Determining Distances from Photographs

Discussions about the history of the ship, technical details, etc.

Moderator: Bill Jurens

iankw
Member
Posts: 196
Joined: Wed Nov 10, 2004 11:41 pm
Location: Rotherham, England

Post by iankw »

Bob, you still seem to be missing what is obvious to me, but then I'm not actually bothered enough to do the maths (especially when there are others that are, or those that say the whole concept is inaccurate because we don't have the original negs). PG did more turns and, hence, lost more speed!!! Surely that could account for the difference?

You don't like vague figures like 2500 - 3000m? The whole idea of using these photos is bound to lead to vague figures, that is how I read it anyway (sorry if I read that wrong Bill). You like one figure with a margin for error? What about 2750 +/- 250 then? Is that any better? What margin for error would you accept? I'm sorry Bob but all I see here is an attempt to reopen the whole "Bismarck didn't turn away" argument. I apologise if I have misunderstood you here.

regards (sincerely)

Ian
Robert J. Winklareth
-
Posts: 116
Joined: Wed Oct 20, 2004 5:58 pm
Location: Woodbridge, VA USA
Contact:

Post by Robert J. Winklareth »

Hi Ian,

By actual measurement of the distance traveled by the Prinz Eugen from 0556 to 0609 on her Battle Sketch, she lost about 200 yards along a vector of her baseline course of 220 degrees as a result of her several turns. That’s a plus 200 yards for the Bismarck. If the Bismarck made any comparable turns, then that would be a 200-yard loss for the Bismarck, bringing us back to square one. If the Bismarck had not made any turns at all, then all we’re talking about is 200 yards difference at the most, not the over 1,200 yards now being discussed.

You certainly have a point about the tolerances mentioned. Unfortunately I have the handicap of being an engineer, and my slide rule never give me a range of answers. Nor do I remember seeing simple algebraic relationships that resulted in a range of values. When I see a range of values like 2500-3000 meters, I think of an off-the-wall estimate that could just as well turn out to be anywhere between 2000 and 3500 yards. Sorry, but that’s my outlook.

Regarding your last comment, that certainly is not the case. In fact, any discussion of my concept of the battle and reversed photo theory is off limits for discussion under this topic. We are required to limit our discussion to the determination of distances from photographs, and while we did have some digression from the subject a short while back, we are now back on track, so let’s keep it that way.

Best regards.

Bob
iankw
Member
Posts: 196
Joined: Wed Nov 10, 2004 11:41 pm
Location: Rotherham, England

Post by iankw »

I'm glad to hear it Bob. I'll go back to watching and reading the thread with interest, hoping to learn a bit of something about trig. Re being an engineer, you surely learned about +/- tolerances? Ok, so your slide rule gives you a set figure, but you have to calculate the errors (by the way I have a slide rule but I was never taught to do anything other than simple multiplication with it - my maths teacher let us down there). When I did my science degree at college I always hated the calculation of errors, but it was an essential part of any practical prosedure. My bro is a high-flying prof of surface science at an English Uni and I remember watching him plotting error bars on a graph of results many years ago.

Ok, back to my corner :wink:

regards
Bill Jurens
Moderator
Posts: 876
Joined: Mon Oct 18, 2004 4:21 am
Location: USA

Off-the-wall

Post by Bill Jurens »

The photogrammetric equations are rock-solid and will give correct answers to any degree of precision one wishes. Unfortunately, accurate photogrammetric analysis requires the precise measurement of very small distances -- sometimes to within a couple of thousandths of a millimeter in size -- and this, in turn, requires both very high quality photography and extremely precise measuring equipment. Topographic mapping via aerial photos commonly employs 9" x 9" negatives exposed in cameras that typically cost in the hundreds of thousands of dollars, analysed on machinery that typically costs tens of thousands of dollars more. This is not mickey mouse stuff, and even then, under ideal conditions, it's difficult to get accuracy exceeding one part in 5000.

The quality of the photographs which we have is, in photogrammetric terms, extremely bad. The photos were (apparently) exposed hand-held, via what were, in photogrammetric terms, fairly crappy non-metric non-fixed focal length cameras, equipped with non-collumated uncalibrated 'consumer grade' lenses, on misalgned bedplates using what was probably consumer grade film. After exposure, the negatives, which appear to have been badly developed, were fed through one-or-more enlargers of totally unknown quality and the resulting prints re-screened or re-photographed by who-knows-who using heaven only knows what sort of enlarging lens, probably for passing around the wardroom.

As a rough guideline, one might note that the typical resolving power of lenses of this period when imaging images of average contrast in short focal lengths using normal type films (e.g. Super XX pan) developed under average conditions was only about 1000 reciprocal angular radians, i.e. about 0.057 degrees. The tangent of 0.057 degrees is about 0.001, meaning that the virtual resolution of a target 2500 meters away is only about 2.5 meters, i.e. items less than about 2.5 meters in size would be pretty much invisible. That's a fair proportion of the absolute size of the target.

Taking precise measurements on such absolutely terrible photography is problematic at best, although approximately accurate results can probably be obtained by using a number of independent methods and checking for correlations (or, more importantly, lack-of-same). The ambiguity in distances expressed in my findings relates directly to the quality of the image, and the difficulty of attempting to measure very accurate distances on what is, in reality, a very blurry and poor quality image.

The results obtained may not be as precise as one wishes, but are all that can reasonably be expected considering the raw material at hand. The resultant figures may not be precise, but they are certainly far from 'off-the-wall' estimates.

For good information on the typical quality of photography in the 40's, one might refer to G.C. Brock's book "The Physical Aspects of Aerial Photography". This defines the typical precision (and limitations) of 1940's vintage aerial photography in detail. The results are equally applicable to the imagery of Bismarck, except that the camera was pointed horizontally in one case, and vertically downward in the other.

Bill Jurens.
User avatar
Antonio Bonomi
Senior Member
Posts: 3799
Joined: Mon Oct 18, 2004 10:44 am
Location: Vimercate ( Milano ) - Italy

Determining photo distances

Post by Antonio Bonomi »

Ciao all,

I agree with Bill Jurens wise suggestion : ... '' approximately accurate results can probably be obtained by using a number of independent methods and checking for correlations ''.

That is basically what we have been doing so far among various persons on the firts salvo photo ( nh 69722 ).

Some examples :

My trigonometry calculation based on '' tgx method '' was not dependent on focal length and provided 2300-2500 meters.

Marty pixels method showing 2400 meters, with no focal lenght dependency too.

Herr Nilsson with is pixels method measuring 2749 meters ( +/- 200 meters ), with no focal length dependency as well.

Bill Jurens several evaluation methodologies explained on above post's determining a distance between 2500-3000 meters.

At this point, I hope we all agree about the evaluation of 2500 meters on the first salvo photo as an acceptable average estimate.

Ciao Antonio :D
Robert J. Winklareth
-
Posts: 116
Joined: Wed Oct 20, 2004 5:58 pm
Location: Woodbridge, VA USA
Contact:

Post by Robert J. Winklareth »

Hi all,

For the time being, I will restate my position that NH69722 was taken with a normal lens of 50mm focal length and that based on my methodology, the distance to the Bismarck in NH69722 was 2700 yards. I believe that this is in agreement with the views held by Antonio and others.

I have to qualify my answer pending resolution of the discrepancy between the results obtained from the photographic analysis of NH69722 versus the time-distance relationships associated with the relative tracks and speed of the two German ships during the battle.

Bob
Robert J. Winklareth
-
Posts: 116
Joined: Wed Oct 20, 2004 5:58 pm
Location: Woodbridge, VA USA
Contact:

Post by Robert J. Winklareth »

Hi all,

I believe that we have now exhausted discussion on NH69722, so why don’t we leave that subject for now and move on to succeeding photographs in the series. If we make any progress with those, we might be able to work backwards into NH69722 and eventually crack that nut.

As early as 1983, over 20 years ago, I came up with a sequence of photographs of the Bismarck in battle by viewing slide copies of published photographs om my slide viewer. The series began of course with NH69722, and that photograph was followed by NH69729, NH69730, and the uncataloged broadside view of the Bismarck as she was passing the Prinz Eugen off the port beam of the cruiser.

I have maintained the same position on the sequence of those photographs ever since, and I still believe that it represents the actual progress of the Bismarck during the first half of the battle. That sequence shows the Bismarck directly astern of the Prinz Eugen at 0556 (NH69722), moving up on the port side of the Prinz Eugen at 0559 (NH69729), off the port quarter of the Prinz Eugen at 0601 (NH69730), and passing the Prinz Eugen at about 0605 (broadside view).

My initial assessment of the sequence of those photographs was later confirmed by the Prinz Eugen’s Battle Sketch, which shows the Prinz Eugen traveling on a straight-line course of 220 degrees until the first of her series of turns at 0603, and the Prince of Wales Salvo Plot, which shows the Bismarck traveling on an overall course of 212 degrees from 0553 until 0602.

Those two documents together establish that the Bismarck was sailing on a course 8 degrees to port of the Prinz Eugen’s baseline course of 220 degrees, and with the Bismarck directly astern of the Prinz Eugen at 0556, the Bismarck would have been coming up on the port side of the Prinz Eugen after that. The first four photographs in the series are entirely consistent with the documentary evidence showing the Bismarck coming up on the port side of the Prinz Eugen during the initial stage of the battle.

Besides being consistent with the photographic evidence, the first four photographs tend to confirm Admiral Lutjens intent to have the Bismarck come between the Prinz Eugen and Prince of Wales to protect the Prinz Eugen from enemy fire. This was in accordance with established German naval operating procedures to place lighter armored vessels on the “lee” side of the action.

Antonio and possibly others have come up with different arrangements of the photographs to support their own points of view. I am not challenging the validity of those other concepts, but merely pointing out the validity of my alternative. I would ask that the others refrain at this time from introducing their concepts and accept my sequence as a viable alternative for the purpose of further discussion of the determination of distances from photographs. They are certainly free to bring up their own points of view after discussion on this arrangement has been concluded.

In my original posting, I indicated that NH69730 might have been taken with a 90mm lens because the results showed the Bismarck to be less than 400 yards away from the Prinz Eugen, which seemed to be implausible. After reexamining the photograph, I now believe that my original assessment may have been wrong and that NH69730 was actually taken with a normal lens having a focal length of 50mm.

Unlike NH69722, there was no indication that NH69730 had been cropped. In fact, all of the published versions of that picture show a smaller area of coverage than in NH69730, including the one in the book “Prinz Eugen in the First Battle,” which proved that NH69722 had been cropped. However, NH69730 did not fill the 2 x 3 format inherent in 35mm pictures, so some cropping must have taken place.

After enlarging the height of NH69730 to conform to the 2 x 3 format, I applied my methodology using a normal 50mm focal length lens. I arrived at a figure of 504 yards, which of course can be rounded off to an even 500 yards, as the separation between the Bismarck and Prinz Eugen at 0601 when NH69730 was taken.

While I don’t subscribe to using raw judgment to determine the distance to the Bismarck in photographs, I do believe that judgment can be used to assess whether the result obtained by more sophisticated means appears to be reasonable. Considering the size of the Bismarck and amount of detail seen in NH69730 compared with the other photographs, I personally believe that my new figure of 500 yards does appear to be reasonable.

A distance of 500 yards represents somewhat less than two ship lengths of the Bismarck considering that she was 823+ feet (275 yards) long. I found that it was a lot easier to visualize the progress of the Bismarck from one photograph to the next in a specific sequence of views by ship length rather than by hundreds of yards in my time-distance analysis. Here again in my judgment the 500-yard distance appears to be reasonable.

I welcome any comments from others on my assessment of NH69730 in the context of it following NH69722 and NH69729 in the sequence of views of the Bismarck in action.

Bob
Robert J. Winklareth
-
Posts: 116
Joined: Wed Oct 20, 2004 5:58 pm
Location: Woodbridge, VA USA
Contact:

Post by Robert J. Winklareth »

Hi all,

In his posting of 24 February, Bill Jurens made some statements with which I strongly disagree, and since they impact on our efforts to analyze the photographic evidence on hand, I feel compelled to respond to his posting for the sake of our continued pursuit of this subject.

Over the past few years, Bill has often stated that no meaningful information can be derived from an analysis of the photographs of the Bismarck in action because of their poor quality. While the quality of the photographs does leave much to be desired, nevertheless some of them have provided us with a wealth of information, including NH69722, which has been the subject of intensive study and discussion on this forum.

By enlarging NH69726 to a magnification of 15 x, I was able to perform an “autopsy” (detailed critical analysis) of that photograph, identifying over a dozen points of comparison with top quality photographs of the Bismarck. The amount of detail revealed by that enlargement is truly amazing, and it led to several irrefutable conclusions regarding the orientation and direction of fire of the Bismarck in that photograph.

For example, the two bright spots seen amidships of the Bismarck at the level of her main deck turned out to be rather sharp images of the center and rear secondary gun turrets on the post side of the ship. Even the barbettes of those gun turrets were readily distinguishable. All of the reflections of the flash from the Bismarck’s heavy guns could be positively identified with specific structural components of the ship.

Now for whatever reason, Bill demeans even the equipment and materials used to provide those photographs. For his information, the 1936 vintage Leica and Zeiss Contax II cameras are still considered today to be among the best quality cameras ever produced. Zeiss optics, which were also used on German military cameras, were regarded as being the best in the world. German-made “Agfa” photographic film also had a reputation of being of the highest quality.

I agree that the processing of the film was sloppy, considering the size and number of spots, scratches, and other imperfections seen in the photographs. That, however, does not seriously detract from our ability to derive significant information from those photographs. All we need is a positive attitude and strong desire to resolve the remaining issues associated with achieving the true scenario of events that transpired in the Denmark Strait on 24 May 1941.

I see very little correlation between aerial photography and the surface photography that we are dealing with here. Pointing the camera downward does not give us results equally applicable the imagery of the Bismarck taken with the camera pointed horizontally, as Bill suggests.

In aerial photography, we are dealing with a flat plane, i.e., the surface of the earth, which is at a specific distance away from the camera. Any measurements of the height and depth of the terrain are relatively small and require the use of stereoscopic equipment to determine variations in elevation.

In surface photography, we are dealing with a very wide range of distances from the camera and a far greater depth of field. Unfortunately, our photographs were not taken with stereoscopic equipment that would give us precise results, so we must use other means to determine distances from photographs.

Bob
Bill Jurens
Moderator
Posts: 876
Joined: Mon Oct 18, 2004 4:21 am
Location: USA

Photogrammetry

Post by Bill Jurens »

You can disagree all you like, but I'll stand by my statements, with a few comments.

I will reiterate that the analysis of these photographs in many cases amounts to a Rorsach test. With apologies to Simon and Garfunkel, "... A man sees what he wants to see and disregards the rest..." You wouldn't believe the types of mistakes in interpretation people make in blind testing situations -- it would curl your hair. If Mr. Winklareth can find any professional photogrammetrist who would claim that the photography we are looking at is capable of anything other than the crudest photogrammetric analysis, I would certainly like to know whom they might be.

I will stand by my statement regarding handheld 35mm cameras in general. I know good cameras, as I run large format Linhof cameras myself, and nobody would consider a Linhof as other than top-notch item. But, good as they might be for documentary photography or photo-journalism, 35mm cameras -- including the Leica and the Contax type -- are simply not very good photogrammetrically. There are several reasons for this, of which I will quote only a few off the top of my head. Commericial' cameras used for ordinary photography never utilize fully calibrated lenses and never mount lenses where the focal length is correctly calibrated. Actually, rather than mounting calibrated lenses in carefully machined fixed mounts, focussing (often by rotating the lens) far from being forbidden, is actually encouraged! this is bad news, photogrammetrically. Further, the film plane in 35mm cameras is generally not flat or held to be flat via a vacuum back or similar device. It's generally pretty flat, but in photogrammetry, pretty flat is not flat enough. Further, no 35mm camera contains fiducial marks which allow reconstruction of the principal point. And the format is very small, rendering the image on the negative very small indeed. Many mapping cameras use roughly 9" x 9" negatives giving a useful negative area approximately 50 times larger than a 35mm film frame. And the lenses, comparatively, suck. The most expensive commercial lens is probably not as good as the worst photogrammetric lens.

All of this means that even a very good 35mm camera is pretty bad photogrammetrically. Many years ago I put the best quality lenses I could find on a very carefully aligned Linhof 4x5 camera to see how it would stack up with an aerial mapping camera. It stank. Of course it did, it cost less than 1/10th the price of a proper photogrammetric job. And as anyone who has bought one can testify, Linhofs aren't cheap! (Very nice for other photography, though...). Are Leicas and Contaxes good cameras? Sure. Are they photogrammetric cameras? Nope.

Obviously only one who has never tried to map places such as Switzerland could make the statement that the surface of the earth is a flat plane.

Hope this helps...

Bill Jurens.
Robert J. Winklareth
-
Posts: 116
Joined: Wed Oct 20, 2004 5:58 pm
Location: Woodbridge, VA USA
Contact:

Post by Robert J. Winklareth »

Hi all,

No one would argue that photographs taken with a 35mm camera are of equal quality as those taken with a professional large-format camera like the Linhof camera, or that a 35mm camera is suitable for aerial photography. All that I said was that a 35mm camera, such as the Leica and Contax II, are capable of taking photographs of sufficient quality for our purposes.

We don’t have to see the rivets or the weld lines on a ship, only a suitable image to determine the orientation and inclination of the Bismarck, the direction in which she is firing, and sufficient resolution to measure the height of the tower mast from the level of the sea for comparison purposes in our analysis of the distance to the Bismarck

I really don’t know why Bill Jurens would want to rehash his old arguments again at this time. It serves no useful purpose and it only detracts from our efforts to derive as much intelligence as possible from the photographic evidence at hand. I would welcome any positive assistance that he can provide.

Bob
User avatar
Antonio Bonomi
Senior Member
Posts: 3799
Joined: Mon Oct 18, 2004 10:44 am
Location: Vimercate ( Milano ) - Italy

Re: Determining Distances from Photographs

Post by Antonio Bonomi »

Hello everybody,

somebody is putting in discussion the photos taken with and on the PG film timeframe showing shells landing from Hood ( based on archives wrong captions ) instead that coming from PoW as they were.

Here some examples for the photo NH 69728


NH_69728.jpg
NH_69728.jpg (71.63 KiB) Viewed 2982 times

Bild-146-1990-081-10A_equivalent_NH_69728.jpg
Bild-146-1990-081-10A_equivalent_NH_69728.jpg (85.24 KiB) Viewed 2982 times

IWM_382_equivalent_to__NH_69728.jpg
IWM_382_equivalent_to__NH_69728.jpg (71.48 KiB) Viewed 2982 times

.... continuing
In order to honor a soldier, we have to tell the truth about what happened over there. The whole, hard, cold truth. And until we do that, we dishonor her and every soldier who died, who gave their life for their country. ( Courage Under Fire )
User avatar
Antonio Bonomi
Senior Member
Posts: 3799
Joined: Mon Oct 18, 2004 10:44 am
Location: Vimercate ( Milano ) - Italy

Re: Determining Distances from Photographs

Post by Antonio Bonomi »

Hello everybody,

continuing ...

Baron-Book-NH_69728_caption_Italian.jpg
Baron-Book-NH_69728_caption_Italian.jpg (90.49 KiB) Viewed 2979 times

and I can continue with many more examples of an incorrect captioning, .... sometimes correct, ... but never so precise anyway.

Here we have the precise caption as it should be :

NH_69728_correct_caption.jpg
NH_69728_correct_caption.jpg (62.64 KiB) Viewed 2979 times

continuing ...
In order to honor a soldier, we have to tell the truth about what happened over there. The whole, hard, cold truth. And until we do that, we dishonor her and every soldier who died, who gave their life for their country. ( Courage Under Fire )
User avatar
Antonio Bonomi
Senior Member
Posts: 3799
Joined: Mon Oct 18, 2004 10:44 am
Location: Vimercate ( Milano ) - Italy

Re: Determining Distances from Photographs

Post by Antonio Bonomi »

Hello everybody,

continuing ….

Now lets see why it can only be like that.

Because that photo was taken when the PG film started few seconds after, and consequently it was after the torpedo alarm issued by Prinz Eugen at 06:03, … and in fact the Bismarck main turrets are rotated aft the beam by 35° as everybody can see.

NH_69728_correct_main_turrets_bearing_aft_35_degrees.jpg
NH_69728_correct_main_turrets_bearing_aft_35_degrees.jpg (45.17 KiB) Viewed 2979 times

and here we have the PG film providing the correct timing of that photo, since PG is turning west on course 270°, ... as the film does demonstrate, ... from 06:03 and 30 seconds until soon after 06:05.

NH_69728_correct_timing_06_03_45 seconds.jpg
NH_69728_correct_timing_06_03_45 seconds.jpg (82.65 KiB) Viewed 2979 times

It does not take much nwo to realize that after 06:03 and 30 seconds battle time, ... the Hood was already under the waves completely.

Consequently that shells can only come from PoW after 06:03, ... so it was a local control shell fired by the Y turret obviously, ... as I have demonstrated above ... and it is confirmed by the PoW gunnery plot and narrative.


Regarding this photo the IWM and Bundesarchiv captions are simply wrong, ... the NH Center and the Baron were correct but not so precise.

End of the demonstration for the NH 69728 photo.

Bye Antonio
In order to honor a soldier, we have to tell the truth about what happened over there. The whole, hard, cold truth. And until we do that, we dishonor her and every soldier who died, who gave their life for their country. ( Courage Under Fire )
Post Reply