Advice please white bg

[sk66]

but you said....

You suggested a reflected reading, which is not good advice. If the angle of lighting to the background is sharp due to space limitations, then just meter back to the lights, not the camera. The best way to render white as white is to measure the light falling on it with an incident reading. I've shot in hugely tight spaces in such a way, and never, ever had a problem.

I can't think of any good reason to use a reflected reading off a white background.

In fact...if the lights were at an acute angle, and you took an incident reading back to camera, that would far more likely result in over exposure of the background, not the other way around.

How do you figure that? You would think a camera could never take a properly exposed image.
The fact is the camera (sensor) doesn't care what the direction of the light source is, it only cares about the light being reflected back to it. A reflectance reading is never "wrong," it's just telling you something "different." An incidence reading can be wrong, *especially* if you were to point it at the light source and they were on a very shallow angle to the BG. Ok. not "wrong," it's just telling you something "different." It's NOT telling you how much light is going to be reflected back to the camera per-se.

You'll see/hear different ways of going about things... Some say make a white BG meter 2 stops over, some say make it meter the same.... Neither is "wrong," they're just different. +2 for a reflectance reading is a good start, +0 for an incidence reading is a good start (*if* the lights are forward or there is a good amount of main spill).

You could also just set the ratios and shoot in TTL if you have that capability. Or you could just set the power levels based on distance if you know how to do that. They can all be just as accurate of a starting point...

That's why any meter reading/method is only a starting point. What matters is what gets back to the camera and therefore, what the histogram/image looks like.

TBH, I haven't used an incidence meter in probably a decade, and I haven't done much studio work since switching to digital. I do have a couple of images taken in small spaces w/ speedlights that I can share.... just to prove I'm not an idiot. Both of these are less than 5ft from the BG. Not that I consider either to be excellent examples, both were quickies and I'd do things a little differently if I were to do them again.

No, neither is of a full body shot of even a baby, but the lighting would have accommodated a small child (not a full length adult).

 

[sk66]

I tend to agree about the 255 for any white subject. But things like hard specular highlights or pure white "blown" BG I expect them to be around 255... and that probably needs adjusted for the display medium...

 

[Hell on earth]

For some reason I've not been getting alerts for this so apologies for not responding before
Thanks for all the posts and advice
H

 

[HoppyUK]

Guys, there have been endless heated arguments down on the lighting forum about exactly this subject. Reflected or incident reading, to the lights or to the camera, cumulative lighting effects, cosine law etc etc.

A light meter is a very useful tool, if you know how to use it and interpret the results. But when it comes to making final exposure settings, the Gods of Digital gave us the histogram and blinkies, and both are as close as it's possible to get to showing exactly what the sensor is recording, regardless of metering method (that can only ever be a good guess, at the best of times). It's all there, under our noses on the back of the camera.

 

[Pookeyhead]

How do you figure that? You would think a camera could never take a properly exposed image.

Point a camera at a completely white wall... trust it's meter reading, and see what happens. It will not be white.

The fact is the camera (sensor) doesn't care what the direction of the light source is, it only cares about the light being reflected back to it.

We're taking about using a hand held flash meter here... not the camera. Most held held light/flash meters with a hemi-spherical invacone are directional... it does matter where the lights are, and it does matter where you point it.

A reflectance reading is never "wrong," it's just telling you something "different." An incidence reading can be wrong, *especially* if you were to point it at the light source and they were on a very shallow angle to the BG. Ok. not "wrong," it's just telling you something "different." It's NOT telling you how much light is going to be reflected back to the camera per-se.

No.. it measures how much light is falling on the surface... It's measuring the luminance and ignoring the objects properties. As above though, most light meters are slightly directional, and if the light is coming from the side, then pointing it at the camera can, and probably will result in over-exposure.

You'll see/hear different ways of going about things... Some say make a white BG meter 2 stops over, some say make it meter the same.... Neither is "wrong," they're just different. +2 for a reflectance reading is a good start, +0 for an incidence reading is a good start (*if* the lights are forward or there is a good amount of main spill).

A: there's no good reason to be using a reflective reading in a studio... B: There's absolutely no need to over expose a white backdrop if you know what you're doing. If you need to, you're metering wrong.

You could also just set the ratios and shoot in TTL if you have that capability. Or you could just set the power levels based on distance if you know how to do that. They can all be just as accurate of a starting point...

TTL can work, yes... if using speedlights etc, but absolutely nowhere near as reliable, or consistent as a manual incident reading with manually set flash. Calculating power based on distance is extremely hit and miss, and why then hell would you do that when you have the previous two options? LOL

That's why any meter reading/method is only a starting point. What matters is what gets back to the camera and therefore, what the histogram/image looks like.

Speak for yourself. I can meter the lights, shoot, and get it right every time. Maybe you need more practice.

TBH, I haven't used an incidence meter in probably a decade, and I haven't done much studio work since switching to digital. I do have a couple of images taken in small spaces w/ speedlights that I can share.... just to prove I'm not an idiot. Both of these are less than 5ft from the BG. Not that I consider either to be excellent examples, both were quickies and I'd do things a little differently if I were to do them again.

Brilliant... now show us your unedited RAW files please

Agree... but also disagree.

The pictorial effect being discussed is the one that customers demand. Call it the 'Venture look' if you like,

White is white... you do not have to over expose it. 251 RGB will look pure white. RGB255 is wrong. Nothing except bright specular reflections should be 255 in a print.

but it's pure white, and plenty of wrap is all part of it. What you, or I, or quite a few other respected posters on here might prefer, is to miss the point.

The white background wrapping round the subject and flaring out edge detail has never been part of anyone's look, especially Venture. That's just crap lighting unless you want something to look all mystical and dreamy, or contre jour-like.

So a small degree of deliberate over-exposure of the background is necessary.

No... it's not. I just posted a straight off camera shot demonstrating that it's just not necessary. It's sloppy lighting, that's what it is.

And in one sense at least it makes things easier because it's very hard to get perfectly even exposure over the background

Well.. doesn't THAT just some up photographers these days then

You want an even background.. then light it evenly, and if you can't, light the part that matter evenly, and then just tidy up the rest in post, but there's still no need to over-expose.

in a typical small/domestic studio, so the trick is to get it 'just' blown directly behind the main subject, and then clean up the grey bits in post processing. That's a heck of a lot easier to do if it's all 255/255/255.

Nothing should be "blown" in a white background shot unless you want the background to look and behave like a light source. White should be white, but still correctly exposed. Pure RGB255 is just crap lighting. Given a choice, I'd rather slightly under expose. So long as the white background is the brightest ting on the histogram and it's relatively evenly lit, then shooting the BG slightly under allows precise control of that white point in post.

You advocate blowing out your background because it's easier. fair enough. I can't tell you how to shoot. I'm from a generation that was brought up shooting E6 professionally. We were taught to get it right in camera. Something that seems to be a lost art these days. People nowadays just can't believe that we used to get results we wanted on a piece of film, and just printed it. LOL You needed to retouch back in the pre-photoshop days you broke out the brushes and the retouching dyes.

Guys, there have been endless heated arguments down on the lighting forum about exactly this subject. Reflected or incident reading, to the lights or to the camera, cumulative lighting effects, cosine law etc etc.

A light meter is a very useful tool, if you know how to use it and interpret the results.

Which you should know how to do. If you own an incident meter... you should know how to use it... if not, sell it, as there's no point in having it.

But when it comes to making final exposure settings, the Gods of Digital gave us the histogram and blinkies, and both are as close as it's possible to get to showing exactly what the sensor is recording

Errrr.....

, regardless of metering method (that can only ever be a good guess, at the best of times). It's all there, under our noses on the back of the camera.

The histogram on the camera's display is not a histogram of your RAW. It's a histogram of your 8bit JPEG preview. Trust it if you want... it's often as accurate, and admittedly better than nothing When shooting tethered and using the histogram in Capture One or Lightroom is a different matter... but do you do that?

 

[Hell on earth]

I've been saying to myself for awhile, to give up doing white bg's unless I have more room, I don't even like them so Don't know why I give in when asked
The cost of a hi-lite is just stupid, unless doing a lot of them which I don't
But still it's interesting to know how to do it in difficult situations but it is a pain in the butt

H

 

[Pookeyhead]

I've been saying to myself for awhile, to give up doing white bg's unless I have more room, I don't even like them so Don't know why I give in when asked
The cost of a hi-lite is just stupid, unless doing a lot of them which I don't
But still it's interesting to know how to do it in difficult situations but it is a pain in the butt

H

It is a nightmare with no room to seperate model from BG, but you only need around a metre and a half or so, and practising flagging off lights can do a great deal of good. Making your own flags to suit each situation helps too. Cinefoil is useful for this.... as is a box of clips, ties etc.

You do it because the client asks you.. and they're paying

 

[HoppyUK]

David, you seem to be blaming me for what customers want, while shifting the goal posts around to suit your way of working. If, for whatever reason, you need a pure white background and lots of wrap, you have to over-expose the background. Not much, and most newcomers over-do it and make a mess, but 251/251/251 is a) not enough, and b) impossible to obtain outside of a large studio and some expertise. Feel free to say you don't like the pure-white-with-wrap look but that's a separate point.

'Get it right in-camera' is a good phrase, and mostly I agree with it, but it's not always the best or most efficient method and digital allows us to do things differently, often better, or in a way that's not possible with film (I started with E3).

 

[sk66]

No.. it measures how much light is falling on the surface... It's measuring the luminance and ignoring the objects properties. As above though, most light meters are slightly directional, and if the light is coming from the side, then pointing it at the camera can, and probably will result in over-exposure.

I feel like you are just trying to be difficult.... I didn't say taking an incidence reading was "wrong" or wouldn't work. I specifically said it could work well in a "normal" lighting setup.

I didn't say to point it at the camera, I said take a reflectance reading. If you know a reflectance reading is going to be two stops under, then why not trust it? Just make it read 2 stops high...
I didn't say "overexpose" a white BG, I said add 2 for a reflectance or add 0 for an incidence reading (with some caveats).
One reason an incidence reading works for a white BG with the lights at an angle is 90+ % of the time there is additional spill that compensates for the angle of the lights, and if not you usually have to meter the BG a little high.... but that's to achieve "correct" lighting, not blow it out. As you said, an incidence reading does not take into account the objects properties. Mostly we think of it being "color independent" but it also doesn't take into account the efficiency of reflectance. There's an exposure requirement difference between vinyl/paper/muslin even if they are all "white" (although it may be minor). And angle of incidence does matter as to that efficiency in directing light to the camera.

Exactly how do you set your lights? Do you set the BG to the same reading as the subject? Or do you set it to meter slightly higher? Typically, a more "normal" lighting setup will require the BG to meter ~.5 high if there isn't a lot of spill.

Yes, once you know the characteristics of your BG/Modifiers/etc it can be very easy/quick to set up, meter, expose and be very close to what you're expecting... And I would even say if the setup remains fairly static (distances/positions/etc) even "right every time."
But if you can get it 100% right, 100% of the time, first time, with things changing, and just by metering; then you are a lighting god.

You recommended she do exactly what she did do, except she did point the meter towards the camera (which you say should have resulted in even more overexposure)...and she set the BG lights to meter almost 1 stop high....and her subject was close to the BG with a lot of spill. Why didn't that result in a white BG? With a "correct" reading that should have resulted in the BG being ~2 stops over.

 

[Pookeyhead]

David, you seem to be blaming me for what customers want, while shifting the goal posts around to suit your way of working. If, for whatever reason, you need a pure white background and lots of wrap, you have to over-expose the background.

I guess I'm just not seeing what you're seeing when I read that sentence. I read "lots of wrap" and I see a flared out background and an image with reduced edge contrast. An over-lit backdrop in other words. Like this...... from the same session actually...

Yes.... that was metered with the background 2 stops over.... of course it was.... but the background is now a light source. Is that what you're referring to? Helen wasn't though.

Not much, and most newcomers over-do it and make a mess, but 251/251/251 is a) not enough, and b) impossible to obtain outside of a large studio and some expertise.

It is possible with limited space. I suppose my biggest issue here is the "some expertise" thing. Being a beginner shouldn't mean making compromises, it should be learning how to achieve what you want through good lighting. If it's possible with practice and expertise as you say (and you're right.. it is), then that should be the goal, not finding ways to fix things in post.

'Get it right in-camera' is a good phrase, and mostly I agree with it, but it's not always the best or most efficient method and digital allows us to do things differently, often better, or in a way that's not possible with film (I started with E3).

If, what you want is this wrap around lighting you refer to, then it's still best to get it right in camera, is it not? The shot above has an over exposed background, but it still required no modification in post. In fact, I'm not sure how you are going to make the lighting wrap around in post if it wasn't already doing so in camera.

Helen however, doesn't seem to be wanting wrap around lighting. She's already said she merely wants a white background, and she's actually achieving it. You do not need to over expose to achieve that if you light correctly. It certainly shouldn't need all the extensive post production people are giving her image in this thread.

If you know a reflectance reading is going to be two stops under, then why not trust it? Just make it read 2 stops high...

Why take a reflective reading, only to have to modify it and add a stage that could conceivably add user error, when you've already got an incident meter in your hand? LOL. I can think of no flash meter available today that is reflective only.... so why choose to meter reflectively? Just makes no sense. Take an incident reading and you don't have to modify it at all.

, and if not you usually have to meter the BG a little high.... but that's to achieve "correct" lighting,

Argue all you want. I've already shown you what I get straight off camera by correctly incident reading the backdrop and not over exposing it. I seem to have no problem making a white background look white by incident metering it.

Exactly how do you set your lights? Do you set the BG to the same reading as the subject? Or do you set it to meter slightly higher? Typically, a more "normal" lighting setup will require the BG to meter ~.5 high if there isn't a lot of spill.

LOL... you speak with such authority about over exposing backgrounds, but I'm the only one in here posting up straight out of camera images to make my point. Everyone else is just making noises. No.. if I want a white background, I meter for the background, as that is what I pre-determine as my brightest point in the image. I meter it and shoot at that aperture so I do not clip it (this will give you a white backdrop if you/re lighting it correctly, but not RGB255), I then light everything else around that... so main subject light will be metered to the same aperture. I've always worked this way... It's quite capable of getting a pure white background without any need to over-expose but that background is not blown out. Nothing in my histogram is clipped.

The only time I would, is to produce what Hoppy keeps referring to, which is heavy backlighting (see above)

Yes, once you know the characteristics of your BG/Modifiers/etc it can be very easy/quick to set up, meter, expose and be very close to what you're expecting... And I would even say if the setup remains fairly static (distances/positions/etc) even "right every time."
But if you can get it 100% right, 100% of the time, first time, with things changing, and just by metering; then you are a lighting god.

So be it, I'm a lighting God then LOL. In reality though... I test. I'm not saying I can always set up lights, take one reading, and then BANG!... it's in the bag (although I often can)... I will test the lighting, and ensure it is not clipping the highlights (the backdrop in this case) and then I'll shoot. Incident readings are very, very accurate if you know exactly how to use your meter correctly.

You recommended she do exactly what she did do, except she did point the meter towards the camera (which you say should have resulted in even more overexposure)...and she set the BG lights to meter almost 1 stop high....and her subject was close to the BG with a lot of spill. Why didn't that result in a white BG? With a "correct" reading that should have resulted in the BG being ~2 stops over.

The biggest problem with Helen's shot is trying to control shadows due to having the little boy sat on the ground, and then it's compounded by the creases in the vinyl where's he's sat. Her actual background is pretty clean actually. Her issue is one of space as the shadows are obviously falling onto the scoop behind through proximity, and I still maintain that can be massively reduced by really careful flagging of both sets of lights, especially the background. You sit someone on a background that gives under pressure, and it's no longer flat... you'll get deeper, uglier shadows instead of a clean, soft pool of shadow that actually looks natural. She could really actually do with finding a way to stop the vinyl creasing like that. If it's on carpet, she needs to put a sheet of something less compliant under it.

I'm here mainly to explode this myth that white backgrounds need to be over exposed, not necessarily to crit Helen's work however.

 

[Hell on earth]

The biggest problem with Helen's shot is trying to control shadows due to having the little boy sat on the ground, and then it's compounded by the creases in the vinyl where's he's sat. Her actual background is pretty clean actually. Her issue is one of space as the shadows are obviously falling onto the scoop behind through proximity, and I still maintain that can be massively reduced by really careful flagging of both sets of lights, especially the background. You sit someone on a background that gives under pressure, and it's no longer flat... you'll get deeper, uglier shadows instead of a clean, soft pool of shadow that actually looks natural. She could really actually do with finding a way to stop the vinyl creasing like that. If it's on carpet, she needs to put a sheet of something less compliant under it.

All of this makes sense and I've taken on board.
My feeling now is that in my small space, the light bounced around the walls making the lighting even, bg to subject, as I have said before and looks like in the image posted.
I could add flags to stop spill to subject but it's not going to stop the light bouncing everywhere else
I guess that is where a hi-lite would be of most advantage, grids maybe another way

Maybe had I used the spot meter in the camera to the subject this would have upped the white too but then may have blown the highlight on the subject

I saw a vid somewhere, where the guy just lite the bg and then spot metered the subject, but the subject was wearing dark clothes and was a head, 3/4 shot not a floor shot
Much like you would do say out side on a dark subject but blowing the sky... err backlighting innit ?


H

 

[HoppyUK]

David, you do seem to be arguing rather for the sake of it No problem, it's all good debate, but I never suggested wrap can be applied in post processing. That would indeed be daft, but cleaning up grey patches, which is the unavoidable reality for most folks, is much more sensibly done in post.

The other thing is your insistence that white should not be pure 255/255/255 white for some reason. I've never heard that before, and after hanging around this lighting forum and others for a few years now, I don't recall anyone else ever mentioning it either. If you scroll down a few pages, you'll see that shooting white backgrounds comes up all the time here, and the look that customers want is pure white, high-ish key, with wrap. The tricky bit is getting that without bleaching the subject outline with excessive over-exposure, and washing things out with flare.

To Helen, I wonder if when you talk about metering levels and exposure, you've been metering background and subject separately, but not cumulatively with all lights on. In a smaller working area, you will indeed get a lot of fill-in from the background bouncing around the room and lifting the main subject exposure. The smart way of setting up is to use a hand meter to get things quickly to working level, then adjust final exposure with histogram and blinkies.

 

[Hell on earth]

To Helen, I wonder if when you talk about metering levels and exposure, you've been metering background and subject separately, but not cumulatively with all lights on. In a smaller working area, you will indeed get a lot of fill-in from the background bouncing around the room and lifting the main subject exposure. The smart way of setting up is to use a hand meter to get things quickly to working level, then adjust final exposure with histogram and blinkies.

Richard
Doh! is me, Think your have hit the nail on the head, it's something I forget to do at times
I guess I've thought, I've set the bg and set the subject so forget the combined but then.. would it not have been brighter ?

H

 

[sk66]

LOL... you speak with such authority about over exposing backgrounds, but I'm the only one in here posting up straight out of camera images to make my point. Everyone else is just making noises.

Just for you. Not raw, but as close as can be.

Stuff by sk66 on Talk Photography

Obviously I edited the crap out of that photo, but I did not cut out the BG or do any heavy editing there. I will grant that the BG isn't entirely even; it was lit with one small softbox...

you speak with such authority about over exposing backgrounds

One more time... I never said anything about "overexposing" the BG. I've talked about making it "white," not blown.

 

[sk66]

Richard
Doh! is me, Think your have hit the nail on the head, it's something I forget to do at times
I guess I've thought, I've set the bg and set the subject so forget the combined but then.. would it not have been brighter ?

H

Yes, the BG (and subject in this case) would have been brighter where the two light sources combine. In fact, it was...

 

[sk66]

FWIW, it's not entirely unheard of for a meter to need "calibrated" to the camera...

 

[HoppyUK]

FWIW, it's not entirely unheard of for a meter to need "calibrated" to the camera...

Possibly.

The more common issues, almost likelihoods, include the difference between f/numbers and T-stops, inaccurate f/numbers, variably inaccurate f/numbers with vari-aperture zooms, inconsistent f/numbers especially higher f/numbers, vignetting that commonly causes the edges to be a stop darker than the centre, and the corners sometimes up to three stops down (!).

I live with these things everyday in my work testing lenses pretty much full time ATM for UK photo magazines. There is so much happens to the light after it hits the lens that no hand meter knows anything about. But the histogram and blinkies do.

 

[Hell on earth]

Well the Lady picked up the images today and was very happy
I edited the bg in Lightroom, the way Hugh showed and they printed up nicely, but it's a get out of jail card I don't really like.

H

 

[sk66]

Hmm, unless you switch between film/cinema lenses I don't understand how the f/t issue plays in... is the t-stop for an f/stop hugely inconsistent between lenses? I certainly believe f-stops are not entirely accurate, nothing is; but I've never found it to be problematic.

 

[sk66]

Well the Lady picked up the images today and was very happy
I edited the bg in Lightroom, the way Hugh showed and they printed up nicely, but it's a get out of jail card I don't really like.

H

Good deal...
FWIW, I agree with others in that the BG almost always requires some editing... wrinkles, edges, marks, sensor crap... but it would be better to be closer from the start.
I'm curious, did you use the exact settings recommended? I'm curious because that puts the BG right at 255, and because I've never heard of 255 printing poorly, it just doesn't print at all (the white level is determined by the paper).
FWIW, I use 245 for white details I want to show in a print because details above that may not be reproducible on a given paper... I guess maybe that's what David meant...

 

[Pookeyhead]

FWIW, it's not entirely unheard of for a meter to need "calibrated" to the camera...

Nope.. it's not. Not something I've ever found the need to do though. I think Richard is nearer the mark though, with inconsistencies in f/stop accuracy. Even that though... it's going to be very minor. We're talking differences that are not really visibly apparent. I've certainly never had an issue and like Richard, I test lenses all the time.

Just for you. Not raw, but as close as can be.

Then close, but no cigar

Looks clean enough to me. You can indeed get a clean background with the model close, yes. It's easier to flag off speedlights too

 

[sk66]

Looks clean enough to me. You can indeed get a clean background with the model close, yes. It's easier to flag off speedlights too

That's why in tight situations I'd suggest using softboxes to light the BG... flags are built in and better/more even spread. It does require you have more of them though... I happen to have a couple of the crappy little ones for speedlights (~24" I think) which aren't any good for much else (I made that mistake when I thought I'd try going strobist).

 

[Pookeyhead]

That's why in tight situations I'd suggest using softboxes to light the BG... flags are built in and better/more even spread. It does require you have more of them though... I happen to have a couple of the crappy little ones for speedlights (~24" I think) which aren't any good for much else (I made that mistake when I thought I'd try going strobist).

One problem there though/. They're big.... which is exactly what you don't want if space is tight.

Sometimes.... you just need space. I think 24" would be the absolute minimum I'd think about to light a background.

I always use large softboxes to light backgrounds.. even large backgrounds. Far more even; larger the light source the less the fall off due to inverse square law.

 

[sk66]

24" between subject/BG? LOL, at that distance they're practically both the subject. Or 24" softbox? They're not good for much...

More even yes; more "wrap," but not less falloff. Falloff is only based on distance and it's easy to calculate/determine.

 

[Pookeyhead]

24" between subject/BG? LOL,

Noooo.... LOL

I said 24" would be the smallest softbox I'd use to light a backdrop.