The Prop Den
http://thepropden.aokforums.com/

Doing Better Photographic Analysis
http://thepropden.aokforums.com/viewtopic.php?f=3&t=1735		 Page 1 of 3
Author:  CSMacLaren [ Sun Mar 23, 2008 12:58 am ]
Post subject:  Doing Better Photographic Analysis
DOING BETTER PHOTOGRAPHIC ANALYSIS

Photoshop is a wonderful tool that has allowed us to do many photo comparisons (or “comps”) of different helmets. It’s important to understand it’s limitations in order to not only produce better and more accurate comps but to also discern when a comp is faulty or misleading. Because Vader is such an iconic work of art, people can unwittingly or intentionally place idealized poses in parallel and have visual results look factual.

There are various topics I would like to address, such as pitch, roll and yaw – as well as camera position and distance. As thoughts develop into writing, I will update the first few posts of this thread.

Dealing with Size Differences

When comparing helmets of similar poses from two photographs, size matters. Many make the assumption that Helmet “A” and “B” are the same and resize accordingly

(Helmet "A" is a Don Post Deluxe. Helmet "B" is a Rubie's Deluxe / Supreme.)

Image

As a result, one might conclude that “B” is a good size match for “A”:

Image

While this is a good analysis of general shape, and the masks appear roughly facing the same direction, this is where the assumption of their sizes being equal fail:

Image

The above shows the guessed size (left) versus the actual size (right). That size difference can throw off even the experts.

In reality, the size differences are as follows (nothing beats a true side-by-side).

Image

Comps are only good if you are aware of the size differences between the helmets being compared. Most people will try to size “A” to match “B” and overlay them to verify a match in size. However, one should not assume that “A” and “B” are the same size. There are various vendor brands and they all vary in size and shape. Some vendors will idealize their products by making the skull larger and more symmetrical to achieve more heroic proportions. This, however, can bias a facemask comparison, especially to those who feel “bigger is better” in terms of lineage.

If you’re not comparing your photo against someone else’s but are photographing two of your own helmets, stand as far back as 9-10 feet and have the helmets face you as accurately as possible (as shown in the photo above).
Author:  CSMacLaren [ Sun Mar 23, 2008 1:02 am ]
Post subject: 
Perspective Distortion

Perspective distortion (PD), according to Wikipedia, is “the appearance of a part of the subject as abnormally large, relative to the rest of the scene, or an apparent lack of distance between objects in the foreground and those behind them.” The closer you are to a helmet, its face looks large but the silhouette of the head will look smaller in proportion. Close up, the lens sees less of the sides of the helmet. But if you were to step back to 20 feet and zoom in tight, you’d see a totally different picture. Note production footage of Star Wars and the distance between Vader and the camera at certain scenes of the movies.

Six feet is a good distance that reduces PD. Most cameras with even minimal zoom can handle this. (If you only have a camera phone, then sorry, please consider investing in a nicer camera!)

When it comes to our hobby, accuracy is important to many of us, and thus we will do a lot of photo analysis. The following are of the same helmet in the same room, photographed at different distances:

Image

For a larger image of the above, click here:
http://i18.photobucket.com/albums/b138/ ... comp_1.jpg

With a line-up like the one shown above, you can develop an eye for photographic distance. Then you can examine a photo like this:

Image

…. and say, “Hey, the one on right looks like it was photographed about 3 or 4 feet. This isn’t an apples-to-apples comparison” and have the person stand back and re-photograph the helmet at 6 feet to produce some better results. Ironically, these are the same brand of helmet, taken at two different distances and two different angles.

One has to be responsible to take photographic distance is not taken into consideration, as perspective distortion can skew or even bias a comparison.

Image

Most often, when comps are done, the one doing the analysis will try to line up the face of pictures “A” and “B”. One therefore cannot assume that because the face lines up that all else must be accurately representative of the shape of the helmet regardless of perspective distortion.

The following illustrates the size and proportions difference created by PD.

Image

Thus you can see the face matches up in size quite well, but the dome, the mask silhouette, and the neck completely fail to match. As you can see, the size differences, if used inappropriately, can be misleading, and therefore we as a community have a responsibility to recognize the shortcomings of each comp (on a case-by-case basis) when judging whether or not the pictures give an accurate story.
Author:  CSMacLaren [ Sun Mar 23, 2008 1:06 am ]
Post subject: 
Pitch, Roll, and Yaw

Because Vader's helmet (both mask and dome, collectively) is a three dimensional object, the concept of pitch, roll and yaw are important when comparing helmets from different photographs. It is important to make sure the pitch, roll and yaw are as close as possible in order to obtain an even comparison. Because the Vader helmet sculpture is an irregular and asymmetrical shape, if the rotation of either mask or dome is off by several degrees, it can create an impression of increased or decreased size, thereby thwarting proper size comparisons.

Let's examine the terminology. Credit and thanks goes to Skip for use of this X-Wing - it's a paper model! (Paper Models)

Image

From a helmet wearer's perspective:

Pitch is like you nodding up and down.

Roll (like when a fighter plan rolls, in fact) is like you tilting your head from shoulder to shoulder.

Yaw is like you turning your head left to right to say "no".

Article forthcoming.
Author:  CSMacLaren [ Sun Mar 23, 2008 1:06 am ]
Post subject: 
Placeholder 2

This space left intentionally blank for future purposes.
Author:  CSMacLaren [ Sun Mar 23, 2008 1:07 am ]
Post subject: 
Placeholder 3

This space left intentionally blank for future purposes.

Thanks! I think I just need 3 placeholders. Please feel free to post feedback. :wink:
Author:  VEGAS VADER [ Sun Mar 23, 2008 1:33 am ]
Post subject: 
verry impresive Mac!! :cool: can we make this into a sticky thread?
Author:  Too Much Garlic [ Sun Mar 23, 2008 11:14 am ]
Post subject: 
That's some profound differences. :eek

The thread has been made a sticky...
Author:  Lambotour [ Sun Mar 23, 2008 3:09 pm ]
Post subject: 
Very informative and great illustrations Mac. Amazing what a little bit of distance will do.
Author:  SithLord [ Sun Mar 23, 2008 3:15 pm ]
Post subject: 
Try the comp with 4 vs 10 ft. It's clear there will be a difference with 3 vs 10.
Author:  CSMacLaren [ Sun Mar 23, 2008 7:54 pm ]
Post subject: 
4ft vs 10ft


Bear in mind that these comps are of the same helmet and same dome positioning, each facing the same direction. Camera height stays relatively the same while camera distance changes. If a dome is tilted back more, results will look different, but due to the fixed mount of the DP DLX, I can't simulate that.

Image

Image

I don't want people discouraged with skewed results to give up on photographic analysis. Again, you have to recognize the strengths and weaknesses of each situation on a case by case basis.
Author:  CSMacLaren [ Sun Mar 23, 2008 8:37 pm ]
Post subject: 
6ft vs 10ft

Now comes the question: is there an appreciable difference between 6 feet and 10 feet? If not, why do we ask people to photograph at 6 feet? As mentioned before, some people have 1-2 Megapixel cameras that have very mediocre zoom. 6 feet is a comfortable standard that minimizes PD and allows them to still zoom in and get an acceptable level of detail without investing in a more expensive camera. However, camera prices are coming down, and you can get a 4-7 Megapixel camera roughly for $150-200.

Image

Side by side, with an image of this size, the areas farther away from the center (i.e. face of the mask) will become smaller. The two images will have a far larger center common area when you overlay them. Still, there is an appreciable difference:

Image

However, the same image made smaller, the differences are difficult to detect.

Image


Thus it depends on what someone is trying to here. Those not knowing the distance differences but are trying to be as detailed as possible might feel that the dome on the right is larger than the dome on the left. Again, we have to disclose camera distances and match things up carefully.


When to 6 feet or more distance

If your camera has a powerful zoom or telephoto lens and you can handle 10-15 feet, doing a lineup of different helmets. However, it's very beneficial to disclose the distance, so that someone who may use your photo for comparative analysis can factor in the PD related to the photographic distance.

Image
Author:  CSMacLaren [ Sun Mar 23, 2008 8:47 pm ]
Post subject: 
An Example of Improper Size Comparison

This is a good example of where comps can fail us. Again, comps are a best-guess presuming that the domes or the masks are of equal size. This is where we get into some troubled waters. Different vendors masks and helmets will vary in size, and it makes comparisons very tricky.

Here is an example of someone who sees Helmet "B" and wants to compare it with his helmet ("A"). He matches the dome sizes exactly and this is what he gets:

Image

As a result of seeing this, people assume that in real life, "A" has a larger facemask than "B". If those regarding it believe size means superior or closer lineage to the original, then the above comp is misleading. [/u]
Author:  SithLord [ Mon Mar 24, 2008 4:07 am ]
Post subject: 
Thanks for showing the 4vs10 and 6vs10ft. If you are zooming in, you would be changing two parameters at once but of course that is what people are inclined to do to have the helmet fill the frame.
But what if they didn't? I don't mean for you to show this as it's a bit more work but I just wanted to point out that photos can be taken further back without changing the focal length of the lens and that will reduce the compression effect somewhat that occurs from zooming in. That's what I will do on occasion and then I crop the image if the resolution is sufficient because I know that zooming in will produce that front-to-rear compression effect. Of course pulling back at the same focal length will make the helmet occupy the center, and smaller portion, of the field for a shorter focal length lens, making it distort less. Unfortunately most non-SLR digital cameras do not have accurate control over focal lengths between the ends of the zoom range...just thinking out loud here...

Keep in mind that when I do comparisons, I've already measured the masks so that's the starting point. We can't always know the distance. It's not like I just overlay them without knowing anything about their sizes...and certain features cannot be explained away by perspective differences if they appear to be distinct from every angle...
Author:  HAL9000 [ Mon Mar 24, 2008 6:35 am ]
Post subject: 
We should also mention the depressing fact that no two lenses are
the same! Unless you have the money and Schneider optics will happily
engineer you a pair that are only 0.0025 mn in difference.

So for the most part - comparisons where the helmets are all shot by the
same lens are fine. But when you start comparing shots from different cameras then you are at a disadvantage even if both shots were taken at
the same appropriate distance and focal length.
Author:  SithLord [ Mon Mar 24, 2008 6:33 pm ]
Post subject: 
Well assuming they produce an optically flat image, all other things being equal, if the focal length is the same it should render a similar image in terms of proportions (yes a 50mm lens for a medium format camera will be different than a 50mm lens for a 35mm film camera because the angle of coverage will be different...and that will come into play with digital camera chip size). Regardless, I applaud Mac...that's a lot of work to show the differences...
Page 1 of 3 All times are UTC
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/
Hosted by Freeforum.ca, get your free forum now! TOS | Support Forums | Report a violation
MultiForums powered by echoPHP phpBB MultiForums