Date   

Re: Color accuracy (Was: scanners to measure colour)

smokeandsteam@...
 

In a message dated 5/17/2003 7:22:48 PM Pacific Daylight Time,
thompson@signaturepress.com writes:

You CAN
achieve a match from any medium to any other with sufficient work
But only under certain circumstances. Different media and pigments may give a
match under one light source, but completely different results under another
-- even between bright sun and overcast. This is called metamerism and makes
a complete mess of any attempt to come up with a precise match -- of you want
it to match under all circumstances you will need to duplicate the original
mix and dragon's blood is very nearly as hard to find as lead based paints.

I honestly think that most modellers would find it instructive to spend a
term or semester studying painting - artists have been dealing with the
problems of colour far longer than there have been railroads

Aidrian
Willows CA


Re: other colour measures

Stafford Swain <sswain@...>
 

I acquired one of these chip swatches too in the CNR color study. It was cheaper than the PANTONE guide but there are far fewer colors and getting matches for anything is you truly want is well, unlikely.

Thanks for the comments on this subject thus far.

Has any one on the list tried to colour match with the Federal Standard 595
b standard colours? These are U.S. Federal government colours, used for
painting contractors. I'm told Canada has abandoned its own colours, and
uses the U.S. system now. I'm told the U.K. has a standard as well, but
these too are being superseded by the US F.S. colours. Apparently, these
chips are used in the model aviation field. As actual paint samples, they
are said to have an advantage over Pantone.

Comments?

Rob Kirkham
--
Stafford Swain
26 Kenneth Street
Winnipeg, MB, Canada
R3T 0K8
(204) 477-9246
sswain@mts.net


Re: scanners to measure colour

Stafford Swain <sswain@...>
 

Well said, Bob, been there, pretty much done it too!

Rob Kirkham said on Saturday, May 17, 2003 at 10:07:16 AM:

A group of CPR SIG types are developing a project to identify CPR paint
colours, including freight car colours. It occured to me that by scanning a
painted object or chip, we would be able to obtain a lot of standardized
information, including hue, saturation, luminosity, and red, green blue
numbers.
Is there a flaw to this - in other words, will the scanner fail to act in a
> standardized way when it makes the image? I was guessing that it would be
similar to the matching machine in the paint store, but without the ability
to provide arecipe for the right paint formula. I'd appreciate comments and
information from anyone knowledgable on the subject.
I had a chance to work with some color mavens a few years ago when I was
chair of the Web3D Consortium's Color & Lighting Working Group. This is
what I learned:

(a) "NTSC" means "Not The Same Color".

(b) "CMYK" means "Can't Make Your Kolor".

(c) The principal use of the Pantone Matching System is to persuade your
printer to knock a couple bucks off the price when he gets the color
wrong.

(d) Windows Color Management, which came in with Windows 98, and which
you probably didn't know you have on your computer, is really very good,
and it's responsible for what you see on your screen looking pretty much
like what came in from your scanner and what gets printed out on your
printer. Pretty much. But then, "pretty much" is more difficult to
achieve than most of us can imagine, and we owe the folks responsible a
big thank you.

HOWEVER! (e) The ground truth is whether dried paint on your model car
is going to look like dried paint on the prototype under typical
lighting conditions (which are different: despite the growing popularity
of garden railroading, the vast majority of models are seen indoors
under incandescent or fluorescent light). EVERYTHING that gets in
between those two paint samples (and lights and eyeballs), such as a set
of magic numbers from a scanner and a graphics program, will act to
PREVENT the two from looking alike.

There's also the question of whether you should be matching the real
prototype or the hideously oversaturated printed Kodachrome on the dust
jacket of the railroading book which has probably taken its place in the
minds of far too many of us. However, that way lies madness. Stick
with the real prototype, not the imagined one, and at the end goose the
saturation a little.

Even then, there are goblins lying in wait in virtually every step of
scanning, viewing, and printing (white point and intensity transfer
functions are only the more easily treatable of them). At the end of
the day, it's the paint chips and the lights and the eyeballs. Forget
that at your peril.
--
Rev. Bob "Bob" Crispen
bob at crispen dot org

"I'm supposed to be rooting for AOL over Microsoft? What's up next,
Union Carbide versus Philip Morris?" - Jamie Zawinski



To unsubscribe from this group, send an email to:
STMFC-unsubscribe@egroups.com



Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
--
Stafford Swain
26 Kenneth Street
Winnipeg, MB, Canada
R3T 0K8
(204) 477-9246
sswain@mts.net


pre-WW1 SAL boxcars

ThisIsR@...
 

Fellow Listers:
Does anyone out there have any info on early I.E. pre-WW1 SEABOARD
rollingstock? I know they had to have boxcar classes prior to the B-3 class.
The
earliest roster I have is dated 1938 and doesn't report any older boxcars.
It does list
some gons built in 1905. Thanks.

Richard Stallworth


AccuPaint Pullman green Recommendations

Andy Carlson
 

Ed Austin, one of many fine modelers who live by
AccuPaint's numerous industry best qualities has the
following mix. I have been attempting to get the
traditional Pullman green by mixing black with
mono-pigmented chrome yellow ( 7-1 ratio, for
starters) and I am not getting the good results as I
get with Automotive lacquers. I believe it must be a
problem with the yellow pigments. When I find a
satisfactory mix, I will share. Anyhow, here is Ed's
mix:

10 parts Brunswick Green
3 parts Oxide Brown
1 part CN Yellow

I remain,
-Andy Carlson


Re: scanners to measure colour

Dave Nelson <muskoka@...>
 

-----Original Message-----
From: Ron Hildebrand [mailto:SteamFreight@hilstudio.com]
Actually, Dave, I think the RGB and luminance data from a scan would be
highly accurate. [snip]
Ron, I was simply recalling all the black & blue marks, abrasions, and other
inflicted wounds I collected when I proposed the same use of technology as
Rob has, and if any of my comments of this morning are overstated in the
negative it might be due to being knocked half senseless in the first
discussion of scanning by the words of rebuttal which were leaping right off
the monitor screen to pummel me about the head.

But in the final analysis I do recommend the exercise and I do recommend to
the exerciser the advise they present their learnings in the form of mixed
pigments on a model, duly avoiding mention of RGB etc. etc, even tho applied
pigment mixture may well have been driven by the RGB output of a scanned
paint chip or photograph. IOW, for almost all, pigments please and numbers
numb.

Dave Nelson


Re: Steel running boards

Chet French <cfrench@...>
 

--- In STMFC@yahoogroups.com, thompson@s... wrote:
When did the various vendors develop and start manufacturing steel
running boards and brake steps for house cars. I am wondering what
might have been available to the Wabash for their 1940 rebuilding
program which produced the 83000-83231, 83232-83481, 84000-84543,
and
85000-85199 series box cars. I have not been able to document what
type or make running boards were used.
Chet, PFE tried out Apex steel boards in early 1939 and promptly
adopted
them as standard. This was a steel grid design. Some earlier ones
were
solid steel with raised patterns (like diamond tread plate, though
I don't
know if that particular pattern was used).


Tony and Richard, Thank you for your input.

Chet French
Dixon, IL


Re: scanners to measure colour

Rev. Bob 'Bob' Crispen
 

Rob Kirkham said on Saturday, May 17, 2003 at 10:07:16 AM:

A group of CPR SIG types are developing a project to identify CPR paint
colours, including freight car colours. It occured to me that by scanning a
painted object or chip, we would be able to obtain a lot of standardized
information, including hue, saturation, luminosity, and red, green blue
numbers.
Is there a flaw to this - in other words, will the scanner fail to act in a
standardized way when it makes the image? I was guessing that it would be
similar to the matching machine in the paint store, but without the ability
to provide arecipe for the right paint formula. I'd appreciate comments and
information from anyone knowledgable on the subject.
I had a chance to work with some color mavens a few years ago when I was
chair of the Web3D Consortium's Color & Lighting Working Group. This is
what I learned:

(a) "NTSC" means "Not The Same Color".

(b) "CMYK" means "Can't Make Your Kolor".

(c) The principal use of the Pantone Matching System is to persuade your
printer to knock a couple bucks off the price when he gets the color
wrong.

(d) Windows Color Management, which came in with Windows 98, and which
you probably didn't know you have on your computer, is really very good,
and it's responsible for what you see on your screen looking pretty much
like what came in from your scanner and what gets printed out on your
printer. Pretty much. But then, "pretty much" is more difficult to
achieve than most of us can imagine, and we owe the folks responsible a
big thank you.

HOWEVER! (e) The ground truth is whether dried paint on your model car
is going to look like dried paint on the prototype under typical
lighting conditions (which are different: despite the growing popularity
of garden railroading, the vast majority of models are seen indoors
under incandescent or fluorescent light). EVERYTHING that gets in
between those two paint samples (and lights and eyeballs), such as a set
of magic numbers from a scanner and a graphics program, will act to
PREVENT the two from looking alike.

There's also the question of whether you should be matching the real
prototype or the hideously oversaturated printed Kodachrome on the dust
jacket of the railroading book which has probably taken its place in the
minds of far too many of us. However, that way lies madness. Stick
with the real prototype, not the imagined one, and at the end goose the
saturation a little.

Even then, there are goblins lying in wait in virtually every step of
scanning, viewing, and printing (white point and intensity transfer
functions are only the more easily treatable of them). At the end of
the day, it's the paint chips and the lights and the eyeballs. Forget
that at your peril.
--
Rev. Bob "Bob" Crispen
bob at crispen dot org

"I'm supposed to be rooting for AOL over Microsoft? What's up next,
Union Carbide versus Philip Morris?" - Jamie Zawinski


CN SIG CPR colours

Rob Kirkham <rdkirkham@...>
 

Thanks, Stafford,

can you put me into touch with John Morris?

Rob Kirkham


other colour measures

Rob Kirkham <rdkirkham@...>
 

Thanks for the comments on this subject thus far.

Has any one on the list tried to colour match with the Federal Standard 595
b standard colours? These are U.S. Federal government colours, used for
painting contractors. I'm told Canada has abandoned its own colours, and
uses the U.S. system now. I'm told the U.K. has a standard as well, but
these too are being superseded by the US F.S. colours. Apparently, these
chips are used in the model aviation field. As actual paint samples, they
are said to have an advantage over Pantone.

Comments?

Rob Kirkham


Re: B & O Gondola Lettering

Jon Miller <atsf@...>
 

13 Great States to be on the car--sic---slogan was added to the logo around
1940-41 as applies to these cars. What changed in 1946 is that the
BALTIMORE AND OHIO moved to the center of the car<

Ok now I realize that we are talking about mill gons and not boxcars but
if I remember I was told that the 13 states emblem was way to late for my
era on the X29s. So are we saying B&O put the emblem on mill gons in the
'40-'41 period but not on boxcars until 5+ years later?

Jon Miller
AT&SF
For me time has stopped in 1941
Digitrax DCC owner, Chief/Zephyr systems
NMRA Life member #2623
Member SFRH&MS


Re: CN SIG CPR colours

Stafford Swain <sswain@...>
 

Stafford Swain says:

We currently have 18 different colors (BTW 4 are CPR colors and 2 are
CN/VIA colors)
I heard about these recently. Any chance of learning your sources for
matching the CPR colours?

Rob Kirkham
Scalecoat for three of them. The crew that did the 2816 restore a couple of years ago caused the fourth to happen.

Basically, Scalecoat had killed off two of its traditional CPR colors (Diesel Gray and Diesel yellow) leaving only its Tuscan in the product line. These three colors were generally believed to quite accurate by CPR key folks I know. There was thus a clamor to bring them back and it was cool with Scalecoat for us include the CPR Tuscan in our line too. Essentially, I view these three shades as the best we have until you fellows (though your study) determine we have to change them (which we should if that's the way it falls out).

The brand-new CPR Steam Gray was re-discovered in the research for the 2816 restoration by CP. John Morris was provided several samples by one that restoration's consults.



--
Stafford Swain
26 Kenneth Street
Winnipeg, MB, Canada
R3T 0K8
(204) 477-9246
sswain@mts.net


Re: B & O Gondola Lettering

Scott Pitzer
 

Jerry,
You're saying you're afraid your period is too early for the 13 Great States to be on the car, right?
As I get it, the slogan was added to the logo around 1940-41 as applies to these cars. What changed in 1946 is that the BALTIMORE AND OHIO moved to the center of the car, allowing "B&O" to return as reporting marks above the number. In that case the herald isn't a problem.
Scott Pitzer


I've got Al Westerfield's great B & O O-27 mill gondola kit and
would like to letter it using the handsome "Linking 13 Great States
with the Nation" logo. The trouble is I think I'm right on the border
line, time wise, for it's use (late 1946). Can someone please help
with the exact date this lettering change came into effect on the O-
27's ?
It wouldn't be a problem to choose the earlier lettering style, as
all the of it is included with the decal sheet. But, I'd really like
to do the "Linking 13 Great States with the Nation" hearld, can some
one help. Chris, Bob, somebody! Thanks, Jerry Stewart




Re: scanners to measure colour

Ron Hildebrand <SteamFreight@...>
 

Actually, Dave, I think the RGB and luminance data from a scan would be highly accurate. Photographers have to deal with such things all the time, and the system in place (with 255 steps from white to black for each of the three colors plus black) yields extremely accurate results. For instance, the problem with monitor visual mis-matches can be taken care of by profiling the monitor, so the actual display characteristics of the system is measured and calibrated for use in compensating the data fed to the monitor, resulting in an accurate visual display. (A Color Vision Spyder with a for generating a monitor profile can be purchased for around $200.)

I would agree that using the resulting data could be very problematic, however, because that data would have to be translated once again to mixes of pigments or base colors. That's what the scanners in paint shops are programmed to do, but from an actual sample, not numerical data.

If one wanted to provide an accurate color sample to, say a model paint manufacturer, it could be done digitally with a photographic print, saving the problem of getting a physical sample from a full size boxcar or locomotive, for instance. A photograph could be made with a digital camera, and have photographic prints made of the sample (with properly calibrated and profiled equipment being used for the entire process, from scanner or camera to print), and have the resulting print used as a sample to be read by either the paint store, or the paint manufacturer if a line of accurate model paints was the end result desired.

As you also mentioned, other problems could be introduced in the process, such as dirt or oxidation altering the true color of the original. Also, the prototype had to deal with variations in color from batch to batch, too, so any given prototype sample is only one of perhaps hundreds of variations of that color as it existed on the railroad in real life.

Ron Hildebrand

At 09:38 AM 5/17/2003 -0700, you wrote:
This was discussed at length some time ago, shortly after the creation of
the list. You may want to check the archives.

There are several issues you should be familiar with -- the both the gamma
and fundamental color accuracy on your scanner will be less than ideal so
you won't necessarilly get an exact match -- close, but not exact.
Essentially neither RGB or CMYK will be numerically sufficent to capture the
true color. Your chips will exhibit many colors when viewed as pixels, only
some of which will come close to anything in the Pantone library -- a
difference your eye will percieve immediately. Your monitor will likely be
unable to represent the same visual color as what your scanner captured,
leading you to draw visually false conclusions when you measure pixels. The
RGB and/or CMYK for commercial paints will not be terribly close to anything
you measure. And lastly when you report whatever you find there will be
dozens of persons who reply with: a) it's not accurate; b) color is the
effect of particular lighting; c) large objects reflect differently from
small models; d) layout lighting is not the same as sunlight; e) tint
(bright light) and tone (dirt) make it all different anyway; and lastly f) I
do things to please myself so the hell with your numbers.

All that said, while it may sound like I'm saying don't bother, I'm not.
The exercise in itself will teach you any number of useful things about
color -- pigments or light, whether it is an understanding of the
differences in two different bottles of paint, why one paint appears as it
does, why a certain commercial freight car red is wrong, and what drops you
might add to a mix to get where you wnat to go. It's worth the effort.
Just don't expect to convince anybody you've got the right RGB numbers for
the XYZ car of 1947. OTOH, you can say "I used a mix of 10:3:1 of paints A,
B, and C on my XYZ car of 1947" and loads of people will lap it up like a
cat over sweet milk. Go figure.

Dave Nelson


Re: scanners to measure colour

asychis@...
 

In a message dated 05/17/03 12:26:53 PM Central Standard Time,
ljack70117@adelphia.net writes:

Even a paint company that mixes a large batch of paint then
wants to mix the same color in another batch can not hit it exactly.
That is why when painting your house they tell you to use the same
batch.
So, the question becomes, why are we worrying so much about this? All
prototype freight cars weather immedaitely, so why not just get it as close
as possible and fire away?

Jerry (out of jail at last) Michels


B & O Gondola Lettering

switchengines <jrs060@...>
 

I've got Al Westerfield's great B & O O-27 mill gondola kit and
would like to letter it using the handsome "Linking 13 Great States
with the Nation" logo. The trouble is I think I'm right on the border
line, time wise, for it's use (late 1946). Can someone please help
with the exact date this lettering change came into effect on the O-
27's ?
It wouldn't be a problem to choose the earlier lettering style, as
all the of it is included with the decal sheet. But, I'd really like
to do the "Linking 13 Great States with the Nation" hearld, can some
one help. Chris, Bob, somebody! Thanks, Jerry Stewart


Re: PRR design covered hoppers

Denis F. Blake <dblake2996@...>
 

I would bet that "rock" in this instance, is dry rock, processed phosphate. On SE RR's rock was the term used for phosphate. For instance on the SAL there were dry rock and wet rock cars. Same can be said for the acl as well.

Best regards,

Denis F. Blake
NS Conductor
Columbus, OH
TTHOTS

Please visit my photo site at

http://hyperphoto.photoloft.com/view/allalbums.asp?s=cano&u=1665499

----- Original Message -----
From: tim gilbert
To: STMFC@yahoogroups.com
Sent: Saturday, May 17, 2003 2:49 PM
Subject: Re: [STMFC] PRR design covered hoppers


Dana and Larry Kline wrote:
>
(snip)
>
> Tim Gilbert, Dave Nelson, Mike Brock, Jeff English and others: What
> do the
> available wheel reports say about observation of PRR covered hoppers
> on
> other railroads?

Larry,

The only PRR H30's that were included in any wheel reports I have parsed
were the ten reported on the SOU's Washington Division in the Fall of
1947. They appear to have been loaded at Tye River VA with "Rock (?),"
for points north of Potomac Yard. The H30's appeared to have traveled
southbound from Pot Yard empty.

Tim Gilbert

Yahoo! Groups Sponsor



To unsubscribe from this group, send an email to:
STMFC-unsubscribe@egroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.


Re: PRR design covered hoppers

tim gilbert <tgilbert@...>
 

Dana and Larry Kline wrote:
(snip)

Tim Gilbert, Dave Nelson, Mike Brock, Jeff English and others: What
do the
available wheel reports say about observation of PRR covered hoppers
on
other railroads?
Larry,

The only PRR H30's that were included in any wheel reports I have parsed
were the ten reported on the SOU's Washington Division in the Fall of
1947. They appear to have been loaded at Tye River VA with "Rock (?),"
for points north of Potomac Yard. The H30's appeared to have traveled
southbound from Pot Yard empty.

Tim Gilbert


Re: Correct Color For B&O house cars in the Mid 1940's

switchengines <jrs060@...>
 

Chris my friend I can show you a color picture, and your right it
looks more of a "brown shade" to me also. It's in Classic Trains,
the winter 2001 issue, on page # 56 of the classic photo section. I
remember B & O's oxide red when freshly painted back in the 1950's
and this is nothing like it.
Ted, the oxide red color when first painted on the car was a very
vivid red color, it turned to a pinkish shade as it chalked white
with age and weathered. Regards, Jerry Stewart


--- In STMFC@yahoogroups.com, CBarkan@a... wrote:
I agree with Ted that Floquil Zinc Chromate primer is an excellent
starting
color for B&O boxcars from the mid-1940s through 1950s. As I said
in another
post, Scalecoat 2 oxide red is also correct. So you have a choice
depending
on your preferences regarding these two brands of paint. I moved
away from
Floquil for basic car-painting because Scalecoat's finish is better
for
decaling, but there are still other applications when Floquil is
useful.

The correct color prior to the mid-1940s is another story.
Unfortunately we
don't really know what it is. The B&O changed their spec in the
mid-1940s
but I have not seen enough good quality color photos prior to that
date to
really know what it was. I think it was a "browner" shade than the
oxide red
referenced above. In addition to boxcars, it was also used on
their
composite gondolas. If anyone can refer me to any good color
photos of B&O
boxcars painted prior to the mid-1940s, I would be eager to see
them.

Chris Barkan


In a message dated 5/16/03 4:21:23 PM, tculotta@s... writes:

<< --- In STMFC@yahoogroups.com, eric thur <erict1361@y...> wrote:

Can anyone give me a tip on what Flowquil or Scalecoat paint
colors to mix
to

acheive the Correct Red-Pinkish color for B&O M26-M27-M15 & M53
Boxcars.


Eric:


The best starting point I know of is Floquil's Zinc Chromate Primer.


Regards,

Ted Culotta >>


Re: Steel running boards

thompson@...
 

When did the various vendors develop and start manufacturing steel
running boards and brake steps for house cars. I am wondering what
might have been available to the Wabash for their 1940 rebuilding
program which produced the 83000-83231, 83232-83481, 84000-84543, and
85000-85199 series box cars. I have not been able to document what
type or make running boards were used.
Chet, PFE tried out Apex steel boards in early 1939 and promptly adopted
them as standard. This was a steel grid design. Some earlier ones were
solid steel with raised patterns (like diamond tread plate, though I don't
know if that particular pattern was used).

Tony Thompson Editor, Signature Press, Berkeley, CA
2942 Linden Ave., Berkeley, CA 94705 http://www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@signaturepress.com
Publishers of books on railroads and on Western history

168741 - 168760 of 187869