Scarfwearer Wooden console plans

[darren79]

Tony, your digital art just goes to show it's not how amazing the software used is but the skill and talent of the user that makes the difference!

Awesome!

[galacticprobe]

I may well be entirely wrong in this, but even though I'm happy to accept that the lower panels appear to be taller than the top ones I'm having a bit of trouble getting my head around the idea that the prop builders at the BBC would have done anything other than cut out 12 panels the same size and shape and fix them on to the hexagon as quickly as possible.  They weren't aiming for a beautiful piece of furniture, they were making something quickly using techniques and build quality worthy of Blue Peter - that it was covered with and probably held together with sticky backed plastic pretty well guaranteed that this prop was always going to have a pretty limited shelf-life

Well, I'm afraid I must disagree with you. A beautiful piece of furniture is exactly what they did build. If you look through the different topics in here (http://tardisbuilders.com/index.php?board=54.0) which show some remarkably close-up views of this console, you'll see the high level of detail that went into building this. (You'll also find some of that detail brought to life, with some discussion, in http://tardisbuilders.com/index.php?topic=284.0 Build Diary.

Tony, I hope all goes well as possible with whatever gets in the way. Unless you're on a trip to Mars I hope you'll get to spend at least a little time here in the coming months.

Tony, the same goes for me as well. Hopefully things will work themselves out sooner than later and you can get back to us. (You will be missed while you're away.)

Dino.

[amos pogle]

GalacticProbe - I expressed myself badly, I was not aiming to imply that the finished result wasn't beautiful, I've studied all these threads closely and am building one myself at the moment because I like it enough to want to have one in the house.

I've taken on board the comment about looking at the evidence that is in front of our eyes, but am not currently seeing it in a way that convinces me that the argument that the top panels are different to the bottom panels, rather than simply raised on two thickness of 12mm plywood on a part where the cameras never seem to have landed.  If I had the mathematical and geometrical skills to take one of the pictures and calculate the angles to see if they are different then I would be doing it.  If anyone here has the skills, time, patience and inclination to do this then I am entirely open to changing my mind. 

I'm honestly not trying to pick a fight with anyone, just attempting to get my head around what I'm looking at.

[Scarfwearer]

This is the image that lespaceplie put up some time ago that makes the difference particularly evident (I talked about it in my post above with my revised plan).

http://tardisbuilders.com/index.php?action=dlattach;k=20;topic=1102.0;attach=102887;image

It's not a very high resolution image, so I've looked at others to confirm. We have higher resolution pictures now, but this early one makes the point fairly well.

The green lines are in the same plane, so you can measure them relative to one another.
This shows that the upper and lower panel heights are in the ratio 3:4, which I don't think anyone had noticed before.

The angle of the panels below works out at 45° because they are as tall as they are wide.
The panels above work out at about 37° because the upper panel supports neatly form a 3-4-5 Pythagoras triangle.

So the lower panels are about 2" taller than the upper ones.

The plinth is pretty much exactly the height of everything above it.


(My own build of this console was to my version 1 plans, so I had the panel heights the same above and below. Am I going to take it apart and fix it now? Probably not...

[galacticprobe]

GalacticProbe - I expressed myself badly, I was not aiming to imply that the finished result wasn't beautiful, I've studied all these threads closely and am building one myself at the moment because I like it enough to want to have one in the house.

Ah; such goes the way of the typed word: no inflection, no facial expression, nothing to indicate emotion. It's easy for things to be misunderstood. (It's happened to me enough times that I've started inserting smileys as needed, or an "outside narrative" to my comments, again as needed <said with the sincerity of new understanding of the previously-quoted post>.)

Anyway, I just couldn't resist using that famous line that Tony has used on me more times than I can count: "I'm afraid must disagree with you on this" (sometimes preceded with "I'm sorry, Dino, but..").

I've taken on board the comment about looking at the evidence that is in front of our eyes, but am not currently seeing it in a way that convinces me that the argument that the top panels are different to the bottom panels, rather than simply raised on two thickness of 12mm plywood on a part where the cameras never seem to have landed.

I see now. Personally, I think when you take into account the "frame" around the top fold-down panel covers and compare that overall measurement to the frameless lower panels, you'd find that they are the same size measured edge to edge. At least that's how they look in the higher resolution images of the console.

If I had the mathematical and geometrical skills to take one of the pictures and calculate the angles to see if they are different then I would be doing it.  If anyone here has the skills, time, patience and inclination to do this then I am entirely open to changing my mind.  

Well, once you ask a question like that, Scarfwearer is usually one of the first (if not the first) to show that geometry. (Me, on the other hand, totally sucked at geometry when I was in school. ) When talking strictly size (as above), Tony is the most astute person I know when it comes to finding an object of known size in an image, and using "pixel counting" to determine the pixel-per-inch (or pixel-per-millimeter) to do some very detailed determination of size... and he's usually not off by much more than 0.5mm!

I'm honestly not trying to pick a fight with anyone, just attempting to get my head around what I'm looking at.

Steady on, pogle; no one is accusing you of that. Disagreeing with what someone posted (even if the post was misunderstood), doesn't mean anyone is being accused of picking a fight. It's called what I like to refer to as the U.S. TV show '60 Minutes' edition of their "Point-Counterpoint" segment. It's a discussion, or as referred to in the scientific community, an "argument". One side poses its argument, and then the other side poses its counter-argument.

Argument doesn't necessarily mean what people commonly think of: the yelling, shouting, finger-pointing, etc. "Argument" is also defined as "a reason or set of reasons given with the aim of persuading others that an action or idea is right or wrong" (i.e. 'There is an argument that the top panels of the console's overall dimensions are the same as those of the console's bottom panels'... or 'There is an argument that the color of <insert name of item here> appears more gold than green on screen due to how the studio lights reflect off the material the <item> is made of, and the filters used on the camera lens during recording')... if you know what I mean.

So, in short: we're all good. We're just discussing how a prop was constructed, and this happens to be one of the more difficult props for any of us to wrap our heads around!

Dino.

[Scarfwearer]

I thought I'd walk through the methodology that led me to the console dimensions I posted earlier.
It's possible that the numbers I got are not accurate. It's also possible that my methodology is not a sufficiently good approximation.
The Gold Standard I think has to be photogammetry: creating a single 3D wire-frame model that fits exactly over a selection of images of the prop seen from different angles. My hope when people started posting 3D model renders of Tardises and consoles was that this would be the outcome, but it's generally been used for artwork rather than measurement checking. Mike Verta did a little of this for the TYJ many moons ago, but never finished it. I've tried going up the learning curve with Blender to do this myself, but I just can't get my head around it yet.

So this is what I have in the interim:



The general idea with measuring off pictures is that if you can find pairs of points that are at the same distance from the viewpoint in a picture that you can measure the distance between them, and compare the distance between other pairs of points at the same distance from the viewpoint. If you know the actual distance between one pair, you can determine the distance between the other pair by ratios. This is well known and much used.
When a plane is not parallel to the picture plane the midpoint of an object face in that plane moves a distance from the midpoint of the picture of it, so perspective distortion results.

I took several steps:

1. I rotated the whole image a small amount to make the plinth straight-on to the picture, so that it's widest extent is horizontal. This is just for convenience in measuring.

2. I determined the centre of the plinth by identifying lines that are known to go through the centre: a line between opposite rail posts, a line across the visible join between the two halves of the plinth (both in black), and a vertical line half way between the widest points (by measurement - in orange). These all agree to within measurement error.

3. I determined the actual diameter as a horizontal line across the plinth (in green). This is behind the apparent widest point because of perspective distortion, and is the only line of known length: 2200mm as marked on the BBC plans. The green line is on the vertical Reference Plane, which is parallel to the plane of the image. Imagine it as a thin wall that cuts the plinth along the green line.
This could be determined algebraically with more effort.

4. I determined the apparent width of the console (purple) by the method of ratios. This comes to 998mm. I got the same result using a drawing tool by scaling the image so that the diameter was the correct size and using the measuring ruler drawing element. The numbers on the diagram are not typed in, they're filled in automatically by the tool when I draw the ruler between two points.


There are a couple of corrections to apply, because the opposite corners of the console are not actually in the reference plane.

Firstly, the console is clearly rotated from its widest aspect, and so appears narrower than it really is. We can determine this rotation by overlaying a regular hexagon on it (shown in red) so that the edges of the console plinth align with the hexagon points. I did this by trial and error, and the rotation comes to approximately 15°.
An algebraic method using measurements of the apparent edge spacing would likely be more accurate, but this is probably good enough.

Knowing the angle, I calculated the corrected width of the console using 1/cosΘ, which is matched very well by the drawing program I used, as shown in the hexagon to the right.
So the width of the console looks like 1033.5mm so far.

Secondly, I noticed that the console is not actually centred in the reference plane at the actual diameter of the plinth, but slightly behind it, so it appears yet smaller than it should.
The leftmost claw foot of the console is very close to the diameter line. Because the console is known to be rotated 15°, we can calculate that the rightmost foot is actually set back about 200mm (8 inches) from it. This means that the centre of the console is set back about 10cm.
This is not a lot, but is enough to be measurably affected by perspective:

If you look at the railings either side of the console, the ends of them have different apparent distances apart (shown as blue lines) by about 15%. This is over a range of distances from the viewpoint of about 100cm (the rail lengths). The offset of the console from the reference plane is about 10cm - a tenth of the difference, so I roughly approximate the apparent reduction of the console size as 15%/10 = 1.5%. Expanding the calculated width of the console of 1033.5 by 1.015 (1.5%) to compensate, we get: 1049mm.

And that is how I arrived at a figure of 105cm for the width of the console.

Even at this good resolution the console is only around 400 pixels wide, so each pixel corresponds to 2-3mm. So that is the error bar on any measurement. Since measurements are usually done at two ends, this doubles the possible error, so we're not going to get numbers with better than about 5mm of accuracy from this image.


I'm not particularly claiming that this number is correct or even that this methodology is correct (it's actually only an approximation anyway - and I've glossed over at least one place where there may be some further perspective distortion that I think is probably negligible). It's just what I used to arrive at a number for the scale, and why I now think the console is larger than I used to. To some extent it's still guesswork.

[oslm]

Hi guys,

With the help of the console drawings Tony did a few pages back, I have done an update to the 3D model using those measurements--







Tony, your measurements have been a great help and thanks for taking the time to create them --

Later

Oslm