Problem solved:
in [Digital Cameras]
|
From: BD on 8 Feb 2010 17:59 > Jump some more to try to justify why you waste your time with PhotoSlop. > This is fun! :-)- Hide quoted text - What's fun - trolling? Trust me - you're not that good. ;)
From: TJ on 9 Feb 2010 09:06 On 02/08/2010 05:45 PM, Too Funny wrote: > You all seem to have a little problem sharing the link to the > charts that show even GIMP is far better than PhotoSlop. > *EVEN* GIMP? *EVEN?* Your language indicates you have a problem considering open-source software as anything but second- or third-class. How pathetically wrong. TJ -- 90 per cent of everything is crud. - Theodore Sturgeon
From: Too Funny on 9 Feb 2010 12:09 On Tue, 09 Feb 2010 09:06:09 -0500, TJ <TJ(a)noneofyour.business> wrote: >On 02/08/2010 05:45 PM, Too Funny wrote: > >> You all seem to have a little problem sharing the link to the >> charts that show even GIMP is far better than PhotoSlop. >> >*EVEN* GIMP? *EVEN?* Your language indicates you have a problem >considering open-source software as anything but second- or third-class. >How pathetically wrong. > >TJ You are correct. A bad choice of words on my part. My use of "even" in that instance was while I was thinking about their cost, not their capability. I use many such programs. Cartes du Ciel, the best astronomy software available. CHDK, the most comprehensive operating systems of any cameras on earth. VirtualDub, a video editor that does more things and does it faster than almost all others. OpenOffice. etc. etc. If I had to choose between GIMP and PhotoSlop, I would easily choose GIMP.
From: Too Funny on 9 Feb 2010 12:57 On Tue, 09 Feb 2010 09:06:56 -0800, Arthur Entlich <e-printerhelp(a)mvps.org> wrote: >Alan, > >I'm only reading this in the comp.periphs.printers Usenet forum, and it >seems I missed the beginning of the "Problem Solved" thread here (some >people apparently are not cross-posting). I know some thing >cross-posting is a crime against nature, but in a case like this were >several newsgroups are involved in the "discussion" it would be nice to >know if someone has come up with a definitive response to this issue. It wasn't a cross-posting issue. It's that the PhotoSlop fanboys were so embarrassed by this that they tried to not let others know the convoluted solutions they were coming up with in trying to explain it all. > >Is it possible for you to post the response which explains this. I already did. That's why I included the information they finally came up with in the first "Problem Solved" reply that you read. Which they consequently snipped again in any further replies so nobody would see the egg on their faces. As, I noticed, you did as well. > >I did correspond with a friend who is a color engineer and he explained >that he seemed to recall that Photoshop uses a splined black and white >gradient as the default black to white selection via the gradient toll >menu, while it is possible other programs do not do so, and this might >explain the Photoshop diversion from other programs. > >His comments (paraphrased) were that: > >1) Using certain features within Photoshop for "testing" might not be >appropriate because not all features in Photoshop are designed to be >used as a "testbed" but instead are designed for practical use for >editing images. As are all the other programs. Ever hear of Canvas? It's been used to publish many magazines, among all the other things it is capable of doing. Including cartography with a geographic coordinate system to sub-millimeter accuracy in reference to the earth-sized GIS coordinate systems. We're not talking about some simple red-eye removal editor that you get bundled with your printer or camera. > >2) Perhaps there would be some advantage to having an optional check box >to provide a non-splined, linear version of the black to white (and >perhaps other) gradients for certain applications. He suspected that >the color gradient default may also be splined in Photoshop. It's not a spline vs. linear issue. Apparently they have a checkbox to flatten their gradients but even that didn't help. Didn't you relay their solution to your "color engineer" friend? You know, the part you snipped out again. When doing my own testing I found that even their PhotoSlop's rainbow gradient was in error. Flattened or not. > >He suggests as an easy (although not fully scientific) way of showing >the splining of the black and white gradient, make the gradient, and >then posterize the results to a certain number, such as 16 or 32 units. >Then take note of the spacing or size of the grey bands, and note they >are not equal. > >Based upon this, I do wonder if the engineers at Photoshop possibly took >human vision, and perhaps display methodologies into account and >adjusted these to make them respond more "linearly" to human vision, >rather than to strict mathematical linearity. Yes, a wonderful justification of why all the other programs in the world work properly and PhotoSlop does not. > >3) To quote him: > >"... let me state that I am not really a fan of Granger rainbows. They >have their place, but like the proverbial wrench being used for a hammer, >they often are not being used for an appropriate task, so I come to wish >they weren't being used at all. They are highly synthetic, and focus on out >of Gamut colors, so they tend to tell you a lot about colors you'll never >capture with a camera and little about ones that you will." How do you know you will never capture them if you remove them in the first place? How do you know your printer can't produce them if you remove them in the first place? It appears that none of you are very bright. But I guess that's why you were so easily manipulated into buying and supporting PhotoSlop. > >As I somewhat suspected when this great "debate" came to the forefront >here, it may well be someone making a mountain out of a thoroughly >explainable molehill. Yes, they did explain it. See the first post in this thread that you replied to with their convoluted explanation, justification, and solution. You know, the part that you too snipped out. > >I don't believe the designers of Photoshop ever intimated or suggested >that the gradient defaults they provide in the gradient menu are exactly >linear. > Good for them. I wonder what else they never tell anyone? Like they never told anyone up until 2008 that their math platform was only 16-bit and could have been run on Windows 3.1 all these years. It was only recently changed into a 32-bit math platform in CS4. Something that all the other programs have had at their core for the last decade or more. Hell, their programmers still haven't even discovered any resampling algorithms beyond lowly bicubic from the last century. They're precisely that archaic. Their emperor has absolutely no clothes whatsoever. And even when shown their emperor's nakedness, they then pursue the habit of poking their own eyes out to make sure they don't notice. Just clap three times with your eyes closed and keep saying to yourselves, "I believe, I believe, I believe..."
From: Paul Furman on 9 Feb 2010 13:11
Arthur Entlich wrote: > > I'm only reading this in the comp.periphs.printers Usenet forum, and it > seems I missed the beginning of the "Problem Solved" thread here (some > people apparently are not cross-posting). I know some thing > cross-posting is a crime against nature, but in a case like this were > several newsgroups are involved in the "discussion" it would be nice to > know if someone has come up with a definitive response to this issue. > > Is it possible for you to post the response which explains this. Arthur, The explanation is that the programs use different math to accomplish the task with the same name. I was able to reproduce the matching effect in photoshop using Screen blend mode for the top half and Multiply for the bottom half: http://www.flickr.com/photos/edgehill/sets/72157623375625518/detail/ I also showed it in lab mode, which is designed to match human vision more closely. Notice the 5th frame in that set how they convert to grayscale and it's pretty obvious what the reasoning is for the odd shapes. This thread also exists in two threads of nearly identical titles but in the other thread, the math was explained. Here's a snip from that discussion: > On 2/7/2010 5:36 PM, Pete wrote: >> ... >> Now, given our horizontal colour gradient with S=100% and H going >> from 359� down to 0� left to right, what will happen when we attempt >> to change the luminance from 0 at the bottom to 100% at the top? >> ... >> Red around one third of the way up. >> Green around two thirds of the way up. >> Blue around a tenth of the way up. |