|
Post by Peter Jenssen on Aug 23, 2016 14:32:50 GMT
I would be very interested in this kind of photogrammetry software as well. The first time I heard about it was probably over ten years ago now. I don't recall who were discussing it or on which forum anymore. What I do recall is that the software already existed then. And the context was to make ship model plans from photos of ships. (I was then toying with an idea for a bottle model of CB Pedersen, and had collected some photos) There should be considerable progress made since then?
Back then, there were hardly any results when searching. These days, the problem is quite the opposite. How to find the appropriate one. As Fred pointed out, it gets tricky if the pictures are not taken with the same camera.
Cheers, Peter
|
|
|
Post by jules on Aug 23, 2016 14:55:14 GMT
Hello Jan (amateur), Now we've come to accept that vertical frames can look tilted in photographs; the Blom model showed that. And now we've come to accept that the angle of the gun ports can be a useful general indicator of the framing orientation; as Fred believes. We can now check if the Hohenzollern model gun ports are tilted. (That's what I've been trying to say all the time: no need to find all the nails anymore, we can focus on the tilt of the gun ports. And therefore can use the long distance pictures of the model. And as long as no one can actually show me that the nails on the model correspond to actual frames underneath, I am not willing to accept that these nail patterns represent evidence of the framing of the model, let alone the real ship.) I included a photograph of the model showing its complete starboard side. (I simply do not understand why I have to look at a composition of five separate photographs, which all show different viewing angles, when we have such awesome photographic material available: the whole ship captured in one photograph. Here is a detail of Peter's composition. How does this show in which direction the gun ports or the nailing patterns were tilted?) Here is the all in one photograph. This photograph is taken from a longer distance from the model, and therefore shows less distortion than the photographs taken from a shorter distance. In my view, this photograph does not show the gradual tilting Peter describes. When we look at the gunports on the lower deck, the vertical gun port sills of most of the ports seem to be vertical, except the ones in the front. But that can be explained by photographical distortion. When we look at the gunports on the second deck, most of the gunports seem to be vertical, except the ones in the back. But again, that can be explained by photographical distortion. When we look at the gunports from both decks combined, we see the following. I included two details of the photograph above; one from the back of the model, one from the front of the model. We see that the gun ports from the second deck do not line up vertically with the gun ports from the lower deck. How can a frame have a certain tilt on one deck, and another tilt on the other deck? This effect, in my view, can only be explained by photographic distortion. But, can all these effects really be explained by photographical distortion? In short: when we assume that the frames in a model tilt, we have to prove that the gunports in any photograph of that model are actually tilted, and not only look tilted. As long as the tilt can be explained by photographic distortion, the photograph kan not be considered as proof for the tilting. So I prepared a couple of sketches in my 3D-drawing program. This program, like almost all 3-D drawing programs, is able to show the three views we all know, top-, front- and right view, and also able to show a perspective view. This perspective view is created by using a device that works like a camera. You can freely choose the camera position, and you can even choose lenses for this camera. With this program I drew some simple solids; one is a rectangular, flat solid, the other is a curved solid, and put them together to represent a piece of ships hull. Here they are. When we take a 'photograph' of a section of these drawings, we can make it look like this:
These two pictures, in my view, show that we are able to recreate the effects we see in the photograph of the model by photographic distortion. This means that we can not prove that the ports are tilted, and therefore can not prove that the frames are tilted.
Photographic evidence simply is not enough to establish if the gunports (frames) are tilted or not. You have to take careful measurements to establish if the ports of a model are vertical or tilted. You have to place the model exactly on its waterline and determine; preferably with a plumb bob, if the sills of the ports are vertical or tilted. Since the waterline under 'stuurlast' of the Hohenzollern model was clearly inscribed into the hull, and since Winter took the effort to make some very good measurements, and since the Winter drawing shows the gun ports to be vertical to this waterline, I incline to say that the ports are vertical and not tilted.
But, again, I think photographic evidence on its own can not constitute enough proof for making statements about frame positioning or tilting. For example, when we look at the ladder hanging from the side of the hull in the 'big' picture: is it vertical or tilted, is it straight or curved? I would say vertical and straight. And you, based on this photographical evidence alone?
Many kind regards,
Jules
|
|
|
Post by amateur on Aug 23, 2016 15:05:22 GMT
I always think (but are by no means sure) that it is straight, and following the curvature of the hull to some extent.
Jan
|
|
|
Post by tromp on Aug 23, 2016 16:08:10 GMT
Aug 23, 2016 14:55:14 GMT jules said: Hello Jan (amateur),
(I simply do not understand why I have to look at a composition of five separate photographs, which all show different viewing angles, when we have such awesome photographic material available: the whole ship captured in one photograph. Hello Jan (amateur),The whole ship captured in one photograph shows exactly the tilting effect you want to see and you want us to see, as in the Otte Blom model picture. You can't and you won't get rid of it. That's why I showed you close-ups composed of four individual photographs that do not have this effect but still Show the tilting. However, this is apparently something you fail to understand (or want to understand) and that's why you're showing it to us. ...But, can all these effects really be explained by photographical distortion? In short: when we assume that the frames in a model tilt, we have to prove that the gunports in any photograph of that model are actually tilted, and not only look tilted. As long as the tilt can be explained by photographic distortion, the photograph kan not be considered as proof for the tilting.This also cannot be considered as proof for non-tilting frames. And you still haven't answered my questions. Peter
|
|
|
Post by jules on Aug 23, 2016 20:21:31 GMT
Hello Rein, I'm still not sure if I made my point about the nail patterns in relation to the frames in the Hohenzollern model perfectly clear. Just suppose the builder of the 17th century Hohenzollern model built a model that was built according this method below. So no frames, just a smooth surface to put the planking on. When positioning the fasteners for the planking, he was not restricted by any constructional objects, he could position them anywhere he wanted to. He could even write his name in fasteners. And now let's just suppose that the builder of the model thought he saw tilted frames on the wharf. These frames weren't tilted, but he thought they were tilted. I think we can guess what pattern he would use on the model: a tilted one. Or let's just suppose that the model builder made a mistake in positioning the fasteners, and that he put in a tilted pattern by mistake. What does this prove? This only proves that the builder of the model used a tilted pattern to distribute the fasteners on the model, not that the actual ship was built with tilted frames. Now, if the model was built on frames, and built on frames that were exactly placed as in the real ship, this would be different. The pattern than is, more or less, determined by the framing underneath, and the nail pattern could form an indication of the frames. You asked me to show some parts of the Gent model. It may be clear by now, that by looking at pictures of a model, we can not with certainty make assumptions on the framing of that model. Here's why, explained in a different way. Underneath I show three details of three separate photo's I took of the same gunport of the Gent model. The first is tilting right, the second is almost straight, the third is tilting left. Mind you all these photo's were taken with the camera on a tripod, with the camera at the same height of the floor, with the camera at the same distance of the model, and with the camera, as far as possible, perpendicular to the center line of the model (I really tried my best). What does this say about the tilting of the frames in the Gent model? Kind regards, Jules PS I forgot to mention in my last post that all the gunports in my 3D-sketches have vertical sills. Sorry.
|
|
ara
Junior Member
Posts: 69
|
Post by ara on Aug 23, 2016 20:43:35 GMT
Hi Jules,
Well, after this lengthy discussion, it seems you can never completely trust photos, and therefore, to gain some amount of certainty, the best way would be really to subject the photo material to the software Fred mentioned. But that of course would be the work of a professional. We should also do this with the Gent model for comparison. I still think the two models are built differently in their frames, based on looking on the photos. Apparently that view could change. But that´s really a thing for software experts. Birnie mentions the "inhouten" in his protocol. So why should the HZM not have these? Do you know something about the fate of the Gent model? Will it stay in the box forever? Is Mrs Baldewijns still in charge? Bye Rein
|
|
|
Post by fredhocker on Aug 24, 2016 7:09:20 GMT
I think we are starting to chase our tails in circles here!
On the ship remains I have worked on with tilted frames, it is not something subtle, it is usually clearly visible. It also tends to be something that starts near amidships and increases towards the ends (see drawing of Vasa frames in an earlier post), but there is a related phenomenon which can be confusing in this regard. The upper futtocks or top timbers at the ends of the ship are sometimes tilted out of the plane of the lower parts of the same frame. This usually involves tilting these timbers towards the ends, but in some cases they may be tilted in towards amidships. On Vasa, the last six top timbers (those abaft the fashion piece in the stern) lean aft, parallel to the upper fashion piece. On the OB71, a passenger ferry built in 1587, the framing is titled as in Vasa, but the upper futtocks at the bow are tilted forward, out of the plane of the floor timbers, and so appear to be vertical or parallel!
Jules, I am afraid that there is no software that will allow you to make an orthographic plan from two van de Velde drawings - I wish there were! The photogrammetry software works because the optical distortion caused by the camera lens is consistent, and the information in each image is the same, so the software can identify points common to two photos and use their relative positions to calculate where the camera was positioned, etc. The van de Veldes were good, but every perspective view constructed by eye, rather than with an instrument, is different in its projection, so there is no baseline data from which to calculate.
Many of the programs which could be used with the photos are readily available and run on a laptop - you don't actually need a pro. A couple of caveats, however: it is best if one has the full negative to work with, rather than a cropped image, and there has to be enough overlap between images to allow the software to calculate the relative positions of points visible in adjacent images.
It would be wonderful if the Gent model could be scanned or otherwise digitally documented, as we have done with our Sheldon model and as the NMM in England has done with several of its early Admiralty -style models.
Fred
|
|
|
Post by jules on Aug 24, 2016 13:22:42 GMT
Hello Fred,
Oh yes, in my opinion the debate has finished, and it's been a bumpy ride, but doubts are still around. Sometimes you have to make a point over and over again, in different ways, to get across what you're trying to say. So I continue, though I indeed feel like I'm chasing my own tail. Ah, a shame there is no Van de Velde software available yet. The software you refer to is probably the Photoshop correction set by Adobe. Isn't it? The caveats you refer to with regard to this software, do not apply for the photographs I took of the Gent model: one camera, one lens, and plenty of overlap in the different photographs (as already shown in the gunport triplet above). So whenever I'm up to it, I will give it a try; see what comes out of that. But I also made a film of the model when it was still in the collection. That shows a lot more than just a couple of shots (though I took over 200); perception is completely different when watching a film. And, in theory, this film contains thousands of shots of the model, and these shots were all taken with one camera/lens setup, and sure these shots overlap. There's still some hope. I would like to suggest that you, with your reputation in the archaeological- and museum world, and your persuasive powers, contact the STAM in Gent, and convince them of the enormous historical importance of the model they have standing in a box in a storage room. Maybe they will make the model available for the scanning process you refer to, when You push their buttons. Like they made the model available when I asked if I could make some photographs, and they, to my big surprise, agreed. Maybe we can set this up together, I volunteer to do the legwork.
Kind regards,
Jules
|
|
|
Post by Peter Jenssen on Aug 25, 2016 3:12:35 GMT
Hello Rein, I'm still not sure if I made my point about the nail patterns in relation to the frames in the Hohenzollern model perfectly clear. Just suppose the builder of the 17th century Hohenzollern model built a model that was built according this method below. So no frames, just a smooth surface to put the planking on. When positioning the fasteners for the planking, he was not restricted by any constructional objects, he could position them anywhere he wanted to. He could even write his name in fasteners. And now let's just suppose that the builder of the model thought he saw tilted frames on the wharf. These frames weren't tilted, but he thought they were tilted. I think we can guess what pattern he would use on the model: a tilted one. Or let's just suppose that the model builder made a mistake in positioning the fasteners, and that he put in a tilted pattern by mistake. What does this prove? This only proves that the builder of the model used a tilted pattern to distribute the fasteners on the model, not that the actual ship was built with tilted frames. Hi Jules, What the original model builder did is probably quite relevant in terms of the goal of creating a copy of his model. In that context, practices and developments of actual shipbuilding methods would possibly serve as an aid in understanding what the model builder is striving to represent. Even if those methods are becoming obsolete(already are extinct?) at the time of the construction of the old model. As Fred was pointing out earlier, shipbuilders of the time would probably shake their heads at our obsession of details.. Cheers, Peter (J) It's been very interesting to follow this discussion and I have learnt a lot. Interesting point about introduction of machine sawing of timber in the Netherlands. It would be interesting to follow up a bit on how the complex shapes of the frames were accomplished. For my model construction especially thanks to Fred for the framing angles on the Vasa. Where I had previously thought only the aft frames where angled, the whole structure looks positively fan shaped. I have not installed my top timbers yet. This may come in handy. Also particularly interesting is the lateral fanout of the gun ports, just in time for me to make sure the gun ports point the right way sideways. I was just about to make them 90 degrees to the ships side like I suspect most people up until now have! (Though I'm now (again) confused as to where Vasa had her long 24 pounders..) After seeing the test shot of the Vasa cannon go through the ships side like it was barely there, it's easy to understand why they might try to turn the whole side into a solid wood wall. But after seeing the slow motion footage of what happens on the inside (plus the remaining energy is enough to continue for hundreds of meters including straight through trees), the futility and counter-productivity of that becomes quite clear; any wood placed in the way of such a cannon ball only turns into so much lethal wood splinters.
|
|
|
Post by jules on Aug 25, 2016 6:41:20 GMT
Hello Peter,
Thanks for your contribution. About the building method of the Hohenzollern model. Since no answer emerged after asking several times, it looks like we will never know for sure which building method was used. But if people insist that it was built on frames, why not make a copy of the model with actual frames? That might add a lot to this discussion. When you think the model shows tilted gunports, tilted frames, tilted deckbeams which were placed between the frames, show us how it was done, by building a model. And it doesn't even have to be an actual model anymore. 3D drawing programs have become cheaper, cheaper than the material to build a model of considerable size, so build the model like that. Or, why not both? Design in digital, build in real.
I do not consider a thesis like 'all Dutch ships were built with tilted frames in the 17th century', a detail. I actually care about this stuff, been studying it for years now...
The machine sawing of timber by windmills is considered to be the starting point of our Golden Age, the 17th century. In a nutshell: we were able to built ships faster, develop our trade all over the world, and became very rich. Not bad for a simple invention.
Kind regards, and good luck with reproducing all this in your model,
Jules
|
|
|
Post by fredhocker on Aug 25, 2016 8:01:35 GMT
Jules, I would not say that the Gouden Eeuw was a result of sawmill technology, although it certainly helped. The Dutch were already the dominant commercial maritime power in northern Europe by the mid-16th century, because they controlled the most lucrative trade in the world, which was not the silks and spices of the Indies, but the timber, grain and fish of the Baltic. The Baltic trade remained the main source of foreign income for Dutch merchants throughout the 17th century, usually generating about twice as much profit as the Far Eastern trades. Superior organizational skill had led them to this,starting in the 15th century, and sawmill technology is more a reflection of that skill than a specific cause. The Dutch already controlled the trade in Baltic and German timber, and had already developed better organizational approaches to large-scale production of ships before they had sawmills. They also took the lead in establishing the modern organizational model which we now take for granted, the independent corporation with a separate legal identity and a lifespan longer than one voyage. Milled lumber certainly increased their efficiency, but I would not call it the central development, especially as Dutch shipbuilders were already the most efficient in Europe before they had milled lumber. Most other countries adopted it rapidly (we had wind-powered mills in Sweden by the 1610s), but no one else had all of the other elements in place to take maximum advantage of the benefits sawmills offered.
Fred
|
|
|
Post by jules on Aug 25, 2016 8:12:31 GMT
Hi Fred,
That's what you get when you try to do it in a nutshell. You're perfectly right of course.
Thought about the Gent venture yet?
Regards,
Jules
|
|
|
Post by fredhocker on Aug 26, 2016 7:43:52 GMT
I would be happy to contact Gent about their model, since it is important comparative material for us. Who is the relevant curator there? Fred
|
|
|
Post by jules on Aug 26, 2016 13:17:05 GMT
Hello Fred,
That is very good news indeed! The curator, when I photographed the model, was Mrs. Jeannine Baldewijns. But, I haven't had contact with her for a while; I'm not sure if she's still there. So I sent her an email this afternoon, explaining the situation and that you want to take up contact with her. If I don't have a reply by tuesday, I will phone the museum and ask if she's still there. If not, I'll ask for the name of the new curator. I will keep you posted, of course.
Kind regards,
Jules
|
|
|
Post by jules on Aug 27, 2016 14:42:57 GMT
Hello Fred,
I got an answer from Mrs. Baldewijns. She's still there, and is looking forward to the contact. I will send an e-mail to your museums address with her e-mail address.
Regards,
Jules
|
|