Scissors

Okay, I've received several requests from lovers (and haters) of both colors for me to do a full professional analysis of the differences between them. What I found surprised even me.

First, some quick background on basic physics/photography. Light is, in many ways, three dimensional; just like space. The primary colors of light are red, green, and blue. The secondary colors between these primaries are cyan, magenta, and yellow. You may recognize the secondaries because they are what printers usually use. Images on computers are saved using light's three "coordinates." This can be either red, green, and blue; or it can be hue, saturation, and luminance.

About Red, Green, and Blue
Forget what you learning about painting, as mixing paints and dyes is different than mixing light. Red and green make yellow, green and blue make cyan, and blue and red make magenta. Orange is between the primary Red and the secondary Yellow. Your television screen and computer monitor both display using very small red, green, and blue elements. Check with a magnifying glass if you don't believe me.

Image files can save image information using red, green, and blue. Quite simply, the file holds information for each pixel (dot) with a range of numbers. For example, you can have a solid image of only one color where every pixel is 255 for red, 128 for green, and 0 for blue. This would give you a vivid orange.

About Hue, Saturation, and Luminance
There is another 3-dimensional method for saving color information of an image. This is Hue, Saturation, and Luminance.

Hue: In this case, hue is the only "dimension" which gives the color. The range starts at red, hits green at 1/3 of its length, hits blue at 2/3, and wraps back around to red at the end. So if your hue is on a scale of 0-255, then red is at zero, green is at 84.3, and blue is at 169.6.

Saturation: This is a measure of the vividness of the color determined by Hue. No matter what number hue is, if saturation is at zero then you have either white, black, or a shade of grey in between. Full saturation gives the most vivid colors.

Luminance: This is the measure of brightness. If saturation is at zero and luminance is at zero, you have black. If saturation is at zero and luminance is at full, you have white.

So a hue of 85, a saturation of 127, and a luminance of 127 would give you a nice, solid green. Not vivid, but not washed out either.

On to the analysis!

What I did was take one of the side-by-side comparison shots and cropped out a part of each color that was most true. This meant choosing an area where there was enough light to see, but no direct reflection. Because of this, I grabbed from the front of the nose of both vehicles.

I then split the two images into their constituent hue, saturation, and luminance for comparison. We'll take these one at a time...

Hue
They are almost exactly the same color! In fact, the Precision Red was slightly closer (2 on a scale of 255) to orange than the Torch Red was. So effectively they are the same (almost) exact color. The Precision Red was 0 on the hue scale and Torch Red was 254. (Remember that the scale wraps back around, so 0 and 255 are right next to each other.

Saturation
Here is where the difference begins. Torch Red maxes the Saturation scale at 255. Precision Red is more muted and averaged to the 225-235 range. What's this mean? Torch Red is more vivid (duh!)

Luminance
The biggest difference by far was in luminance. Torch Red averaged at 100, whereas Precision Red (with the same lighting) averaged to 65. In plain english, this means that Torch Red is brighter because the paint reflects more light.

Summary: Torch Red and Precision Red are almost the exact same color on the spectrum/rainbow. Torch Red is more vivid and much brighter than Precision red, which is in turn more muted and much darker.

Disclaimers:
If the camera's white balance was off, it would affect the ultimate hue of both vehicles. However, it would not significantly alter the hues relative to each other considering how close they are. Luminance is determined partially by lighting conditions but, again, both areas that I compared were receiving the same light level and were from the same picture. The original images were not altered in any way, save to crop and move parts; the actual color information was not changed at any point. Your milage may vary. No C6es were harmed in the making of this thread.

If anyone wants to host the supporting images until the forum's web server is fixed, I would appreciate it.

Scissors

And before someone goes bonkers, I'm not saying one or the other is better or worse. I'm just trying to point out that the "Torch Orange/Precision Purple" crap is stupid.

The only reason someone may think that Torch Red is closer to orange is because most oranges you see have a high saturation and luminance, while most purples you see have a lower saturation and luminance.

This does not, however, mean that they have to be that way. You can have a low saturation/low luminance orange. It's just that it's so ugly, nobody uses it.

TedG

Ok that being said, a computer uses the method you spoke of. But a painter uses a quite different method. For someone to get the colors we see the torch red would use more of a yellow tint. Those cans they mix paint with have no reference to saturation, chroma, luminance etc. So is it stupid to say more orange(yellow) or purple(blue)??

Corvette Junkie

Now if you could do this with Precision Red and Mag Red. To me Precision Red is a lifeless Mag Red.

TedG

It does remind me of a non metalic mag red. The paint chip in the ordering guide showed the Precision red to be almost exactly the same as the mag red. It was a metalflake though. I even took it outsid to see if I could see any difference. That spawned my first post about the C6.

Scissors

No matter what method the painter uses to arrive at various colors, the result can be analyzed by looking at the red, green, and blue that reflects from the end product.

If I were to paint a piece of paper with red and yellow to get orange. The result would still have a hue, saturation, and luminance. The orange light reflected would still be composed of more red than green, and likely no blue.

When painting to get the torch red you see, you wouldn't use much yellow paint at all, if any. You'd use a paint that is quite simply pure, vivid red. The Precision Red would use a different substance which simply reflects less red than Torch Red does.

When I also split the image comparison into Red, Green, and Blue I found that Green and Blue were at almost zero on both colors. Any non-red mixed into Torch or Precision Red is minimal at most.

TRAYC7

It definetly makes some sense what you are saying... although some was over my head.... Im a visual learner so if Im shown it sinks in what you are referencing...

My opinion is a little more simple... They took 3 parts Torch and 1 part Mag Red and made Precision Red....

TedG

Did I fail to say good analysis of the color?

Wadoka

You can never match up personal tastes either, because each of us perceives color in our own personal way.

I looked at PR very closely at the Bash. The shot of Tracey in the car is not a good shot, because the flash makes it look much brighter than it really is. The outdoor/overcast shots are pretty good.

It is a dark, almost dull red. It is the color of drying blood, really. Dark red, with a hint of brown and rust. It grows on you. When I saw it, I went "I don't know about that..." But I liked it a lot when I saw the PR/CR convertible inside the Museum. If I was getting a convertible, I'd get PR/CR/BK.

It is a rich color. It looks very good with the Cobalt Red interior.

It's not going to catch the attention of LE or people who glimpse you in their RVM. It's not a standout red.

Zeppelin

It seems to me that if most oranges we see are high in saturation and luminance and most purples we see are lower in saturation and luminance, then saying Torch Red is more "orange-ish" or Precision Red is more "purple-ish" is a convenient shorthand for saying all the stuff it just took you half an hour to type. Therefore, I would not say the whole Torch Orange/Precision Purple thing is really stupid. It is efficient, albeit imprecise.




[Modified by Zeppelin, 10:36 PM 5/11/2004]

BuckyThreadkiller

Scissors I have the utmost respect for your technical abilities but on this one you are chasing vapor. There's no way to accurately tell actual color from digital photos.

The color that we see in life is reflected sunlight. and color is completely determined by the wavelength of the visible light spectrum that is absorbed by the subject. If it absorbs all of the light and reflects none, it is black. Reflect everything and it is white. All visible color is in between.
To view color on a monitor you are relying light generated by the monitor, but this is color that is captured by the camera digitally, saved to media and then reproduced by only the RGB screen or the CRT.

In short - generated color can NEVER match reflected color so there's no way to accurately tell. All of your calculations, while most probably accurate, are based on faulty baseline data.

GDP

Yes, but is it better?

Zeppelin

I tend to agree. I did find the background information very interesting and educational, though.

Scissors

Yes, there is.

The monitor has absolutely nothing to do with the analysis, so I don't see your point.

The only possible ways that the perceived wavelength of the light could have been altered is with a faulty CCD on the camera, incorrect whitebalance, or during JPG compression. Modern technology yields CCDs that are as sensitive and accurate as the human eye, just at a lower resolution--so that's probably out. This image I used is one that several people stated was the most accurate. The whitebalance issues was covered in the original post. This leaves only JPG compression. JPG compression, among other things, takes blocks of pixels and averages out saturation and sometimes hue. However, this is why I took a large enough area out of both cars so that I could average out the already-averaged blocks. But since it was all averages to begin with, the compression actually helped me!

But this isn't about generated color. This is about analyzing perceived color as it was reflected from the paint.

And it's not true that generated color can never match reflected color to begin with. Apple specifically patented technology to do exactly this many years ago.


[Modified by Scissors, 7:35 AM 5/12/2004]

Matt Gruber

from the post title i thought it would contain the formula from a paint shop. would be interesting to compare all 3 reds that way.

Scissors

Additional info about Saturation:

In addition to what was said above, it is also what basically controls the other parts of the spectrum. Full saturation means that the other parts of the spectrum are gone. Zero saturation means that the other parts of the spectrum are equal to the part specified by the Hue setting.

BuckyThreadkiller

Let me be more succinct -

You are measuring the color in the images. Not the paint.

Any difference in innumerable ambient conditions from color shift on the lens of the camera / accuracy of the white balance, the amount of color cover or smog filtering sunlight, the viewing angle of the bodywork....All that can make one color appear different than it actually is on film/monitor or any repro method vs. reality.

I'm quite familiar with Apple's ColorSync technology and while it is helpful in reproduction it is by no means close to completely accurate. I'm working on an Apple PowerBook with colorSync and trust me there is a significant blue cast to the monitor.

The ONLY accurate way to compare the paint color is to do it under controlled white light (5000k) conditions. It is physically impossible for any mechanical or digital reproduction to match color.




[Modified by BuckyThreadkiller, 7:34 AM 5/12/2004]

Scissors

But that's the thing. I don't need to be exact. I just need to be accurate enough to show the difference that people can actually perceive.

Even under perfect conditions with perfect equipment and methods, I highly doubt we would find a hue difference between the two colors of more than 1%.

Unless you want to argue that the Precision Red is actually brighter and more saturated than the Torch Red. Good luck getting that result in a lab.

Yes, you're right that this is an imperfect analysis. But it's accurate enough for what humans can perceive, backed up by the people who stated that the pic was an accurate representation.


[Modified by Scissors, 9:15 AM 5/12/2004]

Adam Boca

Did you get paid to do this analysis? Are you sure it was professional?

Scissors

I paid myself $3.50