What Do New Kits Have Inaccuracies

[Jmaracing]

There are people on here who scratch up some insane stuff. Whole cars built from sheet and strip stock!

They take all the measurements, divide them down to scale, and recreate them flawlessly.

If one person can do it with no compensation other than the pleasure of building, why can't a huge multinational corporation pay people to do it, too?

I agree!!

[Petetrucker07]

I think it's a case of, I don't know how to your job, but my book says your doing it wrong.

It's people who have no idea what they're doing. They have a book with pictures. To them, that's all they need. Then every so often they get a chance to "look" at a 1:1. Then they say, yeah our drawings and everything look pretty close to it. Project approved!

[Matt Bacon]

Well, I think we can all agree that the Moebius Chrysler 300C is a pretty good kit, made by enthusiasts. Yet even they didn't scale down by 1/25 every component on the car and reproduce it in plastic (not even a bonnet/hood hinge mechanism, I mean, I ask you...;-P). There are always going to be a host of compromises to achieve something that's moldable (no one has EVER made a car kit with panels 1/25th of the thickness of the real things...), that can be assembled by a wide range of skill levels, isn't insanely complex (no pistons, cams, crankshaft and con rods INSIDE the engine) and affordable at the same time. And I strongly suspect that the 300C was as much a "labour of love" as any car kit that's been produced recently. Someone could ask them how the economics stack up in terms of time, tooling costs, rework etc. I think you'll find that rather a lot of sales over rather a lot of years are required to break even on the small part of the margin on that product that finds its way back to the manufacturer...

If one person can do it with no compensation other than the pleasure of building, why can't a huge multinational corporation pay people to do it, too?

Because if you ask that one person how much they'd have to be paid to make it a job, rather than a hobby, you'd find the answer is too rich for most of our pockets; and because the multinational corporation has an obligation to their shareholders to make the best profit they can... and not sink vast sums into developing the perfect kit of a product that will sell to a few thousand enthusiasts.

And another thing -- if I tool up the best 1/48 Spitfire IX ever seen, I can go on selling that to people for decades. If I tool up some obscure hot rod, or the 1957 model of a car that a few people's dad's had, the return on investment is going to look pretty creaky. I'd go as far as to say there is NO single car subject with the longevity and broad audience that you'd find for a Spitfire, Bf109 or Tiger tank. That's just the way the hobby is. Same's true for the ship guys, the figure builders, and the ten people in the world who think that the Dora rail gun is the coolest thing ever built (and before you remind me, I know that there have been two 1/35 kits of the beast -- I've seen them at shows, massively discounted. I don't think they'll be seen as a wise investment decision by either company!)

bestest,

M.

[Deathgoblin]

I just look at most inaccuracies as a challenge to correct. I've seen some real ingenuity on here to correct some minor and even major flaws on kits.

[1930fordpickup]

Sure they can measure and do the math , they have a computer to make it right, BUT then the human element walks in the door and claims that is does not "look" right.

[Longbox55]

And how do you know this? Show us the facts, I want proof, not just you saying it.

I don't recall where I saw it, as it was a while back. I also have not really looked into it myself. Just passing along what I've read about the kit.

[martinfan5]

It doesnt cost more to get it right in the first place

[Ace-Garageguy]

It doesnt cost more to get it right in the first place

Exactly. I wonder just why this seems to be so hard to grasp.

[Matt Bacon]

It doesnt cost more to get it right in the first place

It's hard to grasp because it's not true. It certainly does cost more if "getting it right" means sending people halfway round the world to photograph and measure a real 1:1 in a museum or collection somewhere. I've lost count of how many long-awaited new aircraft kits have turned out to be "utterly unbuildable" or a "caricature of the...", when what seems to have happened is that researchers and toolmakers have relied upon existing plans (in some cases many sets, each reproducing the errors of the earlier ones) and photos rather than looking at the real thing. Interpreting 2D photos into a 3D model is not an easy thing to do. And until recently, the laser scanning technology to capture a real prototype in 3D has been rare and unaffordable. I'm just hoping that the recent trend for collectors and restorers to 3D scan their precious metal so that bucks to make replacement panels and other parts can be kept in storage means that there will be a globally accessible archive of 100% accurate 3D models of some of the greatest cars ever made to help model companies do a bang-up job...

bestest,

M.

[Matt Bacon]

Let me give you a "for instance". I'd like to build a model of a De Havilland DH.91 Albatross:

A beautiful plane, I think you'll agree. I've had a few photos and some three-view plans from a standard book for about ten years. I could have carved something that would look pretty much like the DH.91 out of balsa, and vac-form the main parts, any time in that time. Would it have been 100% accurate? No, probably not. I've been collecting every picture, plan and reference I can find for the rest of that decade. I've been corresponding with the De Havilland museum to find out what drawings they have available for about five years (and this bearing in mind that the "general arrangement" drawings that manufacturers have which show the whole aircraft rather than the parts to be made are often not accurate scale plans themselves -- they were never intended to be). So, if I was actually charging my time, and costing it into a budget for a model, it most certainly _would_ cost more to "get it right" than get it more or less right...

bestest,

M.

Edited April 19, 2015 by Matt Bacon

[horsepower]

Could just be my view, but it sort of sounds to me like the old story of three blind men that come across an elephant and are describing it to each other, no one person has a clear idea of what they are looking at, or how what they're describing to each other has to do with what the finished product is really supposed to be.

In other words what the U.S. designers are trying to convey to the foreign tool makers lose something in translation, especially to someone who's never seen the end product in person. As we have come to discover looking at photos instead of walking around what you're trying to copy can suffer from how each person deciphers their particular angle of viewing.

If we as people who are familiar with the subject matter can't agree if what we're seeing in pictures is correct or not, how can we expect someone who has never been anywhere near the actual subject matter try to get the proportions and angles correct?

And we're as much at fault as the companies are, we're pushing for the release, and complaining to anyone who will listen that it's taking forever to get the product out, and eventually the powers that be are in the position of putting what they have in hand into production, or suffer the consequences of pressure from the consumers, and the bean counters who are screaming to high heaven about loss of time and possible sales by prolonging the sales even more.

Edited April 19, 2015 by horsepower

[Ace-Garageguy]

It's hard to grasp because it's not true. It certainly does cost more if "getting it right" means sending people halfway round the world to photograph and measure a real 1:1 in a museum or collection somewhere.

Really?

Then how was it even possible for companies like AMT and Johan to produce dimensionally and proportionally accurate models before laser-scanning, massive and cheap computing power, digital photography, instant transmission of vast amounts of data, CAD and CAE? They used film cameras, rulers, measuring tapes, calipers and pencils on paper. And there weren't armies of men making prototype models and tooling from them. There were a FEW very highly skilled individuals (who were reasonably well-paid, but not paid anything like what executives got).

And it was pretty obvious on a lot of old models that the guys designing the kit parts actually had some understanding of the functions of the real parts they were modeling. That's not always the case these days.

The more technology advances, the more excuses are getting made about why nobody can do their work to a standard that was established 50+ years ago.

And an American model company really should be able to hire a small team of locals to do accurate measuring of a 1:1 subject for what amounts to chicken feed.

Then transmit accurate data to the Chinese tooling designers, if necessary (though there seem more and more US injection-molding companies and rapid-prototypers online every month).

And are there REALLY no Americans left who can measure a car and make a tool without breaking the budget?

If that's REALLY the case, this country (and the West in general) has gone much farther down the toilet than I'd like to believe.

Edited April 19, 2015 by Ace-Garageguy

[chunkypeanutbutter]

And wouldn't a completely accurate kit bring more buyers than one "good enough," thus leveling the production costs?

Edited April 19, 2015 by chunkypeanutbutter

[martinfan5]

It's hard to grasp because it's not true. It certainly does cost more if "getting it right"

I think you missed my point, if you have the budget to get right , its not going to cost you anymore to get right or to get it wrong, so why not make sure you get it right.

[69NovaYenko]

It's hard to grasp because it's not true. It certainly does cost more if "getting it right" means sending people halfway round the world to photograph and measure a real 1:1 in a museum or collection somewhere.....

And until recently, the laser scanning technology to capture a real prototype in 3D has been rare and unaffordable. I'm just hoping that the recent trend for collectors and restorers to 3D scan their precious metal so that bucks to make replacement panels and other parts can be kept in storage means that there will be a globally accessible archive of 100% accurate 3D models of some of the greatest cars ever made to help model companies do a bang-up job...

bestest,

M.

Without a dought that you have touched upon some excellent points in your responses.

I myself have a passion for aircraft and have tried my hand at building WW ll war-birds.

You are correct, it is challenging to find real surviving 1 to 1 military aircraft and armored vehicles to use as a template to build a scale replica. There are exceedingly few survivors. Hence, kit manufacturers have to send people halfway round the world to photograph and measure a real 1:1 in a museum or collection somewhere.

However, in my humble opinion cars, trucks, motorcycles are exceedingly more plentiful and more readily accessible. Hence, manufacturers should not have to send researchers world-wide to photograph and measure a real 1:1 subject in a museum or collection somewhere.

As you mentioned laser scanning technology and 3D modeling has become far more common and affordable now a days. Since this technology has becomes more commonplace and affordable there should be more accuracy in replicating scaled subjects. :-)

Edited April 20, 2015 by 69NovaYenko

[Matt Bacon]

And wouldn't a completely accurate kit bring more buyers than one "good enough," thus leveling the production costs?

Almost certainly not. The vast majority of kits are bought by people who aren't enthusiasts, but kinda know what the thing looks like. Make a "completely accurate kit" (like, say, the Fujimi Enthusiast Series) and the _average_ buyer is put off by the price and complexity. Revell don't make the wheels on their Ferraris the way they do just to annoy us Styrene Tifosi... they do it so the average 10-12 year old can end up with a car that has wheels that turn...

I think you missed my point, if you have the budget to get right , its not going to cost you anymore to get right or to get it wrong, so why not make sure you get it right.

Nope, I think you missed mine. If you have the budget to get it right, then of course you should get it right. But it IS more expensive to be _sure_ to get it right. With a limited budget you're less likely to get it right. My first carved balsa DH.91 might have lucked in and been correct, but the additional years of research and more sources make it much more likely that (when I get round to it) the version I build tomorrow will be correct. I'll still have to "guess" about the interior of the undercarriage bays, mind you...

As for Johan and AMT didn't they get a lot of support from the car manufacturers? They were, after all, making models of the latest cars off the production lines. It's not as if Johan's kits were stellar accurate, either... they got the body shell bang on and well detailed, but I'm not sure the chassis, suspension and engine detailing stands up to much scrutiny...

In the end, it all comes down to budget and return on investment. At the most basic level, give two guys two or three days with the real thing, with their cameras, measures, triangulation points etc, and you'll get a better result than with one guy looking at photos for half a day. But if the latter is all you can realistically afford, then your model is going to be less accurate.

http://www.airfix.com/us-en/news/airfix-development-news-avro-shackleton-mr2-172/

The skills are still there -- it's just that not everyone wants to/can afford to use them. (Airfix are looking at a "payback time" for that mould that's ten years away, if it sells reasonably for all that time). This will be an interesting "case in point", because Revell are ALSO releasing a Shackleton this year, and they HAVEN'T had access to the preserved airframe or the original drawings that the Trust own. Let's see how the price and accuracy stack up in the Fall...)

bestest,

M.

[ZTony8]

As long as a kit doesn't come out looking like a Palmer or Premier turd I'm happy.It's the parts fit problems that bother me more than some minor body goof.That's why I like to build Tamiya kits.

[Art Anderson]

An awful lot gets said about this, unfortunately by modelers who've never been involved in the process. With any manufactured product (model car kit in this case), whether it's a new injection-molded styrene kit, or a resin aftermarket kit/transkit, there are, and will always be, limitations--beginning with the very start of research (been there, done that, got the T-shirt!):

Take for example the entire body of a real car (from front bumper to rear bumper, from side to side, from rocker panel to the top of the roof: For starters, the surfaces of that real car body were assembled from a number of real sheet metal stampings that are separate parts. Compare that now to a model car kit body: We expect a one-piece body shell, with a separate hood, with often the front and rear valances (or bumper splash pans) necessarily having to be separate parts. But, in all of that, there are injection molding limitations. A mold for injection molding has to have some draft angle, meaning that the plastic part simply has to "pull away" from the steel mold surface without sliding, as that will mar the surface of the plastic in most cases, and in any event, will cause wear on even the hardest steel over time (Considering that polystyrene will dull your razor saw teeth, even needle files over time, it's no different with steel molding dies.) So, with all this in mind, there may well be some inaccuracies that will be necessary IF the modeler/customer expectation of a one piece body shell is to be met at a price point that won't scare most of them away. In all this, making a one piece model car kit body means taking a shape that was made up of more than a dozen INDIVIDUAL panels on the real car, and reducing that to a 6-piece, sliding cavity injection molding tool--there will necessarily be some at-least-small compromises necessary (model car builders voted, with their $$, overwhelmingly against multipiece body shells some 55 years ago!).

Model companies today, in many if not most cases, no longer have the fairly large design/development staffs that they once did (it was a lot easier for AMT Corporation to have such a large staff back 50 years ago when they were producing upwards of 15,000,000 model car kits every year!). In today's world, it's just not financially feasible to have a large staff of people given the much smaller production runs of today's model car kits. And, just as would have been the case back in the day of much larger staffs, money AND time will be major considerations--just as it goes with each and every one of us, model companies ultimately have to keep in mind that "If their outgo exceeds their income, their upkeep is their downfall"! It's all too easy to make a small fortune developing and manufacturing model kits--IF you start with a large one.

Tim Boyd explained, in brief detail, what it takes to get a laser scan of a real car or pickup truck. Lasers being light, light reflects at all sorts of odd angles from the curved shapes and surfaces of a nice, shiny new car, certainly a highly polished restoration--requiring "dusting" that car with non-reflective powders of one sort or another which will allow the laser beam to be reflected straight back at the scanner. Now that's OK if you're scanning that real car for your employer who happens to be a 1:1 automaker--but ask yourself just how many owners of a pristine older car would allow that? Not many, I would venture. In fact, I found, when I was doing my masters for resin casting, that the best, and easiest cars to photograph for reference weren't restorations, but cars whose body surfaces were faded paint, even some surface rust--those showed me far more than a nice shiny show-winning car.

Next is measurements: This is something that simply has to be done very carefully--I cannot believe that many owners of a really perfect example are particularly willing to have measuring tapes and carpenter's rules slapped all over a multi-thousand dollar paint job, even though generally that is how some measurements have to be taken (again, been there, done that). Measure an engine? Even that can be difficult to do--not very often are those readily available without all the wiring and plumbing, so that can be difficult as well.

It's ideal to be able to be able to photograph the underside of a car or truck, but almost always, that requires a lift--and not very often is that possible to achieve--again not many owners of cars being researched are willing to drive them to some garage to put the car up on a lift (I was able to photograph an H2 Hummer at the AM General plant, up on a lift in their engineering and testing area--and even at that, I couldn't easily measure all the details--that was for the Johnny Lightning 1/64 scale diecast--and GM was adamant that it be as correct as possible in that small scale!).

And then, considering the allusions here in this thread to aircraft models, those subjects present the same kinds of difficulties: Only a small percentage of modelers will ever have even seen the thing in question--even P-51 Mustangs aren't sitting at every airport! The same is true with such as '55-'56 Chrysler 300's--only a few hundred of each of those years were produced, and only a percentage still exist, even less within an affordable driving distance from wherever the person is that has to do the model development work. As late as the early 1980's, most of the pictures that were taken of WW-II aircraft were unknown--still stuffed in albums owned by people who had seen and taken pictures that could be invaluable--or in some aircraft company's, even government filing cabinets, with nobody available to go dig them out. That's often the case with automobiles as well.

And, given that computer technology comes into this sort of thread EVERY time one gets started, the first maxim I heard in a computer science class was "GIGO",meaning "Garbage In, Garbage Out". Computers simply regurgitate information that is compiled from information fed to them, end of statement. That is never more true than when dealing with CAD and 3D files. That's not meant to imply a lack of skills or commitment on the part of the operator either--but it sure does help if the specialist involved has that "eye" for realism, a passion for the actual automotive subject--otherwise it seems to be as much a clinical operation as anything.

And then, there's always going to be the human factor involved here: Our eyes see the real car somewhat differently than they do the model, for reasons I'm not going to reiterate. That goes just as much for the people doing the product development as it does for those of us who build the model kit. The smaller the scale of the model, the more difficult that makes things. With that in mind, it's far too easy, and rather disingenuous as well, to throw out the comment/question about "modelers VS kit assemblers", all the while forgetting that the product development human beings have to face the same issues. In 1985, Tamiya made available a rather revealing videotape of their product development and production processes, to hobby shops all over the place (narrated by an American, BTW!), showing the processes they used to create a Porsche 911. One statement in the midst of the tooling mockup phase (back then, tooling mockups were all carved by hand from blocks of wood!) that said a lot, and I can still quote it verbatim: "It is possible to make a model car kit perfectly accurate numerically, and yet it may not look right"! The video then when on to show designers and pattern makers deciding just how to adjust the contours of that Porsche body to make it "look right". The technology may have changed, advanced far beyond hand carved patterns, but the process of making even digitally created automotive shapes still presents the same issues much of the time. No matter how perfectly the model has been scaled, down to say a 1/25 scale fraction of an inch--all the difficulties and nuances still remain--make the thing look right to the vast majority of human eyes with their brains translating that vision into each individual's concept of realism.

Art

[Bob Ellis]

To me, the new kits have inaccuracies that exhibit square tendencies. Once at a meeting I heard this; "It's not a circle, it's an infinite sided polygon". What does that mean? It means a computer sees straight lines and not curved lines. It approximates curved lines. Use AutoCAD or another drawing program and you will see. Blow up the size of the drawing and you will see straight lines connected to other straight lines moving in other directions to approximate curved lines.

To me, the old AMT kit benefited from an artist hand making the masters from 1/10 models. They had a very smooth and round appearance.

Call me crazy (some will) but that is my impression.

[Michael Moore]

To me, the new kits have inaccuracies that exhibit square tendencies. Once at a meeting I heard this; "It's not a circle, it's an infinite sided polygon". What does that mean? It means a computer sees straight lines and not curved lines. It approximates curved lines. Use AutoCAD or another drawing program and you will see. Blow up the size of the drawing and you will see straight lines connected to other straight lines moving in other directions to approximate curved lines.

To me, the old AMT kit benefited from an artist hand making the masters from 1/10 models. They had a very smooth and round appearance.

Call me crazy (some will) but that is my impression.

You are crazy! (L) Seriously, you are correct to a point. When it comes to moldmaking Johan, MPC and AMT products really did have some great bodies and detailing cast. However in order to keep costs down and keep young guys getting into the hobby they made simple assembly a big factor too. Now we have laser precision running gear but as you say "approximated" curves. I think the Revell LX Mustang is a good example. It is a great attempt but off a bit. But you look at the Round 2 stuff and you find great outer shells but those out of scale engines and in some cases 1 piece chassis. That is why I am glad I can be happy with just having a car in scale form to build.

[Darin Bastedo]

Why aren't every kit that we build on our shelves perfect? Did we not have time to perfectly research out builds? did we not have time to correct the flaws? Why isn't the paint perfect? you should have stripped it and repainted it until you got it perfect. Why didn't you replace the 2 speed wiper motor with the single speed? at what point do we say "good enough" and simply call it finished.

The reason our builds are not perfect is because humans aren't perfect. The reason kits aren't perfect is because you have imperfect humans working for a corporation and that corporation has limitations both on time and money. At some point both have to say "good enough".

[charlie8575]

Really?

Then how was it even possible for companies like AMT and Johan to produce dimensionally and proportionally accurate models before laser-scanning, massive and cheap computing power, digital photography, instant transmission of vast amounts of data, CAD and CAE? They used film cameras, rulers, measuring tapes, calipers and pencils on paper. And there weren't armies of men making prototype models and tooling from them. There were a FEW very highly skilled individuals (who were reasonably well-paid, but not paid anything like what executives got).

And it was pretty obvious on a lot of old models that the guys designing the kit parts actually had some understanding of the functions of the real parts they were modeling. That's not always the case these days.

The more technology advances, the more excuses are getting made about why nobody can do their work to a standard that was established 50+ years ago.

And an American model company really should be able to hire a small team of locals to do accurate measuring of a 1:1 subject for what amounts to chicken feed.

Then transmit accurate data to the Chinese tooling designers, if necessary (though there seem more and more US injection-molding companies and rapid-prototypers online every month).

And are there REALLY no Americans left who can measure a car and make a tool without breaking the budget?

If that's REALLY the case, this country (and the West in general) has gone much farther down the toilet than I'd like to believe.

I think Bill really has hit some points here.

We've become so advanced, we're retrograded and can no longer do the job correctly because of over-dependence on computers and general laziness.

I think it's pretty sad, myself. Like people, computers and other machines are imperfect, and like people, their work must be checked.

I'm tired of excuses. Completely accurate, a'la Fujimi Enthusiast kits, is one thing, but to deliver an average-difficulty product with accurate bodies, interiors, undercarriages and drive-train exteriors is something that any basic mold engineer should be able to do without excuses.

Expand the budget. Go down to one release a year if you must. I'd rather one excellent kit than three badly-compromised ones. Find ways to expand your market to make the tooling costs work. Upper management in all the model manufacturing companies needs to DO THEIR JOBS. I don't care if its one man or one hundred, Do the work you're being paid to do, and if that means firing excessively-tight accountants to get more money for better product or re-arranging production schedules, do it.

Charlie Larkin

[martinfan5]

Budget or not, there is no excuses to get stuff wrong , if it is your job to make something accurate to a 1:1 subject, then you should do whatever it takes to within you budget to get right.

[charlie8575]

To me, the new kits have inaccuracies that exhibit square tendencies. Once at a meeting I heard this; "It's not a circle, it's an infinite sided polygon". What does that mean? It means a computer sees straight lines and not curved lines. It approximates curved lines. Use AutoCAD or another drawing program and you will see. Blow up the size of the drawing and you will see straight lines connected to other straight lines moving in other directions to approximate curved lines.

To me, the old AMT kit benefited from an artist hand making the masters from 1/10 models. They had a very smooth and round appearance.

Call me crazy (some will) but that is my impression.

I think you're right, Bob. In fact, I think you may have hit part of the problem.

In the quest for better-faster-cheaper, we've ended up with some things better-some not/slower because we insist on sending stuff hither and yon/it's cheaper until you fix the problems.

I think it might be worth exploring a hybrid approach with the gross and fine measurements done by a combination of computer and analog methods, and the mock-ups produced in such a manner, as well. I'd give serious consideration to hiring someone with demonstrable fine-art background to work with the engineers to make sure the contours and shapes are right. The computerized model can literally shave weeks off the mock-up manufacture, and even if it takes say, a week to make adjustments with putty, plastic, and whatever else, you're still saving possibly a month or two in product development. And, by having someone palm the mock-ups, mistakes can be found much faster and prevent problems like Mustang roofs and mishapen fenders.

If you do this right, it will not only work, but save egg on your face and yield a better product faster, possibly for less development money net.

Charlie Larkin

[Brudda]

I believe everyone here is right. Kits are expensive and the overseas kits are more. I really agree with Art Anderson ,Sometimes a part does not look right at reduced scale. Im building a 1:12 scale GT 500 convertible.I have modified or scratch built more than I wanted. The reason the build is taking so long is that i am making several versions of a part,1 is a scaled down version and part 2 and 3 are modified to look right at scale. Some parts look right on at scale, while others look really strange at scale.Try 14 inch tires on a 65 gto. Believe me it looks strange at scale.

I understand the mfg part of business. Build real planes and you really start to understand what it takes to do it right. On a smaller scale however, It is quite different. Its not life or death and not really rocket science but it is a art form.Try to scale something down from a 1 to 1. It takes time. The more time the more accurate thus more money.

I knew people in the business part of modeling years ago and (not saying names) That the american companies had 1 person doing the measurements on a subject for 1 or 2 days. The japanese companies had several people measuring a subject for weeks. But there was a big difference in price.

The american model companies are doing great for the price.Build a model company,do what they do,and then get a few crybabies saying this and that is wrong with your model. Do it for a locked price and deadline.Not an easy thing.

OK now you have the perfect kit, no complaints, but can you do a perfect build? Can you be Mr. Perfect?There are a small few that can do this,on this sight. And you know who you are,but the rest of us are just builders.So my question is ,if the model is not perfect and you are not perfect, whats to worry about?Just build and have fun.