Bounce flash

[steve_may1]

<p>Hi<br>

Being very new to flash photography I have spent a great deal of time researching the various techniques, one of which is bounce flash. Most of this is logical, with one exception - when the flash is bounced <em>away</em> from the subject (over ones left shoulder for instance) onto the ceiling. I could understand the logic of this if there were a wall directly behind the photographer, but if there is no wall in the immediate vicinity how does the light reach the subject?<br>

Thanks in advance for any replies<br>

Steve</p>

[gdw]

<p>Steve, when talking about bouncing flash over your shoulder they are talking about "bouncing flash" and that implies pointing the flash toward an object. It is assumed that you understand the term "bounce" and not try it if there is no object availabe to reflect the light from the flash back to the subject. This is an excellent technique in small rooms, point the flash at the intersection of the wall and the ceiling or into a corner with the ceiling and two walls to bounce back the light.</p>

[luis_g]

<p>It is aimed at and bounces off the ceiling, and back down unto the subject (though it can be bounced off walls and corners, too).</p>

[bob_sunley]

<p>Most of it won't, unless the ceiling is a cathedral style angling up from behind you. :( Black painted ceilings are even worse.</p>

<p>You have to be aware of the distances and the angles involved. Bouncing at the wrong angle means the light misses the subject. Too long a distance disperses the light so much so that very little hits the subject. When you are bouncing light off the ceiling, you usually want to narrow the beam from the flash so it has more reach.</p>

<p> </p>

[frank_skomial]

<p>Watch out for colored walls, and avoid them for bouncing.</p>

<p>Even modern flash like Nikon SB800/900 will not help you determine the needed power for bouncing. This will come with your experience. When in bounced flash head position the flash will not display correct exposure maximun range, so you will need to experiment, and see how much to boost ISO, or open aperture, to get some dispersed/reflected light that would be meaningful.</p>

[Matt Laur]

<p>I have done the reverse-bounce-behind-me trick many times, but of course only when the environment (walls, etc) make sense. In your scenario, Steve (no wall or other surfaces/objects behind you), it's definitely a matter of throwing flash power away. <br /><br />I have use this outdoors with the flash pointed directly behind me (horizontally) at the side of a white mini-van. This provided a somewhat inefficient reflector, but made for a nice big light source on the subject a few feet in front of me. The laws of physics always apply, though, even when it's damned inconvenient.</p>

[picturesque]

<p>All the folks above are correct. If there is a 'bounceable' surface behind you, the light bounces off the surface and then onto the subject. Also look up the phrase 'angle of incidence equals angle of reflection'.</p>

<p>However, if you are talking about <strong>ceiling</strong> bounce, it is possible to do a backward ceiling bounce. Usually ceiling bounce is forward, meaning the flash head is slightly tilted forward, to throw the resulting spot of light from the flash at a point approximately halfway between you/the flash and your subject, so that light travels the shortest distance possible while still bouncing.</p>

<p>With backward ceiling bounce, you are relying upon <strong>part</strong> of the spot of light thrown on the ceiling to still reach your subject. It works if your subject is not that far from you. It also works better if you manually widen your beam, depending upon the height of the ceiling. The wider you make the beam, the more part of the spot is able to reach your subject. </p>

<p>Backward ceiling bounce is not as efficient as forward ceiling bounce. The reason I sometimes do backward ceiling bounce is so that the light spilling out the front of the flash does not hit my subject. With closer subjects, that 'spillage' is still harsh and you run the risk of overexposing the subject.</p>

<p>You can also do a combination of backward ceiling bounce and backward wall bounce by aiming of the back wall where it joins the ceiling.</p>

[rodeo_joe1]

<p>I have to disagree with a lot of the previous posts and say that bouncing from the ceiling behind, directly above or to one side of the camera makes perfect sense. Using bounce flash is about improving the quality and <em>direction</em> of light. Bouncing in front of the camera usually results in light that comes almost directly down on the subject, which often looks quite unnatural. Whereas light that comes from behind or to one side of the camera can give a much more natural look. Also, increasing the distance of the light from the subject means the degree of fall-off is lessened.</p>

<p>Steve, you need to get away from the idea that light bounces from matt surfaces in the same way that it reflects from a mirror - it doesn't. When light hits a matt surface it diffuses and spreads fairly equally in all directions, and this is the whole principle of using bounced or reflected light. Yes, there'll be less light on the subject because it has further to travel before it gets there, but the losses from reflection should be no greater than if the light was pointed forward toward the subject. IMHO the <em>quality</em> of light is far more important than its <em>quantity</em>.</p>

<p>Having said that, you do need to check on the colour and glossiness of the reflecting surface. If the surface is too shiny then a lot of light will be lost, but a "normal" white emulsion-painted ceiling makes an almost perfect diffuse reflector.</p>

<blockquote>

<p>Also look up the phrase 'angle of incidence equals angle of reflection'.</p>

</blockquote>

<p>That only applies to reflections from a polished or mirrored surface. We're talking about matt surfaces here.</p>

[Matt Laur]

<blockquote>

<p>you need to get away from the idea that light bounces from matt surfaces in the same way that it reflects from a mirror - it doesn't</p>

</blockquote>

<p>Speaking as someone with a large matt surface, I certainly agree. This is also true of <em>matte</em> surfaces, which are sometimes found on things like door mats. But nattering about matters like matte-surfaced mats makes Matt a dull guy, as flattering a notion as patenting my slightly fattened Matt surface as a matte mat may be. I have been known to use myself as a Matterhorn-sized reflector while shooting a patterned platter of sauce-splattered Maltagliati.</p>

[picturesque]

<p>Rodeo--I don't agree that the angle of incidence line only applies to polished or mirrored surfaces. It applies to any kind of bounce in a general sense. It applies to how the bounce happens rather than what happens to the bounced light when it hits x surface.</p>

<p>The angle of incidence 'rule of thumb' still applies when you perform a bounce with matt surfaces--even black, matt surfaces. The fact that the surface quality may diffuse the bounced light does not negate the main angle--or path of travel--of the light. The fact that the surface color may 'eat' the light to the point that you don't have enough returned light from the bounce to affect your subject does not negate the angle.</p>

<p>I also disagree that bouncing in front of the camera causes the light to come directly down on the subject. It <strong>can</strong>, but this depends on how close the subject is. The closer the subject, the less distance to perform the bounce, the sharper the angle of bounce, the more the bounced light is 'downward'.</p>

[Matt Laur]

<p>Agree with Nadine. The angle of incidence is absolute. Scattering of light is a different matter. Even on most slightly textured surfaces (typical flat ceiling paint, for example), the scattered light will still be strongest in direction in keeping with the angle rule.</p>

[picturesque]

<p>Ha ha--Matt. Even I spelled matte wrong. Sorry.</p>

[pete_s.]

<p>There are <strong>two components</strong> to every reflection - <strong>diffused reflection</strong> that will scatter light in all directions and <strong>specular reflection</strong> that will reflect according to the angle of incidence.</p>

<p>Our eyes and mind actually sees what kind of surface something has (glossy, matte) by looking at the type of reflection.</p>

<p>A perfect mirror would have 100% specular (direct) reflection and 0% diffuse reflection.<br /> A perfect matte surface would have 0% specular reflection and 100% diffuse reflection.</p>

<p>Ceilings are usually more matte than glossy IME.</p>

<p>When you bounce backwards <strong>the diffuse part of the reflection will go in every direction including forward</strong> while the specular part of the reflection will only reflect and continue backwards.</p>

<p>So that's why bouncing backwards works but only a part of all light will be bouncing forward.<br>

The take home message is that <strong>light doesn't bounce like a ball</strong>.</p>

<p> </p>

[picturesque]

<p>Yes, Pete, which is why in backward ceiling bounce, part of the light still goes forward, even though most of it goes backward, following the angle of incidence rule. I would say light does bounce like a ball, but depending upon what kind of surfaces it has 'met' on the way, with more or less cohesiveness.</p>

[rodeo_joe1]

<p>With a perfectly matte (Sorry Matt! Please accept my flat-out apology.) surface - look up "Lambertian" reflection or scattering - the light goes in every direction equally. It does NOT go more forward than back, and the angle of incidence is therefore completely irrelevant.</p>

<p>This is very easy to see and prove. Just <strong>look </strong>at how the light reflects from a matte white surface, for example a patch of sunlight hitting a white emulsioned wall. Does the patch of light vary noticeably in brightness as you walk around it and look at it from various angles? No? And if you don't believe your eyes, take some meter readings. Some other surfaces, apart from matte white paint, that closely imitate a true Lambertian reflector are white paper and white cotton or linen. There will be a small specular component, but it only affects or reflects a tiny percentage of the light.</p>

[rodeo_joe1]

<p>Actually, I'll save you all the bother of checking for yourselves and, I suspect, save myself the bother of an extended argument.</p>

<p>Attached are two side-by-side pictures of a flash hitting a matte painted and slightly textured ceiling. The flashgun used was a Canon 540EZ set on manual at 1/128th power, sat atop a lighting stand and angled at around 45 degrees to the ceiling - or maybe a little less than 45 degrees. The lens aperture remained the same at f/5.6 for both pictures, the only difference being that one shot was taken from behind the flash; i.e. as if the flash was pointed behind the camera, and the other shot was taken from directly in front of the flash.</p>

<p>You'll see that the apparent brightness of the pool of light is almost identical in both cases. In fact the "rear" facing flash looks slightly brighter to me, but the camera histogram showed practically the same peak brightness and distribution of brightness. Looks like Mr Lambert was right after all.</p><div></div>

[picturesque]

<p>Rodeo, I don't see how photographing the beam of light itself proves that angle of incidence is irrelevent. I don't intend to give you an extended argument either, because I don't have to prove my theory is right or wrong--I just know what I have observed. I assume the OP can figure that one out for himself, by researching the theory and trying things out for himself.</p>

<p>I will say though that what Pete said above is what I also believe--however, a bounced beam of light off a matte surface does not instantly become so scattered that it loses it's ability to maintain some of it's direction entirely. There are <strong>some</strong> instances where angle of incidence does not come into play--much. I still believe it is not totally gone however, and plays more of a role depending upon other factors. This is why I believe angle of incidence is not irrelevant.</p>

[tom_mann1]

<p>Rodeo is using his camera as a light meter. He did this by effectively setting both his camera and his flash on manual. That way, if one of the two images he presented had been brighter, he can be quite confident that more light was headed from the ceiling in the direction of his camera. If I understand his description correctly, in his "backward" case, both camera and flash were on the same side as the illuminated patch of ceiling, whereas in his "forward" case, his camera was on the opposite side of the illuminated patch of ceiling, but in both cases, the angle to the ceiling was about 45 degrees. Thus, if there had been a lot of specular reflection from the ceiling, his "forward" image would have been much brighter and/shown a hot spot. It didn't.</p>

<p>That being said, Nadine also makes a very good point: Most surfaces are neither purely specular (ie, not perfect mirrors), nor are they purely Lambertian -- they are somewhere in-between. This is well known and is known as a quasi-Lambertian surface. This is the difference between paint finishes such as flat, satin, semi-gloss, high gloss, etc., except stated in slightly more scientific language.</p>

<p>HTH,</p>

<p>Tom M</p>

<p>PS - FWIW, there is a nice, but altogether too short article on such distinctions in Wikipedia: http://en.wikipedia.org/wiki/Lambertian_reflectance. In addition, the results of some carefully done experiments on this topic can be found here: http://www.4physics.com/tn3/lambertian.htm - scroll about half way down the page.</p>

<p> </p>

[steve_may1]

<p>Some very helpful responses here, thanks very much</p>

[lachaine]

<p>Simple answer:</p>

<p>The light has to bounce back in the general direction of the subject. If there is nothing for it to bounce off of, you don't get bounce flash.</p>

[rodeo_joe1]

<blockquote>

<p>.. a bounced beam of light off a matte surface does not instantly become so scattered that it loses it's ability to maintain some of it's direction entirely.</p>

</blockquote>

<p>When does it become scattered. Seconds later? Minutes later?<br>

Joking aside, light certainly does become instantly scattered in all directions when it hits a fully matte surface. Otherwise there would be hardly any point in using bounce flash at all. We use "bouncing" to diffuse light and thereby create softer lighting. If light didn't diffuse in all directions from the bounced surface, then it wouldn't really be soft light would it?</p>

<p>Nadine, you're welcome to believe your own "theory", which is actually just an opinion. Just don't inflict it on everyone else until you've got something, or some accepted authority to back it up with. Like some pictures illustrating how your particular type of diffuse reflection works maybe?</p>

[picturesque]

<p>Rodeo--I never said my 'theory' is anything anyone else has to believe, or that it wasn't an opinion. I certainly am not inflicting it on anyone. Therefore, I do not feel the need to prove anything.</p>

<p>As stated--the OP can figure it out for himself. You can believe what you want to believe. I can believe what I want to believe.</p>

<p>I will say though, that softness of light is not a result of diffusion, but of the relation of the size of the light source to the subject. I do know that diffusion and softness of light are two different things.</p>

[frank_skomial]

<p>Nadine explained in her next statement to the quoted one out of context, what "some instances" she means. She clearly did not mean instance of time, even though it looked like, if taken out of context.</p>

<p>Recent scientific discoveries that the speed of light is not the fastest speed out there in the universe, undermines Einstein's theory. This could also undermine our understanding of photography...:)</p>

[tom_mann1]

<p>1. <em>"Recent scientific discoveries that the speed of light is not the fastest speed out there in the universe, undermines Einstein's theory..."</em></p>

<p>Frank, I hope the smiley face at the end of that paragraph also was intended to apply to the sentence I quoted above. If it was, please just ignore the next paragraph -- I just didn't want any misunderstanding inadvertently spread.</p>

<p>OTOH, if you were even half serious about Relativity being "undermined", I suggest you look at the literature on this subject more carefully. Even the group which originally reported their experiment in this area essentially pleaded for other groups to try to reproduce their result. The jury is out, and will be out for a long time on this. A hot-off-the-press initial report on one experiment (albeit a large, complicated one) doesn't "undermine" a theory that has been validated by tens of thousands of experiments and a century of work. Scientists are cautious about using words like "undermine" for very good reason.</p>

<p>2. The difference between Rodeo and Nadine appears to be arising almost entirely because of a difference in their definitions of "diffusion of light". Any source of light that is not a point source and is not perfectly collimated (ie, all going in the same direction) possesses the two qualities I just referred to, ie, point source vs extended source, and collimated vs uncollimated. </p>

<p>Light bouncing around in a room can start out being nearly a point source and fairly well collimated (say, from a hot-shoe flash zoomed out to 200 mm), but the bouncing makes it become an extended, non-collimated source. It is common among physicists and engineers to call this process of multiple bounces "diffusion", exactly as Rodeo did. Nadine, OTOH, is using photographers' terminology in which she correctly distinguishes between say a completely diffused small source (eg, a Lumisphere -- the light is heading out from it in all directions, but the source is small, so the effect on a photograph is not soft), and the fact that one needs an extended source to wrap around a subject to produce "softness". </p>

<p>There's no need for anyone to get hot under the collar about this because, to me, it's clear that they both are very familiar with this subject but are focusing on different aspects of it and are simply using slightly different terminology. </p>

<p>HTH,</p>

<p>Tom M</p>