Re: Epson canned profiles
Re: Epson canned profiles
- Subject: Re: Epson canned profiles
- From: edmund ronald <email@hidden>
- Date: Thu, 19 Apr 2018 15:13:10 +0200
My feeling is that the 720, 1440 and 2880 settings mostly change the
*intermediate* rasterisation process of the file, ie for a file constituted
of solid blocks separated by empty space like a target, the data sent to
the print heads *micro channels* ultimately is exactly the same (with more
precise offsets for the borders).
Think of it this way: Make a 180 dpi block target in Photoshop, and a
version of same upscaled to 1440, send both to driver at 1440 and they
should print the same. Now do that with a nice mountain landscape image,
and obviously there will be a huge difference. It's just that the 1440
version allows much more *detail* from the original file to be added, not
color information.
Edmund
ᐧ
On Thu, Apr 19, 2018 at 2:19 PM, Peter Harris <email@hidden> wrote:
> Alrighty, then: if the differences in output between 1440 and 2880 are so
> slight, why are these settings available? Assuming all other parameters are
> equal, will prints made at 1440 be noticeably different from 2880? (To me
> they are, but only in extremely rare circumstances.)
>
> Surely the good folks at Epson are not so wickedly prescient that they
> baked these options into their driver to deliberately provoke Andrew and
> Ben into their highly embarrassing flame war. (You guys realize how small
> these ad hominem attacks make you look, yes?) Ditto for the other options.
>
> TIA for your help (and please, don’t stack me because I am ignorant.)
>
> Peter Harris
> Programmer/Analyst
> Staff Research Associate
> Baskin Visual Arts Studios
> University of California Santa Cruz
> Santa Cruz, California 95064
> 831-459-3107
>
> therealpeterharris.com
>
>
> > On Apr 19, 2018, at 4:12 AM, Graeme Gill <email@hidden> wrote:
> >
> > ben wrote:
> >
> >> Also...I've mused in this thread to the effect that a perfectly linear
> printer wouldn't
> >> need a LUT profile, that a matrix (or even shaper) would do the trick.
> Or, if you need
> >> LUT for software compatibility, you would only need the bare minimum
> number of
> >> patches.
> >
> > Sounds unlikely. Ink mixing simply doesn't seem to translate into
> additive light mixing.
> > Even a process as physically straight forward as a Chromalin proof
> (where each color
> > has a separate dot layer on a transparent film) isn't really additive.
> >
> > I guess it's because the path of light reflected from the surface is not
> simple -
> > on average it interacts with multiple filtering elements in the
> substrate.
> >
> > So even if each channel was device value -> linear light (something that
> would be
> > pretty terrible for visually progressive control - displays have a gamma
> characteristic
> > for a reason!), the channels won't mix additively.
> >
> > Cheers,
> >
> > Graeme Gill.
> > _______________________________________________
> > Do not post admin requests to the list. They will be ignored.
> > colorsync-users mailing list (email@hidden)
> > Help/Unsubscribe/Update your Subscription:
> users/email@hidden
> >
> > This email sent to email@hidden
> _______________________________________________
> Do not post admin requests to the list. They will be ignored.
> colorsync-users mailing list (email@hidden)
> Help/Unsubscribe/Update your Subscription:
> users/email@hidden
>
> This email sent to email@hidden
>
_______________________________________________
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