Where Light Meets Dark

Where Light Meets Dark has a new address

2006-09-05T09:41:00.001+10:00

Where Light Meets Dark has a new address.

www.wherelightmeetsdark.com

I moved to a new server about a week ago. Unfortunately I don't have RSS (syndication) enabled yet, but plans are on the board. Apologies if you were subscribed to this service.

Head on over to the new website - there are updates on the Emmerichs photo analysis, analogies with my own experiences sighting a (local) cryptid, and a commemoration to Steve Irwin who passed away yesterday doing what he loved - which was inspiring millions to take a fresh look at conserving and caring for our unique and fragile natural environment.

The Irony - Swamp Wallaby in Oatley Park

2006-08-26T22:08:00.000+10:00

If you've read through my earlier posts, you'll be aware that the prime concern of this website at present is evaluating the credibility of two photographs taken by German tourists Klaus Emmerichs in February 2005 which allegedly show a living thylacine in Tasmania. The thylacine is accepted as having become extinct in 1936.

For a bit of fun, one recent post encouraged you to see how hard it is to capture that conclusive photograph.

In a complete twist of total irony, I today found myself in just that position.

I live in Sydney Australia, in its southern suburbs. There are numerous parks of various sizes. One immensely popular recreational park is the 45 hectare Oatley bush park. This morning I took two toddlers for a short walk along numerous bush tracks - none of which I'd walked before, despite visiting the park dozens of times in the 2+ decades I've lived in the area.

During this adventure, I came across some enormous fungi and an exquisite orchid which I had never seen before. Unfortunately, I did not have any camera with me, but thought about heading back tomorrow morning to photograph them.

As it turned out, I was able to return in the afternoon with just one child. After capturing these and a few other interesting flowers and fungi on camera, we were walking along another track I'd never seen before, heading back towards our car. For some reason I decided to break away from the path and scramble up some rocks to my left, child in arm.

As I neared the top, I heard the unmistakable thumping of some macropod clearing the area at the sound of my approach.

Although years earlier I'd heard that wallabies had been recorded in this municipality, I recall even then I had no clue as to where they could be hiding. Even Oatley Park, so I thought, could not be harbouring wallabies without them being obvious.

At any rate, I now had camera in hand (neck strap on), toddler in the other arm and began to pursue my quarry. Slowly we crept, hopping from rock to rock as much as possible, cursing each stage I had to cross dried gumleaves and twigs for the crunching that made and trying to convince a two and a half year old that whispering would be a far better way to catch a glimpse of this rare creature.

After a time, I again heard the thumping, and then briefly, caught sight of a quite dark - almost black - wallaby bounding up the hill in the distance. I fired off a shot, towards the setting sun, which proved later to be fruitless.

However, I now had the general position of my quarry in sight.

In addition, I knew that straight up the spur there would be a road. On the other side of the road there was a soccer match playing, with all the requisite spectators. (As it turned out, there was also a cycling competition running, which meant a larger than average number of cyclists were circuiting the nearby section of road (which was not a part of the race track). Including the families with children who were making use of the recreational equipment and barbeques, there would have been anywhere between 200 and 500 people in the park.)

Across the gully was an edge of the park which abutted residential dwellings. The creature had two options - to head up over the spur and down the other side into further bushland (in which case my chase was lost), or to come back down the gully in photographic view.

My course of action was clear - head further up the spur before the wallaby did, so that I could keep it between myself and the residential buildings. I would then also have the sun to my back when I took my photos.

The strategy worked, and a wallaby - somewhat larger than what I thought I recalled from the first sighting - headed back down the gully. (Did I see two different animals?)

Balancing precariously on the sandstone overhangs, patient toddler in aching arm, I had my camera at the ready and composed on an open space the wallaby just had to pass through.

Unfortunately, for whatever reason, it never appeared in the frame and again I had lost sight of it. This time, however, it now had the downhill run of me, which led to the less used roadway and more bushland on the other side.

Having heard the thumping stop again I thought I would try my luck again - this time heading back down along the spur so as to retain my height and sunlight advantage. Eventually I ran out of sandstone and had to head down into the gully.

At last, there it was, hopping off further down the gully.

Three quick shots and I was sure I had captured it on film!

Needless to say, in the back of my mind I just couldn't believe that at home I was in the thick of investigating the thylacine photos and I knew straight away that this presented an opportunity for drawing a comparison.

Before I present the images here, I would like to point out the following:

The first I knew of the animal was to hear it bounding away

The wallaby was effectively trapped in an area of bushland about 300 meters long by 150 meters wide (a little less than a quarter mile by a little less than an eighth of a mile)

I had a camera in hand, turned on, and ready because I was photographing the flora

I saw the animal on two separate occassions before finally photographing it on the third occassion

Despite the restricted area, it still took me 10 minutes to get the photographs

The final photos are of far poorer quality than the thylacine photos

Before proceding, I would like to note that some people will interpret this last point by claiming that this makes the Emmerichs photos even more unbelievable, as they actually capture the animal at far closer range and with far more clarity - but recall that my animal was aware of my presence long before I knew it was there, and it was actively avoiding me for 10 minutes (in a confined space) before I captured it. In Emmerichs' case, the animal wasn't aware of his presence until it heard the
camera's electronic sounds.

Emmerichs' photos aside, I'd like to emphasise again the general point on how hard it is to take a photograph of a cryptid! Even given all the restrictions above, it took me 10 minutes to produce less than glamorous "evidence"!

For the record, the Oatley Flora and Fauna Conservation Society notes the last recorded presence of a swamp wallaby in Oatley park as 1997.

I do not know which species of wallaby I saw today. Given the 1997 record, and that the swamp wallaby is called the black wallaby in Queensland, my best guess is that this was the same species.

Altogether I took four photographs. The first one, taken into the sun, showed nothing of significance.

The next three, taken with the sun behind, caught the wallaby in the first two frames.

Following are the original images (resized) and then two corresponding closeups.

Photo 1 (click to enlarge)

Photo 2 (click to enlarge)

Closeup 1 (click to enlarge)

Closeup 2 (click to enlarge)

I know a wallaby is not a thylacine. And I know I'm in the middle of Sydney.

My questions to you - the reader - are: is this the real deal? Could I have gone out of my way to digitally forge these images? Would I have bothered? Do I have an ulterior motive in coming up with these images? Is my story consistent? What do you think of the quality of the images? Do they convince you that I saw a wallaby in Oatley Park, even though none has been seen there for 9 years? Could I have taken the photo elsewhere (after all, wallabies aren't extinct), and made the story up?

To be sure - my story is true. If anyone has genuine concern regarding this, I can walk you to each of the flowers and fungi I photographed, both before and after the wallaby.

Bear in mind also, that the photos taken by Emmerichs which were published by the Tasmanian media were "distorted ... somewhat to guard against reproduction." (as reported by Col Bailey on Cryptomundo after personal communication with Emmerichs. Since this time, Emmerichs has also communicated with me personally to express this same fact.) The Emmerichs originals are far better quality than was seen online, and certainly far better quality than these wallaby photos!

What thylacine???

2006-08-18T11:19:00.000+10:00

Not happy with the photos? (Yes, I know - they're not here right now, but hopefully you saw them before they were taken down)

Reckon someone should have surely captured a decent photo by now???

Think it's that easy?

Find out just how hard it can be to get that magic photo. Keep in mind - you're holding a camera and expecting your quarry to come into view. See how you do trying to photograph a flying super hero at:

Superman Returns!

Thanks to Alex for the link!

As I wrote earlier, it should come as no surprise if the first photos of a 21st century thylacine should be snapped by a tourist by chance.

Copyright Notice (part 9)

2006-08-15T07:57:00.000+10:00

I have just been informed that publication of the original images is in violation of copyright conditions, and have therefore removed the analysis images at Where Light Meets Dark.

A quick summary (part 8)

2006-08-14T10:44:00.000+10:00

A lot has been written here on Where Light Meets Dark, but for those who want the quick rundown, where exactly are we?

As soon as there was mention that Klaus Emmerichs took two photos of the thylacine my hopes went up because if the camera moved significantly between photos, then it should be possible to model the scene that appears in them. That is the main subject of this blog at present.

Once the photos were published at Cryptomundo, an international community of interested parties examined them and started to constructively (and perhaps not so constructively) comment on them.

In particular, Alton Higgins provided a series of images which overlays David Fleay's 1933 photograph onto one of Klaus Emmerichs' 2005 photographs.

At this point, most people felt that the correlation was so great that Emmerichs must surely have produced a hoax photo using Fleay's original.

Next, I began publishing my thoughts, Alton and I began communicating, and he provided a new series of images, published here at Where Light Meets Dark, comparing the position and spacing of stripes on different thylacines.

The result was that many thylacines show the same spacing and positioning of stripes along the spine despite the fact that research shows enormous variability in stripe length, direction (forwards or backwards), number and termination (single line or forked).

In other words, we should not be surprised if a 2005 thylacine shows the same pattern of striping near the spine that a 1933 thylacine does. Emmerichs' photos cannot be dismissed on these grounds.

In the past few days I have acquired further information about Emmerichs' photos. In particular, Col Bailey (who has spent nearly 40 years interviewing both post-extinction witnesses and pre-extinction thylacine trappers) published a piece in the Tasmanian Times clarifying Emmerichs' story.

(In adding my comments to that article, I drew the attention of Tasmanian Times editor Lindsay Tuffin, to Where Light Meets Dark, resulting in a new article where further reader discussion may be found.)

In April 2006, Emmerichs returned to Tasmania hoping to capture video footage of the thylacine. During this visit he spoke at length with Bailey regarding the photos and the location at which they were taken.

One clarification coming from that article is that the Ricoh camera used for the photographs contains 12 megabytes of internal storage for photographs. According to Bailey, Ricoh has confirmed that data cannot be transferred from a PC or other external source onto the memory chip; it can only get there if it was captured by the camera itself.

As Emmerichs' claimed from the beginning, this lends weight to the argument that the photographs are genuine.

Returning to the image analysis presented here at Where Light Meets Dark, I have so far concluded that the scene appears consistently constructed between the two photographs; that is - these do appear to be two genuine photographs of a real forest scene. Note also that Bailey has visited the exact location at which the photos were taken and there has never been any question that the scene depicted really exists.

Conclusion

I conclude that the scene is consistent between shots and that these are genuine photos of a genuine scene,

Bailey (with nearly 40 years interviewing experience) testifies to Emmerichs being an honerable man whom Bailey trusts entirely in this matter,

Ricoh confirms that data can only get onto the in-camera memory chip if it was captured by the camera itself, and

Higgins demonstrated that we cannot dismiss Emmerichs' photos on the basis of similarity in striping pattern near the spine with any other known photo.

What then, is the likelihood of these photos being genuine? What is the likelihood that they are a hoax?

In order to be a hoax, Emmerichs would almost certainly have had to take two real photographs of something resembling a thylacine, in situ. According to a Sydney Morning Herald article, "the Herald's photographic managing editor, Mike Bowers, conducted his own experiments" to this effect.

One question I still hope to answer is whether the thylacine moved between photos, because if it did, then this would be one final detail in the most elaborate and convincing hoax ever conducted regarding this species.

Alternatively, we would have to conclude that it is highly likely that Emmerichs' photos of a thylacine are genuine.

Answering that vital question - "did the thylacine move?" - should certainly be possible so long as there is enough data in the photos.

At this stage, I feel the quality of the images I am working with may well let me down. Bailey has clarified that the images published in April 2006 were intentionally modified in order to protect the integrity of the originals.

Access to the original images would almost certainly improve the accuracy of measurements undertaken in this analysis, but understandably, these are being closely guarded.

My offer to Emmerichs to analyse the original images still stands, although as has been clarified in my communication with him - and now to the Where Light Meets Dark readers here - I am not a professional in image analysis. I have conducted this work so far in good faith, and am open to constructive criticism regarding my process. (Please submit a comment if you would like to do so!)

Finally, in initial media reports, there was mention of Emmerichs' photo closely resembling a well-known 1930s photograph. Whilst various key players in this story have already made this comparison, I have yet to acquire the specific 1930s image in order to do likewise.

Getting close (part 7)

2006-08-05T04:03:00.000+10:00

It's time to return to the task at hand... did the thylacine move between the first and second photos taken by Klaus Emmerichs in Tasmania in February 2005?

Fig 08 (Click to enlarge)
The scaled images were approximately horizontally aligned with respect to the thylacine (as per previous figures). The top image (labelled "B") is the bigger of the two images published at Cryptomundo. The bottom image (labelled "S") is the smaller. Points which can be clearly seen in both images were then labelled numerically. The circles have no significance other than to mark the point. The triangle labelled "6" was drawn over image S, then copied to image B and rotated over image B until the best match was found. It is the same size in both images. The lines labelled "9" were drawn over the darker areas on the log on image S, then copied and placed over image B. The lines are the same size and orientation in both images.

A description of each data point follows:

1 - the lowest point on leaf L1 such that the circle that marks the point remains contained within the boundary of the leaf. This is the same leaf used for scaling up image S.
2 - the rightmost point of the leaf
3 - the lowermost point of the leaf
4 - the lowermost point of the leaf
5 - the lowermost point at the left of the leaf
6 - the basic outline of the leaf
7 - the 2 pale areas between stripes on the rump of the thylacine
8 - a pale region on the log which is bounded on top and on bottom by darker regions
9 - a sequence of lines traced through darker markings visible on the log

Change in relative positions

As can be seen, many of the data points have changed position relative to each other between photos. Some of these changes may be due to the camera being in a different position for each photo. Some of these changes (in leaf position) may be due to the leaves changing position between photos (for example, due to air movement or due to the thylacine moving them). I plan to test whether the thylacine itself changed positions between photos.

What should we expect?

The first thing we need to note before interpreting the changes observed in the data points, is what we expect to happen with regards to relative position, when the camera moves.

Consider the example below.

Fig 09 (Click to enlarge)
Plan (top) view of two arbitrary data points showing the way they appear from two points, A and B.

In this example, we are looking down on the scene. At first, the viewing point (or camera) is at point A. Viewing both dots from this position, you would see the purple dot on the right, and the green dot on the left, as shown by "View at A". In this particular case, the purple dot slightly overlaps the green dot. When the viewing position is changed (the camera moves) to B, the purple dot now appears to the left of the green dot. Therefore, the object in the foreground (closer to the viewing position, or camera), appears to move in the opposite direction to the camera between shots. (When the camera moved right, the foreground object appeared to move left in the image).

In addition, the closer a foreground object is to the camera, the further it will appear to move with respect to a background object, than, say an object halfway between them. This is not demonstrated here.

Finally, the same principle applies with respect to vertical movement. For example, if the first photo was taken whilst crouching, then the photographer stood up and took the second photo, then the camera has moved up between shots. Correspondingly, foreground objects would appear to move down between photos.

In the present analysis, there appears to be considerable horizontal movement between photos, and slight vertical movement (as will be shown below).

But it's not quite that simple

In the case of the thylacine photos, we don't know which was taken first, and which was taken second. For arguments sake, we will assume the lower (S) image was taken first, and the upper (B) image was taken second, although it doesn't really matter - what we're looking for is whether the thylacine moved with respect to its environment between photos. The assumption that the bottom photo was taken first is made simply to make the discussion easier to follow.

Some initial observations (some of which each need testing):

1 - We assume objects in the foreground appear to move in the opposite direction that the camera moves between shots
2 - We assume the bottom photo was taken first; although this is not important for establishing whether the thylacine moved, it aids discussion
3 - We assume the leaves did not significantly change relative position between photos
4 - We suspect that the leaves may have insignificantly changed relative position between photos
5 - We make no assumption as to whether the thylacine moved or not. Thus, arguments based on data point 7 cannot be used to establish facts about the environment around the thylacine.
6 - Leaf 3 overlaps leaf 1 in the first image (S)
7 - Leaf 3 is closer to the camera than leaf 1 (from statement 6)
8 - Leaf 3 moved right between photos (from 2, observation)
9 - The camera moved left between photos (from 1, 7, 8)

Note: at this point we have established that the camera moved left with respect to the scene, between photos (assuming the bottom photo was taken first). As a result, we can now measure the relative distances that objects moved left or right with respect to each other in order to determine the layering in which the objects appear. That is, which object is closest? Which is next further away? Right down to which object is furthest?

In particular, note that anything which appears to have moved right, with respect to leaf 1, must be in front of leaf 1. The further right it appears to move, the closer to the camera it is. Anything which appears to move left with respect to leaf 1, must be behind leaf 1. The further left it appears to move, the further away it is.

The figure below shows the process used for measuring horizontal distance between points. For the example, I have chosen leaf 1 (data point 1) and the thylacine (data point 7).

Fig 10 (Click to enlarge)
The measurement process. The image was zoomed to 400% (4x) and a vertical line was drawn against the left edge of each data point circle such that the line touches, but does not overlap the circle. A horizontal line was placed such that it overlapped each vertical line at each end. Not shown here, focus was moved away from the horizontal line so as to remove the selection points in order to confirm the position of the horizontal line. The properties for the horizontal line were displayed, and the horizontal dimension of the line can be read from the "Width" field in the "Size and rotate" section.

Two important notes; first, the exact measurement is irrelevant - we are interested in relative measurements. Second, this process assumes the circles are placed in exactly the same position relative to the object they represent in each image. Given that this precision is not possible, a 5% difference in measurement will be required in order for the difference to qualify as significant.

Fig 11 (Click to enlarge)

Following is a table listing the relative differences in position between leaf 1 and every other label in the image. For each distance, a value of 5% (labelled "5%") and 95% (labelled "Change - 5%") is shown. Starting with the object nearest the camera (data point 5), the question was asked: "Does the next nearest object fall within 95% of the previous one?". If so, then the next nearest was ordered equally. If not, then the order was increased.

The verdict?

From the table, we have the order of objects as: leaf 5 and the log at 8 (equally), the log at position 9, leaf 4, leaf 3, leaf 2, the thylacine at point 7.

Fig 12 (Click to enlarge)
Reduced scale depiction showing where the horizontal measurements were made.

One very compelling conclusion from this discussion is that the environmental objects (leaves, log), move with a consistency that might be expected from a real scene. Specifically, the log at point 9 is in fact marginally further from the camera than at point 8 (as was measured and as would be expected). Likewise, the leaf at point 4 is in fact marginally further from the camera than the leaf at point 3 (as was measured and as would be expected from the leaf shape).

For this reason, I conclude that the appearance of the leaves and log is consistent with that which would be expected if two separate photos were taken of the same, real, three dimensional scene.

In other words, I believe the leaves and log are real and are not digital manipulations.

What about the thylacine?

At present, all that can be said from the data points examined, is that the thylacine is the furthest object from the camera. This still leaves open the possibility that the thylacine was digitally added to two legitimate photographs of the leafy scene.

The first step we need to take is to ascertain that there are objects in the scene which are further away than the thylacine. If this can be established, then we can reasonably assume that there was in fact a thylacine in the scene when the photographs were taken (so long as all measured horizontal movement remains consistent with what might reasonably be expected).

If this point is established, then we are forced to conclude that something resembling a thylacine was in fact photographed, twice, in situ.

Imagining for a moment that we establish that conclusion, then the next, and final question to ask, would be whether the thylacine changed its own position relative to its environment between photos. In fact, we will get the answer as soon as we identify objects behind the thylacine moving consistently with respect to the rest of the scene.

If the thylacine moved, then we must conclude that Mr Emmerichs either photographed a live thylacine, or he constructed a scene containing an object resembling a live thylacine and paid such attention to detail as to move it slightly between photographs.

This author (with all due respect) cannot at present hold any opinion on the character of Mr Emmerichs, but it is on record at Cryptomundo that Col Bailey, who has "been investigating claims of thylacine sightings for almost 40 years", has "thoroughly investigated [Mr Emmerichs'] claims and interviewed him at length, as well as visiting the location of his sighting and [is] of the opinion he is an honerable man and is telling the truth about this matter" (Bailey, 2006).

The final piece of the puzzle

And so the final piece of this puzzle looms. We shall return to seek to identify background objects and finally answer the question - do we have reasonable grounds to believe Mr Emmerichs' photos are of a live thylacine?

References

Bailey, C. Commenting on "Cryptomundo Exclusive: Thylacine Photos" (On-line) Accessed 5 August 2006 at http://www.cryptomundo.com/bigfoot-report/thylacine-photos

A quick note (part 6)

2006-08-05T00:58:00.000+10:00

Before we get back on track with the main aim of the image analysis, just a quick note about my last post here at Where Light Meets Dark. Thanks to Anton Higgins' analysis, I drew the conclusion that we could not dismiss Klaus Emmerichs' February 2005 photos of a thylacine on the basis that its stripes are similarly positioned and spaced to those of a thylacine in a 1933 photograph.

Leading in to that, in the post before, I was examining an odd marking on the hind leg of Emmerichs' thylacine. A similar mark was found in the 1933 photo, but the dark colouration across the animal's stomach seemed poorly aligned between photos. In that post I implied that if the 2005 photos were a hoax, then perhaps they were created using a different version (or scale) of the 1933 image than that used in my analysis.

In all fairness, and in light of Higgins' striping pattern analysis, I should add that another possible explanation is that Emmerichs did in fact photograph a live thylacine and did not use a pre-existing photograph to produce a hoax.

We will return to the job of reconstructing the scene as Emmerichs photographed it, in the next post.

About those stripes (part 5)

2006-08-02T21:39:00.000+10:00

In my last post I took a close look at the image overlays created by Alton Higgins for Cryptomundo.

Privately and publicly, Higgins has stated that the purpose for creating the overlays was to demonstrate that Mr Emmerichs' photos from Februrary 2005 do in fact show a thylacine.

Prior to Higgins' overlays, many couldn't see the thylacine due to the poor quality of the image, which, it has been confirmed, was a scan of a newspaper clipping of intentionally modified versions of digital photographs.

After Higgins' overlays, some still couldn't see the thylacine, but many others felt the overlaid 1933 image correlated far too closely to the 2005 image, leading to speculation that the 2005 images are hoaxes.

Here at Where Light Meets Dark, this author overlaid both the 1933 and the 2005 images with a series of lines and arrows to demonstrate the relative position of various features of each animal. Perhaps most strikingly, the pale fur between stripes seemed to align almost perfectly. This observation was here contrasted with Robert Paddle's observation that between animals, "enormous variability existed between banding patterns, concerning the total number of stripes, how far down the body and thighs they exteded, and whether they were straight or curved, tapering to a single point or ending in a fork." (Paddle, 2000). Although I noted that comparison between Mr Emmerichs' two photos preliminarily indicated that the animal moved its tail between shots, it seemed the case was set against these photos being genuine.

Analysis between the two 2005 photos is still to be carried further, but in the meantime, Alton Higgins brings us a new analysis aimed at squarely answering the question: "Can we dismiss Mr Emmerichs' photos on the basis that the animal's stripes closely resemble those from an animal photographed in the 1930s?"

And it would seem the answer is "No".

The next section describes in Alton's own words, his process for comparison of striping patterns between various photographs of thylacines.

Thylacine Stripe Spacing Consistency

"Recent photos of what appears to be a thylacine were taken in Tasmania by Klaus Emmerichs, a German tourist. I compared the Emmerichs photos to photos of thylacines and observed what appeared to be close matches of striping patterns. I looked at images of thylacines that were available via the Internet. Comparisons of thylacine stripes revealed that they differed in width, length, and possibly darkness. In addition, some stripes were forked. However, the spacing and positioning of thylacine stripes appeared to be quite consistent.

"I compared the stripe spacing patterns exhibited in what appeared to be seven different thylacines. Photos of the pairs of animals matched up are provided on the left side of each comparison. The stripes of the two thylacines were cropped and resized, placed back to back, then aligned. The top animal’s stripes are shown on the left side of the back-to-back alignments. Results of six combinations are shown. In each case, stripe positions were closely aligned.

"One aspect of possible significance regarding stripe pattern consistency pertains to the 2005 Emmerichs photos. Stripes visible in the Emmerichs photos appeared to exhibit a close match to photos of thylacines, leading to speculation that the 2005 photos were hoaxes based on old photos. However, because of thylacine stripe pattern consistency, as illustrated by the comparisons shown here, I do not think the Emmerichs photos should be dismissed out of hand based strictly on the fact that they compare favorably to known photos of thylacines."

Presented below is a composite of the six comparisons made by Alton.

Figure 07 (Click to enlarge)
Comparison in stripe patterns between different thylacines, a photo analysis

It can be seen from these comparisons of 7 individuals (including the pelt), that whilst variation in striping patterns did exist, their spacing and relative positioning along the spine were significantly consistent between individuals. Bearing in mind that Emmerichs' photos show far less clarity than the images used here (in part by design as they were "distorted ... somewhat to guard against reproduction" when sent to press) (Bailey, 2006), I confer with Higgins that Emmerichs' photos can not be dismissed on the basis that the striping pattern of that thylacine matches (in terms of spacing and relative stripe positioning) any other photographed thylacine.

How many stripes???

In researching this post, I asked the question - how many stripes have been documented on thylacines? There is no clarity in the answer...

"Across the back 20, on the tail 3" (Campbell undated, quoting Paterson 1805)

"There are 13-19 black vertical stripes that run from the midback to the base of the tail." (Ellis, 2002)

"it has 15 to 20 distinct transverse dark stripes across the back from shoulders to tail." (Dixon, 1989)

"with sixteen to eighteen dark stripes on its back and rump" (Wikipedia, 2006)

"19 to 25 dark brown or black stripes" (Rolland, 1997)

Depending on the reference, a thylacine might have as few as 13, or as many as 25 stripes. These may or may not extend from the back to the tail. Despite these references, most of the patterns shown in Higgins' analysis depict about 17 stripes, although the entirity of the tail is never shown. Given this consistency along the back however, it should be safe to assume that 17 to 19 stripes is the norm, and figures such as 13 would be the exception (or the possibly the norm for much younger animals than those shown here).

In other words, the fact that the thylacine in Emmerichs' photo depicts stripes in the same position, and of the same number, as in the 1933 photo used for Higgins' first analysis does not imply that the 2005 photos are digital manipulations of the 1933 photograph (or any other photograph).

This brings us back to the original intent of this series of articles - can we construct a model of the scene as Emmerichs photographed it, based on the change in position of the leaves between photos? And if so, did the animal change its position between photos?

More to come...

... but before we go, please note that it hasn't yet been ascertained that all the animals in Higgins' analysis (above) are distinct individuals.

References

Bailey, C. 2006 commenting on "Cryptomundo Exclusive: Thylacine Photos" (On-line), Cryptomundo. Accessed August 02, 2006 at http://www.cryptomundo.com/bigfoot-report/thylacine-photos

Campbell, C. undated "The Scientific Name" (On-line), The Thylacine Museum. Accessed August 02, 2006 at http://www.naturalworlds.org/thylacine/naturalhistory/history/history_1805_1936_1.htm

Dixon, J. M. 1989. "Thylacinidae" (On-line), in Fauna of Australia Volume 1B
Mammalia. Accessed August 02, 2006 at http://www.deh.gov.au/biodiversity/abrs/publications/fauna-of-australia/pubs/volume1b/20-ind.pdf

Ellis, E. 2002. "Thylacinus cynocephalus" (On-line), Animal Diversity Web. Accessed August 02, 2006 at http://animaldiversity.ummz.umich.edu/site/accounts/information/Thylacinus_cynocephalus.html

Paddle, R. 2000. "The Last Tasmanian Tiger: The History and Extinction of the Thylacine", Cambridge University Press, p 45.

Rolland, W. 1997 "The Tasmanian Tiger The Elusive Thylacine", Book Agencies of Tasmania, p 7.

Wikipedia, 2006. "Thylacine" (On-line), Wikipedia. Accessed August 02, 2006 at http://en.wikipedia.org/wiki/Thylacine

Interlude Continues (... about that spot) (part 4)

2006-08-01T00:14:00.000+10:00

In my last post, we took a look at the 1933 image used by Alton Higgins in his overlay series at Cryptomundo.

There was one peculiar element - a dark patch on the hind leg of the animal, visible in Mr Emmerichs' 2005 photograph, but absent in the 1933 photograph. At the time I speculated that by adjusting the contrast on the 1933 image, it might be possible that that dark patch would become visible.

In Figure 06 below, I've done just that.

Before we get to it, bear in mind that the 1933 photo may not be exactly to the same scale as the 2005 photo.

Figure 06 (Click to enlarge)
The bottom image (yellow background numbers) is the 2005 photograph by Mr Emmerichs. The centre image (green background numbers) is the 1933 image used in the overlay series. The top image (pink background numbers) is the 1933 image with adjustments made to contrast and brightness.

All illustration elements (arrows, numbers, etc) are identical in size and relative position between the three images. They were aligned using the arrow labelled "1" (described below).

The arrow labelled "1" shows the point at which the curve of the hind leg joins the curve of the tail base. This was used as a reference point to align all the other illustration elements.

The arrow labelled "2" indicates the dark patch found in the 2005 photo, but not in the 1933 photo as presented in the overlay series. By adjusting the contrast and brightness of the 1933 image, a similar patch of dark colouration can be detected in the 1933 image. The alignment is not identical, which may be due to the 1933 image being not necessarily to the same scale as the 2005 image.

The arc labelled "3" runs along the top edge of a similar dark patch found in the 2005 image. The bottom edge of this dark patch was ignored as it is likely obscured by the green leaf directly under the "3" label in the bottom image. Although there is a similarly shaped arc visible in the high contrast (top) version of the 1933 image, it is displaced significantly to the left.

The leaves labelled "4" and "5" were traced and filled with 40% opacity, white colour. These were imposed on all images to give a better indication of which areas of the thylacine would be obscured by the leaves. By examining those areas of the 1933 image which are not obscured, and comparing them with the 2005 image, the similarities in striping are emphasised.

The arrows labelled "6" indicate a pale, greenish patch on the 2005 image. During production a similarly shaped pale patch of fur was observed at the top of the foreleg in the 1933 image. The arrows were placed in order to compare the relative positioning of these patches with respect to the other elements examined in this analysis. Although the 1933 patch is similarly shaped, it is displaced significantly lower and slightly to the right when compared to the 2005 image.

This analysis shows that a similar dark patch exists in the 1933 image on the thylacine's hind leg, although the contrast in the 2005 image is greater.

In addition, two other features seem to appear in both images, although in this comparison they are not exactly aligned. If the 2005 image is indeed a digitally modified version of the 1933 print, then the poor alignment may be explained (in the case of label "3") by utilisation of a different sequence of adjustments to contrast and brightness, combined with a slightly different magnification.

The poor alignment in the case of label "6" may be due to a greenish patch (in the 2005 image) being superimposed on the original image and largely (although not entirely) obscuring the pale patch at the top of the foreleg.

This scenario is supported by the two arrows at point "2" in Figure 05 (see earlier post) which correlate to similar patches in the 1933 image. The dark fur between those two light patches in the 1933 image is not visible in the 2005 image due to the green leaf colouration.

The tonality of the leaf colouration at this point is consistent with the leaf labelled "5" in Figure 06 (this image. The colouration can better be seen in Figure 05) and it is also consistent with the greenish patch in question. In particular, the tonality is different to the tonality of the lighter markings between stripes as they appear next to that leaf in Figure 05, and also the two paler patches on the shoulder. Lastly, the shoulder patches do match the pale markings between stripes.

Although the observations presented in these two posts appear to present a strong case that the 2005 image is a modification of the 1933 print which is held by the Archives Office of Tasmania, there are contradicting observations which challange such a conclusion.

In particular, the initial analysis at Where Light Meets Dark observed a change in relative position of the tail between the 2005 photos. This needs to be further investigated and ideally the second 2005 image needs to be compared with the 1933 print. Finally, as noted earlier, the changes in leaf positioning may be able to be used to construct a model of the scene depicted in the photos. Using this model, a conclusion may be drawn about whether the animal moved between photos. If the 2005 images are digital modifications, then the process of creating a model of the scene may itself yield conclusions about the consistency of the images.

Stay tuned...

Interlude (part 3)

2006-07-31T23:01:00.000+10:00

We're between postings here at Where Light Meets Dark and over at Cryptomundo, Alton Higgins has submitted an excellent series of images which overlays the covering photograph from Robert Paddle's book, "The Last Tasmanian Tiger" onto one of Mr Emmerichs' February 2005 photograph.

The results are very interesting.

The match is more than a little perfect. As noted by Col Bailey of the Tasmanian Tiger Research and Data Centrein the original contribution by Mike Williams at Cryptomundo, the newspaper images were "distored ... somewhat to guard against reproduction."

My guess is that this accounts for the somewhat flourescent appearance of the thylacine's stripes. In fact, the greenish "stripes" are more likely to correlate with the usual tan fur colour of the thylacine. If you take a careful look at the photograph on Paddle's book (see centre image in Figure, below), there is quite a high contrast between the dark and light portions in that photo. In particular, the brightest areas tend to follow an arc, similar in shape to a rainbow. This correlates closely to Mr Emmerichs' photo.

Fig 05 (Click to enlarge)
This image shows the 1933 photograph depicted on the cover of Robert Paddle's book "The Last Tasmanian Tiger" (where it is credited to the Archives Office of Tasmania) at centre (black and white), together with two versions of Mr Emmerichs' February 2005 photograph (high contrast at top, and as published at Cryptomundo at bottom).

When considering the following points, note that the centre image may not be presented to exactly the same scale as the other two images. It was taken from the series of overlays presented by Alton Higgins at Cryptomundo.

The curve labelled "1" runs through the bottom edges of the highest-contrast white portions of the animal's stripes. It is identical in size and orientation in all three images.

The two arrows labelled "2" indicate two portions of high contrast pale fur colour on the animal's shoulder. The two arrows are equally spaced and oriented the same way in all three images.

The arrow labelled "3" shows a darker colouration which is visible in Mr Emmerichs' photo, but not in the 1933 print. Further adjustment of the 1933 image may yet indicate that this patch is present in the 1933 photo also. This analysis will be presented here at Where Light Meets Dark.

The arrow labelled "4" shows a significant change in the curve between the animal's hind leg and the thick base of the animal's tail. The deviation is present in all images in approximately the same location.

Arrows 3 and 4 are equally spaced and oriented the same way in all three images.

The correlation between Mr Emmerichs' photo and the 1933 print is fairly high, even with the poor quality reproductions.

Perhaps the most obvious correlation hasn't yet been mentioned here: the number positioning and spacing of the white portions between the dark stripes all correlate strongly also.

Paddle (2000) notes that "observation of photographs of live and dead thylacines demonstrates that enormous variability existed in banding patterns, concerning the total number of stripes, how far down the body and thighs they extended, and whether they were straight or curved, tapering to a single point or ending in a fork".

Given that the norm is for difference in appearance, and similarity is the exception, Mr Emmerichs' thylacine is looking all the more remarkable if it is a genuine 2005 specimen.

However, as previously noted, photographs (no matter how clear) will almost certainly never provide conclusive evidence of the existance of the thylacine. Contrawise - the fact that the thylacine in Mr Emmerichs' photo so closely correlates to a known photo of a 1933 specimen, does not automatically guarantee that the photo is not genuine.

The plot thickens.

References

Paddle, R. (2000) 'The Last Tasmanian Tiger: The History and Extinction of the Thylacine', Cambride University Press, 45.

Let the games begin! (part 2)

2006-07-30T09:31:00.000+10:00

What can we do with scans of newspaper clippings of digital images? Clearly if you've seen the images at Cryptomundo (be sure to click on them for enlargements), their quality leaves a lot to be desired.

However, as I've argued in the commentary at Cryptomundo and here at Where Light Meets Dark, the composition of the images gives us plenty to work with, but only because we have two photos which have been taken moments apart.

[Side note: The analysis presented here is being conducted piece-wise, in that I will publish results as I get them. Analysing images in this manner takes time, but with the discussion running hot amongst cryptozoologists worldwide, I felt it pertinent to get the argument out there for discussion.]

Apples with Apples

The first obstacle we have to contend with is to ensure the images are resized so that they are to scale. In order to do so, we need to find an element which is present in both photos. Fortunately, there is just such an element - a leaf, which I will call L1 (leaf 1).

The smaller image at Cryptomundo will be called image S. The bigger image at Cryptomundo will be called image B. (The term "larger" was avoided to avoid confusion with using the abbreviation 'L', which in this discussion refers to a leaf).

Fig 01 (click to enlarge)
A close up of both images showing leaf L1. A line was drawn on image B which spans leaf BL1. The image was zoomed to 200% (x2) and the line adjusted until its ends were directly over the leaf edges at relatively identifiable points. A copy of the line was positioned over image S. Image S was iteratively proportionally scaled until the line spanned the leaf SL1 in the same way as the first image. Image S was zoomed to 200% (x2) to confirm the ends of the line were directly over the edges of the leaf. The two lines are identical in length and angle. Screen captures at 200% zoom were taken and positioned side by side to produce this figure.

Note that in figure 01 the orientation of leaf L1 is identical between images. This is relevant to the discussion on camera rotation.

The context

Let's step back a little and see exactly where on the images leaf L1 can be found.

Fig 02 (click to enlarge)
This figure was created to get an overview on how the two images look in relation to each other. The smaller image (S) has been enlarged, and leaf L1 has been horizontally aligned.

Note that many of the other features visible in the image have changed relative position. As per Figure 01, the black lines through leaf L1 were used to ensure the accuracy of the magnification of image S.

Double Check

At this point, let's double check the magnification that was carried out. Another area of the image which is visible in both photos is that depicting the two strips on the rump of the thylacine. Well almost. As you'll see from the discussion of this Figure (below), a small portion of the stripes in one image is obscured.

Fig 03 (click to enlarge)
The rump. This image shows the rump area of the thylacine in both the bigger (B, top) and smaller (S, bottom) images. Each image of the rump is shown with measurement lines (at right) and without (at left). The first measurement made was of the width of the two stripes. The smaller image was measured, then the line duplicated and placed over the bigger image. This author finds the comparison inconclusive for two reasons; the top of the right-most stripe in image B is partly obscured by a leaf and the distinction between the somewhat falsely coloured stripe and the leaves in the image background is difficult to discern. As such, the accuracy of the placement of the lines is questionable.

The second (horizontal) line was placed at the widest visible point between the tail and hind leg in image B. The line was then copied and aligned with the hind leg in image S. It is clear the extent of the line protrudes over the tail of the animal but no conclusion can be drawn as it cannot be ascertained that the line has been placed at the same vertical distance from some fixed reference point (such as the ground or the rump); this lead to figure 04

Inconclusive

Let me emphasise that I found the analysis involving Figure 03 to be inconclusive. However, this analysis inspired Figure 04 (below). If we can vertically fix the position of the horizontal line in Figure 03, from some known point, then we can ask whether the tail has moved with respect to the rear portion of the hind leg.

Fig 04 (click to enlarge)
Figure 04 was created to address one shortcoming in figure 03 – namely the unreliability of the horizontal measurement in figure 03. In figure 04 a vertical line was added to image B which joined the end of the rump measurement to the horizontal measurement between the leg and the tail (at its widest visible point in that image). The indicated length of the line (5.95cm) is irrelevant. The same line was copied and duplicated on image S (line 2). Horizontal line 3 was then moved vertically into position so as to touch line 2 in the same way as shown on image B. Line 3 was also moved horizontally into position so as to touch the hind leg in the same way as shown on image B. The horizontal lines are of identical length between images. In image S, the horizontal line (3) clearly protrudes significantly over the tail for more than 50% of its thickness.

Note that for the horizontal positioning of the line, three visible points on the hind leg are indicated by the arrows. The exact point of contact between line 3 and the hind leg is obscured (in image S) by a mark which resembles vegetation, lying between arrows 2 and 3. The arrows provide points of reference either side of the contact point allowing the contact point to be interpolated.

Additional information is yielded by comparing the line subtended by the three arrows in image S (ie the angle of the hind leg) with the angle of the hind leg in image B; the hind leg appears to be “leaning slightly to the right of vertical”; that is, the top of the line is positioned slightly further right than the bottom. Looking at a point on the hind leg in image B which is equivalent in vertical distance from the horizontal line as that shown by arrow 1, clearly shows the leg “curving to the left” as you progress up the leg.

From this figure we can conclude that
• the leg is depicted at a different angle in each image
• and the tail is further from the leg in image B

One explanation for these observations is that in fact the photos depict a real thylacine or a thylacine taxidermy, and the tail moved with respect to the hind leg between photos. In addition or separately, the hind leg changed position with respect to the camera.

However, before a conclusion can be drawn, consideration must be given to alternative explanations. At this point we analyse the possibility that the photograph is of an inanimate object and question how the observed differences between the photos could be achieved by movement of the camera relative to the animal. It is sufficient to exclude all other photographed objects from the analysis as they have no bearing on the way in which the camera perceives the animal. Subsequent analyses will continue to examine the relative positioning of other elements in the photographs with the intent of determining whether or not consistency exists between the image.

(In plain terms, further analysis will continue to investigate whether it’s plausible to assert that the animal moved relative to its environment between photos.)

Tasmanian Tiger Photographed! (part 1)

2006-07-29T04:05:00.000+10:00

The photographs

In February 2005, Klaus Emmerichs and Birgit Jansen allegedly captured a thylacine, or Tasmanian tiger in two digital photographs. The couple was touring Tasmania from Germany via a stop in Melbourne where Mr Emmerichs' brother lives.

The couple claim they did not know the animal was extinct, took two photos, then the animal walked away. Later they realised they'd photographed an animal which hasn't been confirmed alive since 1936.

However, the couple had shortly to return to Germany, via Melbourne. In Melbourne they gave copies of the photographs to Mr Emmerichs' brother and allowed him to decide what to do with the images.

The brother flew back to Tasmania where several experts examined the images and agreed they depicted portions of a thylacine otherwise obscured by foliage. They were clear, however, that they could make no comment on the authenticity of the images unless they could be left for forensic investigation. However, the brother took the images back to Melbourne and began contacting media agencies regarding sale of the images.

Over time, the story drifted out of media attention, the images were apparently never published, and things went quiet.

"Stand by their claim"

In April 2006, the story came back into the media limelight when Klaus Emmerichs and Birgit Jansen returned to Tasmania, this time armed with a video camera and determination to prove their original sighting was no hoax.

The photos were supplied to the Sunday Tasmanian (newspaper) without charge, and published. The photographers took the opportunity to make some clarifications regarding their mysterious disappearance back to Germany after the sighting. It was never their intent to profit from the images, although now that they have been published in Tasmania, they are requesting a $1000 fee for publishing interstate.

In addition, the couple responded to several lines of questioning targeted at the image.

Photographers allege the blurring is not consistent with auto focus problems. Mr Emmerichs explains that a night-vision mode was set which increases exposure time, resulting in blurring.

Photos on the camera taken in January have the text 'JAN' to represent the month in their date stamp. However photographs taken in February, including the thylacine images, use the numeral '02' to represent the month. Mr Emmerichs explained that Brigit attempted to alter the camera's clock from German time to Australian time during the flight to Tasmania, also resulting in the different date stamp format.

It has been alleged the photos resemble a famous photo from the 1930s and to date that allegation has not received a response.

So what are we left with?

This is where it gets interesting.

The photos are available for viewing at the popular cryptozoology website, Cryptomundo. Be sure to click on them to view enlargements.

In typical commentary fashion, much debate has commenced on the merit of the photos, including many passionate comments contributed by this author.

The theme for this blog - Where Light Meets Dark - has today been switched to take a good analytical look at these photographs and to make attempts at ascertaining the viability of the photographs being genuine.

Unfortunately, it is 4.24am and I really must sleep, but stay tuned for discussion on the facts and allegations as we have them.

Despite claims that scans of newspaper clippings of digital photos are so degraded as to be worthless, I feel quite to the contrary. One crucial ingredient is the fact that we have two photos, and that gives us plenty to work with...

References

The Mercury News, 16/4/2006, "Picture pair insists their tiger's no fake", accessed 8/7/2006

Cryptomundo, 27/7/2006, "Cryptomundo Exclusive: Thylacine Photos", accessed 29/7/2006

Further Reading

The Thylacine Museum
Thylacine at Wikipedia