Register · Search · Software · Join Upload & Sell · Hosting

Moderated by: Fred Miranda
Username   Password

FM Forum Rules
FM Forums | Lighting & Studio Techniques | Join Upload & Sell   
Search Used
1
2 end
  

Archive 2008 · Backdrop lights bleeding?
  
 
Craig Yannuzzi
Offline
Upload & Sell: On
p.1 #1 · Backdrop lights bleeding?


Had my niece and nephew over this weekend for some pics and noticed that some of the shots seemed "hazy". Or at least thats how they looked to me. I've posted a shot to illustrate what some pictures looked like straight out of the camera. I think the washed out look of the subject is due to the spill over from my background lights?? Is it some sort of flair?

Can someone explain what is going on here? Are my eyes playing tricks on me?

How can I adjust my strobes to eliminate this effect?

Thanks!

Pic made the cutting room floor and is posted for illustrative purposes only














Dec 15, 2008 at 12:39 AM
k7xd
Offline
Upload & Sell: Off
p.1 #2 · Backdrop lights bleeding?


It does not look that bad to me.

The BG being highly reflective has a tendency to cause lens flare,
lowering contrast. Keep the BG lights down to 1 stop or so
above the key.



Dec 15, 2008 at 01:52 AM
Paratima
Offline
Upload & Sell: On
p.1 #3 · Backdrop lights bleeding?


The catchlights in her eyes show a couple of strobes and she has a very "flat" look. k7xd's suggestion is good, but maybe also let a little shadow in to show some depth? Just an idea.

Dec 15, 2008 at 02:06 AM
Photon
Online
Dedicated FM
Upload & Sell: On
p.1 #4 · Backdrop lights bleeding?


Paratima wrote:
The catchlights in her eyes show a couple of strobes and she has a very "flat" look. k7xd's suggestion is good, but maybe also let a little shadow in to show some depth? Just an idea.

I agree. Judging by the amount of spill on her shoulders and through hair, I don't think the background light is too intense, just a good true white. A little direction to your main light will give shape to the face and texture to the clothing. Try just enough fill to keep the shadows from looking too strong for your taste.

Dec 15, 2008 at 02:36 AM
Gregg Heckler
Offline
Upload & Sell: Off
p.1 #5 · Backdrop lights bleeding?


Keep in mind your BG lights aren't exposing the front of you subject. Your background light was fine. It's your key and fill that's the problem (or maybe your concern). It's pretty flat as evidenced by the catchlights in her eyes. Both lights are a little low and the ratio is pretty even causing the image to look a little flat. It may even be a tiny bit over exposed. This lighting is OK if you're copying a page from a book, but 3 dimensional objects usually looking better with a little more modeling (ratio).

Dec 15, 2008 at 06:39 AM
Craig Yannuzzi
Offline
Upload & Sell: On
p.1 #6 · Backdrop lights bleeding?


Thanks guys for the responses!

As to the flatness of the light, what changes might I make?

I'm assuming placement of the lights needs to change? Here the subject was cross lit with the key (medium octabox) at camera left and the fill (medium softbox) at camera right. Each at 45 degrees to the subject and set between 1/2 and 1/4 power for the key and between 1/4 and 1/8 power for the fill.

That's a one stop difference, correct?

Was the difference between the key not enough to bring in some modeling?

The lights were set a little low and directed straight at the subjects face so I'm assuming that also had an effect on the flatness.

Thanks!

Dec 15, 2008 at 12:26 PM
cgardner
Offline
Dedicated FM
Upload & Sell: Off
p.1 #7 · Backdrop lights bleeding?


Your background is clipped to 255. 255 = pure light. Is your background made out of pure light?

We shoot on plain smooth backgrounds so the brain will tune it out and not be distracted by it. The background will look just as white perceptually, and tune it out just as quickly if exposed correctly in the technical sense, placing its tonal value around 250 in a 24-bit (8/color) JPG file.



This image is copyrighted by the owner




When setting exposure I use a target like the one below with a textured white towel, which unlike smooth targets will reveal visually when detail is starting to be lost in the highlights. Preserving that detail is what defines a correct exposure. The MacBeth color target is useful for judging the effects of flare on saturation of color.


This image is copyrighted by the owner



I'll start with the background lighting off, as above and adjust the foreground lighting until the brightest highlight on the foreground, such as the rim-light on the towel, is about 1/3 stop below clipping in the the over-exposure warning (OEW).

Why 1/3 stop? Because by doing a bracketed series of exposures previously and comparing what the camera OEW showed vs. when detail in the RAW file started to melt away I found the best exposure of the white towel - a proxy for highlight detail in clothing, eyes, teeth, etc. - was 1/3 stop below the point the camera playback started to show clipping.

Once I get the foreground zeroed in I raise the background lighting to the point where it starts to show clipping in the OEW, then back it down until the OEW warning disappears. The OEW reveals where clipping is occurring first so I can use it to guide light placement to make the lighting of the background as even as possible.



This image is copyrighted by the owner




The background isn't captured in the camera as "knock-out" white (255) but rather ideally somewhere around 250.250.250. Note in the shot above that the rim lighting of the towel is still visible? That's something you need to grasp about the perceptual aspects of lighting on white backgrounds. To make the white highlights on the foreground subjects visible and do their important perceptual job creating the illusion of 3D via contrast its necessary to make the "white" of the background a slightly darker shade of white. If you go back and compare the rim lighting on the towel in the first shot, before the background lights were even turned on you'll note that when the background is darker the stuff in the foreground actually seems brighter perceptually. That's because the brain, while it will ignore the background detail by virtue of the fact the eye in person would render it out of focus will key off the background tone to "anchor" the tonal scale.

All of this advice flies in the face of conventional wisdom which says to over-expose the background by a stop to knock it out completely. But consider that if you over-expose the background by a stop you will be bouncing twice as much flare light back into to lens than if you expose the background below clipping. You'll find your foreground subjects are more saturated... Try it both ways and compare the results.

With regard to lighting your subject in the foreground, use butterfly for full face views. Full face is a symmetrical view of the face, so you need a complementary symmetrical lighting pattern to create natural downward modeling. Put one light directly over the camera lens above eye level, but not so high the brow begins to shade the eyes and the fill source below the lens.



This image is copyrighted by the owner





This image is copyrighted by the owner





This image is copyrighted by the owner




That placement of the light will create a highlight down the center ridge of the nose and even shadows on either side, which creates the appearance of symmetry. If you cross lights horizontally they cancel each other and you get the opposite, a darker line or zone down the center of the face and unfilled areas where neither light reaches.

If you were to place one light to the side and the fill over the camera it would create a sideway shadow pattern which when overlapping the full face view would have the net effect of making the highlighted side bigger than the shadow side and the face look asymmetrical. The key-light-to-side "short" lighting pattern is more effective for oblique views on dark backgrounds...



This image is copyrighted by the owner




Compare how that lighting models the face, vs the "broad" lighting from the other side...



This image is copyrighted by the owner




But for an oblique view on white short lighting may not be effective. The goal is to make the face overall contrast with the background but also pull the attention of the viewer to the front of it. You can see in the examples above how the short lighting on the dark background does that by creating compelling contrast your eye can't resist. But put that same lighting pattern on a white background and the brightly lit far side of the face will disappear perceptually into the background. So instead the more effective strategy for oblique views on white backgrounds is broad lighting with light open shadows. Broad light will make the far side of the face slightly darker, allowing it to contrast well against the white background and an the same time make the side of the head facing the camera brighter and blend into the background perceptually. The net effect of the lighting pattering is a contrast gradient which pulls the attention of the viewer to the slightly darker far side: the same dynamic of eye movement as short lighting on a dark background.

Moving the eye of the viewer to what is most important with CONTRAST is what lighting patterns are all about. So when you are moving lights around try to understand the underlying goals of what the lights should be doing, to pull the viewer into the front of the face and model it in a way that looks slim and symmetrical with natural downward modeling of the nose and cheeks, no harsh shadows, and good light and a catchlight in both eyes.

The thing which isn't grasped immediately about white backgrounds is the fact that COLOR contrast of warm skin is often a stronger attractive than tonal contrast, especially if the subject is a blonde. If the subject has dark hair you'll notice the hair first, color contrast of the face second. If the subject is a blonde, it will be color contrast alone which pulls the eye into the face. That's why non-competing, non-distracting clothing is an important consideration on white backgrounds. Little girls are traditionally dressed in pink, but a pink dress will overpower the face. If the goal is to see the face use a white dress with a pink collar.

Click the WWW button and look in the TOC for tutorials on basic lighting configurations and why they work, and how to expose white background shots correctly perceptually.

Chuck

Edited on Dec 15, 2008 at 01:44 PM · View previous versions


Dec 15, 2008 at 01:27 PM
Craig Yannuzzi
Offline
Upload & Sell: On
p.1 #8 · Backdrop lights bleeding?


cgardner wrote:
Your background is clipped to 255. 255 = pure light. Is your background made out of pure light?

We shoot on plain smooth backgrounds so the brain will tune it out and not be distracted by it. The background will look just as white perceptually, and tune it out just as quickly if exposed correctly in the technical scene, placing its tonal value around 250 in a 24-bit (8/color) JPG file.



This image is copyrighted by the owner




When setting exposure I use a target like the one below with a textured white towel, which unlike smooth targets will reveal visually when detail is starting to be lost in the highlights. Preserving that detail is what defines a correct exposure. The MacBeth color target is useful for judging the effects of flare on saturation of color.


This image is copyrighted by the owner



I'll start with the background lighting off, as above and adjust the foreground lighting until the brightest highlight on the foreground, such as the rim-light on the towel, is about 1/3 stop below clipping in the the over-exposure warning (OEW).

Why 1/3 stop? Because by doing a bracketed series of exposures previously and comparing what the camera OEW showed vs. when detail in the RAW file started to melt away I found the best exposure of the white towel - a proxy for highlight detail in clothing, eyes, teeth, etc. - was 1/3 stop below the point the camera playback started to show clipping.

Once I get the foreground zeroed in I raise the background lighting to the point where it starts to show clipping in the OEW, then back it down until the OEW warning disappears. The OEW reveals where clipping is occurring first so I can use it to guide light placement to make the lighting of the background as even as possible.



This image is copyrighted by the owner




The background isn't captured in the camera as "knock-out" white (255) but rather ideally somewhere around 250.250.250. Note in the shot above that the rim lighting of the towel is still visible? That's something you need to grasp about the perceptual aspects of lighting on white backgrounds. To make the white highlights on the foreground subjects visible and do their important perceptual job creating the illusion of 3D via contrast its necessary to make the "white" of the background a slightly darker shade of white. If you go back and compare the rim lighting on the towel in the first shot, before the background lights were even turned on you'll note that when the background is darker the stuff in the foreground actually seems brighter perceptually. That's because the brain, while it will ignore the background detail by virtue of the fact the eye in person would render it out of focus will key off the background tone to "anchor" the tonal scale.

All of this advice flies in the face of conventional wisdom which says to over-expose the background by a stop to knock it out completely. But consider that if you over-expose the background by a stop you will be bouncing twice as much flare light back into to lens than if you expose the background below clipping. You'll find your foreground subjects are more saturated... Try it both ways and compare the results.

With regard to lighting your subject in the foreground, use butterfly for full face views. Full face is a symmetrical view of the face, so you need a complementary symmetrical lighting pattern to create natural downward modeling. Put one light directly over the camera lens above eye level, but not so high the brow begins to shade the eyes and the fill source below the lens.



This image is copyrighted by the owner





This image is copyrighted by the owner





This image is copyrighted by the owner




That placement of the light will create a highlight down the center ridge of the nose and even shadows on either side, which creates the appearance of symmetry. If you cross lights horizontally they cancel each other and you get the opposite, a darker line or zone down the center of the face and unfilled areas where neither light reaches.

If you were to place one light to the side and the fill over the camera it would create a sideway shadow pattern which when overlapping the full face view would have the net effect of making the highlighted side bigger than the shadow side and the face look asymmetrical. The key-light-to-side "short" lighting pattern is more effective for oblique views on dark backgrounds...



This image is copyrighted by the owner




Compare how that lighting models the face, vs the "broad" lighting from the other side...



This image is copyrighted by the owner




So when you are moving lights around try to understand the underlying goals of what the lights should be doing, to pull the viewer into the front of the face and model it in a way that looks slim and symmetrical with natural downward modeling of the nose and cheeks, no harsh shadows, and good light and a catchlight in both eyes.

Click the WWW button and look in the TOC for tutorials on basic lighting configurations and why they work, and how to expose white background shots correctly perceptually.

Chuck


As usual, great stuff Chuck! I knew I had read one of your posts before about trying to avoid 255-255-255 on the backdrop but couldn't find it. Guess I could have found it on your site.....

I'll experiment tonight with toning down the backdrop lights, as well as moving the key higher and directly inline with the camera and subject.

This shoot was a great learning experience given the ages (2 and 3) of these kids. Posing was out of the question, so I basically shot when the kids "orbits" around the basement took them in close enough proximity to the strobes and backdrop!

Thanks again!

Dec 15, 2008 at 01:36 PM
bobbyz
Offline
Upload & Sell: Off
p.1 #9 · Backdrop lights bleeding?


Craig Yannuzzi wrote:
Here the subject was cross lit with the key (medium octabox) at camera left and the fill (medium softbox) at camera right. Each at 45 degrees to the subject and set between 1/2 and 1/4 power for the key and between 1/4 and 1/8 power for the fill.

That's a one stop difference, correct?


I wouldn't go with what the lights were set at. Lot depends on distance between light and subjects, what modifiers you used on each and if you did some feathering or not. To get light ratios you need to use light meter.


Dec 15, 2008 at 03:18 PM
shatterkiss
Offline
Dedicated FM
Upload & Sell: Off
p.1 #10 · Backdrop lights bleeding?


bobbyz wrote:
I wouldn't go with what the lights were set at. Lot depends on distance between light and subjects, what modifiers you used on each and if you did some feathering or not. To get light ratios you need to use light meter.


Agreed. And it also sounds like you've fallen into the classic trap of setting up all your lights, then trying to figure out what each light should be set to.

You should be building your setups one light at a time, testing and metering with each new light and figuring out what you have and what you still need before adding the next light. In your case, I would have set up the background lights, metered them, made sure the exposure was good and the coverage was even across the compositions I might have been planning. Then I would have added the key light, positioned it and set the exposure, then looked at sample shots to see where I still had too much shadow and needed fill. At that point, if I'd been planning images where the subject was looking into the camera, rather than away from it, there's no way I would have left the key at a 45-deg angle - having it positioned high over the camera and just a teensy bit to one side or the other would probably have been much more flattering for a frontal angle on the face. And then I could evaluate whether or not I needed fill and where it needed to end up just by deciding whether or not I needed to fill in the shadows that were still there.

Photographers refer to "building lighting" for a reason - it's a process, one step at a time. If you rush straight to what you think the setup should be, then try to fix it from there, you'll only get lost in it.

Dec 15, 2008 at 04:10 PM
Craig Yannuzzi
Offline
Upload & Sell: On
p.1 #11 · Backdrop lights bleeding?


shatterkiss wrote:
bobbyz wrote:
I wouldn't go with what the lights were set at. Lot depends on distance between light and subjects, what modifiers you used on each and if you did some feathering or not. To get light ratios you need to use light meter.


Agreed. And it also sounds like you've fallen into the classic trap of setting up all your lights, then trying to figure out what each light should be set to.

You should be building your setups one light at a time, testing and metering with each new light and figuring out what you have and what you still need before adding the next light. In your case, I would have set up the background lights, metered them, made sure the exposure was good and the coverage was even across the compositions I might have been planning. Then I would have added the key light, positioned it and set the exposure, then looked at sample shots to see where I still had too much shadow and needed fill. At that point, if I'd been planning images where the subject was looking into the camera, rather than away from it, there's no way I would have left the key at a 45-deg angle - having it positioned high over the camera and just a teensy bit to one side or the other would probably have been much more flattering for a frontal angle on the face. And then I could evaluate whether or not I needed fill and where it needed to end up just by deciding whether or not I needed to fill in the shadows that were still there.

Photographers refer to "building lighting" for a reason - it's a process, one step at a time. If you rush straight to what you think the setup should be, then try to fix it from there, you'll only get lost in it.



Wow shatterkiss! Your evaluation of my process is dead on and will send me "back to the drawing board" tonight. Thanks for taking the time to comment.

When I meter the background lights am I doing so only as a reference so that when I set my key I'll be sure to be one stop under? Aside from assuring even coverage, is there a particular value I should be shooting for?


As always, Thank You!


Dec 15, 2008 at 05:38 PM
shatterkiss
Offline
Dedicated FM
Upload & Sell: Off
p.1 #12 · Backdrop lights bleeding?


Craig Yannuzzi wrote:
Wow shatterkiss! Your evaluation of my process is dead on and will send me "back to the drawing board" tonight. Thanks for taking the time to comment.


My pleasure! I will give you one caveat: the times when I'll build the whole lighting setup in one go before firing a single test frame or metering anything are the times when it's a setup I've done 50 times in the past and know the results of. A 3-light clamshell beauty/portrait setup, for instance: two lights directly in front of the subject with space between them for the camera to shoot through, one light directly behind the subject to cast a "halo" on the background. I know based on my modifiers and their distance from the subject that I can use equal power on both the front lights, so the only question is the power on the background light. But that comes from using that exact setup regularly: what I call my "bag of tricks". Every photographer has a series of lighting setups that they can resort to in a pinch that are safe, reliable and can be constructed in minutes without much thought. It's what you do when you've got five minutes to setup for a shot and five more minutes to get the shot.

When I meter the background lights am I doing so only as a reference so that when I set my key I'll be sure to be one stop under? Aside from assuring even coverage, is there a particular value I should be shooting for?

Assuming a white background, I'll set the background lights for whatever I want my camera exposure to be. If I'm hoping to shoot at f/8, then I'll set the background so an incident reading gives me f/8: that'll actually result in a reflective reading that would be 1-2 stops overexposed, but that helps to counter any unevenness in the light coverage. An incident reading of f/5.6 across the entire background should still give you even white.

At that point I'll adjust my foreground lighting so the key is reading f/8 as well. Depending on my desired look, my fill could meter at anything from f/2.8 to f/5.6 (if I'm using fill at all).

Remember that you don't NEED fill light. It's a matter of what kind of look you prefer and how you're dealing with your key light. Do you like defined shadow modeling and large, diffuse key sources? Then you could use a large white umbrella really close to your subject and close to the camera axis with no fill at all. The classical, conservative approach says you need fill to moderate shadows...the modern, creative approach says that you shouldn't assume anything at all and should light to create the images that you respond positively to, not the images you're told are correctly-lit. Platon's portraits are great examples of breaking the "rules", using key without fill to create strong images.

Dec 15, 2008 at 06:05 PM
DennisW1
Offline
Upload & Sell: On
p.1 #13 · Backdrop lights bleeding?


cgardner wrote:
Your background is clipped to 255. 255 = pure light. Is your background made out of pure light?

We shoot on plain smooth backgrounds so the brain will tune it out and not be distracted by it. The background will look just as white perceptually, and tune it out just as quickly if exposed correctly in the technical sense, placing its tonal value around 250 in a 24-bit (8/color) JPG file.



This image is copyrighted by the owner




When setting exposure I use a target like the one below with a textured white towel, which unlike smooth targets will reveal visually when detail is starting to be lost in the highlights. Preserving that detail is what defines a correct exposure. The MacBeth color target is useful for judging the effects of flare on saturation of color.


This image is copyrighted by the owner



I'll start with the background lighting off, as above and adjust the foreground lighting until the brightest highlight on the foreground, such as the rim-light on the towel, is about 1/3 stop below clipping in the the over-exposure warning (OEW).

Why 1/3 stop? Because by doing a bracketed series of exposures previously and comparing what the camera OEW showed vs. when detail in the RAW file started to melt away I found the best exposure of the white towel - a proxy for highlight detail in clothing, eyes, teeth, etc. - was 1/3 stop below the point the camera playback started to show clipping.

Once I get the foreground zeroed in I raise the background lighting to the point where it starts to show clipping in the OEW, then back it down until the OEW warning disappears. The OEW reveals where clipping is occurring first so I can use it to guide light placement to make the lighting of the background as even as possible.



This image is copyrighted by the owner




The background isn't captured in the camera as "knock-out" white (255) but rather ideally somewhere around 250.250.250. Note in the shot above that the rim lighting of the towel is still visible? That's something you need to grasp about the perceptual aspects of lighting on white backgrounds. To make the white highlights on the foreground subjects visible and do their important perceptual job creating the illusion of 3D via contrast its necessary to make the "white" of the background a slightly darker shade of white. If you go back and compare the rim lighting on the towel in the first shot, before the background lights were even turned on you'll note that when the background is darker the stuff in the foreground actually seems brighter perceptually. That's because the brain, while it will ignore the background detail by virtue of the fact the eye in person would render it out of focus will key off the background tone to "anchor" the tonal scale.

All of this advice flies in the face of conventional wisdom which says to over-expose the background by a stop to knock it out completely. But consider that if you over-expose the background by a stop you will be bouncing twice as much flare light back into to lens than if you expose the background below clipping. You'll find your foreground subjects are more saturated... Try it both ways and compare the results.

With regard to lighting your subject in the foreground, use butterfly for full face views. Full face is a symmetrical view of the face, so you need a complementary symmetrical lighting pattern to create natural downward modeling. Put one light directly over the camera lens above eye level, but not so high the brow begins to shade the eyes and the fill source below the lens.



This image is copyrighted by the owner





This image is copyrighted by the owner





This image is copyrighted by the owner




That placement of the light will create a highlight down the center ridge of the nose and even shadows on either side, which creates the appearance of symmetry. If you cross lights horizontally they cancel each other and you get the opposite, a darker line or zone down the center of the face and unfilled areas where neither light reaches.

If you were to place one light to the side and the fill over the camera it would create a sideway shadow pattern which when overlapping the full face view would have the net effect of making the highlighted side bigger than the shadow side and the face look asymmetrical. The key-light-to-side "short" lighting pattern is more effective for oblique views on dark backgrounds...



This image is copyrighted by the owner




Compare how that lighting models the face, vs the "broad" lighting from the other side...



This image is copyrighted by the owner




But for an oblique view on white short lighting may not be effective. The goal is to make the face overall contrast with the background but also pull the attention of the viewer to the front of it. You can see in the examples above how the short lighting on the dark background does that by creating compelling contrast your eye can't resist. But put that same lighting pattern on a white background and the brightly lit far side of the face will disappear perceptually into the background. So instead the more effective strategy for oblique views on white backgrounds is broad lighting with light open shadows. Broad light will make the far side of the face slightly darker, allowing it to contrast well against the white background and an the same time make the side of the head facing the camera brighter and blend into the background perceptually. The net effect of the lighting pattering is a contrast gradient which pulls the attention of the viewer to the slightly darker far side: the same dynamic of eye movement as short lighting on a dark background.

Moving the eye of the viewer to what is most important with CONTRAST is what lighting patterns are all about. So when you are moving lights around try to understand the underlying goals of what the lights should be doing, to pull the viewer into the front of the face and model it in a way that looks slim and symmetrical with natural downward modeling of the nose and cheeks, no harsh shadows, and good light and a catchlight in both eyes.

The thing which isn't grasped immediately about white backgrounds is the fact that COLOR contrast of warm skin is often a stronger attractive than tonal contrast, especially if the subject is a blonde. If the subject has dark hair you'll notice the hair first, color contrast of the face second. If the subject is a blonde, it will be color contrast alone which pulls the eye into the face. That's why non-competing, non-distracting clothing is an important consideration on white backgrounds. Little girls are traditionally dressed in pink, but a pink dress will overpower the face. If the goal is to see the face use a white dress with a pink collar.

Click the WWW button and look in the TOC for tutorials on basic lighting configurations and why they work, and how to expose white background shots correctly perceptually.

Chuck



Wow.....

Thanks Chuck. I didn't even have a question but that was one of the best mini-tutorials on lighting I've read in a long time.!!
Your work is beautiful, thanks for sharing it and some of your knowledge.
- D -

Dec 15, 2008 at 06:26 PM
 



c.d.embrey
Offline
Upload & Sell: Off
p.1 #14 · Backdrop lights bleeding?


shatterkiss wrote:
The classical, conservative approach says you need fill to moderate shadows...the modern, creative approach says that you shouldn't assume anything at all and should light to create the images that you respond positively to, not the images you're told are correctly-lit. Platon's portraits are great examples of breaking the "rules", using key without fill to create strong images.


+10 on that!!! Do you want your photos to look like they were taken in the 1980s or the present day?

Look at Mark Halper's celebrities http://studiomark.com/celebrity.html Notice that he doesn't have just one style. One size does not fit all.

BTW Halper is a big fan of just banging the flash into a white wall.


Dec 15, 2008 at 06:45 PM
cgardner
Offline
Dedicated FM
Upload & Sell: Off
p.1 #15 · Backdrop lights bleeding?


If you start by metering the background by the time you add key and fill, especially in a butterfly configuration, the background will become grossly overexposed. That's a procedural error I think many who blow out backgrounds make. The meter for the background to be 1/2 to a full stop above the key light, but when key and fill fall-off get added it may wind up 2 or more stops over. If its blown to begin with there's really no way to tell in the photo how overexposed it is. 255 could mean 1/8th stop past clipping or 8 stops.

As you can see in the examples I posted of the test targets before and after background light was added a white background will be rendered light gray with just the fall off from key and fill.

Whether to adjust background or foreground first is a chicken/egg debate: either will work. But if you measure the background first with the meter with background lights alone you should meter and adjust it again after you'd set the foreground lighting. As mentioned I prefer to do it the other way around, foreground first, background second because the background doesn't affect the frontal lighting except to the extent flare cuts contrast, and the effect of flare is easier to see from a baseline of no background lighting.

With respect to relative foreground / background readings just imagine the person in the foreground is holding a piece of the white seamless next to their face. To reproduce it identically back behind on the background you'd need exactly the same exposure, no? So the starting point for matching background and foreground should be to make a meter reading of the front of the face pointed at the camera and a second with the meter against the background up near where the head appears in the photo against it and make the two identical. Then you can adjust the background slightly higher/lower as needed perceptually as noted previous.

Metering is a technical thing, not some creative aspect like where to place the lights. If you want accurate readings, consistently, across all lighting ratios you need to use the tool as designed. To meter it correctly for exposure with an incident meter place the meter in front of the face, taking care not to block or bounce light into it, and point the dome AT THE CAMERA not the key light. Thats how very incident meter manual recommends that exposure readings be made because the function of the dome is to average key and fill together and mimic the way it reflects off a 3D face.

But also realize if you take the meter out of the box and do that it WILL NOT indicate the correct exposure in most cases because the meter ISO and camera ISO are different at the same nominal (e.g ISO 100) setting. When set to ISO 100 the true ISO of Canon bodies is closer to 120. So if you meter as instructed in the manual you may wind up 1/3 stop over-exposed. There is a procedure to compensate a hand held meter. You take a reading then bracket around it using a textured object like the towel as a target. Open the RAW files in Photoshop and look at the detail in the towel. If you metered f/8 but the f/9 shot reproduces the towel better it means a 1/3 stop compensation needs to be added to the hand held meter. On an L-358 its done by first turning the meter on, then pressing both ISO buttons at the same time and turning the wheel on the side. The need to compensate a meter is not unique to digital. It was necessary to do the same thing with film because the actual ISO of any emulsion batch was seldom what was marked on the box.

For metering ratios the procedure is different. First you need to only have the light you are measuring on. The meter is placed near the middle of the subject face and pointed AT THE LIGHT. Sekonic meters have a retractable dome and it recommends in the manual that ratio readings be taken with dome down.

Ratios can be confusing because other fields, like cinematography, use the direct incident strength to express ratios.
The portrait ratio convention is based on how the reflected light looks/measures on the two sides of the face. It starts with the assumption fill is a constant (i.e., always "1" in the ratio) placed behind / over the camera and illuminating both sides of the face evenly. The convention is modeled on nature where the sun is the key light and the sky provides wrap-around fill. Its not possible to create wrap-around fill with artificial light but fill over the camera creates the same "neutral" lifting of the shadows of the key light.

The portrait lighting ratio is based on overlapping the off axis key light over even fill. That's why you never get a 1:1 ratio with that convention. When two lights with identical strength overlap you get a 2:1 reflected ratio:

H:S
1:1 even fill
1:0 key light overlaps
===
2:1 highlights look / measure 2x brighter than the shadow side

If you increase the key light so its 2x brighter than fill (1-stop) the resulting ratio is 3:1

H:S
1:1 even fill
2:0 key light 2x brighter than fill overlaps
===
3:1 highlights look / measure 3x brighter than the shadow side


What is more important than the numbers is recognizing what levels of contrast on the face are age/gender appropriate and being able to set them in advance with the meter to save time. The reason I use a target on a stand is so I can put it where the face will wind up in the photo and get all the lights set perfectly before the subject steps in front of the lights. It makes for a more productive and relaxed session, especially with family members co-opted for testing.

If I seem dogmatic about technical issues like metering its because I spent most of my career as a technical and technical manager doing various forms of process control and approach technical problems by identifying, eliminating or controlling variables which affect outcome. I also read manuals, more than once, try everything suggested, and assume the people who make the tools like meters know best how to use them. In most cases they do..

Chuck




Dec 15, 2008 at 06:50 PM
shatterkiss
Offline
Dedicated FM
Upload & Sell: Off
p.1 #16 · Backdrop lights bleeding?


cgardner wrote:
If you start by metering the background by the time you add key and fill, especially in a butterfly configuration, the background will become grossly overexposed. That's a procedural error I think many who blow out backgrounds make.


That makes a lot of assumptions, though. How far is your subject from the background, and your key/fill from your subject? At what relative angles are your key and fill positioned and using what modifiers? All of this stuff is highly-situational. If your key is a gridded small softbox angled downward toward your subject at a 60-deg angle from 2' away, and your subject is 10' from the background, and your fill is angled from below the camera at a similar upward angle and set 3 stops below your key, then the background might read as a dark charcoal grey before you put any additional light on it. If it's 4-5 stops below camera exposure you aren't going to see a gross overexposure if you'd otherwise set your background lights to be independently-metered at camera exposure. I think it's better to understand the interaction of these things than to approach them with a codified process. Memorizing a process doesn't teach people to understand the interplay, yet that understanding is vital if you aren't just reproducing the same photo under the same conditions time and time again.

Metering is a technical thing, not some creative aspect like where to place the lights. If you want accurate readings, consistently, across all lighting ratios you need to use the tool as designed.

{snip}

But also realize if you take the meter out of the box and do that it WILL NOT indicate the correct exposure in most cases because the meter ISO and camera ISO are different at the same nominal (e.g ISO 100) setting. When set to ISO 100 the true ISO of Canon bodies is closer to 120. So if you meter as instructed in the manual you may wind up 1/3 stop over-exposed.


See, this is where I start to have a problem again...a fraction-stop difference is only a problem if you live and die by your meter and a singular reading. If you take a reading off your subject and assume without any more thought or consideration that that is what your camera exposure should be, then yes, it's a problem. But if you take multiple readings on your subject, know what the light and dark sides of their face read at, know what the shadow under their chin reads at, know what the top of their head reads at, know what the back of their head (and thus your rimlight) reads at, and then make the creative decision of what your relative camera exposure should be...and then take a test shot and LOOK at it, it's not so much of a problem.

This is quite a bit about knowing your camera, much like we used to have to know our film stocks. I knew that Fuji Astia 100 had the nicest skin tones when overexposed by about 1/3rd of a stop. I knew that Fuji Neopan 1600 only had about 1.5 stops of headroom, so I needed to expose it more conservatively even compared to Neopan 400 pushed two stops. I knew that Agfa Scala kept a lot of detail at up to 3 stops over, so it was a great choice for shoots that involved subtle midrange and highlights. And I knew that my Nikon D200 was set to expose conservatively, so I should generally give it an extra 1/2-stop overexposure to get back to what I would expect from a neutral film exposure.

These things are only a problem if we think of light meters, and meter readings, as being an instruction of what to set the camera exposure to. If you treat your meter reading as what it really is, just a measure of the amount of light in a given place, and use it to inform your own decisions of how to expose a frame given all of the other creative and technical variables, then you'll be fine.

A meter is a technical tool, yes. Metering is a technical process...okay, maybe. No more so than sticking a thermometer in your mouth, as far as I'm concerned. But what you do with a meter reading is less technical than creative. Lighting is certainly one part technical, but it's also one part voodoo and magic.

Lighting built technically and a photograph created technically yield a technical photograph. A photograph created by following rules yields a formulaic photograph. This is why places like art school or culinary school teach students to learn the technical basics and history of their disciplines, then throw them out the window and learn to break the rules and be creative. A Jackson Pollock or Ferran Adria certainly knew the classics before developing their signature styles, but they also knew enough to reject the dogma suggested by the classics.

My biggest fear is a community of photographers who think there is a "correct" way to light an image or a "correct" exposure for their camera at any given moment, versus an enormous array of possibilities and options and ways to take a given situation and express their creativity and vision with it. When I've taught classes, in photography or other disciplines, the first thing I've always taught is to know your tools...and the second thing is to throw dogma out the window and experiment, play, walk backwards into things, leave yourself open to happy accidents. Arrive at a process because it works for you and you understand the implications of its individual steps, not because someone else tells you it's best.

Okay, that's my quarterly rant posted, I'm done...

Dec 15, 2008 at 07:55 PM
c.d.embrey
Offline
Upload & Sell: Off
p.1 #17 · Backdrop lights bleeding?


Once again, I agree with shatterkiss. Is this art or paint-by-the-numbers

BTW sometimes I read the key and sometimes I point the ball at the camera. Depends what I want to know. You need to be smarter than your equipment!




Dec 15, 2008 at 08:14 PM
cgardner
Offline
Dedicated FM
Upload & Sell: Off
p.1 #18 · Backdrop lights bleeding?


Teaching technique to the clueless beginner learning how to use tools and using them from a framework of experience are two different things.

Here's the simple logic behind my suggested approach from the standpoint of learning how the tools work on a cause and effect basis.

If you start your way - light the background to the point of clipping to start, a beginner may NEVER see the effect fill and key, regardless how it is placed, on the background. Most beginners don't use large open studio spaces so there is nearly always spill onto the background and probably would not dawn on them that its raising the level of light on the background far more than metered. Net result? Nuked background and flare. Exactly the why the OP started this thread. He's already doing it your way, more or less, and its not working.

On the other hand of you start without any background lights and first set the lights on the subject is will be very obvious, even two the clueless newbie, how the key and fill is spilling onto the background and affecting exposure. Starting with the background darker and bringing it up to the foreground perceptually also requires no metering. All that is necessary its to look a the rim lighting, as in may example and adjust the background until it is visible to the extent desired. So actually my approach is a quicker path to achieving the goal of being able to achieve the desired results visually.

If there is one thing I stress above all else its awareness of the perceptual aspects. Mastering the technical stuff in a systematic way is the quickest way to get the technical roadblocks to creative perceptual vision out of the way.
Chuck



Edited on Dec 15, 2008 at 09:18 PM · View previous versions


Dec 15, 2008 at 08:27 PM
alanwarp
Offline
Upload & Sell: On
p.1 #19 · Backdrop lights bleeding?


Funny quote on working with white seamless, helped me realize some of the challenges, like needing a new house :

"The more room you have the less bad words will come out of your mouth. A 20×20′ room with 10′ or higher ceilings is a great place to start. You can do it with less but you’ll have more challenges to face and make this more of a pain in the arse then it needs to be. Trying this in a spare bedroom with 8′ ceilings is going to drive you mad and you’ll sound like the dad in A Christmas Story as he worked on the furnace. You can do it… you’ll just use more cuss words doing so."

from: http://www.zarias.com/?p=71


Dec 15, 2008 at 09:15 PM
adamdewilde
Offline
Upload & Sell: Off
p.1 #20 · Backdrop lights bleeding?


Simon, how would one go about setting up each light at a time, if one happens to be using an Acute2 pack and heads, each time one adds a new head, the power for the lights has to change.

Dec 16, 2008 at 02:01 AM
shatterkiss
Offline
Dedicated FM
Upload & Sell: Off
p.1 #21 · Backdrop lights bleeding?


adamdewilde wrote:
Simon, how would one go about setting up each light at a time, if one happens to be using an Acute2 pack and heads, each time one adds a new head, the power for the lights has to change.


Not exactly true...that's the case if you have two heads on the B-channel or if you have the two channels set to asymmetrical, but not otherwise. If you've only got one head on each channel and it's set to symmetrical then the two channels are essentially independent.

In a conversation about using a pair of background lights, one key and one fill you aren't going to be working from a single Acute2 pack anyway - only three heads supported, right? In that case I'd put the background lights on the B-channel of one pack and separate the key/fill to different channels on another pack. On a single pack I'd put two background lights on the B-channel, as I could keep them at equal spacing and power, then put the key on the A-channel. With the pack set to symmetrical, each channel is getting up to half the pack's power without affecting the other channel.

With a pack set to symmetrical you can also unplug the head(s) from one channel without affecting the other. So you can meter background and foreground separately, or key and fill separately, depending on your situation. Really, the only time I use the Acute packs in asymmetrical mode is when I need to push a ton of power to one head.

Dec 16, 2008 at 02:33 AM
cgardner
Offline
Dedicated FM
Upload & Sell: Off
p.1 #22 · Backdrop lights bleeding?


c.d.embrey wrote:
Once again, I agree with shatterkiss. Is this art or paint-by-the-numbers


At the stage the OP as at, no it isn't art, its a struggle to get past all the technical bullshit to acquire the skills necessary to be able create art effectively, effortlessly. So lets keep things into the context of original question and don't confuse helping a beginner learn how to control exposure with handcuffs on creativity.

Everyone climbs a learning curve, each at their own speed. Some people just experiment and figure things out themselves eventually: you seldom see those types on forums asking basic technical questions or posting work with basic technical flaws. But when someone does show up at the bottom of the learning curve I've found its better to start with the basics, for no other reason to establish a common understanding of vocabulary and concepts.

Some conventions, like nuking a white background, might be conventional wisdom but really don't make much sense technically or perceptually. Its just a workaround for not being able to light a background evenly which has some pitfalls, as the OP has discovered. In this thread the the root cause of the problem is a lack of understanding whether knocking out the background is necessary and what the downside risks of doing it are. Once the OP knows the cause and effect of over exposed backgrounds and lens flare he can choose to nuke it or not according to wherever his artistic muse takes him. But either way he'll understand how it will affect the appearance of the foreground subjects, and more importantly WHY on a technical and perceptual level - I try to cover all the bases

The technical part of photography and lighting isn't rocket science and can be mastered in a short time if one approaches it systematically with a few lights or a window, a gray card and a white towel: the camera has all the other necessary monitoring tools. Yes, to some anything remotely systematic does looks a lot like "paint by numbers" but that phase passes quickly as each skill is practiced, mastered and absorbed into the sub-conscious realm of "getting it".

Chuck



Dec 16, 2008 at 03:47 AM
adamdewilde
Offline
Upload & Sell: Off
p.1 #23 · Backdrop lights bleeding?


shatterkiss wrote:
adamdewilde wrote:
Simon, how would one go about setting up each light at a time, if one happens to be using an Acute2 pack and heads, each time one adds a new head, the power for the lights has to change.


Not exactly true...that's the case if you have two heads on the B-channel or if you have the two channels set to asymmetrical, but not otherwise. If you've only got one head on each channel and it's set to symmetrical then the two channels are essentially independent.

In a conversation about using a pair of background lights, one key and one fill you aren't going to be working from a single Acute2 pack anyway - only three heads supported, right? In that case I'd put the background lights on the B-channel of one pack and separate the key/fill to different channels on another pack. On a single pack I'd put two background lights on the B-channel, as I could keep them at equal spacing and power, then put the key on the A-channel. With the pack set to symmetrical, each channel is getting up to half the pack's power without affecting the other channel.

With a pack set to symmetrical you can also unplug the head(s) from one channel without affecting the other. So you can meter background and foreground separately, or key and fill separately, depending on your situation. Really, the only time I use the Acute packs in asymmetrical mode is when I need to push a ton of power to one head.




Sorry to Hijack, I thought he got the answers he needed... Simon I'm still confused, and rarely find myself using my Acute pack the way you do, I generally use 4 lights in my lighting setups including the acuteB, I will pick your brain a bit more on PM if thats alright with you. Let me just reread your post a few more times to see if I can make heads or tails of how you works your pack...

Dec 16, 2008 at 05:28 AM
Tareq
Offline
Upload & Sell: Off
p.1 #24 · Backdrop lights bleeding?


very interesting thread.
Too bad i have lights and light meter and backgrounds, but i don't have models to work with, i had shoot my children many times, even some results i really like but i am sure all of them are just normal and not perfect to many lighting experts here, and in fact i don't know how to use the light meter [i saw many videos and read many articles but seems i still don't know the concept of light meter before i read the full explanation and details], and i even was thinking to get a color checker, but with all that if i can't work many times in studio with someone then i will never learn on lighting and studio tools, in addition, i can't bring those fake models or some still subjects to work on, i feel i want to learn on human rather than those dolls, but in all cases, i will keep reading and watching samples, and wish one day to work with real models.

Dec 17, 2008 at 12:05 PM
shatterkiss
Offline
Dedicated FM
Upload & Sell: Off
p.1 #25 · Backdrop lights bleeding?


cgardner wrote:
Mastering the technical stuff in a systematic way is the quickest way to get the technical roadblocks to creative perceptual vision out of the way.


It's also the quickest way to instill habits based on dogma rather than one's own thought processes. You end up a subscriber to a system, rather than someone who has encountered problems and learned to reason through them; when the system doesn't address a problem or situation, the subscriber is left high and dry while the independent thinker has been taught to improvise. The systematic approach may teach you to answer the "what?" more quickly, but it leaves you scratching your head at the "why?" - it creates a technician, not an artist or craftsman.

It's like those people who spent any amount of time learning penmanship in school (and more importantly, being graded on it): does your handwriting still resemble the model you were taught to follow? Ever wonder what it would look like if you'd been given the chance to develop your own writing style? I just find it so telling that the most creative and exciting photographers working today are folks who aren't technically and formally trained at all - they're creatively trained (often from other disciplines), maybe they assisted other photographers for a time, they're armed with the ability to find creative solutions to mundane problems. Even the most technically-capable assistants tend to learn by doing, watching and thinking, not studying. Whereas some of the most bland and insipid photography around, especially on a site like FM, can come from people who have read and followed every tutorial they can get their hands on and can pixel-peep until the cows come home but can't explain why they approached a photo or its lighting a certain way.

You say that there's a specific and correct way to use a meter (dome facing the camera) because that's what the manuals say. Every working pro I've either been taught by or worked with/for has instead metered individual lights with the dome facing the light, often shading it with their other hand to avoid picking up light from other heads. As part of the process they'll probably also take a central reading facing the camera, but that's just one of an array of readings that essentially gives you a map of the light in the frame. No one reading is your guide to a "proper exposure", they're just the range of light and shadow that inform your decision of how to expose the frame and what your photo will look like. There's no right or wrong, just a series of creative decisions that might or might not serve your intent in crafting the photo.

I'm sure there's a point of view between yours and mine that's healthiest and best for the fledgling photographer. You point out that you're an engineer and technical by trade, whereas I went to art school for things other than photography and the areas in which I'm not self-taught were learned by assisting and working as a photographer. I just worry that in your relentlessly-systematic approach you cover the "Light, Science" but have left out the "Magic". Even the most technical book on photographic lighting that I know of, and the one that I recommend first, remembers to include the magic. I'd love to see the point in your tutorials where you say, "now forget everything I've said, throw this all out the window, go out there and break my rules and develop your own approach to photography." Where you tell people to stop being housepainters and start being Degas or Ferran Adria or Jimi Hendrix.

Dec 17, 2008 at 06:52 PM




FM Forums | Lighting & Studio Techniques | Join Upload & Sell
1
2 end
    
 

You are not logged in. Login or Register

  Username   Password  
Lost your password?