AIR Paint Schemes
Somanykits
Getting ready to put some paint on some WWII fighters and bombers ......
1/72 scale ...... how much scale effect for each? Thanks.....
Dave
AHorv43767
>1/72 scale ...... how much scale effect for each? Thanks.....
IMO, none. Scale effect is a crock.
August
Barry Johnston
Opinion noted (I occasionally lighten paints because they don't look
right. Is it scale effect? who knows).
Bejay
Birgit & Burkhard Domke
>Hey Gang;
>
> Getting ready to put some paint on some WWII fighters and bombers ......
>1/72 scale ...... how much scale effect for each?
Hmm, wait, how much scale effect...I'd say...42 would well be in
place...can't tell how much that's for each...you forgot to quote how
many models are involved...
Rama Lama Hoop da la Wee
Hauptstadt Tempel
Berlin, Germany
Lasse Hillerøe Petersen
b.d...@coruscant.b.shuttle.de (Birgit & Burkhard Domke) wrote:
>On 14 Jan 1999 05:58:29 GMT, soman...@aol.com (Somanykits) wrote:
>
>>Hey Gang;
>>
>> Getting ready to put some paint on some WWII fighters and bombers ......
>>1/72 scale ...... how much scale effect for each?
>
>Hmm, wait, how much scale effect...I'd say...42 would well be in
>place...can't tell how much that's for each...you forgot to quote how
>many models are involved...
Well said.
I'd like to add:
NO NO NO!
_Scale effect_, understood as the addition of an amount of white to
colors, based only on the scale of the model to be painted, is WRONG!
Advocates of this effect claim that the atmospheric interference over the
distance to the object modelled warrants such a process. However, it is
easily seen, that if you:
1) assume that the atmospheric effect over short distances (0 to 5 meter)
negligible
2) consider four models of the same original object (say, an Me109), in
scales 1/144, 1/72, 1/36 (scale selected for easy calculation, for a
practical test, a 1/32 scale model would be adequate) and 1/24
3) place these four models accordingly:
Scale Distance to viewer (in meter)
1/144 0.5
1/72 1.0
1/36 2.0
1/24 3.0
Then all these models depict the original Me109 at a 72 m distance, which
means they should all look exactly the same. For this short distance,
model color differences should be negligible, so they must therefore all
be painted the same colors. Q.E.D.
This argumentation assumes that a color looks the same within the 0 to 5
meter range. This assumtion is quite easy to check for validity. One day
when my time permits, I will do a test, using the percentages of white
recommended for these scales by Testor.
Even if it fails, however, the only thing you can conclude is that color
must be adapted to the viewing distance to the model, as well as the
viewing distance to the original. So _scale effect_ understood as a
function of scale only, still doesn't apply.
Compensating for atmospheric conditions, regardless of scale (versus a
pure mechanical addition of a percentage of white according to chosen
scale), _may_ be a good thing to do, but IMO tends towards impressionism.
Taking it to the absurd, you could argue for painting a silver aircraft
black to depict a night scene, or use grays with white highlights (even
painted reflections for mirror-polished surfaces) instead of metallics.
Mind you, I have nothing against impressionism, and every modeller may
chose whatever method he likes. I just want people to stop advocating
_scale effect_ as a scientific method, when it is absolutely unscientific.
I also want people who consider using it, to understand what this is
about.
I find it highly unpleasant that the _scale effect_ method is being
promoted by a company like Testors, because of it's seemingly simple and
very scientific appearance.
My personal choice regarding compensation for atmospheric conditions is as
follows:
In a 2D picture, there is no 3D to cast shadows or illustrate depth,
therefore the artist must compensate for this by _painting_ the shadows.
However for a 3D picture (aka a model) the model has depth in itself, and
this depth is sufficient to cast shadows by itself, thus it is unnecessary
(even wrong) to enhance shadows by painting them.
Further, a 2D picture cannot assume anything about the viewing conditions,
except hope for a brightly lit gallery; therefore darkness also must be
painted in. For 3D pictures, we may choose to add a diorama or a shadow
box that can provide the proper light conditions for the viewer, so again,
these effects can be moved from the painting of the model to the
conditions for viewing the model. Again, the result is that conditions
should be omitted from the model; the model can be made a "purer"
rendition of the original, because the 3-dimensionality allows effects to
be implemented constructively which a 2D artist has no option but doing
pictorially.
Note also, that although possible, I believe few 2D artists have chosen to
represent 2D projections of metal surfaces with metallic paints; and I
doubt it would have the right effect if one were to try. This is to me a
strong indicator that a modeller should forget about most of the
techniques used by 2D artists, because those techniques are mainly used to
compensate for the lack of the third dimension, which we modellers
actually have.
-Lasse
Danger Atom
OK, with all that said, what color would you paint, say, a Cobra or Black
Hawk in Kuwait City a couple of days after the cease fire? Taking
atmospheric conditions into effect. Would Sooty Black overall be about
right, even though they were painted in some sand colors? Or, how about an
F-15E during a sand storm? A nice shade of tan would raise a few eyebrows.
--
D.A.
Rima Luma Chug-a-Lug
Tumple of the Bottumless Beer Bottel
Keepur of tha teMpul keg tap.
Lasse Hillerře Petersen wrote in message ...
Jeff Rankin-Lowe
Danger Atom wrote:
I think another factor has to be considered when adopting scale colour.
Although the model is at an apparent distance representing the real thing at a
normal viewing distance, the brain knows the eyes are focussing on a point
quite close to the person owning those eyes and the brain.
An audio counterpart can be seen on TV and in movies. Characters speaking
inside sound different as they walk out the door. We instictively know that
voices sound different inside and outside.
I think that some allowance must be made for scale colour, but nowhere near as
much as is often suggested.
Jeff Rankin-Lowe
** BOB SIGMAN ***
and linear extrapolation fail to factor in the complexity of the human
eye. Unmentioned are the fact that the rods and cones have different
spectral responses (remember the Purkinje effect).
Read for example "Color depends on intensity" Chapter 35-2 in
"The Feynman Lectures on Physics".
Bob Sigman
--
*** BOB SIGMAN ***
Georgia Institute of Technology, Atlanta Georgia, 30332
uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!ae241bs
Internet: ae2...@prism.gatech.edu
Anders Svennevik
>
> Lasse's dismissal of scale effect based on simplistic assumptions
> and linear extrapolation fail to factor in the complexity of the human
> eye. Unmentioned are the fact that the rods and cones have different
> spectral responses (remember the Purkinje effect).
I think you misread Lasse's post. He argued (maybe not well nor
convincingly) that scale effect is based on much more than the scale of
the model, and rules like 'add 24% white for 48th scale and 36% white
for 72nd scale' are far too simplistic. On this point I fully agree with
Lasse. The best models I have seen in my many years as an IPMS judge,
are those where the modeller has intuitively compensated for size and
colour.
Anders
Lasse Hillerøe Petersen
SIGMAN ***) wrote:
> Lasse's dismissal of scale effect based on simplistic assumptions
>and linear extrapolation fail to factor in the complexity of the human
>eye.
Definitely right, and on purpose. In my original argumentation, posted
quite a while ago, I tried hard to come up with some definitions that
*avoided* making any assumptions or saying anything that would require any
understanding of the eye or any physics involved. I will happily repost,
if you can't find it on dejaNews. Indeed, I arrived at my assumptions by
simplifying these definitions, which assumed _nothing_ about how color
changes over distance (or in fact anything about color at all), but only
formalized that this may happen.
I hope you will agree with me, that my assumptions are a lot less
simplistic than saying "add 15% white to a color for 1/72 scale".
Nothing would please me more than if someone would post arguments for or
against scale effect, based on expert knowledge of the perception of
color.
> Unmentioned are the fact that the rods and cones have different
>spectral responses (remember the Purkinje effect).
> Read for example "Color depends on intensity" Chapter 35-2 in
>"The Feynman Lectures on Physics".
>Bob Sigman
I doubt that text deals specifically with the problems of scale color.
Since you mentioned it, it might not be too much to ask, if you could post
an explanation of how it translates to understanding of scale colors? No,
I don't remember the Purkinje effect, as I have never heard of it in the
first place. I would be grateful if you could explain it, and also, how
"the fact that the rods and cones have different spectral responses"
relates to the perception of color on scale models. I'm afraid my
"simplistic assumptions" are the best I can come up with.
Oh, heck, I'll rehash my old article:
Let c be a color used on a full-size original.
C(x,c) is what we could call the perception of the color at distance x.
Let cs be the color used to paint a model.
CS(x,cs) is again the perception of the color cs at distance x, a
different function, to allow for possible differences in color perception
due to environmental reasons or whatever.
Now, it is clear that if the model is in 1/72 scale, then ideally, we want
a cs so that for any x:
C(x,c) = CS(x/72,cs)
or in general for the scale m, expressed as a fraction:
C(x,c) = CS(x*m,cs)
There is no need to know a thing about color perception to understand the
above argumentation.
If it is in any way possible, it would be desirable to devise a function
M, so that the above holds for any x:
cs = M(c,m)
My simplification results from assuming that CS(x*m,cs) = cs for any
reasonable x*m ("usual viewing distances"), leading to:
cs = M(c,m) = CS(x*m,cs) = C(x,c)
which does not depend on m. It is thus clear that scale does not affect
the color, only the distance to the original. Therefore it is wrong to
talk about "scale effect" on basis of scale.
Now, there is no reason that C(x,c) = c (or, for that matter, any reason
that CS(x,cs) = cs). But so far, noone has brought forth a good candidate
for C, and it is clear that C(x,c) = c is a candidate that fits. Certainly
the existing "rule" for adding white according to scale is _not_ a
candidate for C.
C(x,c) = c is *my* personal favorite until a better comes along, because
it is the only one that *I* can think of, without having to define what
"color" is, or what is actually hidden inside C.
If you have the knowledge required to provide some measuring method for c
and cs, and a better C or M, then I would be delighted if you would share
it.
-Lasse
Jeffery S. Harrison
<major snip>
>
>which does not depend on m. It is thus clear that scale does not affect
>the color, only the distance to the original. Therefore it is wrong to
>talk about "scale effect" on basis of scale.
>
I'll admit that you lost me a little in your post--what is "M" supposed to
represent? It just suddenly appeared in your formula without definition.
Anyway, scale does effect the color but not the way you're assuming (at
least the way it sounds like you're assuming to me). In some of the sources
(such as the Monogram Luftwaffe painting guide) where they say to add 24%
white for a 1/48th scale model or 36% for a 1/72nd scale model they don't go
into a lot of detail of how they came up with these numbers, in the article
written by Ian Huntley several yeas ago he was a little more clear. The
ratios he came up with were different (less white for any given scale and
non-linear--350% white for a 1/700 scale model just didn't sound right when
I first read about scale effect in the Monogram book) but he also told you
the assumptions he made to develop his chart. He developed his chart by
standing out on an airfield at varying distances from planes and mixing
colors until the matched what he saw. The assumption he made was that
typically a model was viewed from a distance of approximately 1 foot so the
ration of white to color for any given scale was the ratio used for that
scale distance (i.e.. 48-feet for a 1/48th scale model and 72-feet for a
1/72nd scale model), so you're right scale is not really the
consideration--distance is. What he didn't really explain very well was that
if you typically view (or display them so that they can only be viewed from)
your models from a distance of 2 feet then for a 1/72nd scale model you
should be using the rations he came up with for a 1/144th scale model and
for 1/48th scale you should be looking at the 1/96th scale line since the
scale distance you'll be looking from is twice as far as the one he assumes
for his chart.
Jeff
IPMS something or other
C A Irving
Lasse Hillerøe Petersen wrote:
> In article <794ahu$6...@acmey.gatech.edu>, ae2...@prism.gatech.edu (** BOB
> SIGMAN ***) wrote:
>
> > Lasse's dismissal of scale effect based on simplistic assumptions
> >and linear extrapolation fail to factor in the complexity of the human
> >eye.
>
>
> if you can't find it on dejaNews. Indeed, I arrived at my assumptions by
> simplifying these definitions, which assumed _nothing_ about how color
> changes over distance (or in fact anything about color at all), but only
> formalized that this may happen.
>
> I hope you will agree with me, that my assumptions are a lot less
> simplistic than saying "add 15% white to a color for 1/72 scale".
>
> Nothing would please me more than if someone would post arguments for or
> against scale effect, based on expert knowledge of the perception of
> color.
>
> >"The Feynman Lectures on Physics".
> >Bob Sigman
>
> I doubt that text deals specifically with the problems of scale color.
> Since you mentioned it, it might not be too much to ask, if you could post
> an explanation of how it translates to understanding of scale colors? No,
> I don't remember the Purkinje effect, as I have never heard of it in the
> first place. I would be grateful if you could explain it, and also, how
> "the fact that the rods and cones have different spectral responses"
> relates to the perception of color on scale models. I'm afraid my
> "simplistic assumptions" are the best I can come up with.
>
>
>
> Let c be a color used on a full-size original.
> C(x,c) is what we could call the perception of the color at distance x.
>
> Let cs be the color used to paint a model.
> CS(x,cs) is again the perception of the color cs at distance x, a
> different function, to allow for possible differences in color perception
> due to environmental reasons or whatever.
>
> Now, it is clear that if the model is in 1/72 scale, then ideally, we want
> a cs so that for any x:
> C(x,c) = CS(x/72,cs)
>
> or in general for the scale m, expressed as a fraction:
> C(x,c) = CS(x*m,cs)
>
> There is no need to know a thing about color perception to understand the
> above argumentation.
>
> If it is in any way possible, it would be desirable to devise a function
> M, so that the above holds for any x:
>
> cs = M(c,m)
>
> My simplification results from assuming that CS(x*m,cs) = cs for any
> reasonable x*m ("usual viewing distances"), leading to:
>
> cs = M(c,m) = CS(x*m,cs) = C(x,c)
>
>
> Now, there is no reason that C(x,c) = c (or, for that matter, any reason
> that CS(x,cs) = cs). But so far, noone has brought forth a good candidate
> for C, and it is clear that C(x,c) = c is a candidate that fits. Certainly
> the existing "rule" for adding white according to scale is _not_ a
> candidate for C.
>
> C(x,c) = c is *my* personal favorite until a better comes along, because
> it is the only one that *I* can think of, without having to define what
> "color" is, or what is actually hidden inside C.
>
> If you have the knowledge required to provide some measuring method for c
> and cs, and a better C or M, then I would be delighted if you would share
> it.
>
Now I know why I just open the bottle and pour, any bottle!
Andy
Mike
> Read for example "Color depends on intensity" Chapter 35-2 in
> "The Feynman Lectures on Physics".
No thanks, my AMS is bad enough!:)
--
Mike Dougherty
Toronto, Ont.
Canada
IPMS C4928
"Uh oh....."
- famous last words
Lasse Hillerøe Petersen
<jsh...@ptialaska.net> wrote:
><major snip>
>>
>>which does not depend on m. It is thus clear that scale does not affect
>>the color, only the distance to the original. Therefore it is wrong to
>>talk about "scale effect" on basis of scale.
>>
><more snippage>
>
>I'll admit that you lost me a little in your post--what is "M" supposed to
>represent? It just suddenly appeared in your formula without definition.
Sorry. cs = M(m,c), where M is what I would call the model color function;
a function, if it exists, that for any full-size color, ideally yields the
appropriate scale color for any distance. Note that I later make
assumptions that result in M being dependent only on the distance to the
original. I then say that I don't know C, so until someone tells me what C
really is, I'll use C(x,c) = c, which fits.
>He developed his chart by
>standing out on an airfield at varying distances from planes and mixing
>colors until the matched what he saw. The assumption he made was that
>typically a model was viewed from a distance of approximately 1 foot so the
>ration of white to color for any given scale was the ratio used for that
>scale distance (i.e.. 48-feet for a 1/48th scale model and 72-feet for a
>1/72nd scale model), so you're right scale is not really the
>consideration--distance is.
Presuming he did each color for one scale on one day. The next day, a
different scale. Now what would happen, if suddenly the air pressure and
humidity had changed from one day to the next? Or if he went away for a
few months, and came back at a time when the sun was in a different angle?
Or if he worked in the morning, went for a lunch break and came back in
the afternoon? Even so, what does this help, when the color you want to
match is the color of Memphis Belle at the airfield in England, the day
they filmed the famous documentary? Sorry, but my personal belief is that
there are too many variables to take into account to make a general, or
even a specific, scientific solution. I stick to what looks right, and the
original color is what looks best to me: My common sense tells me that
C(x,c) = c is the best approximation I can make when I don't know all the
other variables upon which C depends.
>What he didn't really explain very well was that
>if you typically view (or display them so that they can only be viewed from)
>your models from a distance of 2 feet then for a 1/72nd scale model you
>should be using the rations he came up with for a 1/144th scale model and
>for 1/48th scale you should be looking at the 1/96th scale line since the
>scale distance you'll be looking from is twice as far as the one he assumes
>for his chart.
Exactly. Except my guess is that the color shift is _not_ linear, and
depends a _lot_ on atmospheric conditions for viewing the original, so
there can be _no_ general rule about distance = percentage of white. And I
strongly believe it would be wrong to add twice the amount of white to the
color of a 1/48 scale model, just because you intend to view it from twice
the distance. This is because of the non-linearity that I would expect
with very short distances, and the reason I make the assumption that
CS(x,cs) = cs.
I do hope Bob Sigman can provide more info on this from a better informed
viewpoint. I can only use logic and unqualified guesses based on possibly
faulty common sense.
-Lasse