BTS The Star Destroyer bridges of the Original Trilogy

Thanks for clearing that up. Though I don't agree that the differences are enough to consider them different classes. Could simply be the ISDs were upgraded, also called refitting, with improved domes and arrays. The ANH arrary was simply turned 90 degrees for TESB and ROTJ. Also sometimes later built ships in a class can be built a little differently with design improvements learned from earlier ships in the same class. Such as the Iowa's sisters ship, the New Jersey getting a enclosed bridge, The Titanc's sister ship, the Britanic, the 3rd ship in the Olympic class, which had improved taller water bulkheads, double hull skin, more powerful turbine engine, and larger lifeboat davits to double the number of lifeboats. Came in handy since the Britanic also sank during WW1 after hitting a mine. There can be a great deal of varience in the same class of ships.

 

True - but it's been commented that the "ISD-II" has different proportions as well - it is noticeably narrower in the beam.

 

Iron_lord wrote

If Edlund's 500 mile wide Death Star II can be wrong, can't Peterson's 15 or 16 mile Executor estimate, be wrong?

Fair enough, but AFAIK Edlund's job in ROJ was mostly VFX compositing and camera filming while the ILM model shop decided on the sizes of the ships.
I agree that Lorne's statements should be taken with a grain of salt, as the actual large conning tower model made for ROJ is the decisive element that enables us to make accurate size determinations for the rest of the Star Destroyer bodies those conning towers rest upon.

Once Rogue One comes out on DVD - maybe someone can take some screencaps of various other scale-determining factors besides the width of the Bridge?

Guess, what the second thing I'll do once I received my Blu-ray copy will be...

For example, the Hammerheads - established as around 90m long in Rebels - use the Hammerhead collision to determine how big the Star Destroyer needs to be.

I consider the feature film Rogue One to be the one that decides the outcome. The bridge film set dimensions are known and ultimately determine the size of the RO Star Destroyer.
I also thought about the Hammerhead Corvette (strange name choice, IMHO, that's the Blockade Runner, too, isn't it?), but should RO suggest that it's actually smaller than depicted in Rebels then the feature film should have the last word, IMHO (heck, these Star Destroyer conning towers in Rebels look like caricatures - should I really take these seriously? ).

Slicer87

As you can clearly see in my image comparison there ARE conning Towers that noticably differ in size because of differently sized bridge modules:



The one on the right clearly belongs to an Avenger-class Star Destroyer because the bridge proportions in relation to the rest of the conning tower match the dimensions of the Falcon attached to its back in ESB.

If the model makers supposedly did not pay attention to these details, why did they even devote time and money (amidst an already hectic production) to manufacture another bridge module?

 

Who knows? Things get revised, old shots get reused, and so forth.

Saxton points out that the lights on the towers, being generally all the same places on both, and being equal size and brightness - mean that the towers are more likely to be the same size:

http://www.theforce.net/swtc/towers.html

I don't agree with Saxton on everything, by a very long way (as those in the Fleet Junkies thread can confirm). But I do think he makes a moderately good case for mile-long Avenger-type, mile-long Devastator-type, and 11+ mile long Executor-type.

His main argument for Mile-long Avenger:

(he also points out that there are "ambiguities in the size of the freighter") for the Falcon, hence, Falcon on back of bridge tower, should be handled with caution.

 

I always thought the Falcon model on the back of the SD conning tower was too large of a scale. Perhaps the MF scale was enlarged for that one scene so it shows up better on film?

While I agree the closeup conning tower models are different scales. However, they do not match the scales of the full ship models where their conning towers are identical sizes. Perhaps it was a late in production change, or as I stated before, they enlarged the SSD tower closeup just for the sake of extending the shuttle flyby. IE, instead of writing the flyby scene to fit with model continuity established by TESB, they changed the model scales around to suit the scene instead. Much like how the MF interior is changed around to suit the scene, or even the interior of the SDD bridge.

 

They used the TESB lights layout for the Rogue One Star Destroyer as well:

http://comicbook.com/starwars/2016/...-detailed-versions-of-x-wing-and-darth-vader/

Much like the look of Darth Vader in Rogue One, which had to be tweaked slightly to fit that delicate balance, so did the Star Destroyers, the Empire's huge capital ships, including one in particular.
"I think the Star Destroyers look pretty authentic. It's not matching any one particular Star Destroyer because there were multiples for models that were built. There's the three footer for New Hope. There was an eight footer for Empire and then there were high detail close up sections that were built for when the Falcons hiding on the back of the tower," Knoll explained. "What we did is, because of the order of events when this takes place, philosophically I wanted to match the New Hope Star Destroyer, because that's Vader's Star Destroyer and maybe it shows up in this picture," he said with a laugh. "If you compare the three footer to the eight footer, they're kind of the same shape, but if you really look there's a lot of things that are different about them. Just the purist in me said, 'All right well we should be matching with three footer.' But then you couldn't really match the three footer because there's problems with that, that whole upper surface is essentially un-detailed."



Those porthole lights you remember on the Star Destroyers? They didn't exist on that 3-foot model. That presented a problem to Knoll and the ILM team, and so, like Vader, they went for a hybrid option.
"The lights all came with the eight-footer that was built for Empire. What we built was this kind of hybrid; it was the plan form from the three-footer and a lot of the things that you could see as differences between the two models were following the three-footer. But then a lot of the upper surface details curbed form the eight footer from Empire, and then it's got all of those midline lights and the portholes that the eight-footer had, because that's how you remembered it."


Sounds to me like John Knoll regarded the ANH and TESB Star Destroyers as around the same size in-universe - which is why details from the TESB one, can be copied onto their updated representation of the ANH one.

 

Probably what happened is that for TESB, the model makers had a bigger budget so they built better models with more details and made some artistic changes, a revision. Though I do think the ANH model had better main turret designs over the later models.

 

In the context of the ion cannon scene, where both models were used to represent the same ship - that's plausible.

However, in the context of Rogue One, where they lovingly recreate most of the ANH Star Destroyer's details, it does seem like they accept the differences as existing in-universe.

 

Another pair of shots that might provide a rough idea:

they take place in immediate succession - so it's pretty clear that both depict the same ship underneath the Executor, very close.

 

Saxton argued around 12.5x as long as ISD for the ones I mentioned above:

http://www.theforce.net/swtc/ssd.html

In the first picture above, the grand vessel clearly dwarfs the three attending star destroyers. Pay special attention to the destroyer which is in front of the Executor's docking bay opening. If the stern of the Executor is extrapolated to its proper extent beyond the left edge of the picture then the projected lengths of the two vessels can be measured and compared. One can then correct for the orientation of the ships to gain limits on the relative lengths. The most important thing is that the destroyer is closer to the viewer and therefore relatively "magnified" by the effect of perspective. It can thus be shown that the Executor is about 12.5 times the length of a destroyer. Given that both vessels are at a similar distance from camera in this shot, the destroyer will not be greatly magnified by perspective, and this length ratio must be close to the true value. The only qualification is that the destroyer is pointing in a slightly different direction; this may reduce the lower limit by up to about ten percent

between 10.6 and 11.17x as long for this one (though he seems to have confused port with starboard - the more distant ISD is off starboard side):

One of the most valuable sequences for this purpose is the encounter between the Executor and the personnel shuttle piloted by rebel commandos. These images contain two destroyers moving in formation with the Executor. The first is slightly closer to camera than the command ship. The distance of the second destroyer is unproven, but is further away, possibly off the Executor's stern, portside. Length ratios obtained for the escort ships are 10.62±0.02 and 11.15±0.02 respectively. Combining these constraints, the command ship's length must be somewhere between 10.60 and 11.17 times the destroyer length.

and mentions that one Anoat shot yields an upper limit of 25x as long - possibly this one?

 

Iron_lord wrote



they take place in immediate succession - so it's pretty clear that both depict the same ship underneath the Executor, very close.

I concur, but the first thing I notice is that the conning tower of the Star Destroyer underneath Executor is only 3 quarters the width of Executor's or even smaller, depending on how close it actually is.

Now, I wholeheartedly support Saxton's calculations and examinations, but at the end of the day find myself unable to take these (and others) really seriously as long as these ignore a) the explicit ILM chief model makers' statements (admittedly not yet published / known at the time of Saxton's writing), b) the actual size references we do have at our disposal (ROJ VFX conning tower model with different bridge modules) and c) the actual onscreen evidence (e.g. Falcon attached to Avenger's back, balcony bridge on conning tower briefly seen in ROJ).

The evidence speaks for itself: Lorne Peterson adopted the SSD's conning tower from a (larger) Devastator-class / ANH model Star Destroyer, proven by a) the slanted deflector-shield power generator domes on both the ANH Star Destroyer and the SSD miniature conning tower and b) the fact that only Devastator-class / ANH model conning towers to the port and starboard side of the Executor were provided as a size reference during the Hoth fleet arrival shot:



The conning tower width of an SSD conning tower was ultimately determined by the large VFX conning tower model built for ROJ, i.e. 357 meters based on the (known) dimensions of the command bridge width.

Since the conning tower of the SSD had the "the same size as that of the smaller [Star Destroyer] vessel" (Sculpting a Galaxy, 2006) we should have known after ROJ that there is a Star Destroyer 6.28 times the 357 m conning tower width with a length of 2,242 meters, which obviously must be the Devastator-class, because a) two of its class (during the making of ESB) were used as SSD size reference and b) only with that kind of length could there be a main bay 264 meters long that's necessary to justify that we can see the whole Tantive IV inside this bay in ANH (and not just the bottom part of it filling the entire bay), regardless whether its 123 or 148.8 meters in length.

When we add the 2,242 meter overall length figure of the Devastator-class to the 1,270 meter overall length figure of the Avenger-class and divide it by 2 then its 1,756 meters which IMHO is compatible with the imprecise "one mile" figure, which is ultimately revealed to be first and foremost an average value but most assuredly NOT the standard length of ALL smaller Imperial Star Destroyers.

 

If the towers on the two Devastator-types shown are the same size as the Executor's - does that make the Executor 11x 2242m, using your Devastator figure? (24662m)

 

Iron_lord

I don't understand.

The miniature conning tower on the big VFX model of the SSD is 4.2 cm wide.

4.2 cm equal 357 meters in real size.

The big VFX model of the SSD is 277 cm in length. By simply applying the mathematical Rule of Three - https://en.wikipedia.org/wiki/Cross-multiplication#Rule_of_Three - the real size length of the SSD is therefore 23,545 meters or 14.63 miles (based on a conning tower 357 m wide and knowing the dimensions of the Executor VFX model).

 

I was starting from the "11xDevastator" figure - since, when you get to the 4.2 cm figure, minor variation can result in major change when scaled up.

Using the above method instead, produces a slightly lower number - but it also produces 10.5x rather than 11x Devastator length. The EU (and newcanon) erred slightly on the high side rather than the low - their Executor length was 11.875x their Devastator length rather than 11x.

I'm guessing you're having Executor and "Emperor's SSD" as the same length and class, with the only difference being bridge interior?

 

Iron_lord wrote

I'm guessing you're having Executor and "Emperor's SSD" as the same length and class, with the only difference being bridge interior?

I do now, that's correct. In the past I had been under the misimpression that the conning tower of the Executor was adopted from a one mile long Star Destroyer which was incompatible with the large VFX conning tower built for the SSD in ROJ.

Having finally realized that Lorne Peterson adopted Executor's conning tower from a Star Destroyer "smaller" than the Executor but larger than one just being one mile long, I do not think there is a really good reason any more to assume that the SSD's differ in size.

Technically, that's not entirely correct, however. According to the large ROTJ conning tower model the SSD in ROJ is 14.63 miles long, while Lorne Petrerson stated the length of the Executor to be 16 miles, 8 times the length of an "11,000 feet" long Star Destroyer.

Again, this is information from both Sculpting a Galaxy and Rinzler's Making of ESB but neither provides the context of when Lorne's statement was made (early during the production of ESB?). Should I be able to relocate his San Anselmo address (provided he still resides there), I'd like to ask him about the subject.

Until then my theory would be this: For ROJ the ILM model makers sat down together and realized they had a last opportunity to settle Star Destroyer sizes with the large VFX conning tower model they were about to make. The built a large bridge module to retroactively deliver the bridge balcony of the Avenger, matching its 1,270 meters they couldn't feature in ESB.
Next (or first) they determined how big the bridge would have to look for the Super Star Destroyer (according to original storyboards we would have seen the bridge from "Mad Maxx's" point of view!):



Since Lorne had apparently adopted the conning tower size from a Devastator-class Star Destroyer, this was an opportunity to examine whether his 11,000 feet figure really held enough water. They examined the original Rebel Blockade Runner model (simultaneously prepped for shooting in ROJ) and probably arrived at the conclusion that the Devastator-class's main bay didn't actually to have to be THAT big and ultimately arrived at a conning tower width of "just" 357 meters for a 2,242 meters Devastator-class Star Destroyer as a basis to help them to scale the bridge windows on the large conning tower model accordingly:

 

One way to determine the size of the Rogue One star destroyer Dauntless (besides the bridge module) is to extrapolate from the Dauntless hovering over Jedha.





We know they filmed a specific mesa in Jordan, in the Wadi Rum desert, and used CGI to add the walls of the city to that mesa - and gave the city a detailed interior too:


If we can find a map of that mesa, and determine the exact dimensions of Jedha from the mesa map - we can get the Star Destroyer from that.

 

The size of the Dauntless hoovering over Jedha also crossed my mind, but I think the known and visible dimensions of the conning tower's bridge balcony will ultimately be the more decisive "yardstick" to reliably establish the size for this particular Star Destroyer.

IMHO, there is already good reason for doubt regarding the continuity of dimensions regarding of the "new" temple on Yavin IV, because some of the birds-eye-view images suggest a temple base of only 100 meters in diameter, while the panoramic vistas (also published in Cinefex # 151) appear to show a much larger base, almost twice as big.
I will eventually examine these discrepancies (Yavin IV is another one of my "pet subjects"), but I'm afraid they are there.

 

Regarding inconsistencies between lights and other features (and the ROTJ SSD tower looking like a much more detailed version of the Avenger tower), one possible solution that might make your work and Saxton's work much more compatible - setting the Dauntless aside for the moment since Saxton never worked on that - it was long after his time.

Solution is as follows:
Let all Avenger-type ISDs be the same length as the 2242m Devastator - with comparable sized towers (357m) . Multi-bridge, "Invader-type" - and the Avenger itself.

Result - the lights matching up to updated version of Devastator-type, stops being a problem.

The high-end estimates of the "Invader-type" (Lambda shuttle exiting) of just over 2km - those work too.

And the "multi-bridge" scene - with the Falcon coming round the edge of the tower:



stops being a problem - that scene is pretty compatible with a 357m tower, and has been used to argue for the Multi-Bridge ship being larger than the Avenger, for years.

Set aside the Avenger TESB scene itself - as being bound by prop limitations. They couldn't create a Falcon significantly smaller than 5 cm, and still have it have sufficient detail to be suitable for the Avenger Tower scene - so they oversized the Falcon.

And now, all your figures almost exactly match Saxton's - but scaled up.

 

From: https://twitter.com/ThePropDen/status/841227944282472448

Just want to reference this image. It shows that the dome structures are detachable, and are not perfectly symmetric. Somewhere it was discussed how one of the SSDs is based off the SW77 model, which had non-parallel domes. I forget the specifics. More importantly, the detachability of these domes means that they do not inform a necessary sequence in which two different versions of the large scale bridge superstructure were prepared.

The following is subjective. I don't think the small-bridge version of the SSD can be squared quantitatively. ILM could have prepared the 5-bridge version first, and then used it in the scene where the shuttle flies improbably / spectacularly / theatrically close to the command ship. At which point Lucas spotted a discrepancy between the faceless, inscrutable Death Star in the background, and the military ship, in extreme closeup, that offered five transparent portals that allowed a human connection between Luke and his father on that ship... And then opted for a more faceless, more inscrutable, less windowed, less bridged version of this icon, so that the visual impact would be more of Luke between a rock and a hard place, more like the inexorable walls of the trash compactor. Or. Lucas presciently knew off the bat that he needed a falsely scaled superstructure to squeeze or press out the will out of the hero. And after that crucial scene was accomplished, the 5-bridge version was a convenient, but not indispensable, re-purpose. It could show well against the wonderfully detailed Mon Cal cruiser.

Unless there is some Imperial protocol that requires a ship trying to get through the shield to present itself for closeup inspection by flying right under the nose of the command ship bridge superstructure, it is completely daft ludicrous for someone approaching from out of deep space to "fly casual" and casually end up so close to the bridge of the biggest star destroyer type that exists. You cannot casually do that. It would rather be a stunt like buzzing a tower. Transmission strength in GFFA is not so poor that the shuttle has to practically dock in order to get good reception. Without some in-universe reasoning for why Chewy has to fly so close, is a deliberate error in service to a deliberate compositional purpose, and we give it a pass. Until this thread, I hadn't thought too much about the improbably small SSD bridge, because I give it a pass. Strength of the composition overwhelms the question of how that bridge is so small.

(To double check for tricks of the eyes, I did frame by frame through the sequence. The foreground ISD is definitely in front of the SSD, and definitely has the same apparent angular dimensions of bridge superstructure as the SSD itself and the ISD that is implied to be on the opposite side of the SSD.)

 

This is after zooming in on the bridge in the upper left storyboard. The center window of the SSD bridge is an inverted trapezoid. The windows to either side are upright trapezoids. The X-Wing's vector of movement is more centered on the window to the port of the center window (yellow arrows), than on the center window. Most precisely, the X-Wing's vector of movement is centered on the window that is two away from the center - another inverted trapezoid. For the upright trapezoid in question, I put a red arrow on the port window frame and a green arrow on the starboard window frame. The blue lines don't show up well but they were used to show where measurements were made.

1340 pixels bridge super structure width
184 pixels bridge width
260 pixels dome diameter

bridge = 0.1373 of bridge superstructure
bridge = 0.7077 of dome diameter
dome diameter = 0.1940 of bridge superstructure
So this storyboard depiction is consistent with the SSD having a bridge width analogous to a standard ISD bridge width.

Unless that 5-bridge closeup was originally shot in an intention for something that is not well understood or characterized. It is easiest to assume that it represents a standard Star Destroyer like an Avenger class, but it is most decidedly not the same in its facial appearance. The 5-bridge face is also difficult to square with the MF on the back of the Avenger superstructure. The MF composite behind, around and against the 5-bridge is smaller than the proportion of the MF prop to Avenger prop. Holding the MF constant, the Avenger would be smaller than the 5-bridge. But it's easier to assume a simple compositing error, or a deliberate error of scale to increase the tension. (The BR scaled wrongly against the Nebulon B serves zero narrative purpose. The MF scaled wrongly against the BR might increase tension.)

 

That's an outstanding never before published image of the stern of the large conning tower made for ROJ, great find (I think I have to forward you my "rare VFX models images" 'shopping list' some time...that would include on top a front view of this model with a close-up look at the larger balcony bridge module and those other 'windowed' elements )

 

I. I read this once and I won't recall any specific findings or epiphanies. I know that the ESB scene with "first catch of the day" presents a problem to a naive and straightforward assumption. A more stable perspective lies in the ESB scene showing the perspective of the bridge of a presumably standard ISD. But that is valid only until ROTJ. The three ISDs that give chase to the Falcon are presumably Avenger class, and may for that matter include the Avenger. The perspective that is shown is looking out the port bridge windows. This perspective on the set includes the blue screen backing that is draped behind the port forward span of larger windows and the port alcove / recess span of three slightly smaller windows. The post production places a single homogeneous image of a passing ISD waterline behind the entirety of the set. One must first break down the contradiction. The Captain who says "first catch" is to all appearance looking out framed windows that are actual pressure panes in a direction parallel to the forward vector of the ship, as you have illustrated. The immediate environment of these three windows is then seen in the scene where two of the ISDs are about to collide. That very same section of the set that has the three smaller windows now shows an outside view that is homogeneous with the view that is occurring outside the forward port bridge windows.

Without great need for conversation about it, the model makers and the set designers were not on the same page at the end of their works. I don't own Rinzler Making of ESB, and I don't know time lines or who to assign the greatest onus for not striving for consistency with the first worker. The Avenger, as a model, serves its function. The 3-gun battleship turret on the face of the hammerhead serves its function to be the notional location and shape of the bridge, without being representational. No models were built in ESB that made an attempt at representation of what the exterior of that Elstree bridge set might look like. Once the ILM model makers had the means motive and opportunity to fairly represent the ESB bridge set, with its port alcove windows that to all appearances translated to the eye photons from the same plane of perspective as the forward port windows, those ILM workers did not elect to be thoroughly representational. They settled for less. Neither the SSD bridge face nor the 5-bridge face show the proper geometrical forward prolateness of the ESB bridge set, nor do they show the port alcove windows by giving the representation of the bridge set some extra protrusion from the face. They opted for an oblate and shallow protrusion.

The existential crisis is as follows: It is easy to begrudgingly assume that in GFFA, if they have the technology to generate an image out of thin air, then they might as well have the technology to generate an image of Hoth on one of the port alcove surfaces that we were first led to believe were actual windows, but then come to suppose might be displays, once we see that alcove in its context with the forward port windows, all showing the same homogeneous display of a solid moving vista (ISD waterline). Because we have not yet heard of a movie called Revenge of the Jedi, or seen any new ILM models that try to depict what the exterior of this ESB Elstree stage set looks like, we are bound inside an assumption that we are looking through port alcove windows. We are bound to suppose that we are seeing light transmitted through true pressure panes separating a pressurized volume from the cold vacuum of space, and Not light transmitted from the surfaces of three trapezoidal 4K LCD screens inside the pressure hull that are for some reason framed in the same bulky gray structural members as are the real windows. So, prior to May 25, 1983, and prior to select scenes showing the ILM models that purport to represent the exterior of the bridge, we are able to suppose that the port alcove is actual windows that May, on demand, at the Captain's leisure, serve as displays. Like how the captain with 'first catch', to all appearances, could have done. We can do that all the way up till May 25, 1983. As of May 26, 1983, the idea that the port alcoves in the ESB bridge set can be window (pressure panes) must / ought to be thrown out. It retroactively forces the difficult rationality that those port alcove surfaces were actually displays. Not windows, not pressure panes. (Unless there truly is some as-yet undisclosed ISD class that has a prolate, protruding bridge, and there are yet more ISD designs that have a bridge 'set' that is as shallow and oblate as the ROTJ ILM scale studio models would have them be.

To then rewatch / savor ESB slowly in a snifter with the knowledge that those port alcoves cannot be windows, one ought to be struck with a wall-breaking bit of movie magic. The now-LCD-displays, which understandably gave the first catch captain a perspective of Hoth that matched what the ISD saw directly in front, are now giving George Lucas' cameraman's camera lens exactly the perspective, from their LCD displays, that conveniently matches the real transmission of reflected photons coming through the real forward port windows, so as to give the illusion that the port alcove surfaces are pressure pane, light-transmitting windows. So this throws a contradiction. One cannot simply (boromir) measure the ILM studio scale models and their particular bits that purport to represent the exterior of bridges. Because they always and already are not valid representations, IF one holds that the Star Wars universe is not supposed to know that George Lucas is watching it, so that it changes its behavior in response to this observation. It's a contradiction and error Ought to propagate very generously into the calculations anyone is making.

II. Further slow reviewing of the 'fly casual' scene. I can now lower the improbability with which the shuttle flies close to the SSD superstructure. It requires that the SSD has detected the shuttle, and is serving as the bouncer outside the door to the bar, and may even know that a particular bad apple is about to show up. The SSD, if it has the maneuvering ability, places itself between the approaching shuttle and the cylindrical column of shield-protected airspace immediately between the DSII and the surface (generator). Furthermore, after Han's demonstration of knowing Imperial procedure, it is reasonable that he has the experience that there is a standing expectation of scrutiny - not just in the release of a pass code. Chewy's path is to get to the base of the DSII where the air space starts. The SSD maneuvers its bulk to impede that direct line. The SSD maneuvers its hammer head to the closest proximity to investigate this newcomer shuttle. There is nothing that Chewy or Han can do differently but to perform the confident face that they have business to attend to, and that business starts at the base of the DSII, and that is where they are headed specifically because it is a standing logic that those who lower the shield are not going to wait for the shuttle to get close to the surface before opening it. No, the shield will be opened for a short time only, and the shuttle needs to get its ass inside that protected airspace promptly. So it must take the most direct, straight line it can to that airspace, not to the surface. And so the shuttle is not able to make a maneuver downwards towards the surface Until it has passed the bulk of the bouncer SSD. So that is what, I think, the scene is doing mechanically.

 

The closest approximation to the protruding bridge above:

which doesn't include the alcoves as part of the protrusion, but does project forward in an outright semicircle, are the Star Destroyer bridges in Rebels.

http://vignette1.wikia.nocookie.net...-Bridge.png/revision/latest?cb=20141024175248