discontinuities, striations, halftones and pixels
Colin Berry in his most recent posting used the term half-tone and I was reminded of an email I wrote to Ray Rogers that had I posted to the Shroud Science Group on March 5, 2004. Pixelated and pixels were words being used a lot for the halftone effect back then and Ray, if I remember correctly, was annoyed. Here is a somewhat edited version of my email to Ray and his response.
Scientists are good at being cautious, discriminating and precise in the use of language to describe "facts." But those of us who are not scientists and try to describe the world and its things will often call a raisin a grape or worse yet a raison.
I decided to tackle the word pixel, a word so often used to describe an optical quality of the Shroud. I realized, that early on in my research on the Shroud, I encountered the word pixel used often by "authorities" (Ray’s meaning). It conjured up notions about the images that were simply wrong. I settled on an erroneous notion about the images because of how I understood that word. Much is made of pixels to speculate, sometimes quite wildly, about how the images were formed. Such mental images as I had made it difficult to appreciate the chemistry possibilities being advanced by Ray Rogers.
To think this through, I borrowed three facts from "The List" and decided to discuss them with a team of graphics software and hardware experts. Why graphics experts? 1) They live and breathe pixels, 2) they understand visual representation of information at a very granular level and 3) they will humor me to discuss the subject because I agreed to buy donuts and coffee. Ray’s Review was a pre-requisite reading.
[The list was in the works by the Giulio Fanti and several of us in the Shroud Science Group. It was first presented to a rump group in session during Dallas 2005.]
I refer to three facts in List Version 10:
A29) The color of the image-areas has a discontinuous distribution on the entire facing surface (Pellicori and Evans, 1981).
A30) All the colored fibers are uniformly colored, i.e. an exposed fiber is either colored or not colored (Adler 1996, 1999).
A32) All the image shows a uniform straw yellow coloration yielding less than 2% variation in the absorbance of the individual colored body image fibers (Adler 2000, 2002).
One might read these facts and say `aha, pixels.’ I did. I asked the graphics group if my assumption was reasonable.
"You might easily conclude this but you would be very very wrong," was their answer.
Some comments from the donut bunch follows:
Pixel is a good and bad word. It is good in the sense that it means "PICture ELement"and good in the sense that it implies that "perceived shades" of color result from "visual blending" of "bits" of color density. Pixel is bad in the sense that, in common usage, it implies uniform shape and size spots as well as a uniform pattern of placement. Depending on the implementation, pixels are either on/off, as with images on paper, or vary in intensity as is the case with many display devices. The erroneous implication for the Shroud is on/off.
Discontinuous distribution is good precise language. Spot is a good word to use.
Fact A30 seems to clarify A29 by suggesting that the discontinuity is between fibers that are "on" and fibers that are "off" – for their entire `exposed’ length? Note that `on’ and `off’, as binary states, is implied by the word uniform. Is that what it means?
Fact A32 clarifies the on/off state by explaining that the "on" state varies by less than 2% variation in absorbance (Of white light? On a scale of 1 to 0, black to white?). Does this mean that there is little difference in the shade of a color of "on" states and thus the condition is presumably binary enough to be considered binary? (In Rogers’ "Review" he wrote: "The color density seen in any area of the image appears primarily to be a function of the number of colored fibers per unit area rather than a significant difference in the density of the color of the fibers. This observation was puzzling, and we called it the `half-tone’ effect.")
BTW: Halftone effect does not mean it is a halftone.
If we are talking about a 2% variation of what has been observed on very very faint (close to white) images where there is very little variation in the visual blended shades, then the 2% variation could be much more significant than implied. Let us say that sample A absorbs 3% of the white light (more blue obviously) and sample B absorbs 5%. Is this a 2% variation or a 60% variation? (Is there a margin of error that allows the 2% to be 0% or 4%?). Or do we (the donut bunch) simply not understand what 2% variation means?
[ . . . ]
We should not use the word pixel unless we define it explicitly. Spot is a much better word. The use of the word pixel can only lead to absurd conjecture by some people. As one person put it: "To say that grassy hill over there is green because a bunch of green pixels are growing out of the ground is absurd."
A better interpretation might be to say that it appears that much, but not necessarily all, of what we perceive as different shades of color is due to visual blending from concentrations or densities of spots of color that are closely uniform in color.
Ray replied. On the definition of pixels:
Dan: Thank you very much for clarifying the word "pixel" for the group. It has already caused massive confusion.
On the facts used for the analysis, Ray wrote:
A29) "The color of the image-areas has a discontinuous distribution on the entire facing surface (Pellicori and Evans, 1981)." Before making assumptions on the basis of this statement, please look at the photomicrograph of the tip of the nose that Mark Evans took (ME-29).
A30) "All the colored fibers are uniformly colored, i.e. an exposed fiber is either colored or not colored (Adler 1996, 1999)." That is a somewhat misleading statement. Some image fibers have thicker coatings than others, consequently a darker color. Many fibers are darker on the ends where pendant drops of washing liquid probably formed (I see the same effect in dye experiments). I can send photomicrographs to illustrate these facts. HOWEVER: all of the image areas show exactly the same visible/uv spectrum.
A32) "All the image shows a uniform straw yellow coloration yielding less than 2% variation in the absorbance of the individual colored body image fibers (Adler 2000, 2002)." I would like to have asked Al how he made the measurement. The microscope I used in his lab used incandescent illumination, and I assume he used the same exposure meter he used for photomicrography. Dan was correct in interpreting the importance of a 2% variation. But, perhaps Al simply missed the darkest fibers. Anyway, I can’t agree with the statement, and I will send anyone who asks some photomicrographs.
Clarifying shades of color in term of spectrum and density.
Dan said: "A better interpretation might be to say that it appears that much, but not necessarily all, of what we perceive as different shades of color is due to visual blending from concentrations or densities of spots of color that are closely uniform in color." Right. But they are essentially identical in color (spectrum) and somewhat different in color density.
How good are out definitions of discontinuities, spectrum and density? How good is our understanding of this with regard to the shroud’s image?
Here is ME-29 as it appears on ShroudScope. Click on the image to link to ShroudScope for a better, larger image:
Just ask Hugh, he will tell you he cannot see the nose image either… yet it is really there hidden in plain sight just because of magnification (negative pareidolia).
I meant the tip of the nose image.
Which does demonstrate (if it still needs be)that it is noot as plain as the nose in the middle of the face some may think when it comes to correctly detect and identify body or object imprints on the TS. Correct detection mostly depends on appropriate light, magnification, computer enhancements (use of pseudo coulours included).
The Titulus Damnationis Greek inscription (around the head), Pilate coin (over both eyes), Crown daisy or Rock rose (around the head and on the face)recognition skill is not hardwired in most people’s brains. Far from it.
“Pixels” are contingent.
Even when a qualified observer can detect something that is really there, most of the time it is more sensed than really identified as the observer (whether an optical engineer, a paleographer, a botanist, a professional or scholar numismatist) tends himself to combine both negative and positive pareidoliae or fail to really convincingly highlight his finding.
How does one become ‘qualified’ – who provides the qualification – for this kind of observation?
Apparently, by stealing the thread and changing the subject.
This is subject related. There is no stealing at all, this is pure mental illusion…
To David Goulet (and Dan),
One becomes qualified when, he is able to discriminate between positive and negative pareidolie and misreading whatever the field of expertise (Botany, Numismatics, Paleography, Epigraphy, Bloodstain pattern analysis etc)and is able to convincingly demonstrate what he can see is really there and what he cannot is really not there and convincingly correct misreading.
When you say “to convincingly demonstrate what he can see is really there and what he cannot is really not there” — to whom is this demonstrated? Oneself? Others?
And sorry, Dan, to encourage the sidetrack but trying to follow Max’s thread is akin to Sudoku. Keeps the mind sharp. :)
To both oneself and others.
Dan, did Rogers sent you the images he was talking about because I can’t detect that much of a colour variations in the nose photomicrograph image you posted.
Sorry; microphotograph. LOL
One of them, as I recall, was
Thanks Dan
This is ME-02 from Evans’ collection. It is over one of the eye. See
http://www.sindonology.org/shroudScope/shroudScope.shtml?zl=3&image=3&lon=1060&lat=2863&photomicro=t&popups=ME-02+276+52
Also, ME-20 is very similar (coming from the same area):
http://www.sindonology.org/shroudScope/shroudScope.shtml?zl=3&image=3&lon=1060.0&lat=2862.0&photomicro=t&popups=ME-20+306+27
Interesting, Dan. Thanks.
Lo and behold! Invisible interpolation! Check out Mario’s links above, or visit Shroud Scope for the pictures. ME-20 shows two warp threads twisted together, either by mistake or because one ran out and was replaced, gradually, by another. For a short while (guess – 6 inches or so?) the two threads run alongside each other – this is clearly visible by microscopy, and is entirely similar to any invisible mending technique. No such double weaving occurs by the C-14 corner. And look at the variation in thickness of the warp threads in ME-02 and ME-14. Is that evidence of interpolation?
Hugh & Jenx, I don’t think this is an example of “invisible” weaving, at least not the one described in the Marino/Benford paper. It’s a “visible” imperfection (passage error) in the weaving, as would be expected from any manual weaving process. The same with the variable thickness of the warp threads, both are considered evidence against a medieval weaving. Please check pages 11-15 of this reference;
http://editionsassailly.com/books/The%20Shroud.pdf
It’s a beautiful, very balanced, read and is available online for free. I find the textile portion very informative.
Thanks for the reference. A new one on me. However it consists entirely of dogmatic statements with neither evidence nor references to back them up. In particular, the statement on page 15, “This kind of weaving was practiced throughout the Roman Empire,” which appears to refer to herringbone, is not justified by archaeology. My point about the Evans photos was not really to suggest that the face on the shroud has been repaired, but that any kind of reweaving should be as obvious as what we see there, and there is no evidence of anything similar around the C-14 area, in particular on the “Jull” sample photographed by Barrie Schwortz.
The point I was trying to make is that these are imperfections and errors due to the manual weaving process rather than a reweave.
Also a quote from textile expert John Tyrer (taken From the resurrection of the shroud by Mark Antonacci)
” I am very interested in the comparison that can be made between the altar cloth used to patch the shroud, the shroud itself, and the backing cloth…the altar cloth and the backing cloth are plain woven, and are much better products than the shroud. They seem to contain less weaving faults whilst the shroud is a very poor product by comparison. It is full of warp and weft weaving defects, many mistakes in drawing-in the impression I am left with is that the cloth is much cruder and probably earlier fabric than the backing and patches. This I think lifts the shroud out of the Middle Ages more than anything I have seen about the textile”
BTW, An invisible weave would definitely look (if at all it can be detected) much different than a mistake in the weave. These are two completely different things and I don’t understand the comparison.
Ehboy. I know one fact, I need to invest in coffee and doughnut shop! Ha!
I see the evidence yes. But with all due respect, I’m just an intelligent lay person finding it very difficult to program, reprogram, program…. based on so much information, inconsistencies and variables from one source to the next. It seems to me many push opinions as facts based on spiritual beliefs or lack thereof.
I do appreciate your work. I think I need to absorb this.