Notice: Purchasing, Trials and Accounts Have Changed. What You Need To Know: Part II >>

Topic - Orion Space Plane: 2001 A Space Odyssey

Page 7 of 11
First  Previous  4  5  6  7  8  9  10  Next  Last 
Conscious again. The first time I measured the reference image (middle if the previous page and below), I used a 12" wooden ruler held against my screen. This time I used the ruler tool in Photoshop. My readout is in 3 place decimals of an inch. Because the ship is angled away from us, there is a slight difference in left/right measure, than in up/down measure, relative to accuracy, but for this test model, I will add a small percent amount to these longitudinal (left/right) measurements to compensate for this effect.



For ease of communication, I will refer to the area being measured as a Segment, while the rings at the start and end of the measure are Sections. I will refine the borders of this segment relative to the bowl and recess.

There are a number of changes I see are needed to the test model above. The bowl is too tall and the recess too shallow, and as already mentioned, the space between the bowl and recess should be slightly less than it is. This necessitates some recalculations, searching through reams of notes and working drawings...all time consuming. I'll return to this series when that is done.

I'll be editing this post as I add information. Back to the drawing board.

Message edited by LeonArtO3D on 2/17/2011 - 10:02 AM

TOP OF MY WISH LIST

This is really stretching things (no pun intended), but as the images below show, merely resizing a small image to a larger size is useless for measuring purposes. The enlargement shown of the same area is too fuzzy to distinguish where a feature begins or ends. It would have to be a hi-rez scan of the original photograph but zeroed in on the area marked in magenta, much like the image of the cockpit area detail of this same image.



Fuzzy Imagery


Don't get me wrong, I really want the entire image scanned at the same resolution as the cockpit image by Douglas Trumbull, because this image is the nearest we have to a perfectly centered side view, with a 1/8 rotation (22.5 degrees). I fear I'll have to await THE BOOK for anything more of this. As you can imagine, a clear view of this area would solve the measure problem of this feature.

Speaking of the BOOK, I just had a thought, if the pages of the book are 17" wide, then the full double page image of this photo would be 34" wide, which would be very close to a one to one view of the original 36" model. That is why it will become known as the 2001 BIBLE. I suppose the only person I could get this from is Douglas Trumbull himself, but his generosity has already been stretched beyond the breaking point. I'd be too embarrassed to ask him for anything more. Directly anyway.

Message edited by LeonArtO3D on 2/18/2011 - 4:18 AM

I got this note over on another forum:

" re: discrepancy between instrument panels in the film and in the production photo:

Clarke visited the Orion set and announced that the interior looked like a Chinese restaurant, prompting Kubrick to re-vamp of the interior.

(Here's the link to that article on Clark):
http://books.google.com/books?id=fU78LdDClHUC&pg=PA284&lpg=PA284&dq=kubrick+2001+chinese+restaurant+set&source=bl&ots=uzGeL6G7o4&sig=94cAEyA7HTZ9fNpXfFnLcM92cxw&hl=en&ei=MURNTeKUMIfagAfwyPTUDw&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBYQ6AEwAA#v=onepage&q=kubrick%202001%20chinese%20restaurant%20set&f=false

Leon, thanks for putting the photo's back up on your site. Your site has become the definitive reference for Orion source material. Can't wait to see the progress. And, eventually, other spacecraft. BTW, I've been dabbling in Lightwave for a few years, but was inspired by your work to pick up a copy of Modo. The interface appears to be much more intuitive.

Thomas, Harry Lange is no doubt having a chuckle over this discussion somewhere in the pantheon of science fiction set designers. "Zee symmetry of zee asymmetry", as he called his screwing with expectations."

That forum is:

http://2001aspaceodyssey.ning.com/photo/albums/blueprints-1?xg_source=msg_com_album&xg_pw=&id=543076%3AAlbum%3A21583&commentPage=3#comments


So, welcome Dale, to Luxology and modo.

Message edited by LeonArtO3D on 2/18/2011 - 4:59 PM

Quote from LeonArtO3D :


These models were done by B.G. West, and can be had by visiting the web site sited above by Mike James.





Actually, only the Orion was done by B.J. West.

The Space Station was done by Michael Powell... about 15 years ago using Lightwave Modeler on the Video Toaster.

Credit where it's due.

SMS
Quote from Spaceman Spiff :



Actually, only the Orion was done by B.J. West.

The Space Station was done by Michael Powell... about 15 years ago using Lightwave Modeler on the Video Toaster.

Credit where it's due.

SMS


Thanks Spaceman Spiff, and welcome to this thread. I'd completely forgotten about the other models, and even that some of the Orion conversions were done by others. I need to go back and totally familiarize myself with all that data. There is so much data to all this, it has my head spinning. I suppose I ought to put a link over there to this thread. Thanks Michael Powell for your contribution to the 2001 TSO projects. Feel free to post your comments here, anytime.

Sorry, lunadude! I'm having trouble keeping up. But then I've just had another break through!

Message edited by LeonArtO3D on 2/21/2011 - 2:30 AM

The series of step by step operations at the top of this page is on hold for the moment. For that, i need to take many steps backward, and in that process, I made another breakthrough in another area, an area on which this operation depends. So, I'll be working on that for awhile, and once sorted out, I'll start this idea up again, only from a number of steps back. Sorry for the wiplash effect, folks, but that is just a taste of what this entire project has been like for me. Still, it is a technique that is working. Little things here and there add up to much greater perfection in the long run. Even this interruption is valuable, as the more early problems get eliminated, the less of them I carry into the next operations. Obvious good logistics.

However, on those images at the top of this page, I failed to clarify the boolean operation I'd used to make my cuts. From the 3D Edit tab, I choose Solid Drill/stencil. Also, after many problems that kept popping up with every step of this operation, I finally got in the habit of running Senecas script, Mesh cleanup.pl, between every step. That seemed to eliminate all those problems.

Back to the drawing board, friends.
Welcome to all you messiah people, especially from Setuptab!

Hurrah! Character rigging!
NASA Images has some great images for our space backgrounds. You can get very high rez pix there and registration is free.

http://www.nasaimages.org/index.html

I had a great idea for an illustration, so I downloaded a good picture of earth from space to drop into modo behind Orion III.

Here's a screen capture of the small image:



I'll post the render here shortly.
I've used B. J. West's model for this with some alterations of my own. It's just a start at this idea of how Orion was conceived to separate from the booster engines. there is much modeling yet to do, for instance, part of the wings go with the engine. This is merely my first test render, playing with the lighting.





This will make an interesting short animation, once I have one of my own models complete and cam make side alterations. This is derived from sketches by Harry Lange.

Harry Lange's early drawing from Filming the Future by Piers Bizony (p. 84), illustrates the purpose of the explosive bolts, "....to break away from it's winged booster". Notice the different configurations of this concept in both drawings, as Orion evolved. The contours of this vary completely from the final version, which will require modeling another unique version of our baby. Using the BJW model to make little changes to get back to this earlier concept is like the directions the guy gave the traveler, "You can't get there from here."



This concept would have made the passenger section considerably more comfortable with it's increased width. The "probe" does indeed look like an antenna in this version. The engine cowling doesn't get nipped off either.

All in all, I like the final whole version best. Can't wait to see the piggyback launch vehicle model coming out as a kit, or so it is rumored.

Rest In Peace Harry

Message edited by LeonArtO3D on 2/26/2011 - 9:17 PM

Anyone want messiah Pro 5? I bought an extra license and no longer need it. It would be $1200.00 if you paid full price but I got mine on sale for $40.00. Those people sure know how to rake in the customers. I'm giving mine away to the first person who asks, that's right, it's free, and this is not the only place I've advertised this offer. It's a great way to get into character rigging and that type of animation. I'll need your email and real name to transfer my license to you.

He who hesitates is lost! Petrograde won!!! The licemse has been spoken for! Thanks to Rich Levy for getting this thing started, or MD who gave away 10 of them!

Message edited by LeonArtO3D on 2/28/2011 - 4:14 AM

Just noticed this, have you seen it?

http://www.squadron.com/ItemDetails.asp?item=MW20012

Dave
Quote from DavidEnders :
Just noticed this, have you seen it?

http://www.squadron.com/ItemDetails.asp?item=MW20012

Dave



Thanks Dave. Here at last.

The Moebius 13" 'Spaceclipper'. "Sleek commercial orbital spacecraft strikingly similar to the Orion as seen in the motion picture '2001: A Space Odyssey'." Congratulations to Adam K. Johnson, design consultant, and to Moebius Models founder, Frank Winspur.

Message edited by LeonArtO3D on 3/3/2011 - 2:32 AM

Here's one I've had for ages but can't remember where I got it. It's a print version of what appears to be one of Douglas Trumbulls.

Message edited by LeonArtO3D on 3/10/2011 - 5:56 PM

I've been running a new series of experiments to help me zero in on the true shape, and proportions of the original model, with the available evidence we have at present. I've decided to include these in this work-in-progress report, in case the technique might be of interest to anyone doing this sort of reconstruction project.

Essentially, It is a comparison of all the photos of the Original Studio Model (OSM), with the models I am building. Since none of my test models have textures yet, I've started with the model by B.J. West, to which I have only done a minimal amount of desecration:

First of all, While I will be modeling Orion in true scale to the actual ship, if it existed, that is, 175 feet long, The OSM was I am told, 36" long, not counting the probes. And since the photographs are of that sized model, and I'll want to reproduce the actual camera settings in modo, lens focal length, camera type, distance to the model, angle of direction etc., I scaled BJW's model down to 3 feet. To do so, I chopped off the probes. Then what was left didn't have the same shape as the OSM so I did a little remodeling at the tail to bring it to a point. I wasn't trying for any major degree of accuracy with that, just something to get started with while I work out the technique.

I don't know why BJW chose the size he did for modeling Orion, but his model is 2636', 11" long; 1253', 11" wide wing tip to wing tip; and, 275', 7.35" Tall. That's with the probes still on it! So, my version had to be, as a fraction 1/878 (+/-) the size of his, or 2636 to 3. Very tiny!

Now, I was going to use Absolute Scale, or Scale Absolute, however they call it, but when I looked it up in the old and disreputable documentation, it was no straight forward method that I could grasp intuitively, as they like to describe modo, and I was in too much of a rush to start yet another side project of experiments to figure out what they were talking about, so I did it the old fashioned way, with percentage formula. I did it in two stages. I scaled it down to about 26' and used my formula from there. Even so, I didn't get it to exactly 36", I got 36.0004". Must be to the closest pixel or something.

OK, so from there I popped in a Backdrop Item, loaded the image DT-01A.jpg; that is, this one:



I set it to Camera Projection, and I set my render size to the same dimensions as the photo. I've written to Douglas Trumbull about the camera data for three images, but in the mean time, I started playing with poses and settings.



Close, but no cigar, by Layer three. Here you see layer 3 at 50% opacity, with the image above showing through. Actually, this is my second attempt at running this experiment. Here's a little background on my struggle with this idea:

At the time I started this thread, I considered myself a beginning Intermediate Level modonaut, as opposed to a rank Newbie, which is how I came to view myself after a very short time wrestling with the complexities of this task. This project is completely different than anything else I'd tried with modo, and required, in many cases, learning things I had never attempted. A case in point: With all the thousands of renders I have done from the beginning of my introduction of modo and things 3D, I had never really learned how to use the Camera, or even looked at it's Properties.

Not only that, historically, I've only been a dabbler with cameras and photography. Oh sure, I knew more than most people first getting into 3D, but that was only because my best friend of many years was a photographer and I was around it a lot, played in a darkroom, read about cameras and photography, used his camera to try things, just out of amateur curiosity. I'd never owned my own 35mm camera, nor did a first hand study of the various lens's I'd played with. I knew a little about many things photographic, but not a lot about any of it.

I knew about perspective distortion because I'm a draftsman, or was in an earlier life, 35 years ago, the pen and ink variety. And I'm an artist in the greater context, and have been drawing all my life. OK, so I was a dabbler who had forgotten more than I ever knew about all this when I started this project.

So, when I started this series of experiments, I did have some knowledge of perspective distortion and Lens Distortion, but it was the difference between having read about a subject, vs many days of experimentation with a slow and painful education in it.

As posts are limited to how many characters may be used, I'll start a new one to continue.

Message edited by LeonArtO3D on 3/14/2011 - 5:15 PM

So it was that in my confusion about how to set lens data up in Camera Properties, I got a bit fuzzy in my thinking about how I might fix things. That's when I started another thread, asking for help:

My Lens Distortion Thread

OK, now that you've had a good laugh at me tripping over my own mixed up contexts, I'll try to go on...

When I started this series of experiments, I began with image DT-02, not -01. I did 31 renders, and for each one did a number of tweaks to the settings, operating pretty much in the dark as I struggled to understand what the transform tools did for the camera Item with Set Target turned On and turned Off, enabled or disabled. At times I got close to a good match, and slowly the control got better. I'm not saying I have mastered the Camera settings by any means, far from it. But it has gotten easier to get the effect I'm trying for.

I am still in the dark with some of these features and why they behave the way they do. That is one big reason for doing this series in the open as it were, to get advise from other users when I get into something I don't understand, and can find no reference to in the documentation. So please step in at any time with advise and information.

For the moment, I'm merely trying to catch you up on how it went and where I am at this stage. As I write this, I hit submit periodically, but go on editing it, as an ongoing thing. So, you might hit reload to update the page if things are not changing that often. If I'm taking a break, I'll indicate that at the end of what I'm saying, so you are not left hanging.

Overall, I started with the two color images of Douglas Trumbull's OSM, but then added the black and white image posted above with a good rear view. That one really gave me the fits, and caused me to expand my study of all the elements I'm trying to grasp, the principles of photography and the controls and settings within modo. I did 31 renders of DT-02 before switching over to DT-01 and did 12 renders there before switching to OSM-04.

After doing some study and thinking about all this, I went back to DT-01A and set my Focal Length to 210mm. I was thinking DT may have shot the models with an 8X10 camera, and a friend tells me these cameras normally use a 210mm lens. 210mm may do the job, based on one factor that I observe. I'll give you an example...

The first time I worked with DT-01, my last render was at a much lower focal length and here's a render of that:



Now, here's the 3rd layer, my latest render using the 210mm lens:



Compare the curvature of the rings around the fuselage. Admittedly, the camera position is different in each, and that will be my next experiment. That will speak volumes. Notice that with the wide angle lens below that there is much less curvature evident in the rings than in the image above. BTW, I don't recall the focal length for the top image, but it's small.

210mm is way too big. 35mm more like it.

Message edited by LeonArtO3D on 3/20/2011 - 5:00 PM


Here are some images from Model Design & Blueprinting Handbook by Charles Adams, that illustrate the problems with perspective distortion:


From The Model Design & Blueprinting Handbook, Volume 1, page 14. (C) 2010 Charles Adams. All rights reserved. Used by permission.


From The Model Design & Blueprinting Handbook, Volume 1, page 16. (C) 2010 Charles Adams. All rights reserved. Used by permission.


From The Model Design & Blueprinting Handbook, Volume 1, page 16. (C) 2010 Charles Adams. All rights reserved. Used by permission.


From The Model Design & Blueprinting Handbook, Volume 1, page 16. (C) 2010 Charles Adams. All rights reserved. Used by permission.

Thanks Charles.




If I can find a work around for the problems I've been having with Camera Properties, I just may have found a way to measure the model in this perspective photo. Wish me luck, I'll need it!

Message edited by LeonArtO3D on 3/15/2011 - 11:37 PM

Hopefully, some photographer, like Dan Ablan can clear up some of my confusion. I'm still using 401sp5, so features that have been added to the camera in 501 are not an issue at the moment.

Much of my confusion has abated but some clarification would help. Each lens is a certain focal length (28mm, 80mm, 210mm etc.). If you know what lens was used to take a photo, in an operation like the one I'm doing, or if you want to set up your render as if it were done by a particular lens, this number is what you enter into the Focal Length entry field in Camera Properties...Right?




When I 'Set Target' in Camera Properties according to instructions, I start in Items mode with Camera selected. Then press Ctrl and select the model layer you want to be the target. modo uses the Center of that layer to center the target on but it is way off-set. I tried to move the Center but it also moves the spaceplane model.

"Roll: Provides the ability to offset the items rotation angle away from the target." from the 501 Help files. Near as I can tell you have a single axis to offset with and it's X or Z and the vital offset needed the most is along the Y axis, at least in this particular problem of mine.

Now, Tobias told me he thought you could select a polygon to use as the target, that you didn't have to be in Item mode. So far, I have been unable to do this by any method I've tried.

I don't want to have to reposition my model in the scene to make this feature work, and after futzing with this for weeks, I'm a little fried.




When the target is set, the Rotation transforms don't work, so you have to use the unintuitive Position Transforms, which should be easy. Maybe my intuition is broke after all these years. Maybe Oldtimers Disease is finally settling in.

Then, the Target Distance gets set just right and you have to disable Set Target to do something and the focus of the model goes to Hell and you can't alter the Target Distance. After days and days of this crap, I finally found the Focus Distance down under Depth of Field, and was able to make this adjustment.

Again, this is why a modo Bible would be of value, rather than a half explanation, with important key strokes left out.




I need to make very fine adjustments to my zoom function, and the magnifying glass doesn't cut it. A digital input would be nice. Maybe there's a script that someone wrote to accomplish this very thing. Anyone know of a method?




This one was a mind blower! I did a render and saved it as a .psd file and then as a .tga file. I did this because to wanted a transparent background around my model, so I could see if the backdrop image pokes out around the model, and even though I've disabled the environment, the psd file still has a black background. Now I know I can get rid of that in Photoshop, and will have to, I guess, because the targa render save was not the same as the psd save:



You can tell the targa because the background is transparent. It looks to me like the Targa image has centered the model in the image. Anyone ever seen this before? Color me surprised! Tobias is looking into it.




Well, if I get this figured out I may be able to test my theory that measurements can be made from perspective photos, if you have an appropriate model to measure against...or something like that. LOL

Here were my very early attempts mad years ago in Lightwave. Nothing accurate just for fun.


Bravo Tony!

Thanks for bringing the fun back into this for me, which was why I started this in the first place. It's great to see someone else's work in these pages. Hope you've started a new trend. I think maybe I'm sticking too close to all the technical details, working and solving one, only to jump to another one, just as minute and yet important in the long run; but that sort of fun is very different than some lighthearted playing around with finished models that I'm satisfied with. There is a wealth of material from just the first of these movies, to keep polishing one's skills on for years, if one is looking for perfection. But I do long for the time when I can be retexturing merely a few panels for a close up render of some isolated area of Orion, or doing starscapes to my hearts content.

Yeah, I say that, but while thinking it, I can't wait till i get around to the landing gear for Aries, and that super cool landing scene as they articulate and absorb the shock of landing. Should be a breeze in modo!

Nice work Tony (3D) DeLuca.
Thanks. 2001 was my favorite movie of all time. Stanely was a masterful director. Every frame of his films were paintings. I remember paying to see 2001 in Chicago at the Cinestage In Cinerama 23 times during it's road show. My brother was at the world premier in DC in 68. I actually had 14 email conversation with Arther C Clark over my models. It was funny, everytime I asked him something about Kubrick, he would change the subject. I have download the Modo demo, but I'm finding it very difficult to learn after such a long time using Lightwave.

Message edited by Tony3DD on 3/16/2011 - 11:59 AM

Quote from Tony3DD :
Thanks. 2001 was my favorite movie of all time. Stanely was a masterful director. Every frame of his films were paintings. I remember paying to see 2001 in Chicago at the Cinestage In Cinerama 23 times during it's road show. My brother was at the world premier in DC in 68. I actually had 14 email conversation with Arther C Clark over my models. It was funny, everytime I asked him something about Kubrick, he would change the subject. I have download the Modo demo, but I'm finding it very difficult to learn after such a long time using Lightwave.


That is quite the History and must be some kind of record! Clark and Kubrick, two very different types of genius, potentially explosive...I'll bet there are many war stories that have not been told. I thought the one about the sets looking like a Chinese Restaurant was hilarious. Especially when Kubrick went right out and had them redone. I bet it wasn't his funny bone that got tweaked.

23 times in Cinerama! Oh, to be able to do that today, go in with a big bucket of popcorn and have the lights go down and Zarathustra start shaking the rafters! You need lots of Coke from sitting with your mouth hanging open for 139 minutes.

I had access to a Yamaha X-1 for a couple of years, a shiny white plastic synthesizer built along the lines of a Hammond organ, that is with dual keyboards + foot pedals, a moog strip, and one very sophisticated computer built in with recording & playback effects on 64 channels...a full orchestrator. I used to wake my friends up with Zarathustra. Sounded just like sitting in the orchestra pit at the Berlin Philharmonic. That musical toy could do anything, and cost $40,000.00. It looked like a part of the Discovery bridge. It was beautiful.

Funny, I just realized this project is a lot like what I did with that toy. I used to take Stan Kenton tapes and play them one measure at a time, then set up the same instruments on the Yamaha and make it sound like the recording. I'd work my way through the whole piece until what I was producing alone at the keyboard was what a full orchestra produced for the album. Here, I'm actually doing one that,s better than the original since it's a virtual model without the limitations of a single sized physical model. I can make mine any size I want and detail the leather of the seats if I want or the scratches on a knob at high resolution. Actually anyone will be able to since I'm giving away the model to all comers, to do as they like with it. We may end up with a whole warehouse full of high resolution textures for every panel in this and the other ships. Who knows where this will go.

Edit: Stick with it in modo, it's wonderful, and getting better all the time. (In spite of all the bitching we do about the steep learning curve. It's worth it.)

Message edited by LeonArtO3D on 3/16/2011 - 1:23 PM

Well, for this image anyway (DT-01), it was shot with something on the order of 35mm. This is probably the same for DT-02 but not for the big b/w image up the page a ways. That, I'll have to play with.


The double image shown above was set for a focal length of 36mm, and looks fairly close to the original model. But some fiddling is still necessary.


Speaking of fiddling....

Setting the Target for the Camera

To set up the model, the Camera and the Backdrop image for this experiment, it is much simpler to set a target for the Camera, especially if you do not want to move your model around in the scene to get its angle just right. That does not mean it's easier in all cases. Case in point:



Both the model and it's Center are "Center Selected All". When you Set Target for the camera, it aims at the Center, which in top view would be fine, and even side view would not be too bad. In this case, however, the reference image is at an angle, something like this:



The Backdrop Image has the OSM (Original Studio Model) centered with an equal distance of space beyond the wing tips, but as soon as you hit Set Target, the Orion model is shoved off to the bottom. It's because the wings are so far below the Center line, they act like a huge pendulum.



What to do, What to do? I did move my model around for awhile, just to see how far it needed to move, but the obvious became clear very soon. For Days, I tried to get along without Set Target, and learned why it was invented in the first place. After awhile, I got an idea. Why not alter the reference image to compensate for this difference to center? So I started experimenting with that:



I added some filler at the top, got the new size and changed my render size to that also. Did that about 20 times to get it close:



Now, all I have to do is fiddle around with the position of the camera for a couple of weeks, and various Focal Length settings. Think I'll go watch a movie....

I was really fried when i finished the above message.

Message edited by LeonArtO3D on 3/18/2011 - 2:19 AM

Well, I was hoping to get some feedback here on these problems, so that others following this who were getting similar problems could get quick answers, but I'm not even getting replies to technical problems in this thread. So, what I'll do is create individual threads for these as they come up, then, when I get some help, I'll give you the link to that thread, with some info of what it's about. Logistically, this seems like a better way to go. Individually posted problems get faster attention. On the other hand, 7 pages full of long drawn out descriptions of weird problems have to get lost in the shuffle. Too much data. It gets confusing.
I have not read the whole tread, but was this model originally made in another 3D app?
Page 7 of 11
First  Previous  4  5  6  7  8  9  10  Next  Last