Do you like your Corel PaintShop Pro X7?

<p>Quite a lively discussion about color temp! I wonder whether either of you could tell me/us whether PSP has finally developed a full complement of 16 bit tools, or whether operations like cloning, dodging, softening, etc. can be done only after converting a 16 bit image to 8 bit. I would go to Corel's promotional literature for answers, but they've lied before (and wasted quite a bit of my time and money). I'm also reluctant to install such a huge program just to find out for myself. </p>

<p>Thanks for your help!</p>

<blockquote>

<p>ACR 3.7 only supports Adobe Standard Profiles; there is no Camera Calibration tab with a Camera Profile pop-up menu.</p>

</blockquote>

<p>My god man, you’re on CS2, that’s 10 years old? As I said, pointless, it doesn’t matter that the facts I illustrated with a <strong>modern</strong> version of Lightroom (that you obviously missed among other salient points, you had to ask despite the labeling on the presented images: <em><strong>Andrew, I don't know the program you are using</strong>. </em>You’re not paying attention. This isn’t a tea party debate Frans, we’re talking about science based facts dude)<em>. </em>There is zero reason you need to take us down another rabbit hole with your ancient version of ACR. You could download a demo of a current version of LR and see EXACTLY what I illustrated with differing CCT values from just the alteration of a profile, just one minor point about how the CCT numbers are not the same nor consistent! <strong>The facts above prove that the numbers differ and don’t match the scene.</strong> It shoots holes in your criticism of the raw converter in question due to numbers you didn’t’ expect because you don’t understand this process a lick! WHY the numbers change isn’t the same as the fact that <strong>they do</strong> depending on a large number of variables in the process I’ve outlined, you can’t understand or accept. Either way, like your review of Solux lamps or ideas about CCT values, you’re simply not connecting the dots based on actual observable processes (science). So it’s pointless to continue with you other to point out:</p>

<ol>

<li>You are using a really old software product which is telling, then expecting out of your ignorance that CCT values for <em>As Shot</em> should match in <strong>all</strong> converters. They don’t, that’s been proven, you can try the same tests yourself. As such, your beef about PRO X raw converter is based on ignorance and other’s here should <strong>dismiss another of your flat earth color theories</strong>, that’s the bottom line. </li>

<li>You make a lot of opinionated observations about color that have no basis in fact and when someone who knows a bit more than you tries to <strong>assist</strong> you in your understanding you get pissy as I’ve outlined then ignore the prime mistakes in your thinking by hoping we’ll ignore it while you press on with another route (this time, DNG profile which do alter CCT values, only one tiny point among others I’ve made), you’ve latched onto to ignore the fact that your assumption of the numbers provided in differing raw processors is flat out WRONG). </li>

</ol>

<p><em> </em></p>

<blockquote>

<p>I wonder whether either of you could tell me/us whether PSP has finally developed a full complement of 16 bit tools, or whether operations like cloning, dodging, softening, etc. can be done only after converting a 16 bit image to 8 bit.</p>

</blockquote>

<p>Well in terms of rendering raw data it’s doing all the operations up till you export a TIFF or rendered image AND it supports saving high bit, dollars to doughnuts it’s doing everything high bit internally like other raw processors. Raw data is high bit, but if you feed a raw converter a rendered image, and that data isn’t high bit, it’s converted to high bit for processing but there’s no real gain here. It’s not like you actually fed the converter high bit data. Now high bit may not be 16-bit per color! And few capture devices produce true 16-bit per color anyway (many are 10-12-14-bit) yet the benefits of editing is still there; less rounding errors from processing the data you feed the converter. </p>

<p>Don’t ask Frans . He still needs to connect the dots about CCT values and how they are broad indicators of moving an image from warm to cool appearance via it’s settings. </p>

 

<p>Thanks, but my question wasn't really about RAW conversion or even whether RAW data are good for a true 16 bits of color depth. My question was more to do with whether one is forced into 8 bit processing and must suffer the rounding errors you mention. Restated, I would ask this: Assuming you generate a 16 bit tiff from whatever your preferred software might be (even if some of those 16 bits might be superfluous), and assuming you open that 16 bit tiff in PSP for editing, are you forced to drop the color depth to 8 bits before you can conduct operations like cloning, dodging, burning, etc." Or has Corel FINALLY implemented those tools in 16 bit?</p>

Sarah, seems Coral isn't answering:

http://community.coreldraw.com/talk/coreldraw_graphics_suite_x7/f/811/t/46377

 

Again, in terms of raw, it better and should support high bit editing.

<p>Sarah, PSP allows you to import 16 bit files. I did several basic operations with a 16 bit file and it didn't tell me to convert to 8 bits. However, I didn't do extensive editing because I had had enough after a while; the misreading of the year info in the metadata, the questionable white balance as-shot temperature (which, as it now turns out appears to be a bug, but I'll address that in a separate post) and the awkward preview updating I referred to previously disqualify this program for me.</p>

<p>Sarah, I've done extensive editing on 16bit files without any issues. I can't vouch for every effect but I've used a lot of them and have no complaints. I couldn't care less about the preceding debate about the metadata and really don't think most people do, or would. I find it to be a solid application, I've been using X7 for a couple of months and am still waiting for the first crash.</p>

<p>Andrew,</p>

<p>That's much better: attack, vilify, destroy. But on a more constructive note, let's see what Adobe has to say: "A digital camera records the white balance at the time of exposure as a metadata entry. The Camera Raw plug-in reads this value and makes it the initial setting when you open the file in the Camera Raw dialog box." This is the case for the February 2007 ACR 3.7 when used with CS2 as well as the November 2014 ACR 8.7 when used with Elements 13. The age of CS2 has nothing to do with it.</p>

<p>When applying different camera profiles in ACR 8.7, the white balance as-shot temperature and tint values don't change one iota, as I would expect it to be the case.</p>

<p>On the issue of Corel's PaintShop Pro X7: Corel just confirmed to me that the white balance information in their Camera RAW lab, when opening a RAW file, is supposed to be taken directly from the metadata without any modifications. It appears that there is a bug in PSP and they are now working with me on this issue as well as on the issue of misreading the year information in the metadata.<br>

<br /> So, what say you?<br /> </p>

<blockquote>

<p>That's much better: attack, vilify, destroy.</p>

</blockquote>

<p>What I’m destroying is your credibility in understanding basic color image processing, thanks in large part to your writings here! Just like your take on display calibration, it’s an opinion without any facts to back it up.</p>

<blockquote>

<p>"<em>A digital camera records the white balance at the time of exposure as a metadata entry. The Camera Raw plug-in reads this value and makes it the initial setting when you open the file in the Camera Raw dialog box."</em></p>

</blockquote>

<p>As you prefer to write so often, you still don't get it! Nothing in the above text is incorrect expect your understanding! Of course it records white balance, where do you think the numbers in the EXIF come from? Are they correct values? Nope. I illustrated that already. The actual illuminant as measured from a proper device shows a difference 262K CCT from what LR and ACR report. And as illustrated, differing converters will treat the data and end up with different values. But we’ve been over that, you still don't get it.</p>

<blockquote>

<p>When applying different camera profiles in ACR 8.7, the white balance as-shot temperature and tint values don't change one iota, as I would expect it to be the case.</p>

</blockquote>

<p>Wrong again Frans but at least you are consistently showing us you really don’t understand the subject. Case in point (same file and same profiles shown already in LR):<br /> http://digitaldog.net/files/CTT&DNGprofiles.jpg<br /> <img src="http://digitaldog.net/files/CTT&DNGprofiles.jpg" alt="" /><br /> First issue is you’re unable to use the tools correctly, secondly you can’t accept a submission of a raw showing differing CCT values for ‘<em>as shot</em>’ which are clearly and correctly presented to you now using multiple screen captures. You didn’t even know what software product I was showing to illustrate the differences in CCT from the same raw so it’s pretty obvious you’re not willing to pay any attention.</p>

<blockquote>

<p>On the issue of Corel's PaintShop Pro X7: Corel just confirmed to me that the white balance information in their Camera RAW lab, when opening a RAW file, is supposed to be taken directly from the metadata without any modifications. It <strong>appears</strong> that there is a bug in PSP and they are now working with me on this issue as well as on the issue of misreading the year information in the metadata.</p>

</blockquote>

<p>It <strong>appears</strong> it’s a bug or someone specifically at Corel told you it’s a known bug? Who would that person be Frans? It’s supposed to take the metadata and provide identical values from the same raw, while two other NON COREL converters also show differing values from the same raw. Explain that Frans. If you can, I’m sure like the other posts here, they will be humorous.</p>

<blockquote>

<p>So, what say you?</p>

</blockquote>

<p>I say you still don’t get it, your science is still flawed (<em>When applying different camera profiles in ACR 8.7, the white balance as-shot temperature and tint values don't change one iota, as I would expect it to be the case.</em>) they clearly can, and you’re probably not the person to take at his word that indeed Corel has confirmed you discovered a bug. Lastly and most importantly, trying to educate you is pointless but pointing out your lack of understanding and ability to accept facts is so entertaining.<br>

Even <strong>if</strong> there IS a bug Frans, it is perfectly reasonable and expected to find the same raw file providing <strong>different CCT values</strong>, by several hundred K or more in differing converters. Something you didn’t know but bitched about in your <em>review</em>. It’s been shown above, and it doesn’t even have to be linked to differing camera profiles as shown between LR and Iridient Developer. I should waste more of my time showing you different CCT values in even more raw converters? Here’s a new rabbit hole you can start digging....</p>

<p>Sarah, see:</p>

<p>http://howto.corel.com/index.php/Corel:What_commands_and_features_work_with_16-bit_per_channel_color_images_in_PaintShop_Pro?family=corel&uselang=en&kb=true&product=Category:PaintShop&version=&topic=&sKey=16-bit</p>

<p>With the introduction of 16-bit per channel color support in Paint Shop Pro X and Paint Shop Pro Photo XI, some of the features support 16-bit, while others do not.<br /><br />The following commands and tools support full 16-bit per channel images:<br />Adjust -> Add Noise<br />Adjust -> Backlighting<br />Adjust -> Blur<br />Adjust -> Blur More<br />Adjust -> Brightness/Contrast<br />Adjust -> Channel Mixer<br />Adjust -> Clarify<br />Adjust -> Color Balance<br />Adjust -> Colorize<br />Adjust -> Curves<br />Adjust -> Digital Camera Noise Removal<br />Adjust -> Fill Flash<br />Adjust -> Gamma Correction<br />Adjust -> Gaussian Blur<br />Adjust -> Highlight/Midtone/Shadow<br />Adjust -> High Pass Sharpen<br />Adjust -> Histogram Equalize<br />Adjust -> Histogram Stretch<br />Adjust -> Hue Map<br />Adjust -> Hue/Saturation/Lightness<br />Adjust -> Levels<br />Adjust -> Negative Image<br />Adjust -> One Step Photo Fix<br />Adjust -> One Step Noise Removal<br />Adjust -> Red/Green/Blue<br />Adjust -> RGB<br />Adjust -> Sharpen<br />Adjust -> Sharpen More<br />Adjust -> Smart Photo Fix<br />Adjust -> Soften<br />Adjust -> Soften More<br />Adjust -> Threshold<br />Adjust -> Unsharp Mask<br />Adjustment Layer -> Brightness/Contrast<br />Adjustment Layer -> Channel Mixer<br />Adjustment Layer -> Color Balance<br />Adjustment Layer -> Curves<br />Adjustment Layer -> Hue/Saturation/Lightness<br />Adjustment Layer -> Invert<br />Adjustment Layer -> Levels<br />Adjustment Layer -> Posterize<br />Adjustment Layer -> Threshold<br />Blend Mode -> Normal<br />Clipboard -> Clear<br />Clipboard -> Copy<br />Clipboard -> Cut<br />Clipboard -> Paste<br />Effects -> Black & White Film<br />Effects -> High Pass<br />Effects -> Infrared Film<br />Effects -> Posterize<br />Image -> Canvas Size<br />Image -> Count Colors<br />Image -> Crop<br />Image -> Decrease Color Depth<br />Image -> Flip<br />Image -> Free Rotate<br />Image -> Greyscale<br />Image -> Increase Color Depth<br />Image -> Mirror<br />Image -> Resize<br />Image -> Rotate [Counter Counter] Clockwise<br />Image -> Split to RGB<br />Image -> User Defined Filter<br />Layer -> Promote Background Layer<br />Palettes -> Histogram Palette<br />Tool -> Blemish Remover<br />Tool -> Eyedropper<br />Tool -> Flood Fill<br />Tool -> Freehand Selection<br />Tool -> Magic Wand<br />Tool -> Move<br />Tool -> Pan<br />Tool -> Pick<br />Tool -> Perspective Correction<br />Tool -> Raster Deformation<br />Tool -> Red Eye<br />Tool -> Selection<br />Tool -> Straighten<br />Tool -> Toothbrush<br />Tool -> Zoom</p>

<p>Andrew,<br>

I just downloaded Lightroom 5.7 and the WB As Shot Temp and Tint don't change one iota when I apply any Camera Calibration Profiles, which is exactly what happens in Elements 13 using ACR 8.7. And you say?</p>

<blockquote>

<p>I just downloaded Lightroom 5.7 and the WB As Shot Temp and Tint don't change one iota when I apply any Camera Calibration Profiles, which is exactly what happens in Elements 13 using ACR 8.7. And you say?</p>

</blockquote>

<p>You certainly are a Glutton for Punishment. It’s simply, amazing you’re still not connecting the dots. Why it’s the profile of course. But that doesn’t matter and why you continue to belabor the point is a typical MO of yours. If you would examine the screen captures you’d know something about the profiles selected, heck, they are named. Next, you’d realize that the profiles are simply one tiny data point to illustrate that the same raw data, with the same ‘<em>as shot</em>’ value CAN and DOES change, both in ACR and LR.<br /> Don’t tire yourself trying to understand this bit about profiles, you’ve not yet figured out the simpler facts I presented to you: <strong>the identical raw file showing differing as <em>shot values</em> in two different converters.</strong> And you say?<br /> <br /> You expected based on your misunderstanding of this topic that each raw converter should examine and produce the same as shot values. That’s wrong. You should know by now it’s wrong, I’ve provided a screen capture showing this from two products accessing identical data. And I’m not fixing this for my benefit, you’re welcome to the DNG and associated profiles but what’s the point? You’ll get the same CCT values as I did. And you say? Actually whatever you say, it will unlikely be that you’re understanding of what a raw converter reports for CCT can and does change in some situations.</p>

<p>Nice try, Andrew, but the fact is that my copies of CS2, Elements and LR show the identical As Shot Temp and Tint information and when I apply a different camera profile in LR or Elements the numbers don't change. Adobe says that they show the metadata as it was generated when the shot was taken (what other meaning could the term "As Shot" have?) but somehow your copy of LR does something that is in direct conflict with what Adobe says and what my copy of LR does. Are you sure your copy of LR isn't corrupted?</p>

<p> </p>

<blockquote>

<p>Nice try, Andrew, but the fact is that my copies of CS2, Elements and LR show the identical As Shot Temp and Tint information and when I apply a different camera profile in LR or Elements the numbers don't change.</p>

</blockquote>

<p>As do *some* profiles I have, so what. I have profiles that behave as illustrated. So because you are <strong>unable</strong> to figure out how to use other DNG profiles which DO affect the values, then no such differences can ever exist? That’s flat out wrong Frans. So because you are unable to select a DNG profile that does affect CCT As Shot values, therefore it’s not possible and what I’ve provided is what, science fiction? <br>

Look, you’re being idiotic. <em>Some</em> profiles will <strong>not</strong> alter CCT, <strong>some will.</strong> The same raw using the same profile WILL change CCT values in differing converters. Those are facts. That you can’t replicate this due to your lack of understanding hasn’t and will not change the facts presented above. <br>

As I’ve written, as you continue to ignore, even without dealing with DNG profiles, <strong>the same raw produces differing CCT values for <em>As Shot</em> in two converters.</strong> Using the same profile. I know you want to ignore this but it proves your misunderstanding that all raw converters should produce identical ‘<em>as shot’</em> values is flat out wrong. You simply can’t focus on that critical, simple and proven reality can you? </p>

<p>Explain how some profiles will alter the Temp and Tint values that are coded into the metadata at the time the shot was taken.</p>

<p>Yes, I've experienced myself how two different converters (Adobe's and Corel's) can come up with different answers and by the looks of it Corel's converter has issues, that's why. But all three Adobe converters used in CS2, Elements 13 and LR 5.7 show me the same Temp and Tint numbers, regardless of camera profile.</p>

<p>Go ahead, explain.</p>

<blockquote>

<p>Explain how some profiles will alter the Temp and Tint values that are coded into the metadata at the time the shot was taken.</p>

</blockquote>

<p>The <strong>why</strong> isn’t important, you’re not ready for that yet. The fact is <strong>they do</strong>. Which again has nothing to do with the other <strong>fact</strong> you simply can’t get me to stop shoving in your face and has nothing to do with profiles: <strong>Two different converters treated the identical raw data differently in terms of the CCT values.</strong> Why don’t you think about that before you confuse yourself with how a particular raw processor deals with CCT values with a DNG profile. IF you can.</p>

<blockquote>

<p>Yes, I've experienced myself how two different converters (Adobe's and Corel's) can come up with different answers and by the looks of it Corel's converter has issues, that's why.</p>

</blockquote>

<p>And so does the Adobe converters differ from Irident Developer and if you looked even father without my aid here, you’d find others do as well. A point I made in my first post to assist you in your misunderstanding of CCT numbers. Backed up with facts of which you simply can’t accept. Or understand, or both. Again, this is all so pointless expect it’s typically kind of fun seeing you dig a big hole for yourself.<br>

Worse, you’re SURE Adobe’s number are right, Corel’s is wrong yet you have no way to test this. I’d put my money on Adobe too but I’d measure and examine this all before I put the money down, something that clearly separates us in terms of understanding how these things work. </p>

<blockquote>

<p>But all three Adobe converters used in CS2, Elements 13 and LR 5.7 show me the same Temp and Tint numbers, regardless of camera profile.</p>

</blockquote>

<p>Regardless of ALL profiles? Wrong again. That the ACR converters behave as you describe (with the very limited testing you’ve done and without the proper DNG profiles to use) believing that <strong>ALL</strong> other converters behave the same, despite the facts presented to you, really make you appear quote foolish.</p>

<p>Oh Frans, here’s one more converter to look over that once again, gets the same, identical raw file right or wrong (you’re so good at picking) but one thing is clear, different!<br /> <strong>Same data, same raw. Three different converters. Three different CCT numbers. </strong><br>

The correct CCT value is 5262K, this converter is 4550K, LR says 5000K, so now, we have a whopping 450K among JUST three raw converters.<br /> Go ahead, explain.<br /> http://digitaldog.net/files/photoNinja.jpg<br /> <img src="http://digitaldog.net/files/photoNinja.jpg" alt="" /></p>

<p>Andrew,</p>

<p>When I ask you to explain something that you claim to be an expert in, you weasel out. And yes, I've had enough of your "shoving stuff into my face" as you call it. Don't bother to answer if you're not interested in an honest, constructive discussion.</p>

<blockquote>

<p>When I ask you to explain something that you claim to be an expert in, you weasel out.</p>

</blockquote>

<p>When you stop weaseling out, admit that CCT values vary from the same data from differing raw converters, hundreds if not more values (I’ve shown simply three examples), <strong>THEN</strong> I <em>might</em> explain further why (and no, it isn’t necessary and yes, I’ve done this in as simplistic a way as possible considering the person asking). In fact, I already give you a number of reasons <strong>why</strong> the numbers vary, you’ve decided to ignore them. Start here (READ my 1st post again, it all in B&W)):<br /><br /><a href="/photodb/user?user_id=361342">Andrew Rodney</a> <a href="/member-status-icons"><img title="Subscriber" src="/v3graphics/member-status-icons/sub10plus.gif" alt="" /><img title="Frequent poster" src="/v3graphics/member-status-icons/1roll.gif" alt="" /></a>, Dec 10, 2014; 10:48 a.m.<br /></p>

<blockquote>

<p>And yes, I've had enough of your "shoving stuff into my face" as you call it. Don't bother to answer if you're not interested in an honest, constructive discussion.<br /></p>

</blockquote>

<p>It’s been answered while every question I’ve asked of you hasn't which again is your typical MO in these forums. As such, it doesn’t appear to me you’ve ever had an interest in an honest and constructive discussion. Bottom line (again, I keep having to point this out to you):</p>

<blockquote>

<p>For the same RAW file <strong>I would expect</strong> any RAW converter to be able to calculate the CCT within a couple of 100K...</p>

</blockquote>

<p><strong>Your expectations are wrong</strong>, baseless, not shown to be true!, That’s been illustrated to you now with three different products (actually 5). As you said early on in your so called desire for an honest, constructive discussion, as usual, then setting the mood of the discussion: </p>

<blockquote>

 

 

<p>This color temperature is NOT what the raw converter calculated from the image content; it's the "image as shot" data in the EXIF file! Got it?</p>

 

 

</blockquote>

 

 

It seems it is you sir who doesn’t get it despite my efforts to again educate you (which as I predicted was pointless but entertaining in the process).

 

What I get is you have some very flat earth theories on color that have no basis in fact that you can’t prove.

 

 

<p>Here’s some math** for you to chew on Frans but RGB data is necessary making this raw value weaseling a fact:<br>

http://dsp.stackexchange.com/questions/8949/how-do-i-calculate-the-color-temperature-of-the-light-source-illuminating-an-ima<br>

<em>Also, bear in mind that <strong>CCT is only an approximate metric for most colors</strong>, since only a single curve in the color space actually represents color which can be obtained from a real world black body radiator. Thus for all other colors, t<strong>he calculated color temperature is simply an approximation</strong> of the black body temperature it most closely represents. Thus, for some colors (especially greens) it can actually be<strong> a somewhat meaningless value</strong>, at least in a physical sense. This is illustrated well in the following image.</em><br>

<em>Also, since your question specifically references Adobe Lightroom, I found <a href="http://forums.adobe.com/message/3936298" rel="nofollow">this</a> while searching around:</em></p>

<blockquote>

<p><em>The sliders [in Adobe Lightroom] adjust not the black body temp of the light, but the compensation applied to the image to compensate for the black body temp of the light. This goes the other way round.</em></p>

</blockquote>

<p><strong><em>So bear in mind that the color temperature you see on the Lightroom slider will not be the same as those calculated from the above formulae.</em></strong><br>

**Your calculations Frans, do let us know HOW you’ll calculate RGB values from the raw data, kind of affects the math and resulting values:<em><br /></em><br>

1. Convert the RGB values to <a href="http://en.wikipedia.org/wiki/CIE_1931_color_space">CIE tristimulus values</a> (XYZ) as follows:<br>

<em>X</em>=(−0.14282)(<em>R</em>)+(1.54924)(<em>G</em>)+(−0.95641)(<em>B</em>)<br>

<em>Y</em>=(−0.32466)(<em>R</em>)+(1.57837)(<em>G</em>)+(−0.73191)(<em>B</em>)=<em>Illuminance</em><br>

<em>Z</em>=(−0.68202)(<em>R</em>)+(0.77073)(<em>G</em>)+(0.56332)(<em>B</em>)<br>

2. Calculate the <strong>normalized</strong> chromaticity values:<br>

<em>x</em>=<em>X</em>/(<em>X</em>+<em>Y</em>+<em>Z</em>)<br>

<em>y</em>=<em>Y</em>/(<em>X</em>+<em>Y</em>+<em>Z</em>)<br>

3. Compute the CCT value from:<br>

<em>CCT</em>=449<br>

<em>n</em><br>

3<br>

+3525<br>

<em>n</em><br>

2<br>

+6823.3<em>n</em>+5520.33<br>

where <em>n</em>=(<em>x</em>−0.3320)/(0.1858−<em>y</em>)<br>

Which can be combined to form the following equation:<br>

<em>CCT</em>=449<br>

<em>n</em><br>

3<br>

+3525<br>

<em>n</em><br>

2<br>

+6823.3<em>n</em>+5520.33<br>

where <em>n</em>=((0.23881)<em>R</em>+(0.25499)<em>G</em>+(−0.58291)<em>B</em>)<br>

/((0.11109)<em>R</em>+(−0.85406)<em>G</em>+(0.52289)<em>B</em>)</p>

 

<p> Maybe an equation you can’t fight from someone other than I will start to convince you how wrong you’ve been.</p>

<p>Andrew, I've just had enough of your insults. That's it.</p>

<blockquote>

<p>Andrew, I've just had enough of your insults. That's it.</p>

</blockquote>

<p>Excellent plan to r<strong>un away and ignore the math and facts</strong> provided to you that illustrate your unfounded assumptions about color and CCT values. Maybe you actually learned something although I’d never expect you to admit that. If anything, perhaps in the future you’ll actually study the subject before making conclusions that have no basis in fact. </p>

<p>argue.net Don't let him get to you, Frans</p>

<p>Andrew,</p>

<p>Maybe you could use this thread as a teaching moment. Instead of only stating that different RAW converters come up with different numbers for as-shot WB temperatures, without any explanation as to why, apart from your meaningless "it's all over the place", and insulting me repeatedly in the process, you could have made a positive, constructive contribution by presenting the real reasons why:<br /> 1. There are no standards for translating RAW file metadata into WB temperature<br /> 2. Different camera manufacturers use different algorithms and metadata formats<br /> 3. Camera metadata can change with firmware updates<br /> 4. Adobe doesn't have access to much of the camera metadata<br /> 5. Adobe decided to apply their resources to make Photoshop as useful as possible, rather than try to design different temperature conversion algorithms for a long and growing list of digital cameras<br /> 6. Adobe's website is in error where it claims that the camera's temperature data is the initial value of the WB as-shot temperature<br /> 7. Different RAW converters use different algorithms</p>

<blockquote>

<p>Maybe you could use this thread as a teaching moment.</p>

</blockquote>

<p>Some people like yourself Frans would rather disagree than learn! You were so involved arguing that your incorrect expectations that the values <strong>should</strong> be the same/similar from my first post, it’s incredible you <em>now</em> expect to learn! You got the math, you got examples that the values change and why. Above in this discussion and from your old post here and on LuLa about CCT values for Solux. If I and others have to provide the same lesson over and over again, and you still can’t get it, it isn’t our fault. <br>

Here’s more for you to learn below, but the bottom line again is this, you wrote: </p>

<blockquote>

<p>For the same RAW file <strong>I would expect any RAW converter to be able to calculate the CCT within a couple of 100K</strong></p>

</blockquote>

<p>You’ve come back now, days later, do you have the same incorrect expectations especially after points 1-7 you just posted? <em>Or</em> those points you <strong>learned recently (?)</strong> in the intermission are an admission that what you originally wrote <strong>isn’t</strong> correct and now you’ve come back to <strong>agree with me</strong> that your expectations were incorrect as I attempted to point out to you? <br>

For the same raw file do you expect any raw converter to calculate the CCT value within a couple of 100K’s? Because that isn’t not going to happen and it has been illustrated to you.<br>

More data points to back up my attempt to explain the numbers are a range and your expectations were based on faulty misunderstandings: </p>

<blockquote>

<p><strong>https://forums.adobe.com/thread/312584?start=0&tstart=0:</strong><br>

> <strong>ACR doesn't read the 5D white bal metadata.</strong> <br>

Actually, ACR does. <br>

> The camera K setting is used for in-camera processing and by Canon utilities. <br>

There are no camera K settings. There are scalar multipliers that control the gains required to the R, G, and B camera values to achieve white balance. This is related to the correlated color temperature, but the CCT itself is generally not stored.<br>

> But note that the raw data are white bal agnostic results are subject to a camera profile which may differ between OEM and ACR. <br>

True. <strong>Every raw converter applies its own math given the camera multiplier values to obtain the white balance.</strong> <strong>Similarly, each raw converter has its own math to compute CCT Kelvin values, assuming they are shown and displayed to the user in the first place. Since each raw converter has separate math used, the resulting displayed Kelvin values may be different. </strong><br>

> Contrary to what G Sch writes above, the same K setting ought to give the same results if a "Kelvin" setting is to have real meaning <br>

<strong>That's assuming the camera profiles applied are the same in all cases. But they're not the same, since each raw converter has its own set of profiles. So the results won't be the same. </strong><br>

Eric (Chan of Adobe)<br>

-----------<br>

http://www.lrc.rpi.edu/programs/nlpip/lightinganswers/lightsources/whatisCCT.asp:<br>

CCT values are intended by the lighting industry to give specifiers a <strong>general indication</strong> of the apparent "warmth" or "coolness" of the light emitted by the source. <br>

Another weakness of CCT is illustrated in Figure 8 by points A and B, representing two light sources with the same CCT (3000 K). <strong>Although lights A and B have exactly the same CCT they have very different chromaticities and will look very different to the eye. </strong><br>

-------------<br>

<strong>http://en.wikipedia.org/wiki/Color_temperature:</strong><br>

Although the CCT can be calculated for any chromaticity coordinate, the result is meaningful only if the light sources are nearly white.^[31] The CIE recommends that "The concept of correlated color temperature should not be used if the chromaticity of the test source differs more than [<img src="webkit-fake-url://14579F48-81D0-4D4B-8AB3-9CEE6FED2A1F/5fdd6443b3f7ea11ed85ff14c60f017b.png" alt="5fdd6443b3f7ea11ed85ff14c60f017b.png" />] from the Planckian radiator."^[32] Beyond a certain value of <img src="webkit-fake-url://14579F48-81D0-4D4B-8AB3-9CEE6FED2A1F/5a692f27a4815f93188752873700f15a.png" alt="5a692f27a4815f93188752873700f15a.png" />, a chromaticity co-ordinate may be equidistant to two points on the locus, <strong>causing ambiguity in the CCT.</strong><br>

-------------<br>

http://www.pro-lite.uk.com/File/colour_temp.php:<br>

<strong>It is in the exact definition of CCT that we can be badly tripped up.</strong> While the chromaticity coordinates of a true blackbody source must (by definition) fall exactly on the Planckian locus, the chromaticity coordinates for an LED of a certain correlated colour temperature can fall anywhere along a so-called “ISO-CCT” line that intersects the blackbody locus at the equivalent (true) colour temperature (see Figure 2) . In other words, a CIE standard illuminant A incandescent lamp with a true colour temperature of 2856 K will have chromaticity coordinates of exactly x = 0. 4476 and y = 0.4075. In the CIE 1960 colour space, these coordinates become u = 0.2560 and v = 0.3495. A light source with a correlated colour temperature of 2,856 K can have actual chromaticity coordinates which deviate from the blackbody source by up to duv = ± 0.02. <strong>Given that in the 1960 uv colour space a difference of just ± 0.001 in u or v is generally considered to be noticeable, the definition of CCT permits the colour of white light sources to deviate more than 20 times beyond the point where an observer would start to notice the difference.</strong><br>

What this means in practice is that if you only define the colour of your white LEDs by means of their CCT, <strong>you will potentially end up with a whole variety of shades of white.</strong></p>

</blockquote>

<p> So Frans, after your post last week, I was really hoping you moved on. And now your are back and I’ve again provided <strong>more</strong> data points that attempt by illustration of other’s that your assumptions about the CCT numbers are incorrect. What more do you expect? What I’d hope is you <strong>now</strong> expect what I wrote in my very first post to you about the CCT numbers vary. Far more than 100K. Got it now?</p>