Sep 21, 2023, 03:32 pm


New, New TardisBuilders!

Mike Verta's s18 TY-J

Started by mverta, Nov 14, 2018, 12:03 am

Previous topic - Next topic


I was just looking into this on wikipedia, and it appears that fibreglass is sometimes made with thermoplastic (which melts) rather than thermoset plastic (which doesn't). So perhaps that's what they used. It's still fibreglass - just the melting kind.


Jan 16, 2019, 10:57 pm #31 Last Edit: Jan 16, 2019, 11:00 pm by mverta
Yes some ambiguity of terms was copped to by at least my contact at Dow, who basically said, "Glass shatters; 'glass' can bend with the right resin."  In the end, it's definitely not what we would normally refer to as fiberglass, which, for me, actually has some negatives to it.  Because I'll be building all that wonderful distress and weathering into my build, much of it has to exist in the wood itself, because the fiberglass can't be "worked" after the fact.  It can be ground, and re-glassed, which we'll be doing, but ultimately it takes the form of the wood former underneath.  This is why those draft angles under the steps and sign boxes are so crucial - with 90-degree angles or undercuts, the fiberglass can't separate from the former.

We're shooting a documentary on the entire process, so you'll be able to see this actually happen and what we have to work with.


Interesting. I've always been puzzled by the amount of sagging in these "fiberglass" boxes as it simply didn't make sense, as you point out typical glass would shatter long before distorting. If they were thermoplastic, I'm still puzzled as to what the heat source could have been in the UK LOLOLOLOL.

As for the draft angles being different sides to top/bottom that's unusual not sure engineering wise if that was a necessity of a mistake. The amount of angle on the sign boxes looks pretty pronounced along their length.

Great work.

Makes me view it all in an entirely different light.

On your build are you glassing in support frames to the panels, posts, etc?



All those angles are conclusively proven by the simulation - I could've made the video an hour long just showing more of the same. In fact, I think I'll do a follow-up for you guys, 'cause it's pretty cool to see.

Also yes there is steel tubing providing a structure for rigidity and mounting!


And to think 12 years ago you called me up to ask for measurements from my Codsteaks Tardis :)))

Like you'd have been satisfied with that.



Blown away by your work! Its' crazy to see how close, yet how ever-so-slightly off Tony's plans were. The fact that you went so far as to use Davison's cricket ball for lighting reference astounds me. I'd never have thought of something like that!  ;D

Also, I absolutely loved the music in the video. The soundtrack works so well with what you're talking about and showing, not only because of them both being from S18, but because of the nature of the video itself. The ending, with the "It's the end..." cue and the lead into the main theme, gave me chills. Honestly, your musical work impresses me nearly as much as your building/rendering work. You truly are a man of many talents.

Can't wait to see the physical build come together!
"An apple a day keeps the... no, never mind."


Jan 18, 2019, 02:25 am #36 Last Edit: Jan 18, 2019, 02:37 am by mverta
You have to remember that I have the advantage of actually building and testing dimensions ad infinitum. The environment may be virtual, but the physics of light transport just happen to be one of those things we are good at simulating. If anything, I have an even deeper respect for dimensions derived purely by logical deduction and reasoning. By contrast, my process almost looks like cheating. :)

Speaking of cheating, I actually started to write the music for this little piece myself - that's my primary career - and decided to preserve my energy instead. This project is a long way from the end yet!  Plus, I always did like those cues!  :)


It will take awhile to disseminate all the info I gathered during the research, but here is one little bit of info which may be of use:

All the MK I walls/doors were built from the same single former, and it has two telltale features:


The upturn in the upper-left corner of the window, and the downturn on the top edge of the 2nd panel down on the right-half.  In truth, every panel is unique, and there is the strongest likelihood they were individually routed out, versus made in the typical way builders do it (angled rails with cross pieces).  None of the edges are square in any dimension, and the draft angles vary in shape and depth. But these two features (sometimes much harder to see than at other times) have so far shown up in every frame of s18 where they're visible, and on every reference photo I have.  They're prominent, universal, and distinctive enough that I've made them must-have's on my plans, and part of why I'll likely be having my own door master CNC'd from my digital files, in order to follow all the various contours and imperfections exactly. 


Jan 20, 2019, 11:49 am #38 Last Edit: Jan 20, 2019, 11:50 am by mverta
I haven't really thought much about this yet, but I thought I'd do a quick test on the sign - the first one is mounted on the inside face of some 1/2 diffuse plexi,


and the 2nd test is with it mounted on the outer face.  


I haven't worked on definitively determining which it was for s18, but one thing that I'm noticing is that whatever the original blue "fill" is, it does not photograph like vinyl; it's something else, or a very different formulation than what we use today.  Interesting that the letterforms here seem so precise, when they're actually pixel-accurate vectors traced from the original sign.  Anyway, more to come.


Can't-Unsee-It Part 2 (of 3,613,237)

The rightmost window frames are especially wonky, with the bottom edge being considerably thicker on the left than on the right, leading to an obvious downward slope of the fame.  Again, because a single former was used to make all the walls and doors, this detail is visible from every angle of the police box. Also highlighted again is that downward turn on the 2nd right panel.



Jan 21, 2019, 12:33 am #40 Last Edit: Jan 21, 2019, 01:13 pm by Theta Sigma
The time has come to send the digital files for the walls and doors to the fabricator.  No turning back now, and much is riding on it.  Within these imperfections and asymmetries lies the x-factor; without getting them right it will never truly look like the TARDIS; it will never look like it's come right off the screen.  But I think we're ready.  Fingers crossed!



Jan 21, 2019, 01:18 am #41 Last Edit: Jan 21, 2019, 01:19 am by Scarfwearer
This is particularly intriguing for me because I'm in the midst of building a TYJ Tardis from wood. And I know I'm not going to get it even remotely this close. The little deformations you're pointing out as "considerable" are probably below the threshold that I can manage to reproduce. You need pretty high res photos even to see them. I've felt if I get within a 1/4" I'm doing well. Maybe I'll have a go at those panel corners with a file and some wood filler...

Has anyone ever done anything like this before? Of this size? With this detail? With this precision? From photos? I can't imagine this level of detail and precision is common in CG either...


Jan 21, 2019, 01:53 am #42 Last Edit: Jan 21, 2019, 09:33 am by mverta
Well remember we're essentially decoupled from the photo resolution thanks to the sun.  And actually, this level of detail is not only well within the limits of our perception threshold, but if it weren't so, we wouldn't be able to navigate 3D space or recognize the faces of our parents. We think we've never seen this detail but in truth we always have. Go and check any episode of s18 where you can reasonably see the TARDIS and you won't be able to miss these details.  Where we weren't consciously aware of it our brains always were. Just walking around, the amount of data that we would have to deal with consciously is overwhelming, so our brain does a sort of importance sampling as we move through 3D space, so we only have to deal with what it thinks are the most relevant pieces of information - the approaching car and not the motion of the leaves through the trees. But we still see the leaves. If we froze the motion of the leaves we would instantly feel something was wrong even if we couldn't immediately pinpoint what it was.

It is not necessary to replicate every contour. It is important to replicate the generally warped and distressed nature of the original prop and put in our own contours and asymmetries of a similar displacement in nature to the original, at least if we want it to truly feel authentic.

Not only is this level of detail common in CG, it is actually essential. It holds the linchpin between photoreal and fake. In fact photoreal CG artists tend to be more aware of these things than 99% of the public because we have to be, in our attempts to make people believe something that's fake is real. More often than not, people can't tell us exactly why something doesn't look real, but their brain knows it isn't so they say it looks fake. It's our job to figure out what it is that the brain is really aware of and how much of that information we need to feed into an image for people to recognize it as real.

These sorts of details and asymmetries do not only hold the secret for authenticity, they determine just how interesting an object is to our brain in the first place. As I explained in a video on my CG version, since the brain is aware of all of these little variances, the amount and density of those variances determines how engaging and satisfying it is for the brain. It's the difference between a snack and a meal.

While portions of my police box are being CNC'd out of wood because I am extremely particular about the exact contours, whole other sections of it will be made by hand with wood. They will just likewise possess the sort of displacements and asymmetries and rough hand-tooled look of not only the s18 police box, but all police boxes.

Have no fear, be bold, and put away your levels and t-squares. :)


I've been lurking in this thread and can't disagree with anything, really, as I've experienced it firsthand myself.  But anecdotally, I've also experienced the other side of the coin:  At first my TARDIS wasn't particularly accurate.  Indeed, this was by design.  I was trying to keep the build simple and also inexpensive.  It was blue.  It had panels.  There was a light at the top.  Pretty much everyone that saw it loved it and said "it looks just like the TARDIS!"  But I was never happy with photographs of it for some reason.  And by then I'd become a part of this band of madmen.  So when I figured out a way and found an excuse, I did a refit to bring it more in line with the 1963 box (albeit with a 1970s paintjob and more authentic police box signage).  I loved it so much more.  And everyone here loved it more BUT... 

On more than one occasion I showed a friend a photo of the original build and then a photo of the remodel and got "Eh.  I guess.  What's different about it?" 

So I understand your point and actually buy into it because I've experienced it myself, but I've also seen it have no effect.  I don't know, maybe it's because a cell phone lacks the resolution to really generate the effect. 

Anyway, the level of work going into this brings it on par with the guy building the exact police box replica (and I really need to see him finish that build!) and a few other builds here.  Very fun to follow.
"My dear Litefoot, I've got a lantern and a pair of waders, and possibly the most fearsome piece of hand artillery in all England. What could possibly go wrong?"
-The Doctor.


Well it's not the people who can't tell the difference that we're worried about, it's the ones who can, right?  For some people, blue rectangle with light on top = TARDIS.  That's the low-hanging fruit.  Convince the die-hard'er and now you've got something.

We have the same thing in CG - I call it the magic 70%.  If we get an image about "70% photoreal," then about 70% of the general population will buy it. But we don't target the ones who can't tell the difference.  We target fellow VFX supervisors; the most discerning who instinctively look for "tells."  Convince them, and you've got a future, kid.

In the end - and this is just personal philosophy - the approval I'm after is the hardest one to get: my own.  I've set out to make the most accurate s18 I can, and I have a personal definition for that.  And I'll know if I have or haven't done my level best.  So I have to do my level best; I don't feel I have much choice.  After all, this is what I do for a living, and I do it for a living because it's what I do naturally.  And as much as I enjoy doing it, I equally enjoy sharing the process and learning from others.  When I first joined the board, I saw Timerotor's TARDIS and thought, "Well, that's the bar to top."  We're not actually in any kind of competition - that's silly - but it gave me a target; a perspective; something which really motivated me and was hugely inspiring. And I hope that all this obsession and documenting of the process inspires others, too.

We're all fans, and everyone who picks up a hammer and wood and actually goes about getting one of these together deserves respect, in my view.