Re: fine art reproduction questions
Re: fine art reproduction questions
- Subject: Re: fine art reproduction questions
- From: Ben Goren <email@hidden>
- Date: Sun, 2 May 2010 15:38:06 -0700
On 2010 May 2, at 2:34 PM, Robin Myers wrote:
> On May 2, 2010, at 2:01 PM, Ben Goren wrote:
>> On 2010 May 2, at 1:20 PM, Robin Myers wrote:
>>
>>> After processing the result is a TIFF image with a custom profile.
>>
>> This would be a profile specific to that image? If so, is it usable as a working space or must it be converted to something else first? And, if the latter, are you aware of any particular reason why they didn't just make it a Lab TIFF instead?
>
> The profile is image specific. The original artwork image is a raw image, 16-bit/channel linear tone curve. The output image is a copy of the original image with a custom profile for that image only and is designed to go straight to the printer.
That makes sense. I've considered doing something similar, using Photoshop just to crop, set resolution, and the like (avoiding all operations that would alter color). I need to experiment, but I'm sure Argyll will produce better results going direct from input profile to output profile without first having to go through a working space profile.
That actually makes me think...does EquaLight require images to be in a suitable working space? I would imagine so, but it'd be fantastic if not....
>> Any chance there's more information on the process somewhere? Perhaps, say, enough for somebody to figure out a way of doing the same basic thing manually?
>
> HP has been very quiet with details of the process.
Shame about that. It's been my observation that companies gain more in the long run from sharing their knowledge than they lose in the short run to competitors releasing similar products.
> Manually doing the same thing would be very difficult, given that the process requires processing lots of spectral data.
That which can be done manually can be automated if it becomes sufficiently tedious, of course. Shell scripts and Photoshop actions can do wonders.
I'm thinking...maybe take the picture, use Photoshop to both sample and mark the RGB values you want spectral samples of, create an action to assemble that into averaged RGB values for export, measure those same spots on the original with an i1, and treat the results as a custom chart. With a bit of up-front work, if the basic premise is sound, you can probably get similar results with not much more time spent in the workflow.
> The operator samples the artwork attempting to get the entire range of color represented. Often the problem has been finding enough areas to measure given that there must be a minimum of 50 spectral measurements. Many artworks have such a limited number of colors that often duplicates occur in the measurements.
Hmmm...I'd think that closely-spaced measurements should provide for better precision. And images with a small gamut generally don't give me much trouble (assuming, of course, the colors are within the gamut of the rest of the chain). I'd think a Rothko, for example, would be one of the easier works one could try to reproduce.
But never mind that...how does the system know that sample Z measured with the spectrometer corresponds with coordinates X, Y in the picture? I'd think that'd be the most important part of the process, and the one most worth spending money on.
> It is my experience that there is no such thing as an idiot proof device. The very nature of idiots ensures they will find a way to defeat such a device or system. ;-)
Too true, too true.
And what's worst of all is when you discover yourself to be the idiot! I just hate it when that happens....
Cheers,
b&
_______________________________________________
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