. . . or even a second face on the back of the cloth . . .
Undoubtedly, “Perception of Patterns After Digital Processing of Low-Contrast Images, The Case of the Shroud of Turin” by Paola Di Lazzaro, Daniele Murra and Barrie Schwortz provides a long needed scientific clarity to an important subject.
The abstract reads:
We discuss the link between visual perception and the potentially misleading effect of software techniques for elaborating low-contrast images. In particular, we present the example of the stains embedded into one of the most studied archaeological objects in history, the Shroud of Turin.
We show for the first time that image processing of both old and recent photographs of the Shroud may lead some researchers to perceive inscriptions and patterns that do not actually exist, confirming that there is a narrow boundary line between image enhancement and manipulation.
I have frequently discussed the subject of pareidolia in this blog. This paper, however, provides the scientific explanations, lucidity and focus I could not achieve. Bravo! (One of my most recent postings on pareidolia was from last August, Les Fredette’s Crucifixion Nail: Why I am Skeptical About It).
Not very convinced by their argument in favor of the absence of the second face.
Ok, we obtain a global score of 0.4.
But when you read Fanti, Maggiolo, 2004, I see no unexpected differences between the correlations of the fs and the bs. (table 1, p.499). 6 anatomic details (on 8) have a score higher than 0.6 ( for Ghiberti 2002), 5 on 8 for Enrie and Cordiglia, with a score up to 0.91!
Is it just a wonderful coincidence?
Unfortunately, Table 1 lists eight image sections that are not described in sizes and locations. Some of these sections appear to have small surface (e.g., eyes, nose tip) which has a tendency to increases the score S (depending on how the matching process was done, which is not even described in Fanti and Maggiolo’s paper). Moreover, the matching process of each section appears to have done independently of each other and by searching for a maximum match. This whole process cannot be reproduced and is dubious. Please, see http://www.sindonology.org/reviews/doubleSuperficiality.shtml for more details. Indeed, it is likely that the authors of the ENEA paper could only do a global score S for the entire face since, due to a lack of details for the image sections and matching process, there was no way for them to reproduce the results of Table 1.
Bonjour Mario, je vous souhaite une très bonne année ! J’ai lu votre papier.
Each section should have been precisely described in sizes and location. I agree, it could be misleading. Fanti can provide us all the necessary details. Have you ever asked him? But for sure, di Lazzaro et al. should have discussed at greater length this issue.
You speak of a “tendency to increase” for small surfaces. Not always… I see a difference of 13,3% between “nose” (0.80) and probaby smaller “nose tip” (0.86) but almost nothing between “eyes” (0.61) and probably smaller “eye right” (0.63). And how can you explain that “eye left” is lower to 0.6 if there is a significant “tendency to increase”?
When I look at the comparison between fs 02 and Enrie and I don’t see a significant “tendency to increase”. “Nose” (0.88) and “nose tip” (0.92).”eye right” (0.88) is lower than “eyes” (0.89) and “eye left” (0.96).
Bonjour Cazab, mes meilleurs voeux pour cette nouvelle année.
What I meant for the “tendency to increase”: If the surface of the sections (from the back side) are small, it becomes easier to find corresponding look-alike sections on the front side. For example, if a small section of 10 pixels by 10 pixels is selected to represent the tip of the nose from the back surface, the very bright spot that is visible can be selected, then a similar section on the front surface can be selected where the nose is. Also, the tendency to increase would apply for one specific photograph that was used to find a maximum match, not necessarily for all of them. This is the case for the first column of Table 1. As I wrote in the previous comment, it depends on the matching process used. But they had lots of freedom in the way to search given several small sections. Obviously, by taking the entire face, you are then quite restricted in your matching process since you can only search once. This is what should have been done, but apparently this was not done that way which is very surprising. Note that the paper of Fanti and Maggiolo does not explain how the matching was done which opens up all kind of possibilities from which I selected one, namely by independently searching each section. Not to have explained this process and not providing the locations and sizes of these sections are major missing parts of this study.
Dan, as far as pareidolia and the coin-on-eye issue is concerned, you wrote on 11th December 2012 (see your post « Coins on Eyes Issue Again »):
Firstly, “Is the cloth too coarse [to record Pilate coin partial blood decals]? […]Maybe. I don’t know.”
– There is no ‘maybe’ as the TRUE fact is most definitely, IT IS NOT.
(Your opinion is biased as long as you ONLY rely on NON-professional numismatists, NON-archaeological/ancient bloodstained pattern analysts, NON-archaeological image analysts/cryptanalysts and NON-palaeographers such as Lynn, Gonella and Schwortz who most obviously confuse NON-body tiny blood images with body images).
– Rodante’s and Moroni’s experiments DO prove that (via blood transfer) a 3:1 twill linen fabric (soaked with a watery/oily solution) such as the TS is not too coarse to record small/Pilate coin partial patterns. To assert it is too coarse to record such patterns is totally misleading and nothing less than a pseudo-scientific and pseudo-archaeological untruth.
Secondly, “Is the nature of the film inadequate? Maybe. I don’t know.”
– Which film/picture of the Sindon face close-up are you referring to? The TRUE fact is the same intriguing tiny patterns do appear (more or less out of focus) whatever the photographic procedure not only in 1931 Enrie’s (orthochromatic) and 1978 Miller’s (traditional silver) but also in 2002 Durante’s (extensive digital). The intriguing tiny patterns may completely or partially disappear only under specific lighting conditions.
– To assert that the nature of the film is inadequate in se is to ignore the same intriguing tiny patterns DO appear on genuine Sindon face photographs whatever the photographic procedure.
Then Dan you added: “My problem is perception.”
– I have no doubt about it since it is a very tricky case we have here.
As a final comment you wrote: “I suspect that some of us have different, indeterminable acceptance thresholds where I think I see becomes I see.”
– When e.g. it comes to “ancient coin PARTIAL TINY pattern recognition”, it is even more tricky for the uninitiated brain-and-eye system as the latter may be not only the victim of ‘I think I sees’ (i.e. falsely positive perceptions) BUT ALSO ‘I think I don’t sees’/‘I don’t think I sees’ (i.e. falsely negative perceptions) AND misreading (fragmented letters, letter-sequences or image) AND even a mixture of falsely positive perceptions, falsely negative perceptions and misreading. So much for the “different, indeterminable acceptance thresholds” where I think I see becomes I see, I don’t think I see/I think I don’t see becomes I don’t see, I think I don’t read becomes I don’t read, I misread becomes I read etc.
– The TRUE fact is the Pilate coin pattern recognition issue is completely spoiled by such biased approaches by NON-professional numismatists, NON-archaeological/ancient bloodstained pattern analysts, NON-archaeological image/inscription/text analysts/cryptanalysts and NON-palaeographers.
– Di Lazzaro, Murra & Schwortz’s paper is brilliantly desinformative as it ONLY tackles/sees ‘ONE side of the issue’ as far as pareidolia is concerned. The FACT is It only tackles the ‘I think I see [what is not really there]’ side but most misleadingly overlooks to tackle/see the TWO OTHER sides of the optical illusion issue namely ‘I think I don’t see/I don’t think I see [what is really there] and falsely positive perceptions mixed with falsely negative ones in conjunction or not with misreading.
– It is now up to the professional numismatist, palaeographer/epigraphist and/or archaeological image analyst/cryptanalyst (ideally working in close association with an optic engineer) to REALLY help the uninitiated eye to see what is really not there and what is really there in terms e.g. of partial tiny coin patterns or inscriptions around and on the TS face.
Max I totally agree with your assertion that the work can be a very
polarizing depending on the image science methodology used to quantify your conclusion. The fact is I’m learning to understand the protocol on peer review work and dismiss my own bias when I see the main foundation of the premise flawed compared with my own perspective. However, I’m still mystified why after 30 plus years we’re still wrestling with this topic about the Lepton coin.
Giorgio, don’t you forget when it comes to ancient coin patterns, minute details such as accidental characteristics observable both on the receiving surface and the source object (as extant coin type), can make a crucial evidence not unlike a fingerprint and help to dating the Sindon image.
What would be good, from the point of view of an unconvinced bystander, is for a professional numismatist/paleo-whatever, to give us an example of some other situation in which the identification of difficult-to-see images in one photo were indisputably verified by subsequent events. I can think of Google Earth images of Roman forts in the Egyptian desert, and of Roman ports now underwater, which on exploration were indeed there, but in those cases the original images were pretty unequivocal in the first place. Identifying archeological remains in the colour of growing crops is perhaps another example, but all these are rather large scale. Can anyone give any example of the discovery of coins or other small artefacts achieved by enhancing photos which were at first view unpromising?
Yes. See e.g. my paper (in French) on La Pierre du “Nasseur Menaceur”, une pierre à graffiti templiers (dans la tour du Coudray du Château de Chinon), Actes du IVe colloque national de Loches de 2006 sur les graffiti anciens, éd. ASPAG, 2010 (I succeeded to decipher more than 20 “pierres à graffiti templiers énigmatiques” (2003-2006) and am writing a book hopefully to be released in 2014).
In 2006, I was asked by both the Director of La Forteresse royale de Chinon and the Director of the Ancient Graffitti Museum to present a paper on my research on the Fort du Coudray graffiti.
That would be a most interesting paper to read (although I can’t find it on the internet) and I look forward to your book. Did you use similar photographic enhancement techniques to the ones currently being carried out on the shroud to assist in your identifications?
You can order the proceedings/Actes du IVe colloque national de Loches de 2006 sur les graffiti anciens, éd. ASPAG, 2010 to: MUSEE DE LA MEMOIRE DES MURS SERGE RAMOND – Place de Piegaro – 60550 – VERNEUIL-EN-HALATTE
Then I used both orthochromatic and/or digital snapshots (overalls and close-ups) taken under different raking light conditions (both natural and artificial). I also recur to mechanical and digital squeezes. Hopefully this year, I gonna use 3D laser scanning.
Addenda: I also recurred to contrat enhancement, negative solarisation (with and/or without pseudocolours). Besides I devote a nearly 60 page chapter to the Sindon Turin in conjunction with 4 Knight Templar graffiti (among which that of an epitaphios with a totally naked Christ and a Sindon-like face only visible at a specific hour of the day in summer and in conjunction with the Third hour or Trece and the Lord’s prayer).
Typo: “a totally naked Christ with crossed hand”
+Typo (sorry): “and in conjunction with the First hour or Lauds and Psalm 5:
“Give ear to my words, O Lord,
Consider my meditation.
Hearken unto the voice of my cry, my King, and my God:
For unto thee will I pray.
My voice shalt thou hear in the morning, O Lord;
In the morning will I direct my prayer unto thee, and will look up.
For thou art not a God that hath pleasure in wickedness:
Neither shall evil dwell with thee.
The foolish shall not stand in thy sight:
thou hatest all workers of iniquity.
Thou shalt destroy them that speak leasing:
The Lord will abhor the bloody and deceitful man.
But as for me, I will come into thy house in the multitude of
thy mercy:
And in thy fear will I worship toward thy holy temple.
Lead me, O Lord, in thy righteousness because of mine enemies;
Make thy way straight before my face.
For there is no faithfulness in their mouth;
Their inward part is very wickedness;
Their throat is an open sepulchre;
They flatter with their tongue.
Destroy thou them, O God;
Let them fall by their own counsels;
Cast them out in the multitude of their transgressions;
For they have rebelled against thee.
But let all those that put their trust in thee rejoice:
Let them ever shout for joy, because thou defendest them:
Let them also that love thy name be joyful in thee.
For thou, Lord, wilt bless the righteous;
With favour wilt thou compass him as with A SHIELD.”
The epitaphios graffiti with a totally naked Christ with crossed hands and the Sindon-like face graffiti are two different stones namely “La Pierre du Christ Nu” et “La Pierre à Maints Visages”.
Reminder: Those ancient stone graffiti are not in open access areas in the Coudray tower.
I can certainly give a readily available example of “I think I see”. The last few years I have has as my desktop wallpaper the standard MS picture of Stonehenge on Salisbury Plain. I’ve frequently noticed that under certain conditons of room lighting, or the angle of the sun coming through the window, that unmistakeable faces, sometimes even of recognizable personages, appear on the faces of the standing stones. I’m also reasonably certain that neither the Stone Age people who carved them, nor a retouching of the picture have intentionally produced this impression. Other readers might like to check it out. It is one small piece of evidence why I remain prudently cautious about claims for seeing various images that may or may not be real objects, notwithstanding Max’s insistence, but I fully understand where he is coming from.
There was an earlier discussion some time back concerning Robert De Clari’s claims concerning seeing the various stages of Jesus’s life on the Shroud cloth. Clearly De Clari would be reporting what he believed he saw. At the time, Max produced some coarse images which certainly seemed to support De Clari’s perception. I recall that he mentioned he would endeavour to enhance them with clearer images. If Max has developed those images further, I for one would like a reference to where they might be viewed
Another example in the importance of lighting angle, are the wrinkles and folds on the Shroud cloth which become clearly visible only under conditions of raking light, as demonstrated in a John Jackson video posted here some months ago. Clearly the wrinkles and folds are certainly always there, but only become visible under the proper lighting conditions. There are any number of other examples in the type of light, such as whether it’s visible light, UV, or IR. As well as the direct lighting, ambient lighting conditions also must have an effect.
Dave, sorry again to have to correct you but you got it all mixed up again here. I was NOT AT ALL referring to “Robert De Clari’s claims concerning seeing the various stages of Jesus’s life on the Shroud cloth” (sic) but to a pre-10th c. CE Edessan ritual! Could you give comments on this blog DUE ATTENTION before passing your own comment, PLEASE?
Max, Thanks for the correction. I note your paper, link posted here last October. You wrote:
“This would be of no historical and archaeological consequences were not the image
misperceptions seen here in conjunction with the TS mature man’s face and the fact they are
directly referable to the Image of Edessa’s deliberate mystique of secrecy then prevailing in
the Nestorian ancient liturgy and in particular to Easter ritual description that follows:”
“[…] On Easter [the acheiropoetic Image] used to change its appearance according to
different ages: it shows itself in infancy at the first hour of the day, childhood at the third
hour, adolescence at the sixth hour, and the fullness of age at the ninth hour, when the Son
of God came to his Passion […] and […] cross. ” [Reference provided]
I believe I may have come across the reference at some time previously, and incorrectly attributed it to Robert de Clari. A check on the Clari reference shows this not to be the case. I’m afraid I’ll just have to admit to yet another senior moment – they seem to be piling up these days. Any progress with your enhancement of the images?
The first thing I will say is that I do not know whether coins or flowers are present on the Shroud. No solid evidence has been provided yet and in this sense, the burden of proof relies on those claimining that those images are present. In this blog, I have often mentioned that the I think I see game is not scientific since, so far, authors and have never released the original images they have used for independent confirmation and verification of their methodology and results. This is no science and currently there is no ground to support the idea of images of any kind.
However, I completely disagree with Dan when he says that this paper provides the scientific explanations, lucidity and focus I could not achieve (to explain paraedolia effects in previous posts)
The authors, just like in any ”I think I see” paper do not provide free access to the images they have used. The software they use (Paint Shop Pro X4) is a medium level software and not appropiate for a thorough identification of any tentative letter, coins or flowers. In my view, a very little scientific part of this paper, is comparing Marion’s results using software filters and contrast enhancement when he used a completely different, much more powerful and accepted technique (PCA, just like in the most interesting paper by Morgan of last July) as a way to combine and extract the meaningful information from different wavelengths (red, green, blue, IR) and sources. A scientifically honest approach to reject Marion’s work would require the use of the same technique (PCA, and not a much more simple one) and see then if the ENEA letters appear. Everything should be done with freely available material for open scrutinization and reproducibility. Sorry to say but, despite the claims, I see no science here.
However, the poorest part of it, is the focus on human perception when state-of-the-art techniques, allows moving to objective probabilities of appearance, something that the authors seem to ignore.
Nevertheless, my Bravo! goes to the public relations department of ENEA for reselling again another paper, in this case presented at Frascati Conference 2010, and as usual, in Christmas or Eastern.
I note Gabriel’s comments above, criticism of particular software used, and in particular the comment: “… the focus on human perception when state-of-the-art techniques, allows moving to objective probabilities of appearance…”
The ratio ‘S’ Pearson Product-Moment Correlation given in Equation (1) looks very like a Correlation Coefficient most Stat students will be familiar with. The calculated ratio of ‘S’ = 0.4 is dismissed on basis of the subjective pictorial comparison of the violin and the accordion. I suggest that there is an objective Test of Significance that ought to have been carried out by the Stats investigator in this case. In the Standard Correlation test for two variables (x, y), the coefficient of correlation ‘r’ would be transformed into a ‘z’ variable where z = atanh (r) (inverse hyperbolic tangent of ‘r’). The sample value of ‘z’ will then be normally distributed with standard deviation 1 / sqrt(n – 3) where n is the number of data points. This then enables confidence levels to be determined for the true value of ‘z’ and hence ‘r’. If the data are completely uncorrelated, the hypothesis that ‘r’ = 0, can then be tested. I should think a very similar test can be devised for the two-dimensional Pearson coefficient, but allowing for the appropriate degrees of freedom.
Oups !Sorry for the wrong % and the bad end! I guess you got the idea.
Daveb is right when he mentions the confidence interval issue for the Pearson correlation coefficient. Due to the great correlation between pixel values in adjacent pixels throughout the Shroud, the degrees of freedom must decrease dramatically for r and most probably, an overall value of 0.4 is nothing but an artifact not truly different from 0 at -let`s say- a 95% confidence level. Again, no original image whatsoever is provided for independent confirmation.
Professors Fanti and Maggiolo have not been contested by a scientist writing a peer-reviewed paper. More research is needed on the other images, which are highly controversial, and from the information given to me, with no permission to reveal details, people in the Church have turned a deaf ear to requests.
This release is quite premature and should not have happened. Eventually the final product will be available, most likely at no cost. Don’t waste your money or time reading a rough draft.
Any statistician worth is salt does know the Pearson test is not reliable and can be easily manipulated (especially when the value is around the threshold of statistical validity).
Typo: worth HIS salt
I don’t agree that the Pearson test is unreliable at all! It yields an objective parameter according to a set formula connecting sets of data. False conclusions may be made by the inexperienced or inadequately informed amateur who fails to understand what a statistical test is about, and who fails to probe further in not measuring its significance. I frequently applied it, along with regression analysis during my working life in researching sets of engineering and financial data,. It is more commonly applied when a linear relationship may be assumed between two variables, but it can be extended to multivariate analysis, and also by means of appropriate transformations to include non-linear situations, A text-book I still have by Paul G Hoel Professor of Mathematics in California and published as long ago as 1962 validates these applications.
Karl Pearson 1857-1936 was one of the key founders of modern statistical theory. His correlation test was only one of many contributions he made to the science.
“Neither regression nor correlation analyses can be interpreted as establishing cause-and-effect relationships. They can indicate only how or to what extent variables are associated with each other. The correlation coefficient measures only the degree of linear association between two variables. Any conclusions about a cause-and-effect relationship must be based on the judgment of the analyst.” (Encyc Brit) Despite this EB quote, non-linear and multivariate correlation can still be tested by transforming the data. A paragraph in my comment at #17, indicates the appropriate method of testing levels of significance.
Although the paper provides some fascinating observations on optical illusions, referring to Gestalt theory as an explanation, and includes citations and references, it is somewhat devoid of the experimental data. I feel the authors have leaned a little too heavily on the aspect of illusionary experience as a reason for dismissing pareidolia phenomena. I dare say that it may well be possible for confirmatory statistical tests to be devised to establish the presence or absence of alleged images, but I suspect that the matter is far more complex than is covered in the paper or even in comments published here.
Dave, why don’t you ask Michael Tite (C 14 test coordinator) and the Oxford statistician how they manipulated the Pearson test instead?
The Pearson r works well in the fields Daveb has mentioned but since the works by Willmot its shortcomes and limitations for many other applications are well known. In any case, a plain value of r does not mean anything at all if it is not given along with its confidance boundaries. Coming to the paper, I think that most probably r, would not pass the test of being different from 0 due to the dramatic decrease in the degrees of freedom.
However, the first thing author should do is a thorough review of the scientific literature and see which are the standard indicators used to compare images (certainly not r). This goes to the authors claiming that there is a second face and also to the authors of the current paper. If everything is going to be scientific, all the claims in one sense or another, should be done according to widely accepted methodologies and indicators.
A further concern is the format used for the images after scanning. JPEG, PNG and many other (typical if Paint Shop has also be used for scanning) should not be used because they distort the pixel values.
I hope the authors in their new revised version can address all these points, including the use of PCA if they want to refute Marion’s results.
Thank u Gabriel. I was also about to refer Dave to the works by Wilmot showing the Pearson Test shortcomings and limitations. I can also refer Dave to a famous bookcase: the Turin Sindon C14 dating just in case Dave has (again) forgotten….
…in terms of threshold of statistical validity.
Typo: “textbook case”
Gabriel you wrote: “I hope the authors in their new revised version can address all these points, including the use of PCA if they want to refute Marion’s results.”
Reminder: Marion JUST misread several ancient Greek letters from ghost writings on and around the Sindon face. Ghost writings are really there. Reminder: a paleographer, an eminent optic engineer and an archaeological image and writing cryptanalyst just don’t need to believe in ghosts to see them ;-).
The shortcomngs in the C14 tests had absolutely nothing to do with any inherent failures in the Pearson correlation test, notwithstanding the perpetrators quoting of it. It had everything to do with an inadequate or non-existing sampling protocol. In the trade we call it GIGO = Garbage In, Garbage Out! Find me another example, if you’ve a point!
January 11, 2013 at 8:53 am | #30 (repeated)
Dave, why don’t you ask Michael Tite (C 14 test coordinator) and the Oxford statistician how they manipulated the Pearson test instead?
BTW,Dave what do you make of the works by Wilmot?
Dave, with Michael Tite and the Oxford lab statistician, we are miles away from your contention “False conclusions may (only) be made by the inexperienced or inadequately informed amateur who fails to understand what a statistical test is about, and who fails to probe further in not measuring its significance”, don’t you think?
BTW, in 2007-2008, I wrote (in French) a counter-investigation article on the C 14 dating as a scientific and archaeological fiasco.
Reminder for Dave: the initial issue was “can the Pearson test be manipulated by scientists?
Answer: IT CAN.
Dave, you challenge me to “find you another example, if I’ve a point!”. The “Second Sindon image issue” is another textbook case if you carefully read Fanti & Maggiolo’s paper vs Di Lazzaro Murra & Schwortz’s.
The very fact the taking away of wire fragments coming from various parts of fabric was abandoned by Michael Tite et al, would have offered to the C14 dating a truly representative extensive sample.
Typo (sorry): The very fact the taking away of wire fragments coming from various parts of fabric was abandoned by Michael Tite et al, caused the C14 dating to be performed from a totally unrepresentative sample of the whole cloth.
In other words, the very fact the taking away of wire fragments coming from various parts of fabric was abandoned by Michael Tite et al was most unfortunate.
+ typo: the improper word “wire” should be read THREAD.
The criticisms of the Pearson Correlation Tests attempted above, or for that matter any other generally accepted method of statistical testing, make as much sense as criticising the accepted methods of structural analysis applied to buildings, when incompetently applied. We have buildings that fell over in the February 2011 Christchurch earthquake designed as recently as 1986 because of engineering incompetence. Check http://en.wikipedia.org/wiki/CTV_Building , Refer paragraph on Royal Commission of Inquiry. That is not to criticise the proper methods of Structural Analysis. The same principle applies to any method of statistical testing or any other scientific technique when incompetently applied!
Dave, you wrote: “The same principle applies to any method of statistical testing or any other scientific technique when incompetently applied!”; Do you imply Michael Tite, the Oxford lab statisician are/were incompetent?
Do you imply Michael Tite and the Oxford lab statisician are/were incompetent? Methinks they MANIPULATED the value.
The Pearson test CAN be manipulated by scientists. Period.
… as they were agenda-driven.
It was an act of scientific incompetence not to adhere to a correct sampling protocol. Any correlation ratio they may have calculated would correctly reflect the data input if correctly done. But the data was not properly representative of the whole, You claim MANIPULATION which is a more serious charge than I would make. The engineer in the CTV building design did not need to make an error in his calculations. He only had to get his structural concepts wrong, which unfortunately he did resulting in the loss of some 100 lives! There is more to correct scientific techniques than mere calculation, whether its statistical sampling measures or anything else!
Methinks Tite and the Oxford lab statistician MANIPULATED the value of chi-squared (χ2). Once Tite knew about the Zurich and Arizona results, his vulnerability consisted in wanting a ‘5 star’ final result that could demonstrate beyond the shadow of a rational doubt the Turin Sindon was a medieval fake.
Hence the values of chi-squared (χ2) for the same 1988 samples of the Turin Sindon range from 6,4 (Tite) to 4,3 (Van Haelst). Do notice they are all +/- close to the threshold of statistical validity.
Addendum: Whether consciously or unconsciously, Tite and the Oxford lab statistician MANIPULATED the value of chi-squared (χ2).
Being privileged to have in my possession the actual Enrie photo Father Filas used to extract the image formation of the Lepton Coin, I’m confident with the conclusion derived by the three authors. However, I do agree with all of your concerns that paper itself seems more like branding rather than a true scientific research and their quantifying methods can definitely be manipulated. But as Andy pointed out, it’s still premature so lets wait and see.
Giorgio,
For my research purpose on the coins-on-eyes issue, it would be JUST GREAT if you could sent me by email a HD digitized copy of the actual Enrie photo Filas used to extract the image formation of the dilepton lituus coin. If you’re OK, please ask Dan my private email address and… THANK YOU.
Typos: “For my research purpose on the coins-on-eyes issue and pareidolia” … “if you could spend me”
+ typo: “could send me”
I think I already have your g-mail account let me know if you received my test e-mail addressed to you.
Thank you
Reminder: Di Lazzaro & Murra first “prematurately” released a paper on pareidolia in 2010. Today, this is the second “prematurely” released paper on the same subject. Methinks, they don’t really know their stuff to say the least.
They still need to do SOME MORE homework about pareidolia…and the right way to detect it.