is not "rounded". I have a 100 8
bit precision colours being printed (plus paper white).
The second issue arguably is an issue no matter the number of observations
in the profile. If "noise" is something of a constant across the range then
don't more observations likely still increase the "signal" to "noise" ratio?
> From: Roy Harrington <email@hidden>
> Date: Sat, 17 Dec 2005 12:29:06 -0800
> To: ColorSync <email@hidden>
> Subject: Re: grayscale settings for offset printing
>
> There are several issues that can be problems when you go to lots of
> steps.
>
> Consider a 101 stepwedge using a standard 8 bit grayscale. The gray
> values
> are actually 0-255 and creating K values in Photoshop will result in
> some
> steps being 2 units and some being 3 units which is a pretty big
> variation.
> A 21 step gives 12s and 13s which are probably just fine. There are
> two solutions
> to this -- pick all 3 units steps which is 86 steps or all 5 units
> which is 52 steps.
> You can also design a wedge in 16 bit mode -- its tricky to get the
> exact values
> but I think Posterize works correctly. The wedge can then be converted
> to 8 bit
> and Photoshop with dither the intermediate values so they are still
> evenly spaced.
>
> The second issue is more insidious. There's a fair amount randomness
> in patch
> density due to the way inkjet printing works. This density noise is
> less than visible
> to the eye but with the precision of spectrophotometers I think you can
> pick this up.
> This can end up introducing artificial noise into your prints and
> soft-proofs
> by having a noisy transfer curve.
>
> Roy
>
> On Saturday, December 17, 2005, at 11:10 AM, Roger Breton wrote:
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden