Home · Register · Software · Software · Join Upload & Sell

Moderated by: Fred Miranda

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist

FM Forums | Lighting & Studio Techniques | Join Upload & Sell


Archive 2013 · Another DIY Diffuser (Mk3)
• • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · Another DIY Diffuser (Mk3)

I'm always on the lookout for DIY diffuser materials and today I found some table place mats which are black on one side and matte silver with a textured pattern on the other. I used them to create a "Mark3" iteration of my diffuser. I don't sell them, sharing my experiments is just my way of encouraging others to try DIY and better understand the cause and effect of how speedlight diffusers work. The materials are shown below, except for the roll of 2" wide velcro used for attaching.


The white mesh on the right is used for cross-stitch projects. I used it in my Mk1 design between the sheets of foam as a stiffener. I decided to use the full size placemat but it flopped over. To prevent that I first tried adding wires, but then got the idea to use the the strip mesh to hold it up. It's just stapled to the mat on top and wraps around the flash to form a tunnel. Here's how it looks lying flat:


Here's a comparison with my "Mark2" design. It is made out of a plastic cutting board sheet:
It was simpler to construct than the earlier foam design but I wasn't able to find white cutting sheets at the time and had to paint translucent ones. The paint has started flake off, which is why I had my eye out for new material.

Here's a comparison with the "Mark1" design made out of "fun foam" and the earlier "Mark0" design out of mat board I created around 10 years ago:

In 2003 when I started making the DIY diffusers had been shooting for nearly 30 years with a pair of Vivitars using a bracket and off camera flash without any diffusers and getting results I considered acceptable for candid speedlight shots. I only started experimenting with diffusers I created out of mat board and tape because I thought they were pretty worthless and someone on another form posted a template from a popular commercial model which made it easy to copy out of mat board. I wanted to see what the differences were vs. my baseline of direct flash. I discovered not much if compared outdoors at night, which is what I suspected and why I never spent $20 on a commercial diffuser.

The reason there wasn't a lot of difference in my results was because the diffuser didn't make the source much bigger and using a bracket I didn't see many shadows on the faces with direct flash. If you don't see any shadows there's no compelling need to make the shadows lighter and the edges fuzzier is there? That's why I suggest those wanting to improve single flash shots to buy a bracket.

The main benefit I saw vs. direct flash came indoors from the fact the flash tilted 45 degrees bounced more light off the ceiling than directly changing the modeling vector from eye level to downward and natural. But I was already getting that modeling vector with the bracket and compared to just bouncing the only benefit of the small diffuser was putting catchlights in the eyes a bounce shot lacks.

Since ceiling bounce isn't efficient and because I use two flashes in a OCF key over bracket-centered fill configuration I modified the commercial clone to make it more "bowl" shaped and bounce more light forward, winding up after several prototypes with the Mark0 mat board creation above.

What is different in concept from caps and scoops designed for single flash is that I wanted most of the light bounced forward, which is more efficient. I didn't need ceiling bounce for modeling, I had the bracket for that.

Since day one using flash I've always created a downward modeling angle in single flash shots with a bracket, and with dual flash shots by placing the OCF above the heads of the subjects. One of the benefits of adding the diffuser to the flash on the bracket was that with the flash head aimed straight the center of the diffuser was about 6" higher and created more natural modeling than the direct flash on the bracket.

I shared my early experiments on another forum and a guy seeing my Mark0 design created one using "fun foam" sheets, a material I'd never seen before. It solved the problem of storage by folding flat. But when I tried it I found the foam was too flexible and the "scoop" too tall and narrow. I solved that problem by putting the plastic mesh used for cross-stitch projects between the two sheets of foam. A also made the top flap bigger and folded it down more to more light went forward vs. up.

Folding top flap fold over to cover the flash head produced a pattern more like a soft box, but I also had the option to open it and bounce most light off the ceiling creating results similar to a StoFen when a ceiling was available .

But I found that with the top of diffusers on the bracket a foot or less from an 8' ceiling the footprint overlapped and created "spill fill off the ceiling and walls. What I got with my mostly forward design was a direct "key" vector that modeled the face directly from the diffuser like a soft box does and "wrap-around" fill from the spill. That's illustrated in this single flash shot with a bracket and foam diffuser:

It's easier to see the "key" vector pattern created with the direct light from the diffuser when it is blurred:
That "mask" of highlights created by the downward angle of the light is how our brains recognize the 3D shape of the face. The more the dominant vector of the flash modeling matches angle of natural light the more "normal" the flash lighting looks on the face. That's why I advise people the best investment for making single flash shots more flattering isn't a diffuser it's a $50 Stroboframe camera flip bracket to get the flash above the head of their subjects.

What is also important to understand when using speedlight diffusers is that the apparent "softness" in the lighting isn't going to come from "wrapping" the key light unless you want to drag around a big umbrella. If you want to stay mobile you should learn how to create that downward modeling vector AND a lot of spill fill at the same time with one flash. If you do use that big umbrella what will create the light shadows on the far side of the face opposite it? Not wrap. The big ass umbrella will spill 10x more light around the room than any of the diffusers pictured here. That's good, but not practical for "run and gun" PJ style shooting.

Here are some shots which illustrate the cause and effect of using a bracket with single flash and what happens when it is used as fill with an OCF "key" light to the side. Friends stopped by with their 4 year-old who was usually a fussy and whiny. But that day she was wearing her nice dance outfit and acting like a diva. Having one the camera and single flash on the bracket w. foam diffuser handy I put her in a white chair and grabbed these two quick snaps:
They show the cause and effect of using the bracket. Raising the flash straight up hides most of the shadows and those that are seen like under the chin in the first shot while dark are naturally places and not really noticed much. The contrast of the shadow under the chin helps to reveal it's shape. It's just not very flattering because it's unfilled and too dark even with the "spill fill" bouncing off the light walls and the chair. There wasn't much ceiling bounce because it's 12' in that room In the second close-up I was so close and the flash so high relative to the face the upper lip is shading the teeth but the angle hides the dark shadow the chin was also casting.

Seeing she was in a mood to cooperate I ran and grabbed the slave flash on the stand and flipped the switch on the 580ex in the bracket to Master and dialed in a A:B =1:1 ratio. That's equal INCIDENT key (B) and fill (A):
Same modifers at a similar distance with the same "spill fill" factor. Why does the lighting look "softer" preceptually? Not because the key modifier is wrapping the face but because now the flash on the bracket is acting as fill and eliminating the dark unflattering unfilled shadows.

I know this defies conventional wisdom that you need a big key light modifier to get "soft" light, but consider that in the 40's everyone used direct fresnel lighting for stills and movies and the lighting was made to look very soft, or hard with those same direct sources. How? Those guys understood how to use fill. So did the guy who taught me to use flash 40 years ago because he started with flash bulbs in a small reflector on Speed Graphic and direct studio lighting.

Perceptually "flat" light and "soft" light have the same shadow clues (very subtle). This shot with a single flash on bracket with diffuser had very little "spill fill" because the ceiling was high and dark.
It doesn't look "hard" because the vector of the light isn't very high and creating any unfilled hard shadows you see in the foreground.

What did the diffuser add to that shot over direct flash on a bracket? The flash is a bit higher and the slightly larger size created larger more appealing catchlight refections and larger less specular reflections off the skin. The specularity that is seen doesn't look odd as in a hot shoe mounted flash in portrait mode because raising the source on the bracket + diffuser moves the reflections up on the cheekbones to were natural light puts them when it models the face. Where the highlights fall on the the face are also clue the brain using to discern 3D shape in a photo. Absent many shadows as in the shot above the placement of the highlights on the face are the primary clue the brain uses to determine 3D shape in a 2D photographic rendering.

The take away here is to realize any small speedlight modifer alone is not going to put the highlight clues in the "right" (i.e., natural looking places) on a face unless it is also raised above the face. The small diffuser also isn't going to create "wrap" directly. The "wrapping" which makes the shadows lighter and softer looking comes from all the indirect "spill fill" vectors off the walls and ceiling from the parts of the light's footprint that isn't hitting directly. The modifier splits one source into a dominant "key" vector and lots of different "fill" vectors hitting the shadows from different directions because part of the footprint hits ceiling and walls.

A gradient from highlight to shadow is created by light from the source falling off with distance. If you aim a "flat" eye level flash at a face the nose will be lighter than the ears not because the light "wraps" but because it falls of in intensity front>back on across the side of the head. If the flash is moved closer to the nose with the same modifier the gradient gets steeper because it falls off nose> ears at a greater rate. Went exposure is adjusted to keep the highlights on the nose the same the shadows will appear darker than before. What will make those shadows lighter and the overall gradient more gradual is when fill from a different direction reaches the ears. To do that it need to come from a different direction, which is why bouncing light around the room with a speedlight modifier is a good thing. But so is controlling that cause and effect.

StoFen and Lightsphere are excellent at creating soft lighting in a small indoor room. So soft in can make a normal room look as flat and 2D as an overcast day outdoors. That's soft lighting but is it always the look you want in a photo?

The reason I use two flashes isn't to get more light it's to control where the light comes from and the apparent softness via the lighting ratio. That's what I learned from Zucker many years ago. Put a direct flash 45 degrees to the side and 45 degrees above the eye line on a face and it will be natural, as natural as a face in direct sun at 2PM. To make the lighting with direct flash natural and FLATTERING requires fill to lighten the dark shadows which aren't unflattering because they are poorly placed but because they are simply dark.

Progressively add fill from the camera and the shadows get lighter and perceptually it looks "softer" because the overall highlight:shadow gradient isn't as steep. The fill falling off from nose > ears creates it's own gradient over the cheeks via front > back, nose> ear fall off. The shadow on the nose will be lighter than the ear further from the fill source, if the fill source is over the camera closer to the nose.

Why don't more photographers understand that? Most wouldn't think to put fill anywhere near the camera because they never saw good results with a flash placed there. They probably never used a flash bracket either.

With a diffuser on single flash with a diffuser on the camera hot shoe most of the spill fill comes from the direction of the camera because that's where the light source creating the spill is. If you walked behind the subject you'd see spill fill on the back of the head too from the back wall, but not as much, unless your subject is 6" from the wall like my co-worker Ray was in the shot above. The cause and effect of using the flash on bracket as will when a second is used off camera is similar.

What I get when using two flashes is the ability to move my key light off axis to create more shadow modeling while having even more control of the fill and shadows than when bounce one. As much or more light is bouncing off the walls and ceiling as spill fill but I can control the direct front> back gradient of the fill on the face with the power and distance of my flash on the bracket.

The diffusers? They have less to do with the results than the fact the key light vector models in a naturally flattering way and the fill strategies, direct and bounced, keep the shadows the key light vectors don't hit light in tone.

What's the difference between my Mark0 to Mark 3 designs? In terms of cause and effect not much. When one flash is used all the softening effect seen in the shadows comes from the spill fill with all of them. What the bigger sizes does is create a bigger footprint, which creates more spill fill indoors. Outdoors? The bigger size will just create bigger catchlights with little little change seen in the shadows because there will be no spill fill.

I switched from Mk0 to Mk1 for logistics - being able to fold it flat and stick it on my bag. I tried the foam and it worked better but doesn't hold up well.

I switched from Mk1 to Mk2 because my Mk1's were worn out and the plastic sheets were lighter and simpler to construct (once painted) and not as fragile. But the larger size is less practical. They don't fit in my shoulder bag.

I've switched to the Mk3 design to eliminate the painting step, but had to add the mesh because the table mats are not as stiff. It remains to be seen if I stick with them or go back to the Mk2. I also use still the Mk1 design because they fit into my camera bag on top of the camera and lenses.

How did I find out which works best and which is most practical? By experimenting with different designs, what I encourage everyone to try by posting threads like this

Jan 27, 2013 at 10:20 PM
• • • • •
Upload & Sell: On
p.1 #2 · p.1 #2 · Another DIY Diffuser (Mk3)

I admire your DIY gumption. I would call that a bouncer, not a diffuser though.

Jan 28, 2013 at 02:02 AM
• • • • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · Another DIY Diffuser (Mk3)

ChrisDM wrote:
I admire your DIY gumption. I would call that a bouncer, not a diffuser though.

Call it what you like, just understand how it works and where it bounces the light.

It works indoors by bouncing a lot of light around the room making the character of the light hitting the subject more diffuse (i.e., hitting from more different directions and intensities) than with direct bare flash. But the differences between my Mk0 and Mk3 aren't that different because the direct and bounced vectors are similar.

What affects the results seen in the photos isn't so much how collimated or diffuse the light rays are when they leave a source but rather the vectors of light that wind up hitting the subject. The environment the source is used affects what the footprint hits and that affects diffusion relative to the subject.

Would you call a softbox a diffuser? A small deep one used close to a subject without any of it's footprint creating spill fill will create distinct dark shadows. Keep the size the same but make it shallow. Making it shallow will make the footprint bigger. Indoors it will bounce off the room and create spill fill vectors to fill the shadows. The net effect on the subject? More diffuse lighting. Outdoors the same vectors of light reach the subject as the deep one, but fewer photons because the less of the total footprint hits the subject.

A "bouncer" will produce the same light on the subject as a your definition of "diffuser" if the direct and bounced vectors wind up hitting the subject at similar angles and intensities. The trick in using either effectively is understanding how the key and fill vectors it creates hit the subject and controlling the ratio between them predictably. With speedlight diffusers because the environment constantly changes it becomes more of a situational variable than using a set of lights in the same studio space all the time.

A pro with a huge studio with little or no "spill fill" will get different a different character of light on subject than a hobbyist with the same lights and modifier using them in their living room. The pro would actually need larger sources and more fill light intensity to make up for the lack of spill but still wouldn't be able to match the look of two shoot through umbrellas bouncing light around the living room and lighting the wife like an overcast day

Jan 28, 2013 at 03:21 AM
Zenon Char
• • • •
Upload & Sell: On
p.1 #4 · p.1 #4 · Another DIY Diffuser (Mk3)

Very cool Chuck.

Jan 28, 2013 at 04:46 AM

Search in Used Dept. 


• • • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · Another DIY Diffuser (Mk3)

ChrisDM wrote:
I admire your DIY gumption. I would call that a bouncer, not a diffuser though.

I'd call it a bounce diffuser.

It not only redirects the light, it also enlarges it; and light-source size relative to the subject is what cause a source to be classified as diffuse or point-source.

Jan 28, 2013 at 04:50 AM
• • • • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · Another DIY Diffuser (Mk3)

Part of the problem with terminology, "terms of art", and conventions in any craft is they get into common usage without any understanding of the underlying cause and effect.

For example any flash on camera is referred to as most as "fill" flash, especially outdoors. But in terms of actual cause and effect it actually the "key" light on the foreground because it, not the ambient, create the highlights. The distinction between "key" and "fill" isn't the position of the two lights but which creates the foundation of shadows and which creates the highlights on top of those shadow values.

A "diffuser", precisely, is a physical object placed in front of a light source to scatter it's rays. Sun / overcast sky is the original model. God said "let there be light" and created the Sun. Adam looked pretty good in the Sun because he was a stud dude. But when God took the rib and created Eve she looked down and said, "Gee that hard light isn't flattering Eve at all" and then created overcast to diffuse the light.

The sun isn't a point source, but in terms of relative size from anywhere on earth it looks and acts like one from the POV of the observer . When the sky gets overcast the size of the apparent source gets bigger. The actual cause and effect underlying the more diffuse light overcast produces it that instead of having one strong vector from a single direction it comes more or less equally from all directions. On an overcast day when an incident meter is used dome down in single source measurement mode there's only about 1/2 stop difference seen between pointing at where the sun is behind the clouds and the opposite direction. On a clear day that comparison produces a 3 stop difference between sun and skylight opposite.

Even on a clear day the single source sun creates more than one light vector. The direct rays of the sun that reach the earth are parallel (from the POV of the observer) and created distinct shadows. But we are able to see detail in those shadow because that same single source bounces light off water molecules and other stuff in the atmosphere creating many different secondary "fill" vectors.

But wait.... as they say on the Slicer commercial... If you put a face in open shade on the side of the building you'll usually see the eyes are shaded. Shadows indicate that one vector of light is more dominant than others. The dominant vector in photographic lighting parlance is the "key" source creating the modeling over the "fill" which is revealing the shadow detail.

In open shade the modeling result from the fact skylight is stronger from directly above than the sides. Like flash used on camera many think of the shaded side of a backlit subject as being it "flat" light but in terms of cause and effect there are "key" modeling and "fill" vectors in the skylight: stronger key vector from above and "wrap-around" fill vectors from all directions.

In open shade photographers using ambient light alone will consciously see the modeling clues because they are obvious as when posing a person by a window indoors. But it backlight the contrast overwhelms perception so much the subtle clues are difficult to see and photographers are more likely to miss the important step of turning the face to the light to highlight the from of it and raising the eye sockets up to the light as they would by a window's "key" lighting. As a result of that blissful ignorance of how the skylight is shading the eyes they wind up with shaded eyes when adding flash to backlit subjects.

Indoors with high ISO capable cameras of today flash often isn't necessary for correct exposure as it was with film, but because the direction of the ambient KEY vector is shading the eyes. It's the same problem as open shade or backit subject outdoors but indoors it is possible to cancel out the ambient with shutter speed because it is not as strong as outdoors.

Bounced flash indoors is similar in cause and effect to open shade if the dominant vector is the one coming off the ceiling. How "diffuse" and "soft" the light looks depends on how many different "wrap-around" fill vectors the same source creates from the sides or off the floor. Like open shade if the angle of the light is too steep relative to the faces when it comes down off the ceiling the brows will shade the eyes.

Many seeing dark eye sockets indoors or out think, "Hmm eyes are dark, I need to add fill flash." But in terms of actual cause and effect any flash added to lighten the eyes will also hit the cheeks below changing both by the same amount. The eyes will be lighter, but compared to the highlights on the cheeks they will remain darker. What I see in a lot of shots that the photographer does not realize is dark eyes raised with "fill flash" to the point the eyes look "normal" with blown red channel in the skin highlights giving them an odd yellowish waxy look because all the detail in the red channel is lost.

The ONLY solution to get THE SAME light in eyes and cheek is recognizing the source of the problem is not a lack of fill in the eyes it's a lack of key lighting and raising the face into the key light so it will get past the brow into the eyes. But that cause and effect does not occur to many beginning photographers judging from the number of portraits in shade I see with dull dead looking eye sockets darker than cheeks.

I see the same thing in many indoor flash shots. The flash creates catchlight reflections that give them sparkle but the brow shades both the ambient and the key vector of the flash creating the highlights and the orbits surrounding the catchlight remains dark because they are only illuminated by whatever "spill fill" was bouncing around the room.

The cause and effect of shaded eyes and adding "fill" flash is more easily seen outdoors on a face is cross lit by direct sun. The incident ratio on cheek and shaded eyes is about 8:1 because direct sun is 3 stops brighter than the skylight creating the fill. The reflected ratio is usually less because the cheeks act like a reflector bouncing fill into the eyes (why athletes put those black strips under their eyes to avoid glare).

8:1 Incident ratio cheek:eye sockets

What happens when flash 1-stop under the intensity of the sun (4 units) is added?

12:5 = 2:1 rounded to the nearest integer

The flash changes the contrast of the overall highlight:shadow pattern to make the shadows lighter and "softer" / "flatter" (dimensionally) but the eyes will still be darker than the cheeks. Why? that variable is controlled by the orientation of the face and eyes to the key light, not the contrast of the lighting.

The only way to get highlights in the eyes is to get them into the KEY vector of the lighting. That boys and girls is why it's important to develop situational awareness of both the ambient and flash lighting in terms of the vectors of key and fill it creates.

First thing I do entering a room is look at the eyes of the people in it. Is the ambient source in the room shading them? If so I know I'll need to cancel it out nearly completely with shutter then light the space and face with flash to avoid the same shaded eyes in my photo.

Outdoors shooting a portrait in open shade I do the same thing, look at the eyes, but I can't cancel out the ambient with the shutter. So outdoors I need to find a way to get the faces looking up into the sky to get the light in the eyes. That's why I bring along a ladder on an outdoor portrait session. If shooting candids I look for a rock, bench or some other higher POV knowing the subjects will look up at me to see what I'm doing up there at some point.

If shooting in ambient lighting indoors without flash I do the same thing for the same reason; stand on a chair and catch the faces looking up to get light past the brow into the eyes. Not any light the KEY VECTOR of the light.

What will determine how "diffuse" the lighting is holistically (what affects what it hits)? The ratio of KEY to FILL intensity (the relative amplitude of the vectors) and the number of different directions the KEY and FILL vectors come from.

Nearly all discussions of diffusion / diffuser I read here focus only on whether the KEY vectors are collimated or diffuse (distributed outward from source) but what influences the character of the lighting more holistically are the FILL vectors: where they come from relative to the object or face and how strong they are relative to key vectors.

A good example of this is stadium lighting. All the sources are small but they are equal intensity and ring the field so the net effect holistically is very dimensionally flat modeling. If the lights on one side were twice the intensity of the lights on the other the brighter side would create the key vector the other the fill.

When using speedlight diffusers indoors they will all bounce some of the light off a low ceiling. But whether that vector off the ceiling acts as the key or fill vector varies situationally with the height of the ceiling, and how much light device (call it what you like) projects or bounces forward.

If ceiling bounce > direct you get natural looking modeling because natural key vector comes from overhead and key vector comes from the ceiling.

If ceiling bounce < direct the direct vector becomes the "key" and the light off the ceiling the "fill". The forward vector from the flash is "Key" because it creates the highlights, but it doesn't look "right" (natural) because the angle is unnaturally low and it overpowers the "Fill" bouncing off the ceiling. That describes the cause and effect of what you see when speedlight diffusers are compared with direct flash outdoor. The flash can't create the natural downward vector outdoor by bounce, it must be physically raised over the head of the subject to create it on a bracket or stand.

You'll understand this better when you start to think of lighting holistically in terms of key and fill vectors from the POV of the face it is hitting rather than what directions the light exits the source.

Edited on Jan 28, 2013 at 01:53 PM · View previous versions

Jan 28, 2013 at 01:36 PM

Upload & Sell: Off
p.1 #7 · p.1 #7 · Another DIY Diffuser (Mk3)

Chuck, I'm always learning from you, thanks.


Feb 16, 2013 at 12:09 AM
• •
Upload & Sell: Off
p.1 #8 · p.1 #8 · Another DIY Diffuser (Mk3)

GGcat wrote:
Chuck, I'm always learning from you, thanks.


+1, Thanks you Mr. Chuck.

Feb 16, 2013 at 01:55 AM

FM Forums | Lighting & Studio Techniques | Join Upload & Sell


You are not logged in. Login or Register

Username     Reset password