Skip to main content
news

Re: Lenses and sharpening

Eric Stevens
SubjectRe: Lenses and sharpening
FromEric Stevens
Date09/19/2014 07:05 (09/19/2014 17:05)
Message-ID<d2en1alo7740f01tm5vn7nb05s65jfll68@4ax.com>
Client
Newsgroupsrec.photo.digital
Followsnospam
Followupsnospam (1h & 48m)

On Thu, 18 Sep 2014 22:55:30 -0400, nospam <nospam@nospam.invalid> wrote:

nospam
In article <id5n1at4fms8lm49l2sno0usts2u1qcbab@4ax.com>, Eric Stevens <eric.stevens@sum.co.nz>wrote:

Eric Stevens
This discussion started when in response to Kevin McMurtrie Floyd wrote in Message-ID: <87bnqh1mby.fld@barrow.com>

">The digital form of unsharp mask is the inverse of a blur.

nospam
There's both a frequency (diameter) and an intensity.

Eric Stevens
Not the case. It is the high pass sharpen tool that is the inverse of blur. They can use the exact same algorithm with different parameters. Using one and then the other virtually reverses the results.

UnSharpMask is not reversible."

You completely failed to understand what Floyd was talking about and have added your inestimable contributions ever since.

nospam
once again, in a non-destructive workflow, unsharp mask along with everything else *is* reversible. this is a fact no matter how much you and floyd argue otherwise.

Eric Stevens
It's not a reversible process as it is conventionally defined.

nospam
yes it is.

someone can unsharp mask today and remove it tomorrow and put it back the day after that.

You don't know the formal definition of a reversible process, do you?

the following week, that same someone can remove all colour (convert to b/w) and the week after that, can reverse that, exactly how it was in the original image, because it *is* the original image.

that's what just about everyone would call reversible.

I seem to remember that you once claimed to be familiar with Claude Shannon's seminal paper "A Mathematical Theory of Communication". If that is correct you should be able to tell me what he called a reversible process in that context. --

Regards,

Eric Stevens

nospam (1h & 48m)