| Author | Thread |
|
|
02/02/2005 04:02:31 AM · #1 |
I another thread Vfwlkr asked why B/W was so popular in DPC, and also asked the following question: "Why are centred compositions considered a no-no? I've seen quite a few comments on photographs (not mine) that said they were voting down because it was 'too centered'." Faidoi responded that yoiu have folks downgrading pictures because they don't follow the "rule of thirds", and vfwlkr asked what that was.
The thread since has devolved into a discussion of B/W vs color. I prepared and posted a lengthy discussion of composition, and realize that the title of the thread may lead it to go undiscovered, so I have moved it over here to a new thread.
****************
"The Rule of Thirds" is a close approximation of the ancient Greek concept of the "Golden Rectangle", a proportional rule on which they based their architecture and which is used to this day by architects and designers throughout the world because it is in effect genetically coded into organisms.
A "Golden Rectangle" is one in which the ratio of the short side to the long side is the same as the ratio of the long side to the sum of the short and long sides. If you derive a series of increasing rectangles in this manner, rotating each ccw 90 degrees as you create it, and draw a curve through the relevant vertices of the rectangles, you will inscribe a very specific spiral that is absolutely omnipresent in nature. The most common illustration of it is the sectioning of a chambered nautilus shell.
The Italian Mathematician Fibonacci, a long time ago, created the "Fibonnaci Series" of numbers, where each number in the series is equal to the sum of the previous two numbers in the series; 1,1,2,3,5,8,13,21,34,55, etc etc. You see these numbers defining such spirals as the helices of sunflower seeds and many other things. The reason it shows up so commonly in nature (this proportion, ratio, or numerical derivation) is that it is the most efficient "packing" sequence possible.
Another example of it, less obvious, can be found in the amount of rotation of leaves around the stem of a plant as they rise as the plant grows. Using this ratio/series, the plant assures that the maximum possible amount of sunlight falls on the lower leaves as the upper ones grow. It doesn't do this consciously, of course; the process is coded into its DNA.
To apply the "Rule of Thirds" to an image, divide its area into 3 equal parts both horizontally and vertically. This is done by drawing 4 lines, 2 vertical and 2 horizontal, at the appropriate positions along the 2 axes of the image. You get 4 intersections where these lines cross, and according to the rule of thirds the "focus" of the composition should center on/derive from one or more of these intersections. For example, a "well composed" seascape with a lighthouse might have the horizon line on the bottom horizontal line, with the lighthouse centered on the right vertical line and the top of the lighthouse ending at the top horizontal line. You see a great deal of formulaic art that is obviously created by "gridding" the surface of the canvas and sketching the compositional elements according to this "rule".
The rule has become so pervasive as to be a cliché; and like all clichés, there's a major foundation of truth underlying it. It's always "safe" to make an image according to this rule. As artists, IMO, we need to seek to push outside that boundary a little bit, make our compositions a little edgier, a little different. Sometimes this means going back to currently "unfashionable" rules and revisiting them.
One of these rules, or perhaps "model" is a better term, is the "symmetrical model." I think it's getting an unjust bashing from the voters in DPC a lot of the time. There's nothing inherently "wrong" with making an image that's in a state of relative stasis, self-contained, inward-looking, serene. Not everything has to move and sweep.
Look at the recent "Best of 2004" challenge winners:
Heida's "Summer Breeze" is a strong, dynamic, sweeping image that is classic "rule of thirds."
Nordic's "Winter Web" is a strong example of "leading lines."
Briphoto's "Lightning Strikes" is basically rule of thirds; the lowering sky defines the top third, and the space between the lightnings, to which our eye is drawn, is on the right third.
In "Faceless", Typologic's "Unplugged" follows the rule of thirds pretty much. There is somewhat more foreground than the "rule" would dictate, and the shadow of the head, a strong compositional element, is centered.
Jonpink's "Early Morning Runners" would follow the rule of thirds if it were extended into a rectangle to the right; by cropping this area out of his image, by making a square image, he creates a paradoxical tension that draws us to the "wilderness" we cannot quite perceive but expect to see to the right. This is one of the reasons it works so well.
CheekyMonkey's "Lazy Sunday Afternoon" is fundamentally rule of thirds, with a symmetrical overlay; the sky defines the top third, the vee between the man's left shoulder and the grasses is on the left third (it's a strong compositional element), but the highlit right side of the girl is very nearly centered and anchors the composition strongly. This is a very contained image, much more static/serene than the usual DPC ribboner.
In the "Old and New" challenge, third place is a strong, diagonal composition anchored upper left and lower right by nexuses that hover at "rule of thirds: interstices.
It's very illuminating to compare the first and second place images, which happen to be of the very same building. Gabriel's winning image is classic rule of thirds, with the lit pyramid on one vertical axis and the tower on the other, while the foreground and the sky along the right edge also divide the image very neatly into the other thirds; the whole is capped off, (as is Heida's image) by a strong diagonal sweep from right to left, in front of which the lit pyramid hovers and beckons magically.
The other Louvre shot, Itsimring's second place winner, is very much in the symmetrical mode, with a square framing and the apex of the pyramid hovering smack int he middle of the image. It's a very powerful composition, obviously, as it garnered the second-best score in this challenge. Within this strongly symmetrical image, the lone human silhouette slightly off-center to the left provides a perfect counterpoint that gives both a compositional and human dynamic to the whole.
Arguably this image is technically and emotionally superior to the first place image (not everyone would agree, of course, as evinced by the voting) but the blue ribbon scored considerably better (the gap was substantial) and I think this is because people are more comfortable with the sheer dynamic sweep of these strong diagonals.
There...
That's my musings on composition. Thanks for listening :-)
Robt.
edited to add names of ribboners and thumbs.
Message edited by author 2005-02-02 11:28:49.
|
|
|
|
02/02/2005 04:48:04 AM · #2 |
Thanks, that was helpful!
|
|
|
|
02/02/2005 06:52:47 AM · #3 |
| Hey Robert, Think I'll print this out and refer back too it when I have BOTH eyes open, it's too early in the morning for this right now! Thanks for the lesson, really appreciated. |
|
|
|
02/02/2005 08:05:59 AM · #4 |
Using Fibonnaci's formula you can accurately predict the number of beavers that will multiply from a breeding pair if all the offspring mate and breed at maturity and none are lost to predidation, disease, accident or other outside factors. Using the sequences of numbers one breeding pair can be responsible for a potential population 2048 beavers after a ten year period of time. In 20 years the number is 2,097,152 beavers. That̢۪s a hell of an extended family. Unless greatly modified this formula doesn̢۪t work on other rodents such as rats or rabbits because they mature in less then a year, breed more then once a year and have litters of offspring larger then 2.
One must understand that this doesn̢۪t happen to this extent in nature because of other outside factors such as alligators, coyotes, disease and accidents.
There are many things in nature besides the common example of the nautilus that follow Fibonnaci's formula. The multiplication of many single cell organisms can also charted using this formula.
Message edited by author 2005-02-02 08:06:59.
|
|
|
|
02/02/2005 08:57:30 AM · #5 |
wow, I haven't ever seen that covered in so much detail. thanks a lot
|
|
|
|
02/02/2005 09:13:48 AM · #6 |
| Thanks a lot!!! This is really illustrative. |
|
|
|
02/02/2005 10:03:45 AM · #7 |
great work, robt.!
to go along with bear's technical description, jodie coston has a visual overview that you might find interesting (i liked her whole piece, but if you want to get right to the ROTs part, you'll have to scroll through it). click here |
|
|
|
02/02/2005 10:41:44 AM · #8 |
That's great Robt, a wonderful analysis which will help many. I'd love to see more such discussions.
One issue close to my interests in particular is the subject of landscapes. Landscapes follow different rules for compositions. Most often, there is no focal point (though there can be, an especially powerful technique I've seen and been trying to master is to use a wide angle lens and a crisp foreground object to lead into the picture.)
I often hear people comment on landscapes that there's no focal point, so both education to the voters and photographers would be great. I get and read outdoor photographer for several years now, but I can't recall ever reading a very good general guidelines for landscapes, or for that matter seeing any article in there that was a memorable set of rules/guidelines for composition.
So I'd love to hear the writings and thoughts of professional or well read landscape photographers here. (Hope this isn't a thread hijack--perhaps I should post this separately and just leave the thank you part!)
Regards--Neil
Message edited by author 2005-02-02 10:41:58. |
|
|
|
02/02/2005 10:58:07 AM · #9 |
Thanx for the compliments. I used to teach this stuff, so I can ramble on.
Robt.
|
|
|
|
02/02/2005 11:14:15 AM · #10 |
Hey Robert (bear_music) - Any chance this could be incorporated into the tutorials? If the photos you refer to from the challenges could be included it would be even better. I think John Setzler (jmsetzler) had one in there a couple years ago, then it was pulled. Nothing in the tutorials now for this subject and it could be helpful to many and easy to find.
|
|
|
|
02/02/2005 11:15:52 AM · #11 |
It would be fine by me, though to be frank it's just a quick gloss of the topic in my own opinion. I'll add thumbs now, though...
Robt.
|
|
|
|
02/02/2005 12:46:34 PM · #12 |
Originally posted by bear_music:
The rule has become so pervasive as to be a cliché; and like all clichés, there's a major foundation of truth underlying it. It's always "safe" to make an image according to this rule. As artists, IMO, we need to seek to push outside that boundary a little bit, make our compositions a little edgier, a little different. Sometimes this means going back to currently "unfashionable" rules and revisiting them.
One of these rules, or perhaps "model" is a better term, is the "symmetrical model." I think it's getting an unjust bashing from the voters in DPC a lot of the time. There's nothing inherently "wrong" with making an image that's in a state of relative stasis, self-contained, inward-looking, serene. Not everything has to move and sweep.
|
Perhaps it was well to point this part out as not all situations fit the or are even enhanced by the rule of thirds. I personally despise the term when it incorporates the word rule. I think it hinders those starting out in design, photography or any of the visual arts to come under the impression that this as a rule is somehow set in stone.
|
|
|
|
02/02/2005 12:57:03 PM · #13 |
"The Rule of Thirds" is what it's universally known as. I far prefer the concept of "compositional models" to "rules of composition" myself.
Robt.
|
|
|
|
02/02/2005 01:05:41 PM · #14 |
Originally posted by nsbca7:
Originally posted by bear_music:
The rule has become so pervasive as to be a cliché; and like all clichés, there's a major foundation of truth underlying it. It's always "safe" to make an image according to this rule. As artists, IMO, we need to seek to push outside that boundary a little bit, make our compositions a little edgier, a little different. Sometimes this means going back to currently "unfashionable" rules and revisiting them.
One of these rules, or perhaps "model" is a better term, is the "symmetrical model." I think it's getting an unjust bashing from the voters in DPC a lot of the time. There's nothing inherently "wrong" with making an image that's in a state of relative stasis, self-contained, inward-looking, serene. Not everything has to move and sweep.
|
Perhaps it was well to point this part out as not all situations fit the or are even enhanced by the rule of thirds. I personally despise the term when it incorporates the word rule. I think it hinders those starting out in design, photography or any of the visual arts to come under the impression that this as a rule is somehow set in stone. |
It's certainly not a set in stone rule, but there are a large number of photos that would be more dynamic or engaging if not centered. I would hope that when people make comments suggesting the rule of thirds, that it's because they think it would really help that photo, not because they think every photo should be that way. I personally suggest the rule of thirds a lot (when I actually think it would help) because I know there are a lot of less experienced photographers on the site. When a lot of people start out in photography, they don't realize that uncentered subjects are an option. They shoot a frame so that their subject is in focus and since most P&S cameras only have one focus area (in the center of the frame) they leave their subject centered. I know my photography improved significantly when I discovered the option of having things uncentered, and since this sight is about helping each other and sharing knowledge, it's important to point out things that we think would help a photo, such as the rule of thirds. Some photos are better off centered, but there's also a large number that aren't. We also all have different taste. So if someone suggests the rule of thirds for a photo, the photographer should keep in mind that that someone suggested that because they feel it would help, not because they're bound to a cliche rule.
edit: typo
Message edited by author 2005-02-02 13:07:21. |
|
|
|
02/02/2005 01:10:05 PM · #15 |
Kearock has a point, of course; for those new to photography, "composition" often means "center the subject." For sure, the FIRST compositional "model" to teach beginners to the game is the "rule of thirds". Unfortunately, many people DO take it as if it is written in stone, and automatically downgrade any image that does not fit the model because they are uncomfortable with it. So another "goal" of this site would be to educate folks to the available options.
Robt.
|
|
|
|
02/02/2005 02:17:03 PM · #16 |
Photography has been said to be a subtractive art form, where the objective is to remove everything that does not convey the desired impact until only the impact is left. Simplify, is often the comment I see -- remove the distractions so the image's impact is stronger.
As Robt mentioned, the 'Rule of Thirds' is an approximation of the Golden Ratio described by the ancients and the Fibonnaci Series of numbers that describes the most efficient means of packing elements into something. This holds true regardless of what those elements are, seed on a sunflower, curves in a shell or even elements of a photograph. In a medium in which simplicity is king, the Golden Ratio is its sceptor of power.
I think people get too stuck on the obvious use of the Rule of Thirds, and fail to realize it is simply an abstraction (an approximation) of one aspect of the Golden Ratio in application. The simplification of an application of the Golden Ratio brings dynamic power to the composition.
Of course, there are times when that simplification is not desired; but how how can those instances be recognized without knowing what would simplify the image fist?
As was Robt's point in posting this thread (I believe) the Rule of Thirds is a first step, but it is not the only step to take. It is a powerful tool that is easily learned, but it holds no power of it's own -- it's power comes from the elements of composition it approximats.
David
|
|
|
|
02/02/2005 02:29:14 PM · #17 |
Originally posted by Britannica:
Of course, there are times when that simplification is not desired; but how how can those instances be recognized without knowing what would simplify the image fist?
|
The golden rectangle is essential for the student to learn as a requisite to most forms of visual art. Once past the stage of beginner it would be wise to un-learn it as a rule, and rely solely on the trained eye as an almost intuitive step towards good composition.
In other words, once an artist is proficient with the classic compositional techniques, he or she, over time, should be able to see how a composition works best without following any set rules or even consciously thinking about it. It should become second nature.
|
|
|
|
02/02/2005 04:11:35 PM · #18 |
Originally posted by bear_music:
A "Golden Rectangle" is one in which the ratio of the short side to the long side is the same as the ratio of the long side to the sum of the short and long sides. If you derive a series of increasing rectangles in this manner, rotating each ccw 90 degrees as you create it, and draw a curve through the relevant vertices of the rectangles, you will inscribe a very specific spiral that is absolutely omnipresent in nature. The most common illustration of it is the sectioning of a chambered nautilus shell.
|
I did a bit of calculating and came up with an approx 5x8 rectangle complying with the rule. This means that the filesize of my Rebel - 3076x2052 is about 6% narrower than a "Golden Rectangle".
Does that sound about right?
|
|
|
|
02/02/2005 04:24:54 PM · #19 |
Originally posted by jemison:
Originally posted by bear_music:
A "Golden Rectangle" is one in which the ratio of the short side to the long side is the same as the ratio of the long side to the sum of the short and long sides. If you derive a series of increasing rectangles in this manner, rotating each ccw 90 degrees as you create it, and draw a curve through the relevant vertices of the rectangles, you will inscribe a very specific spiral that is absolutely omnipresent in nature. The most common illustration of it is the sectioning of a chambered nautilus shell.
|
I did a bit of calculating and came up with an approx 5x8 rectangle complying with the rule. This means that the filesize of my Rebel - 3076x2052 is about 6% narrower than a "Golden Rectangle".
Does that sound about right? |
8x12 or 16x24 are perfect. 5x8 is close enough, but just a little wider. 5x7.5 is the correct formula. |
|
|
|
02/02/2005 06:51:48 PM · #20 |
Originally posted by nsbca7:
Originally posted by jemison:
Originally posted by bear_music:
A "Golden Rectangle" is one in which the ratio of the short side to the long side is the same as the ratio of the long side to the sum of the short and long sides. If you derive a series of increasing rectangles in this manner, rotating each ccw 90 degrees as you create it, and draw a curve through the relevant vertices of the rectangles, you will inscribe a very specific spiral that is absolutely omnipresent in nature. The most common illustration of it is the sectioning of a chambered nautilus shell.
|
I did a bit of calculating and came up with an approx 5x8 rectangle complying with the rule. This means that the filesize of my Rebel - 3076x2052 is about 6% narrower than a "Golden Rectangle".
Does that sound about right? |
8x12 or 16x24 are perfect. 5x8 is close enough, but just a little wider. 5x7.5 is the correct formula. |
5x7.5 doesn't fit the formula.
x = short side = 5
y = long side = 7.5
x/y (ratio of short side to long side) = y/(x+y) (ratio of long side to the sum of the short plus long side)
x/y = ratio of short side to long side = 5/7.5 = 0.67
y/(x+y) = ratio of long side to (short side plus long side) = 7.5/12.5 = 0.60
0.67 does not equal 0.60
By the same math, neither 8x12 nor 16x20 fit either.
My 5x8 works out this way:
x/y = 5/8 = 0.63
y/(x+y) = 8/13 = 0.62
Not perfect, but closer than 5x7.5, 8x12 or 16x24.
Message edited by author 2005-02-02 18:52:58.
|
|
|
|
02/02/2005 06:54:13 PM · #21 |
Originally posted by jemison:
5x7.5 doesn't fit the formula.
x = short side = 5
y = long side = 7.5
x/y (ratio of short side to long side) = y/(x+y) (ratio of long side to the sum of the short plus long side)
x/y = ratio of short side to long side = 5/7.5 = 0.67
y/(x+y) = ratio of long side to (short side plus long side) = 7.5/12.5 = 0.60
0.67 does not equal 0.60
By the same math, neither 8x12 nor 16x20 fit either.
My 5x8 works out this way:
x/y = 5/8 = 0.63
y/(x+y) = 8/13 = 0.62
Not perfect, but closer than 5x7.5, 8x12 or 16x24. |
...and I told my college algebra prof that I would NEVER use that stuff in the real world... ;o)
|
|
|
|
02/02/2005 07:07:12 PM · #22 |
Forget the math and trust your eyes for a change :)
Message edited by author 2005-02-02 19:07:29. |
|
|
|
02/02/2005 07:14:59 PM · #23 |
Originally posted by xion:
Forget the math and trust your eyes for a change :) |
Amen!
I don't like rules.
Edit: No offense :)
Message edited by author 2005-02-02 19:15:28. |
|
|
|
02/02/2005 07:26:01 PM · #24 |
Rules, guidelines, whatever, are very useful tools for learning. One has to develop the confidence to throw them away, later. And hope desperately that one isn't jumping the gun.
E |
|
|
|
02/02/2005 07:30:10 PM · #25 |
Originally posted by jemison:
[Not perfect, but closer than 5x7.5, 8x12 or 16x24. |
Your right. I was just figuring that in my head and scratched. I was an lit/art major. I haven't taken a math class since 8th grade.
Rounded off closely:
15x24
10x16
6x10
|
|
