A "Rivet Counter" speaks up!

[gtx6970]

  On 9/26/2012 at 1:50 AM, Harry P. said:

Ditto, ditto, amen and hallelujah!

I agree completely

Hello, I am a rivet counter and I need help

[Art Anderson]

Mr Obsessive wrote:

With CAD, 3D scaling, Internet, a plethora of car shows, clubs, magazines, etc-------there’s no reason why a major model company can’t get the details right.

I know I've said this before, but here I go again:

All the modern technology in the world is great, when it comes to creating plastic kits of scale models, but in all the talk about CAD, 3D scaling, and what other high-tech stuff I fail to include, in the end, the bottom line, there still needs to be the human factor in creating a scale model of anything, particularly cars. However, one simply must understand that CAD was not around when say, the 1955 Chevrolet was being styled and engineered.

Why? Very simple. While we view both the real thing, even photographs, and of course our model cars with the same pair of eyes, those eyes see each version (real car, model of it, pictures of it) differently! Let's take a look at the issues here.

Given whatever spacing each of us have between the center of the pupils of our eyes, we see the real car much differently than the model. For example, if one looks at an accurately done 1/64 scale model of any car, one will see MORE of that car than is ever possible to see of the real thing. It's called "visual perspective". Our eyes can see, if looking dead straight on, down BOTH sides of that 1/64 scale model, because it fits (when viewing straight on from front or rear) between the centers of both our eyes. But, viewing the real car in the same manner, we can barely see past the headlight bezels on say, a '53 Chevy, so right there is one of the issues. Additionally, with even a 1/25 scale model car, we can see that model at angles, and at heights which can only be achieved for example, by borrowing the fire department's aerial ladder or snorkel rig, or by having the real car on a body rotisserie so it can be spun in front of us at any angle we choose. In short, we DO get to see more of the model than of the real thing.

This can really play tricks on us "rivet counters"! In 1984, just as I was opening my own hobby shop (The Modelmaker here in Lafayette IN), Tamiya released a very informative video titled: "The Design And Development Of A Plastic Model Kit". In it, the viewer sees the development process (Tamiya actually had a Porsche 911 in the lobby of their headquarters in Shizoka City Japan). Kit designers with notebooks, measuring tapes and rules and cameras were all over that thing. Next came the drawings (before CAD had reached down to the plastic model industry). Next you could see the making of tooling mockups (back then, a 1/24 scale car was mastered in 1/12 scale, in wood). Finally, with the body shell "done", the patternmaker along with a team of designers, pored over that beautiful 1/12 scale master for the Porsche 911, studying the subtle curvature and contours of every body panel, marking all over the thing with Sharpies or something similar. The narrator described the entire process in English, climaxing with the strong statement (and I remember it vividly!) "It is possible to make a model car numerically accurate, and yet it may not look right! And sure enough, in the next scenes, one could see the patternmaker slightly adjusting some of the shapes, to give the impression of accuracy, as seen by human eyes accustomed to viewing the real thing.

With photographs, it's also often difficult to judge accuracy of shapes and contours, as photographs are by their nature, just 2-dimensional. The camera is no different than a human looking at the subject, with one eye completely covered--NO DEPTH PERCEPTION. In addition, photos in magazines, even photos taken by us ordinary mortals at say, car shows, tend to be what are called "beauty shots". Seldom do we get the chance to really capture every little detail, every subtle curve, and certainly not many of us carry measuring sticks or tapes to car shows to photograph with dimensional data, any car. When developing a model kit, any savvy product development specialist will take hundreds upon hundreds of pictures of the subject. In addition to a full walk-around series, he or she will shoot a myriad of pictures taken at angles which show curvatures, a particular line, the shape say of a roof from enough angles to show the curves and contours. Additionally, a product development specialist often will lay an old fashioned carpenter's folding rule up next to a detail, could be a grille (or section thereof), a taillight lens, the door for length, all of that (the carpenter's rule having been marked before hand--every other inch blacked out so the inches show up in the pics clearly!).

Now, modern technology has given us the "laser scan", which can "see" depth, and even the curvature of items giving that depth, and the price of laser scanners have come down, WAY down in price, making them attractive for those charged with the development of a new kit.

But in the end, making that last bit of realism, it still takes a trained eye, to ensure that once done, the mockup and the final production kit, turn out not only right, but that they look real to our own eyes.

Art

[Art Anderson]

  On 9/26/2012 at 9:37 AM, peekay said:

"It's just a hobby"?

As Bill said, his comments are directed more at the manufacturers and for them it's NOT a hobby. It's their bread and butter and their reputations are at stake.

I sometimes wonder why Japanese kits tend to be superior. I have family ties to Japan and I've come to the conclusion that the Japanese simply put more care into their work. You can see and feel that when you open a Japanese kit.

Unfortunately I'm a fan of American and European cars.

Not always has this been the case! The annals (and the archives of kits from every country where plastic model car kits have been designed, tooled and produced) will have its share of real klinkers! There is one big difference though, from what I've learned over the years: Japanese modelers seem to prefer accuracy and perfection way over low price.

Art

Edited September 26, 2012 by Art Anderson

[Greg Myers]

I remember reading about this "human perspective" in one of Gerald Wingrove's books many years ago.I have to believe getting the angle of a piece of side trim so radically off has to be the antitheses of this.

[moparmagiclives]

  On 9/26/2012 at 2:49 PM, Art Anderson said:

Not always has this been the case! The annals (and the archives of kits from every country where plastic model car kits have been designed, tooled and produced) will have its share of real klinkers! There is one big difference though, from what I've learned over the years: Japanese modelers seem to prefer accuracy and perfection way over low price.

Art

Is it that or do the Japanese company's cater more to adults? And the American company's are still making toys for kids?

I see this as a lamplighters vs traxxas type of battle. Toy vs hobby grade.

[MrObsessive]

  On 9/26/2012 at 3:18 PM, Greg Myers said:

I remember reading about this "human perspective" in one of Gerald Wingrove's books many years ago.I have to believe getting the angle of a piece of side trim so radically off has to be the antitheses of this.

Aaahhh-----the much maligned (and rightfully so) '58 Belvedere. This is one of the models I had in mind when I stated in my original post that certain cars can evoke certain memories. My Dad had this car when I was a kid (a '57 actually), so I'm VERY familiar with how this model should've appeared. It's not that the trim is misaligned per se.........it's that it follows the body lines which are incorrect. The problem starts right behind the doors (where the tailfins actually start), in which the beltline should start a gentle slope upwards (take a look at the 1:1's), and then bump upwards for the tailfins.

The trim should be in a straight line from nose to tail with the exception where it flares up towards the fins.

RC2 decided that the beltline should taper downwards (inexplicable to be sure) and then bump up to the fins. The result is a body that appears to be banana shaped and not like the1:1 at all. Another chance at greatness blown. One look at the box art when this model first turned up and I knew it was wrong.

Not to mention a roof that's too low, C pillars that are too wide, (take a look at Johan's '59 Dodges which were 100% correct) and you have a real mess of a model. It can be worked with for sure, but not without a LOT of work! Not something everyone can or wants to do------or if RC2 had really done their homework, should have to do.

[Agent G]

Long ago a reknown Japanese company examined, measured, photographed and thoroughly studied a 1:1 example of a certain US armored fighting vehicle. Said kit was highly anticipated, and considered state of the art when released 30+ years ago. That was until someone who had quite a bit of experience with the real tank noticed it sat way too high. Seems our esteemed collegues from Japan examined a tank from which the engine was removed. No one noticed that at the time, therefore no one caught the fact the suspension was "unloaded". The kit in question did not have an interior nor did it have an engine so these weren't part of the initia lexam.

My point, things are only as good as the engineers who make them.

G

[sjordan2]

These thoughts are part of what makes this site so valuable. You learn what's right and what's wrong, and there are many talented builders to show you how to fix it if you're inclined to do so. Along the way, you learn a lot about certain cars and kits and building techniques, from people who are true enthusiasts and are very knowledgeable.

Long live the rivet counters, and may the kit companies heed them.

[Monty]

  On 9/26/2012 at 6:45 PM, sjordan2 said:

Long live the rivet counters, and may the kit companies heed them.

Well said! The best example of this is the amount of input Moebius sought in order to get the Hudson correct vs all those expensive automotive subjects Trumpeter released that ended up needing lots of plastic surgery.

[W-409]

I just read through the first page. Haven't read the second page yet, but few thoughts came into my mind.

Someone pointed out the Monogram '69 Camaro. I agree, the body looks incorrect. It's not the best kit when looking at the interior, chassis and engine, but the body looks really odd. I'm building one currently, but it's coming slowly-just because of that.

To me, scale accuracy is, well of course, important. I'm trying to build as good models as I can, and I think that many times, little things make the model "Pop Out". Sometimes I like building fast builds, between more serious projects. I don't like to do corrections to the body, if it's even close, but especially for engines, interiors and chassis parts, I do those corrections if I see that there's a need for them. I don't like major body work, and if the body looks good already even if it's not perfect, is many times "good enough". But it depends.

If you count me into Rivet Counter's group, count me in, and then I'm a rivet counter & proud of it.

[moparmagiclives]

  On 9/26/2012 at 6:21 PM, Agent G said:

Long ago a reknown Japanese company examined, measured, photographed and thoroughly studied a 1:1 example of a certain US armored fighting vehicle. Said kit was highly anticipated, and considered state of the art when released 30+ years ago. That was until someone who had quite a bit of experience with the real tank noticed it sat way too high. Seems our esteemed collegues from Japan examined a tank from which the engine was removed. No one noticed that at the time, therefore no one caught the fact the suspension was "unloaded". The kit in question did not have an interior nor did it have an engine so these weren't part of the initia lexam.

My point, things are only as good as the engineers who make them.

G

That is AWESOME !!!

[Harry P.]

Just in case anyone is wondering what the issue with the '58 Belvedere is... (I added the yellow lines):

[sjordan2]

Belvedere? Based on these shots, there are a lot more issues, such as trunk shape, fins, bumpers, etc. - but I wish all kits had such small problems. Hard to tell about the roofline because of the different lenses and "eye" perspectives that Art was talking about.

[Art Anderson]

  On 9/26/2012 at 3:02 PM, plowboy said:

I'm not a rivet counter myself, but I DO want the body to look accurate. It's kinda odd that a lot of kits/promos made almost fifty years ago are much more accurate representations appearance/body wise of the real car than kits made just a few years ago. The newer kits may have a better detailed chassis and interior (which they should have) but, they always seem to fall flat when it comes to the body proportions. To me, it's easier to update an older body with newer chassis than to work over an inaccurate body or sometimes I will just build them as a curbside.

Even a good number of those "iconic" kits/promo's of the 1950's and 60's have their accuracy problems. SMP/AMT 1959-62 Corvettes with grilles and front pans more representative of a chipmunk with its mouth full than the appearance of the actual car; 1959-60 Chevies that had a lot of accuracy problems; JoHan '58-60 Cadillacs that were deliberately way out of proportion lengthwise and an arbitrary actual 1/8" too narrow (3 scale inches)--due to a requirement to fit a standard size promo box from GM themselves. Those are model cars that come to my mind.

A lot of those so-called inaccuracies (and every one of the model companies pretty much went through the same thing) happened, more than likely as patternmakers had a learning curve, particularly in those days of hand-carved wooden tooling mockups. Some inaccuracies happened due to the simplified nature of most model kits back then--the long-popular AMT Trophy Series '32 Fords are full of them, even their '36 Ford grille is not correct (the actual grille curls inward at the bottom, the AMT grille doesn't). I took the opportunity at this last weekend's Goodguys at Indianapolis Motor Speedway to do some walkaround pics of a '40 Ford sedan delivery--a lots rounder, fatter body than the AMT kit represents (but then, AMT made that body shell to drop on their existing '40 Tudor Sedan fender unit).

Several 1960's kits also suffered, to some degree from asymetry--in a few cases it was almost like "The left side knew not what the right side was doing".

Now, modern-made, 21st Century state of the art model car kits and their body shells should be a lot more accurately done, no question. But again, it's a matter of the "learning curve" where shapes and contours are concerned--a computer knows only what information is fed into it--GIGO (Garbage In, Garbage Out was how I heard it expressed in 1969, Computer Science in college back then).

Then, there was the matter of buyer interference, the "buyer" being the mass retailer, even hobby wholesale distributors. When model companies had to deal with enforced retail price points for their product, AND very tight (and often shifting!) deadlines for delivering production kits of new subjects, then development and tooling costs, along with extreme pressure to get the new kit into the stores really got in the way. Now that the model companies really aren't "tied" to the often unyielding demands of say, Walmart, they can take the time to get it far more right than say, 15-20 years ago.

One more thing to consider here: Our memories of model car kits of years, even decades ago, while they might not have been perfectly done by today's expectations, were exciting enough to our collective younger eyes and hands that we viewed them as perfect when looking back nostalgically. That's why I frequently maintain that the only way to judge the accuracy of a model car kit is to go all the way back to the real car, compare to that, but given that few pictures of real cars in magazines and books really give all the information with which to objectively judge accuracy of a model.

Art

[eizzle]

If we could build the kits without really looking at the 1:1 subjects, we would probably love a lot more kits, we study a subject top to bottom and when we get to the kit we can see the errors easily. I'm just thinking back to when I was a kid and building the "Heavy Chevy" 70 Chevelle. It probably has all kind of errors, but being a kid, I had never seen a real chevelle, I wouldn't have known if something was right or wrong.

[moparmagiclives]

It's interesting to hear and see the things you guys are talking about (I'm not sure about the impala though) I guess I'm out of the "counter club" if I were to build either one of those, chances are, it would be some kind of custom and some of the small things would be changed, covered, or drawn into the model by other features.

[Chuck Kourouklis]

  On 9/26/2012 at 2:47 PM, Art Anderson said:

...In 1984, just as I was opening my own hobby shop (The Modelmaker here in Lafayette IN), Tamiya released a very informative video titled: "The Design And Development Of A Plastic Model Kit"... The narrator described the entire process in English, climaxing with the strong statement (and I remember it vividly!) "It is possible to make a model car numerically accurate, and yet it may not look right!

And yet, what's the pattern we see ever more frequently with current releases? That a manufacturer's own offerings can vary from one another depending on the availability of the 1:1's CAD data. Revell's kits of contemporary subjects - presumably with factory-supplied files - are generally more accurate than their kits of vintage subjects lacking that data. The edge the Moebius Lonestar has over its Hudson and Chrysler kits is pretty stark, in pictures as well as in the 3D presence. And then we have the new Polar Lights '66 Batmobile which is documented as being developed from Mattel's 3D scan data. There are certainly details to nitpick in the kit, but the overall fidelity in proportion and contour is some of the best currently available, far as I can make out.

Where Tamiya shows mastery is in figuring out how to tease their masters in a way that's flattering to the subject, and I've long maintained if the deviations don't offend the aesthetics of the 1:1, they're more likely not to be noticed. One of the clearer examples is their Ferrari F40. Look it over carefully, and you can see how the front end has been gently pulled out and the cowl has been tugged up a bit for a more voluptuous overall shape. The overall effect is more graceful and flowing than the actual car.

But this numerical accuracy they describe is basically a comprehensive breakdown of a three-dimensional shape in many discreet one-dimensional measures - so all the linear dimensions may have scaled spot-on, but there may yet have been contours not perfectly reduced in scale, and perhaps THAT is why those masters didn't look right. A dice cube may be exactly 1/20 the scale of a bowling ball in length, width, and height - but there's clearly a vast difference in surface expression between the two. This is why I think the old Monogram 1/24 '69 Camaro, supposedly exact in every scale dimension, went so horribly awry.

The necessity of a human touch goes without saying, far as I'm concerned. The sharpest 3D printers to date produce masters needing some refinement; and even when we get to a stage where originals are reproduced to near-duplicate quality, you'll still have a master that won't get anywhere without a human imagination to plan its breakdown as a kit. But as you have pointed out, 3D scanning brings an unprecedented advantage: it will not only scale every linear dimension, but every curve and contour. What I'm really talking about is a hyper-sophisticated pantograph here, and that's how I advocate its use.

I'm not so sure how necessary interpretation of a shape will be once collecting reliably precise and accurate three-dimensional data from a scan becomes standard practice, but then, that could be because my frame of reference is different. My acid test is how easily a model can be confused, at least in proportions, for the 1:1 in a photograph. And I find that models which pass that test well reliably look more accurate in my hands.

[peekay]

Good words Art and Chuck. Looks like there will always be "the necessity of a human touch". I guess it depends on the human who's doing the touching.

Also, I personally would rather pay a premium for a model that looks right and assembles well straight out of the box. Those are my two basic expectations of a kit. I have a relatively tight hobby budget but I can be inticed to open my wallet wider than comfortable if the manufacturer routinely gets those basics right.

Edited September 30, 2012 by peekay

[Greg Myers]

[moparmagiclives]

  On 9/27/2012 at 3:15 PM, Greg Myers said:

??

[southpier]

neither are prototypically correct

Edited September 27, 2012 by southpier

[moparmagiclives]

umm, correct to what? Tom Daniels designs ?

[Guest]

While I do not expect 100% perfection, I feel things could be done better than some have been. I agree with you Bill.

[Deano]

  On 9/27/2012 at 8:05 PM, southpier said:

neither are prototypically correct

Um ... they weren't supposed to be. Prototypical Pintos and Vegas don't really fire the imagination (although, I must confess that a certain 1974 Pinto wagon stirs up fond memories).

[jaydar]

I consider myself a rivet counter. I add detail and correct manufacturer's mistakes where my modeling skills allow. The rest is what it is. I don't criticize anyone else's style or belief in this case.

I do hold the manufacturer's to this ideal however. someone else said it best, it does not cost much if any more to do it right.

I will gladly pay more if their is greater detail and it is correct.

joe.