|
Post by Ted Talks Stamps on Nov 22, 2023 18:43:18 GMT
We're talking a difference in cv of $17.50 for the yel-buff vs $240 for the org-buff, so it's important that you verify for me that it's the latter (although I paid a yel-buff price for it). My question: First picture is a scan looking, as close as I could make it (meaning, cooling the color temperature that VueScan insists on warming). The paper does not, to my eyes, have any discernible yellow or orange cast. Would cranking up the saturation, as I did in the second image, to see which color saturates, be a valid way to determine this? On the third image, of the backside, just to show I'm not looking at the world through org-buff colored glasses, the upper left margin of the super-saturated image does look like it may be more yellow than orange. So, I ask you: Is this stamp the org-buff paper variety?
|
|
|
Post by khj on Nov 22, 2023 23:49:28 GMT
This is my non-expert opinion. I would defer to experts such as PostmasterGS. Would cranking up the saturation, as I did in the second image, to see which color saturates, be a valid way to determine this? I would say no. It's one thing to filter, it's another thing altogether to change color saturation. You can make it look like other shades when you change the hue/saturation. The paper does not, to my eyes, have any discernible yellow or orange cast? Then, in my opinion, it is almost certainly yellow-buff (and the lower catalog value stamp is normally the default choice anyway if there is no obvious color differentiation). While the orange-buff may not be obviously discernable, it is nevertheless still clearly discernable from yellow-buff. If you can't even ascertain it might be yellow-buff, then it is almost certainly not orange-buff. Again, that's my operational opinion based on experience -- not based on my ability to eyeball an orange-buff by single stamp picture alone. What brings out the yellow-buff is comparing to same-era definitive stamp that you know was not printed on colored/tinted paper. But, if you have this available, here is a better approach if you lack a genuine 30pf on orange-buff to compare. Use the cheaper 50pf, which also exists on both yellow-buff and orange-buff paper varieties. I might have the 50pf on orange-buff. If my Saar is not in storage, I'll check when I get home (assuming I can find my Saar pages at home!).
|
|
|
Post by Ted Talks Stamps on Nov 23, 2023 0:39:15 GMT
Was Catch-22 your Logic text book in Uni? Your reasoning sounds awfully similar to, Major Major never sees anyone while he's in his office. So, if you’re looking for him, you’ll have to catch him when he’s not there!
|
|
|
Post by khj on Nov 23, 2023 1:34:59 GMT
It's my excuse for not being able to find things. Catch-22 and Gremlins is all the 'splainin' needed.
|
|
|
Post by ClassicPhilatelist on Nov 23, 2023 3:30:16 GMT
Ted, It's never a good idea to tweak the color of a scan to help "see" the color better. Your only real way to be sure of the color is to have a known example to compare against. This may be possible to indicate if you have experience in the particular paper colors, by having only one known, and the other being different enough to demonstrate that. There is a big problem with trying to determine color on a photo or scan, that is then displayed on a digital monitor. The problem here is the lighting that the images were taken/scanned will affect the visibility of the color, the lens and camera (specially digital) can affect the color, and then the monitor that it's being viewed on again affects the color. I have 4 monitors of all same make, "calibrated" the same, and when I move any image around between the four I still see slight variations in the color. Paper colors in my experience are even more susceptible to these effects, because the variations are much more subtle in many cases than they are in a printed stamp. (Though there are certainly some serious bugbears out there in that regard as well). If I'm just taking at face value the first image, it looks more like a yellow paper to me.
|
|
|
Post by Ted Talks Stamps on Nov 23, 2023 15:25:50 GMT
I know exactly what you mean. On top of that different papers and/or inks seem to reflect the scanner light differently even for stamps that appear the same color (mainly in the yellow-orange spectrum, I've found). The yellow Bergedorf #3 is a straight scan, and the yellow came out spot on. However this yellow Azores 56b comes out ochre. I verified all my VueScan color settings were either Off or None. The right-hand image is pretty close -- as close as I could manage, but I'm still not satisfied with it. This stamp is on enameled paper. I wonder if that has anything to do with it. I had the same problem when I was selling books. I wonder if scanners are designed with a built-in less sensitivity to yellow, much as black and white photo printing paper was insensitive to red.
|
|
|
Post by khj on Nov 23, 2023 20:22:33 GMT
It depends on the scanner. I have a book scanner that had terrible sensitivity to certain colors, including yellow. I scanned one book cover that had a series of color bars on the semi-glossy cover. Some of the colors came out super-light, off shade, or even almost completely missing! The reality is, book scanners are designed for speed at an OCR-friendly dpi. They are not designed for accurate color reproduction.
If you want reasonably accurate colors, scanners must be calibrated individually. I used to do a lot of different types of spectroscopy back in my research days, including infrared and visible light spectroscopy. The 2 primary factors are the laser/bulb spectrum and the detector (typically CCD) sensitivity/response spectrum. There will be noticeable variations between lasers/bulbs/detectors, and we were spending big bucks for research grade. Consumer grade bulbs/detectors have a much higher variance.
You can't really calibrate the bulb and detector separately in consumer grade scanners. Which means intensity for a given color can't be calibrated super-accurate. But that doesn't really matter, since you are primarily interested in the actual color/shade rather than the intensity. At most, you can only calibrate the raw output of the scanner and compensate to adjust the detector response to the color spectrum.
The real elephant in the room is the "human" element. Because of this, there are 2 other major issues that need to be addressed: 1. Monitor calibration. The same scanner output may look like different shade because our monitor output is different. So some people will calibrate their monitors (I don't, consumer monitors just aren't that accurate nor stable). 2. Human eye response spectrum (differs for each person) and the degree/method of optical processing done by brain.
#2 is actually the hardest to deal with. Tetrachromats have a leg up on most of us, but they still have to deal with the optical processing done by the brain. The latter is truly mind over matter. It allows us to read words even though some letters are damaged/incomplete/missing, or see things that are partially/mostly obscured by other things (I'm good at spotting parts of potentially valuable stamps in pics of piles/box of mixed stamps). But it also often causes us to see things that aren't really there.
Our brain/eye does a lot of optical processing (smoothening, contrast, and even color change/replacement!). One of the talks that I give involves fooling the audience into seeing colored balls that don't exist, seeing a brown tile and a yellow tile even though both tiles are yellow... What this means is that the colors or absence of colors in/around the stamp actually change how we "see" the color of the stamp! This is why the higher end color keys will have a "hole" in each of the color tiles, so that you can hide the rest of the stamp when you compare. During my research presentations/papers, to minimize this type of problem, we would primarily present data or grayscale images. Nowadays, every one uses computer-generated color-coded images. Not a big fan of that, because the colors are set to help visualize you what they want you to see, rather than what is actually seen. You can't see any errors in method/conclusions if the colors are faked.
Ideally, the best way is to scan an isolated part of the stamp, so there are no effects from the background.
I find that camera pics in consistent defined natural light seems to give the most accurate color reproduction. We all ask for scanner pics because we are normally interested in the details & dimensions rather than accurate color reproduction.
|
|
|
Post by khj on Nov 23, 2023 20:23:37 GMT
I forgot to check for my Saar collection. My wife said something about needing to celebrate Thanksgiving, so I am back at the office.
|
|
|
Post by Ted Talks Stamps on Nov 24, 2023 3:34:02 GMT
I find that camera pics in consistent defined natural light seems to give the most accurate color reproduction. We all ask for scanner pics because we are normally interested in the details & dimensions rather than accurate color reproduction. Now this is an eye opener. I took this picture with my iphone to show how right you are. On my phone the picture looks virtually perfect in color rendition, like the yellow half of the side by side, above. But after uploading and viewing on my monitor, the color has shifted significantly toward orange. Now I'd like to know how I can calibrate my laptop screen and scanner, though somehow I doubt it's possible on these cheap consumer models.
|
|
|
Post by khj on Nov 24, 2023 4:39:57 GMT
I think angore has some experience calibrating scanners. Calibrating a monitor is a separate beast. Most of what is available bundled with hardware/software is predetermined setting options, rather than calibrated specifically for the serial number of the monitor purchased. Back in the research lab, when we needed something calibrated that we couldn't do ourselves or didn't have time -- we would typically have to pay 2x-5x the price. For example, a digital thermometer back then might have run $20-$30, and an ISO calibrated one might cost $60-$75. The problem with monitor calibration is that we are looking at our uncalibrated monitor, not your calibrated monitor. One of the things I notice when looking at the monitors in our office or recording studio, is that colors aren't reproduced exactly among monitors, even same brand/model. If you have a laptop/computer that supports multiple monitors, connect 2 and compare and you will see a difference. When we use a 2nd monitor, most of us use it as extended screen, or remote display for someone else; so we never really see the color difference. Duplicate the monitor side-by-side, and you will see differences in some color shades.
|
|
|
Post by khj on Nov 24, 2023 4:49:04 GMT
My wife recently upgraded my iPhone 6 to iPhone 12, because she didn't want me being 10 generations behind everyone else (now I'm only 3 generations behind!). The camera is not only better resolution, the lens has amazingly little distortion, the CCD/display both have remarkably accurate color reproduction. I've found taking camera pics is so much faster if the dimension accuracy is not important and the super fine details are not needed.
|
|