S18 Via Photogrammetry

[mrbarry]

Who let this guy in? I mean, come on. That CGI R2D2 on the left of that pic is just awful...

Welcome, Mike, to the land of Who. Correct me if I'm wrong, but I think the prop Tardis in the shot on Brighton beach, from The Leisure Hive, is the TYJ Mk I fibreglass prop. While the other shots of it at the Beeb are the Mk II prop, which ostensibly came from the same moulds (English spelling , but exhibits differences in construction and finish.

I look forward to seeing your finished result. If you attempt the New Series Box, I have the original artwork files to hand for the signage.

Barry

[mverta]

Yep, I realized the differences along the way.  Most of the core proportions are the same, but I couldn't figure out why the roof stacks were slightly misaligned.  The Police Box signs too are frustrating because no two are mounted at the same height the same way...

_Mike

[mverta]

Just a quick render test to the geometry thus far - definitely closing in on it.  The accuracy of this process is always limited to the resolution of the source material, which in this case is relatively low-res, but I decided to plug in a real-world measurement and see if the rest of the measurements seemed, "blueprint-y."  I notice that a 6" square post is generally agreed on, so I tried that.  When doing so, virtually every other measurement fell precisely within .25" increments - that's a very good sign.




On to the base and lamp...


_Mike

[anita]

This is incredible stuff.
I'm dying to see the finished results!

[DoctorWho8]

Here are some more photos of the MK1 Y-J box in Colin Baker's time.  If you need a physical reference size, according to Purple the door opening (from the bottom of the three horizontal stiles under the sign box to the top of the base) is about 6 feet 6 inches.




Bill Rudloff

[geminitimelord]

This is incredible stuff.
I'm dying to see the finished results!

WOW DITTO THAT!

[protokev]

This makes me glad ive procrastinated on my model build. Great Work as usual

[mverta]

If you need a physical reference size, according to Purple the door opening (from the bottom of the three horizontal stiles under the sign box to the top of the base) is about 6 feet 6 inches.

Oh dear... If that's true, then there is a very, very deep flaw in my simulation that I can't explain.

Let me clarify...  actually, I think I heard The Doctor explain something like this once: in a photo, size and distance can be confusing.  Something small can look big close-up.  That's why you need multiple photos to work from - you have to see it from many views to recreate the real 3D space in the computer.  In the photos you see me working with, when I place an object, it shows up in all the corresponding photos, and the idea is that when each object appears in the correct position and proportion in all references, then the virtual cameras are correctly calibrated, and the object's proportions are correct.  Then there's no guesswork - once the simulation is accurate, then you're just basically "tracing" in 3 dimensions. 

One thing which makes the TARDIS gloriously friendly to simulate is that it has a lot of right angles, and square dimensions.  As you can see, my simulation is getting closer.  In fact, when I plugged in 6" square as the absolute dimensions of the corner posts, the simulation "accurately" defined the door bevels angles at 21.05 degrees.  In any case, whether it's 21 degrees or 20 degrees or whatever, my simulation is clearly approaching an accuracy tolerance of about .25 - .5 inches.  For the scale of this object, if I was off by more than that, you couldn't miss it.

Again, if you were working with one photo - MAYBE  two - you could be deceived about sizes and distances.  But with 4 or more (I'm actually using 6) references, if everything lines up, then you can start to have confidence in it.  At this point, the greatest deviation I've seen between references is .5" (if we're to assume 6" posts.)

That's a long way of going about saying that if the posts are 6", then the door height below the 3 steps to the base is 6'3", not 6'6".

Now, a .25" I could buy; .5"... would be real hard to miss, but 1"? 2"? 3"?!!  That would reveal some bizarre suspension of physics in a way that would depress me.  Or it would reveal some legendarily elusive challenge in the process for me to face   Just depends on how you look at it.  But I'm the new guy here, so I'm loathe to challenge the wisdom of people who've spent years before me.

So I built a 6'6" door to go along with my 6" posts to test the theory.  I tried for 5 hours, but I could not make those dimensions line up with any of my references, no matter how I twisted the camera, adjusted focal length, or played with extreme lens distortion.  If someone can tell me 10000% that's a real dimension, then I'll go back to square one and rethink the process.  But I honestly can't see how, given how consistent the other proportions hold up across all references, and the potential accuracy of the bevel angle, etc.

Love these challenges!!



_Mike

[mverta]

One other question -

Can someone confirm for me that the Mk 1 sometimes had the lamp mounted right to the apex of the roof, and later had an extra "mount" box on top of which it sat?  I've seen plenty of photos of both cases, but I want to make sure it applies to the Mk 1.


_Mike

[Scarfwearer]

Oh dear... If that's true, then there is a very, very deep flaw in my simulation that I can't explain.

Gene Fender estimated the original prop's door height at 78", but the S18 is nominally 76" (6'4").
Here are the plans that we have access to:
http://groups.msn.com/TheTARDISbuildersGuild/tardisplans.msnw.
Bear in mind that if you scale up, any errors are also scaled up. If you scale down, then difference between even 78" and 75" amounts to a bit less than 0.25" on 6". The difference between 76" and 75" scaled down to the post width would be difficult to spot, I would imagine.

Crispin

[mverta]

Hmm.. well 6'4" would make more sense, anyway.  But I'm still hammering out proportions, which are the same regardless of absolute scale...  I'm rapidly approaching the limit of what I can do with the relatively low-res references we have access to, and plus every shot features differently warped walls, signage that's hung crooked, etc. 


_Mike

[mverta]

Base and roof with lamp mount added. 




I think I like it with the lamp mount more than without...


_Mike

[DoctorWho8]

Just a note that if the door opening height was 6' 4", then Tom Baker would have to constantly slouch to avoid hitting his head as he was 6' 4".  If the door opening height was 6' 6", it would then give him about a 2" clearance.  Too bad the MkII prop isn't still out in BBC TV centre, maybe someone could do a loose measurement using their height with shoes included and see where their head ends up in relation to the door.
Bill Rudloff

[mverta]

That's an excellent point...  here's a shot from Logopolis, though it's a different box, obviously.  But still. Tom's got room:




And here's a shot of Peter Davison (6'1") standing in the doorway of the MK2 (?), and even though he's leaning, he's clearly got more than 2" of clearance above his head if standing upright:




Given that, see what you think of this:




_Mike

[timerotor]

well if this helps to get a rough idea of the hight, I am standing in the middle at 5 ft 11 inches. I had flat shoes on and my hair was flat. I remember thinking that the windows were in line with my head and the top of the doors weren't that much taller than myself. We must bare in mind, the Tardis was on its base (2 inches high?) and I was standing on the flat floor just infront.
Mark
P.S this is fascinating stuff!

I estimate that the door top is about 5.5 inches above me- so scarf wearers 6 ft 4 sounds right to me...