fredmiranda.com
Login

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
1       2       3       …       5      
6
       7       8       end
   

Archive 2012 · Strobist
  
 
RDKirk
Offline
• • • • •
Upload & Sell: Off
p.6 #1 · Strobist


That isn't actually the concept of 'expose to the right'. With expose to the right you never clip important highlights, you expose so that the histogram is as far to the right as possible without clipping those highlights. In your example you would expose as much to the right as possible while making sure the gloss is fully preserved.

That's what I said.

This should put your cat around the centre of the histogram

No, it would not. Most of the black cat would be to the left of the center.
Jan 24, 2012 at 07:44 AM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #2 · Strobist


Beni wrote:
Whether you buy into ETTR or not is of course dependant on yourself. You can't teach the zone system to a tranny shooter but that doesn't make the choice of slide film wrong.

I was shooting 5x4 and 10x8 tranny after shooting the zone system and retaining highlight detail is paramount to both. There was no inhouse Photoshop going on with 5x4 or 10x8 tranny, so you needed to get the exposure right before it left for the agency. Digital is not much different to Tranny, but negative shooters struggled without a lab to rescue them.
Jan 24, 2012 at 08:24 AM
dmacmillan
Offline
• • • • •
Upload & Sell: On
p.6 #3 · Strobist


Beni wrote:
Those who preach a 'one fits all' solution to any type of photography but especially this kind of thing are to be kept away from as much as possible. They are a danger to any photographer.

Amen
Beni wrote:
There is no one strategy which will work in every or even most cases. Those who claim that there is should get out more and stop pretending that they are educators.
Amen
Jan 24, 2012 at 08:56 AM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #4 · Strobist


UK: So say my image is "underexposed". What is your definition of optimally exposed? Gray card image in print matching actual gray card when laid side-by-side?

In technical terms on a digital sensor optimally exposure occurs then the full reflective range of solid objects hit by the illuminant record digital values between 1 and 254 with 0 reproducing voids and 255 specular reflections.

On a digital camera 18% grey card with a digital value of X will fall somewhere within the range of 0-255 when exposed about 2.5 - 3 f/stops below clipping. That is to as if one where to overexpose the 18% card to the point of clipping and bracketed to make it darker, took those files and made prints of them, then laid out the prints and compared with the actual card, one of those prints of the card would reflect the same 18% of light when measured with a densitometer. The card and print would have densitometer reading of log(1/.18) = .745

On a Canon camera when a white object is exposed accurately, with detail, the large spike created by an 18% gray card will not fall exactly in the middle of the histogram representing the sensor range. Thus if you define "correct" exposure as one that centers an 18% card on the histogram that is an incorrect assumption.

The 18% value dates back at into the 1920's or 30's. Apparently even Kodak doesn't know when it first started selling them. Years later when there was a better understanding of human perception it was determined with testing that 12% was a more accurate "perceptual" mid point in the overall tonal range. That is to say of you put a white and black card in front of a subject and asked them to pick a the gray tone that was in the exact middle most would pick 12-13% not 18% which perceptually appears a bit too bright. The ANSI standard used for meters and cameras adopted 12% as the "middle".

Photographers were not aware of this until digital cameras put histograms on the playback. Thinking that a centered 18% card spike was "correct" exposure they found doing that would blow the highlights. Why? The camera sensor response is engineer to put 12% in the middle of the range. That is why boys and girls my white towel exposed just below clipping IN FLAT LIGHTING will result in the spike from the card winds up to the right of center. 18% is lighter in tone that the 12% the sensor will render in the middle.

This is a typical exposure confirmation shot — not rendering a gray card 18% on my print — but as a check to see I'm getting the highlights and shadow end point exposed correctly...

http://super.nova.org/TP/WhiteTowel_18percent.jpg

The card spike winds up right of center on the histogram because the center of the histogram is based on 12% and my 18% card reflects more light than that. Had I used a WhiBal gray card or other more recent products designed specifically for digital exposure the spike would have fallen in the middle of the histogram in that same shot. Why? Those cards are manufactured to reflect 12% not 18%.

This has confused photographers, even pros, for years.

The reason the Kodak card has remained at 18% over the years is, according to legend, because Ansel Adams badgered Kodak to keep it that way. Kodak which participated in the research leading to the new 12% perceptual based midpoint value was at some point considering changing the value on the card from 18% to 12%. Adams by then had created the cult of Zone V = 18%. Adams according to anecdotal accounts traveled to Rochester and over several weeks lobbied Kodak and convinced it that changing the 18% card would cause confusion when those reading his Zone System books based on 18% card = Zone V tried to use it with new 12% cards. Ironically the Kodak card staying at 18% but meters and camera sensor switching the 12% mid point of the response curve / histogram is causing today's generation of digital camera users exactly the confusion Adams feared because they assume "correct" exposure is when the card is center in the histogram. It is, if you use a 12% card like WhiBal but not if you still use an 18% card as shown in the shot above. If I center the 18% card on my camera histogram the resulting exposure blows the highlights.

Why don't I buy an 12% card? Because I don't set exposure per the card. It's just used for setting Custom WB and then in the next step of my workflow as a convenient way to use the towels I do use to set expose so scene overall matches sensor. After achieving that I have the subject hold the card for a post processing baseline shot so I can verify that that exposure is in fact correct on both ends on the towels and rest of the foreground. The color chart is for evaluating camera styles and guiding the tweeking of color selectively without changing WB or tint globally. If you don't understand the difference or why you'd want to do the things I do you really don't understand the process well enough to challenge it. The best way to do that and convince me to do things differently is SHOW ME A BETTER WAY by posting photographs you've taken illustrating your methods.

Adams used 18% as his perceptual mid-point for reasons he explains in his book "The Negative". The zone system as originally conceived consisted of ten zones 0 - 9, not including the base white of the print because in Adams words, "A gray scale of 10 steps seems to be most convenient." In later editions Adams assigned Zone 10 to specular highlights making it an 11-step scale from max black on print/black void in scene to pure white paper base/specular highlight of scene.

In the next paragraph he wrote: "The term 'middle-gray' as I use it is more of an emotional value than a quantitative physical value." Further in the same paragraph he explains, "A surface of surface of average tonality — such as the middle step of tone of a full range glossy print — would have an average reflectance of about 18% and relates to Zone V of the exposure scale. Of course the actual brightness of this gray tone would depend on the intensity of the light falling upon it. But we are concerned here with the proportionate value of tone. 'Middle Gray' represents the geometrical mean of the extremes of useful values."

Source: Ansel Adams, The Negative, 1968 Edition — Fifth Printing 1971, page 16.

Adams was gilding the lily there. In practical terms he used18% because the 18% Kodak cards where the only reliable and widely available process control process tool available for exposure at the time. That is to say you could take your Weston meter, point it at the card in the same light as the scene, and the middle tone values in the resulting print would wind up similar to those seen by eye. So the card compared to the scene was a perceptual baseline, and the exposure was set to the card so the mid-tones on the print matched what was seen by eye in the scene. All of then wound up on what called "Zone V" in flat lighting. But as Adams say the the rendering of the card would depend on whether it was in sun or shade.

Also to put Adams use of 18% and Zone V as a benchmark it is necessary to understand how in his system the overall tonal range — shadow and highlight values on the print — where controlled and matched to the scene by development of the negative to fit the scene-to-print. There was no "middle Levels" slider with the zone system or any film based capture. The best we could do AT CAPTURE is try to record detail in shadows and highlights at the same time. As Adams said what happens in the middle is this .... "But we are concerned here with the proportionate value of tone. 'Middle Gray' represents the geometrical mean of the extremes of useful values."

Again Adams is gilding the lilly. A more straight forward way to describe the Zone System is this:

Expose so the negative has density in Zone 1 shadows above voids (Zone 1). Develop your film so specular highlights in the scene (Zone 10) wind up reproduced by just the paper base and a solid white object like the hood of a car you see the specular highlight on (Zone 9) is the lightest gray tone of silver on the print you can discern the specular highlights against. Where an 18% scene value (Gray Card) will wind up in the print will depend entirely on how it is lit. If hit by sun it will, on a B&W print, wind up looking to a card seen in sunlight. If the card is held half in sun and half in shadow the sunny part of the card will be rendered similarly per comparison of actual 18% card on but the shaded parts of the card will not. They will be rendered darker with the tone in the shadow from the sun is controlled by the amount of fill they receive, outdoors from the indirect skylight or indoors from fill hitting from directions different from the "key" source creating the shadow.

If you did the zone system correctly and made a proof sheet on #2 print paper exposed so the clear film was maximum black on the print you would get results that looked like this:

http://super.nova.org/TP/FirstPortraitsSQ.jpg

The reason I would make proof sheets very systematically on #2 paper exposed to render film base = max. black was because that was the baseline in the Zone System for judging if the results were "correct" or not. When a proof sheet is exposed in that manner any frames which were underexposed would be seen to have no detail in the shadows. On inspection you would find no density in the shadows in those areas — the film didn't get enough light there to record and image above the "fog" left behind after the undeveloped silver crystals were dissolved by the fixer. If the shadow detail was fine but the highlights on the print were blown out it meant the film had been developed too long for that scene contrast, which in that artificially lit shot was controlled with the OVERLAP of key on top of fill coming from behind the camera, per the advice of the Kodak "How to Shoot Portraits" book I was using as my guide in that my first every artificially lit portrait session. If on inspection the highlights of the proof sheet were gray, not white, it was an indication that the film had not been developed long enough to build up density in the highlight areas on the negative. By the time the shadows were completely black on the print too much light would pass in the highlights over exposing them on the print which made the highlights gray.

The significant thing to note above is that the appearance of the midtones wasn't a criteria for a correctly executed workflow with B&W negative film. What matters was: 1) enough exposure for shadow detail, controlled with aperture/shutter, and 2) the negative range, controlled with development with B&W, matched the paper.

How the midtones are rendered is a function of the lighting ratio used. From that "fit range to print" starting baseline seen on the proof sheet a B&W practitioner would look at it and if they decided the clothing should be darker or shadows a bit lighter on the final print they would dodge and burn with paddles and cards with holes in them to give areas of the print less or more net exposure than the proof sheet. The manipulation of the midtones was done in post processing.

As today with digital some would argue the not dodging and burning was a "purer" form of the art of photography and what the camera captured should be the "benchmark" of correctness in technique. Adams didn't subscribe to that philosophy. He would dodge and burn very systematically taking notes. Once he had made a print to his satisfaction he would take that print and mark it with a tissue overlay with instructions tell his assistants where and how much to dodge and burn. That was the more holistic goal behind Adams "System" approach, to allow him to predictably and easily get to the baseline proof stage on a proof print, and from there create a blueprint of how to alter the image from that "fit scene to sensor baseline" to the final print he envisioned with dodging and burning.

I found Adams method matched my goals of spending as little time in the darkroom so I could spend more time capturing images. The proof sheet above is evidence it worked for my even 40 years ago because any of the frames on that proof sheet would make a very nice print with minimal effort because range fits #2 paper and the lighting ratio rendered the girl next door in a flattering way with correctly exposed highlights and flattering light shadow tone on the face with detail in her dark hair. The lighter clothing blends into the background and the dark hair acts as a frame for the face. Apart from refining the facial angles there's not much I'd change if shooting it today.

Did I understand completely what I was doing back then? No. But I got that far, that early by reading books by Kodak, Adams, and many others and following their advice. Along the learning curve I've come to understand the things suggested, like the 3:1 ratio used for those portraits worked. 3:1 happened to be what fit the overall scene range to the #2 print. I was shooting in my parents livingroom (I had just graduated high school) and there was a good bit of extra "spill fill" bouncing off the 8' ceiling contributing to the fill. Everything in the middle fell into place by itself and would up looking "perceptually" correct. Would an 18% card if used in that shot reflect 18% in the printed image. Don't know and don't care any more than I do with digital.

In that shot I set ratio by distance because I was using two 150W bulbs in aluminum clip-on shop lights attached to step ladders as stands as my lighting with the "key" light closer than the centered fill, at distances suggested by Kodak. One of the reasons Kodak suggested that approach was that in those days all a hobbyist could afford was tungsten lighting. That was still the case back in 2002 when I started giving lighting advice and suggesting speedlights. It wasn't to convince people that my speedlight approach was better than Hobby's or anyone else but to try to convince them not to waste money on a set of 500W shop lights on stands from Home Depot and to buy a pair of the Vivitar flashes I was using at the time instead.

40+ years ago I learned to fit scene overall to print and adjust the appearance in the middle with the lighting ratio. The reason I advocate using flash outdoors is because without it you can't control the lighting ratio. Sky fill is three stops darker on a clear day, an 8:1 incident ratio. The more overcast the sky the lower the ratio becomes until on an overcast day the contrast is even less than a 2:1. There is direction that models shape but not very strongly.

The lighting on the face in highlight and shadow wound up as it did in that first B&W session of mine because I exposed and developed for a full range of tone and used the 3:1 lighting ratio. Had I instead opted to use 1:4 lighting ratio by making the key light brighter it would not change the exposure in the the fill only shadows. If I used the same development time as before the highlights wouldn't fit the print because I had pegged exposure to the shadows, and the range of the negative, developed for a 3:1 scene no longer fit the range of the #2 paper. This would be very obvious to me from my standard baseline #2 proof sheet because the shadows would be fine but the highlights would be blown out with just the paper base showing. If I had bracketed exposure there would be frames that looks fine in the highlights, but would have a loss of detail in the shadows because the negative was clear.

The equivalent of "fixing it in Photoshop" would be to use a different grade paper. By making a proof sheet the same way for max. black on clear border on #1, #3, or #4 paper or by changing paper contrast with colored filters I could flnd the paper which would render the highlights correctly.

Some like RDKirk who I've known on the forums for years understand my approach because they understand its zone system roots of fitting range to sensor then manipulating what that produces in the middle tones with a combination lighting and post processing.

Indoors it's as simple as finding the ratio of key / fill that will put detail in white and black towels at the same time...
http://super.nova.org/TP/TowelGary.jpg
then tweeking the middle slider if that "sensor full range baseline ratio" isn't producing the desired look on the face..
http://super.nova.org/TP/TowelGaryLighter.jpg

Using rim light changes perception of the lighting and shifts the Zone 9 / 245 from the front of the towel and face to the back hit by the sun. There is also the option to leave the front on Zone 9 / 245 and just blow the highlights on the back to Zone 10 specular / 255. Or even blow the highlights on the front to 255, but I don't choose to do that.

By using examples I already have on line, such as the HSS shots, or putting targets I don't use for what I am measuring when I took the shot is apparently confusing UK and others. But the big disconnect is different definitions of "correct" exposure. Like Adams my definition for correct exposure has always been that it starts with getting the end points correct, not reproducing a gray card accurately. If you can get past that notion and understand that's not my goal you should understand my approach.

Edited on Jan 24, 2012 at 12:29 PM · View previous versions
Jan 24, 2012 at 12:08 PM
Beni
Offline
• • • • • •
Upload & Sell: Off
p.6 #5 · Strobist


ukphotographer wrote:
I was shooting 5x4 and 10x8 tranny after shooting the zone system and retaining highlight detail is paramount to both. There was no inhouse Photoshop going on with 5x4 or 10x8 tranny, so you needed to get the exposure right before it left for the agency. Digital is not much different to Tranny, but negative shooters struggled without a lab to rescue them.

Unlike with negs however where you expose for the shadows and process for the highlights, with trannies you don't get that choice. It was just an example at any rate . I managed a lab for two years, most pro photographers would expose for the shadows, usually by rating the film as slower than it was, we did the work on the other end, did take longer than 'normally' rated negs to print though.

With digital you pretty much expose for the highlights, process for the midtones then dodge and burn because you're doing the develop/print process all in one go. Choice of highlight detail is of course dependant on usage.

RDKirk wrote:
This should put your cat around the centre of the histogram

No, it would not. Most of the black cat would be to the left of the center.

I've no doubt that you're right, never done it myself Depends on how specular the sheen is I suppose or when the last time the cat was washed. Seems that we are in agreement at any rate.
Jan 24, 2012 at 12:28 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #6 · Strobist


What many photograhers didn't realize then going from B&W to color neg shooting in the studio was how different the range was. In the studio if you set a ratio that records a full range on B&W and shoot with the same ratio on color neg. also exposing for shadow detail the shorter range winds up either clipping highlights or losing shadow detail depending how the lab exposed the print.

The lab could print a color neg. in contrasty light for the shadows, midtones, or the highlights depending on where the most important content was, usually basing it on the highlights in the faces in portraits and not blowing them. The solution to get a full range of tone was adjust the lighting ration and contrast of the light sources (factoring in the effect of spill fill on the shadow tone). What the incident meter says a ratio is and what it actually winds up when it hits the sensor. That explains why you can fit scene to sensor with a LightSphere or even bare bulb in small room — spill fill produces a very low "net" lighting ratio. Compare any lighting tool indoors in a small room vs outdoors at night and during the day and you will see s big difference and better understand the "wrap" making shadows soft, lowering the net ratio and fitting the scene to the digital sensor is coming in part from "spill" fill indoors and skylight fill outdoors.

What transparencies did by forcing photographers to exposure on the highlights was make them better realize how important fill and lighting ratios were for fitting range to sensor, at least in the foreground of outdoor flash shots. In the 70s when I had white and black dogs I shot them all the time with sun at the back and two flashes in front and had no problem recording a full range of tone on both on Kodachrome doing the same thing I suggest for digital: use fill and overlapping key flash to control the ratio on the foreground so it matches the range of the recording medium.

The recording medium range has changed over the year required changes in lighting strategies, not for modeling but simply to fit full range of scene to the print. No flash was needed for that with B&W. About the same amount is needed for neg film and digital, Transparencies require flatter (ratio) light because they have a shorter DR.
Jan 24, 2012 at 12:45 PM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #7 · Strobist


cgardner wrote:
By using examples I already have on line, such as the HSS shots, or putting targets I don't use for what I am measuring when I took the shot is apparently confusing UK and others.

I'm not confused.

A grey card is an absolute value. Its a measure by which all is judged and on which will provide a correct exposure unless extremes are encountered. You shifted that absolute value to whatever value was convenient just to illustrate your point and it clearly shows.

In your first instance the overexposure is because you have reflected the sky or other lighting off it.

In the HSS example you have underexposed it resulting in a dark subject because your flash is inadequate to illuminate the subject to keep it within range of your sensor containing the highlight you introduced. Either the highlight wants reducing by increasing your shutter speed or the flash value wants increasing. Since it's HSS you need to increase your flash value. This renders your deduction of working distance incorrect.

In neither case were there extreme dark or extreme light needing any compensation, merely your own eagerness pushing you and forcing you to arrive at a incorrect conclusion.

No confusion on my part whatsoever.
Jan 24, 2012 at 01:12 PM
dmacmillan
Offline
• • • • •
Upload & Sell: On
p.6 #8 · Strobist


cgardner wrote:
What many photograhers didn't realize then going from B&W to color neg shooting in the studio was how different the range was.
What transparencies did by forcing photographers to exposure on the highlights was make them better realize how important fill and lighting ratios were for fitting range to sensor, at least in the foreground of outdoor flash shots.
Who are these photographers of whom you speak? How do you know what they knew or didn't know?
If you are inferring that they were professionals, especially commercial photographers, I'd like to see the data on which you base the claim that they "didn't realize".

I cannot speak for other professionals, but based on my experience as a professional, Kodak was very good about publicizing changes to their film and the resultant effects it would have on its professional customer base. I remember when they changed from CPS to VPS. I was provided a wealth of information discussing the difference in characteristics between the two film stocks. I spent many a day at the Kodak offices in Atlanta, along with numerous other photographers, attending technical and marketing briefings.
Jan 24, 2012 at 02:29 PM
dmacmillan
Offline
• • • • •
Upload & Sell: On
p.6 #9 · Strobist


williamkazak wrote:
A black cat. Available light. Go for it.
Cute cat! Poor fellow, looks like his/her left eye is clouded. Is this an older cat?
Jan 24, 2012 at 02:33 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #10 · Strobist


The card itself has an absolute value of 18% / .745 density units but how is rendered in a photograph depends entirely on how it is exposed. That's a matter of choice.

The tonal value of the card in test shot is irrelevant to me for exposure. It is there in my test shot only as a color reference. Before including that card with the MacBeth and other targets would have used an identical plain Kodak card for setting Custom WB. Because Custom WB is set I can judge the skin tone when viewing the shot of the subject holding the card, or the overall scene balance from a neutral baseline.

I've posted the RAW shots for the tests and in the 2-flash shot the darkest black in the MacBeth target in shadow is 14. That is above clipping. So how is that underexposed? Underexposed would be a value of 0 in the foreground.

How much fill would I need to add by moving the flash at 1/1 power closer to fit your definition for correctly exposed shadows? 20, 40, 60 eyedropper units?

Yes there are underexposed areas in the background of the shot, because scene is greater than sensor, but it is physically impossible to record the ambient shadows because scene exceeds the sensor. That's why the flash was needed in the foreground.







ukphotographer wrote:
I'm not confused.

A grey card is an absolute value. Its a measure by which all is judged and on which will provide a correct exposure unless extremes are encountered. You shifted that absolute value to whatever value was convenient just to illustrate your point and it clearly shows.

In your first instance the overexposure is because you have reflected the sky or other lighting off it.

In the HSS example you have underexposed it resulting in a dark subject because your flash is inadequate to illuminate the subject to keep it within range of your sensor containing the highlight you introduced. Either...Show more →
Jan 24, 2012 at 04:12 PM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #11 · Strobist


cgardner wrote:
I've posted the RAW shots for the tests and in the 2-flash shot the darkest black in the MacBeth target in shadow is 14. That is above clipping. So how is that underexposed? Underexposed would be a value of 0 in the foreground.



For your reference the black swatch should be 37,37,37. That makes 14 underexposed.

Here are the full details for a Macbeth chart converted from Lab to ProPhotoRGB just so that you know.

http://www.accoladephotography.co.uk/DPR/Macbeth_ColorChecker_500.jpg
Jan 24, 2012 at 04:30 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #12 · Strobist


OK, now at least I know what baseline you are using for comparison. Had you mentioned that earlier your criteria was a 37 eyedropper reading there it would have saved some talking past each other. Most of the black patch is in shadow so it I wouldn't expect it to read 37 there but the other sunlit patches are lower than your listed values. The upper left corner that patch is barely in the direct sun measures 20. It is the shaded part that measured 14.

I just checked and according to the data sheet that came with the Mini target I use the sRBG value actually an even lighter 52,52,52 tone for the 1.5 black patch. When I got the target I recreated it in sRGB from the values, which along with Lab are all that are listed.

http://super.nova.org/TP/ColorChart480.jpg

Here I've taken the gray scale from that file and pasted into the cropped version of the flash test shot..

http://super.nova.org/EDITS/_MG_5035_Cropped2.jpg

Visually the patches that are in the sun are similar and well within my expectations of accuracy visually. That, not an exact numerical match was my criteria for the test. I tested from closer camera / fill distances and found 10ft shot, by comparison better—not by the numbers on the shadow patch, but in overall of the foreground.

I expect a 1.50 reflective density (Zone 2) patch to look Zone 2 when in direct sun, but when in the shadow of a 3:1 lighting ratio of the flash I expect it to look a zone darker (Zone 1) just as a black suit half in shadow and sun would appear by eye.

I see your point now and agree that per your criteria it is underexposed. But seriously, does the fact that a test shot taken per my different criteria doesn't match your criteria discredit my overall approach? I think not.

The fact I have an approach that is frequently challenged is something I find quite interesting. First of all it is not a unique approach, or a new one, to have a full tonal range as the technical goal underlying recording and reproducing any scene I saw. After all I see a full range of detail with my eye in the scene between Zones 1 and 9. So in my view seeing an less in the final image is a failure when it is within my power (and the range of my flashes) to prevent clipping on either end of the tonal scale.

Recording the full range seen by eye was the technical goal when I learned the Zone System from Adams' books. He didn't blow his highlights or shadows on his prints and that seemed like a good approach to me so I adopted it, then adapted it to the roll film I was using.

Recording detail in a groom's suit and brides dress at the same time was the goal technically when I shot weddings for Zucker. Same goal as Adams, but done mostly indoors with color prints that had only about half the range and required not one, but two flashes in an even fill / overlapping key light arrangement to meet the goal.

Recording a full range of tone was also a goal when shooting my black and white dogs on Kodachrome. The same fill/key lighting strategy worked: add fill until shadow detail is recorded accurately then raise the key light until it overlaps, creates highlights and they are exposed correctly.

Digital is exactly the same goal for me. Full tonal range. I know I can get it at will with the same key over fill strategy that has worked for me since my first portrait session guided by a "How To" book by Kodak over 40 years ago.

If your goal isn't a full range image, when possible, then we aren't are never going to agree on technique.

Do I execute every shot I take with technical perfection? No. But who does?


Jan 24, 2012 at 07:44 PM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #13 · Strobist


cgardner wrote:
I see your point now and agree that per your criteria it is underexposed. But seriously, does the fact that a test shot taken per my different criteria doesn't match your criteria discredit my overall approach? I think not.


You might think not, but inaccurate tests only create inaccurate results and because you can't match to a standard the variance cannot be accurately calculated.

Your approach was inconsistent with the samples shown. You were reflecting source light off your targets, you are not using the targets the way they are intended and even with your HSS sample the lighting on the grey card caused a variance of 50 points over its surface. (over a 10" card thats a lot - and where is right?). If you are going to make statements based on results using targets, the targets do need to adhere to their published values in order to be useful. If they don't their inclusion is pointless and the results are meaningless.

If you create a system which works for you, then thats fine, but you'd expect that it would be robust enough to fix the standard targets at the right levels.


Jan 24, 2012 at 08:23 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #14 · Strobist


It amuses me to be criticized for being too technical and not creative enough by some and not being technically precise enough by others.

My approach UK is perceptual, not by the numbers. My goal in that test was not to render the 1.5 target 37, 52 or any other specific target as an exact numerical match. I never said I used the target that way. I said I match target with the results I see in the final product by side to side VISUAL comparison, not absolute numerical comparison. If i wanted to do that I'd own a densitometer like those I used at work for many years and measure the reflected density of the actual target I use and the prints I make. But I don't often make prints and any comparison is between target and screen VISUALLY.

I will not change my approach based on the feedback, but as stated previously the feedback regarding how it was misunderstood was valuable. In the future when explaining it I will include disclaimers that I do not seek 100% numerically accurate rendering of targets I simply seek not to clip any non specular highlights or lose any detail I can see if I can prevent it. In test shot I didn't lose the separation between the 1.5 patch and frame so that met my criteria. In an actual photo the viewer would know or care if it was numerically correct, they would simply expect to see detail.




Jan 24, 2012 at 09:03 PM
williamkazak
Offline
• • • • •
Upload & Sell: Off
p.6 #15 · Strobist


dmacmillan wrote:
Cute cat! Poor fellow, looks like his/her left eye is clouded. Is this an older cat?

This is a cat that the Art Gallery owner, David, found in the streets all torn up and malnourished. I guess it "lost" it's eye or whatever it is called but the cat is gentle and appreciates good food and the run of the gallery. When David got a donated couch, "Ducket" immediately laid claim to one end of it. No more sleeping on the floor in a box", he said.
Jan 24, 2012 at 10:30 PM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #16 · Strobist


cgardner wrote:
It amuses me to be criticized for being too technical and not creative enough by some and not being technically precise enough by others.


I pointed out the discrepancy in your grey scales and sought no reason to criticise, but if you're offering statements of facts based on your equipment tests - you better make sure your house is in order.

My approach UK is perceptual, not by the numbers. My goal in that test was not to render the 1.5 target 37, 52 or any other specific target as an exact numerical match. I never said I used the target that way. I said I match target with the results I see in the final product by side to side VISUAL comparison, not absolute numerical comparison.


Your persistence to illustrate your experience with other tenacious connections didn't ratify your failing, but simply aggravated matters further with you then using the targets for reference.. You can't use a target for reference unless its created correctly.

Your HSS results are underexposed, therefore your results are at fault, that means they are wrong. VISUAL comparison is really too late after you find out that the results are wrong, and just looking at the target should have told you they were wrong or something was amiss by the histograms when you were twiddling with them in Photoshop. You didn't, yet still pasted inaccurate text on them.

If i wanted to do that I'd own a densitometer like those I used at work for many years and measure the reflected density of the actual target I use and the prints I make. But I don't often make prints and any comparison is between target and screen VISUALLY.


How you hope to reproduce a colour in print which is incorrect on screen is totally beyond me since every step in the process is reliant on the previous being correct.

I told you about reflecting your lights off your targets around two years ago. All this could have been avoided.
Jan 25, 2012 at 05:16 AM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #17 · Strobist


UK: My goal in life is not to please you, nor should yours be to correct what you see are to errors of my ways. I don't expect you to follow my approach, and you shouldn't expect me to follow yours either. Let's agree to disagree and move on shall we?
Jan 25, 2012 at 05:25 AM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.6 #18 · Strobist


Frankly, I don't care that you don't want to please me, but when I'm being lectured by a self qualified expert using hairbrained perceptual methods and ignoring regular professional practices then I'm not going to shut up and go away and I will point them out. So yes, I'm happy to disagree, and I'm absolutely fine with that.


Jan 25, 2012 at 06:01 AM
Beni
Offline
• • • • • •
Upload & Sell: Off
p.6 #19 · Strobist


ukphotographer wrote:
but when I'm being lectured by a self qualified expert using hairbrained perceptual methods and ignoring regular professional practices then I'm not going to shut up and go away and I will point them out.
Jan 25, 2012 at 06:10 AM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.6 #20 · Strobist


ukphotographer wrote:
Frankly, I don't care that you don't want to please me, but when I'm being lectured by a self qualified expert using hairbrained perceptual methods and ignoring regular professional practices then I'm not going to shut up and go away and I will point them out. So yes, I'm happy to disagree, and I'm absolutely fine with that.

The broader theme of William's "Stobist" thread was why out of all the approaches used over the years did the "Strobist" approach gain traction, and he asked who did it gain traction with?

Hobbyists. What is my approach designed to do? Address the needs, budgets and skill levels of HOBBYISTS.

Non-profession hobbyists coming to here for advice are for the most part have limited budgets and are totally ignorant about artificial lighting beyond a flash in the hot shoe. Being a hobbyist I face the same budget constrains, have chosen my equipment accordingly, then adapted my knowledge and previous professional experience to best use that equipment.

I didn't always use white towels. Prior to buying into the Canon DSLR system I had always used manual flash using power and distance to control ratio and exposure. I didn't need to meter the lights because my manual flash approach, learn from Zucker and put into practice professional for a couple years shooting wedding for him did not require it to achieve consistent professional quality result on color negative film and prints or transparency film. As mentioned, after "retiring" from the wedding grind I shot my white and black dogs on Kodachrome using the same "professional" approach learned from a top professional (who paid me while learning) and had no problem recording detail in both.

Switching to digital gave me three new tools for judging exposure: playback, histogram, and clipping warning. Having used Photoshop since V1 in the early 90s and used digital cameras since buying an Apple QuickTake in 1995 by the time I bought my Canon 20D and 580ex flashes in 2004-2005 there wasn't much I didn't know about color management or digital image capture and digital reproduction. That is what I did for a living "professionally". I don't know what you were doing in 1984, but I was getting paid to lecturing to other graphic arts professionals, including photographers, on how best to shoot their photographs and design their publications for optimal cost effective printing...

http://super.nova.org/TP/Poster1984_1.jpg
http://super.nova.org/TP/Poster1984_2.jpg

I did the same thing back in the late 70s as production manager at Judd's Inc. Once a year the company would hold a 2 day seminar for customers whose magazines we printed like McGraw-Hill, Architectural Record, AIA Jounal, AMA Journal, AOPA Pilot, etc. Not to teach them design, but how to design cost effectively by understanding the limitations of the production process.

I don't know what you were doing "professionally" in 1990-91, but I was preparing for my second overseas assignment managing the publishing center in Manila and buying, installing and training staff to use digital scanners and imagesetters to convert the mostly analog workflow to digital, and going to places like Hong Kong, Tokyo, Bangkok and Rangoon to teach them how to use Mac based DTP for their locally edited 4/C magazines we printed for them in Manila.

We had started using Mac based DTP and graphics in 1987 and bought any software that hit the market for R&D. Photoshop arrived on my desk around 1991 on 800K floppies. In 1995 we bought an Apple Quicktake 100, but it wasn't until late 2000 while back in Manila for my third tour there, this time as Director of the facility I spent my own money for a Kodak DC290 digital.

I was already on the net back then and participating on a Yahoo Filipino photography forum. I already knew many of the people on it from working on the Internet launch there in 1994-95 when I was web master at the only commercial ISP and taught seminars about the Internet jointly sponsored by USIA and the ISP. The people on the group ask me to hold a "face-to-face" class and needing a venue I called Kodak Philippines who knew me already as a customer of their graphic arts products at work and got them to sponsor it at no cost to participants. I was asked by one of the attendees to repeat the class at the Graphic Arts Show she managed and be the keynote speaker. Since all the participants had net access I did the class notes in HTML and posted them on-line to save the cost of printing handouts. It's still on line at: http://super.nova.org/PhotoClass/ In 2001 in the Philippines most were still shooting film due to high import duties on equipment. In the first class I only covered the first three parts — in was a one-day six hour seminar with a lunch break.

Like my approach now it the advice was targeted primarily for hobbyists using a camera typical of what they might buy at the time. The main theme was while good gear helps, what I more important is understanding the underlying cause and effect. Once that is understood one can objectively decide what gear it the best value for their goals and free from technical stumbling blocks competently follow their creative muse to the limits of their imagination.

For the second class at the Expo which was attended by mostly graphic designers vs. photographers I focused more on the post processing reproduction aspects such as color management that wasn't even in the vocabulary of photographers and designers back then.

I was paid quite well throughout my career — once I left "professional" photography and could afford any gear I wanted. I bought a med. format camera, lenses and color enlarger in the late 70s when I had space for a darkroom again, and in the 80s and 90s did mostly underwater photography with flash. When switching entirely to digital in 2000 I could have afforded a "pro" body or Canon D30 that hit the market at the same time understanding the underlying technology and only having the needs of a hobbyist I knew first generation gear would be a bad investment.

Had I been a professional photographer I would not have switched entirely to digital in 2000. Remember, what I did for a living since 1974 is reproduce the work of professional photographers in magazines and in 2000 digital wasn't up to professional reproduction standard for anything other than small (file dimensions / 300) images. My 2.1MP DC280 could make a 4 x 6 about on part with my film camera. The 5MP D7Hi I bought two years later could make a 5 x 7. I waited until 2004 and the 20D meet my personal "buy it" benchmark of 8MP and under $1,500 with an ultra wide angle lens available. I knew, again from 25 years of work experience in "professional" reproduction and the fact I rarely make a print larger that 8x10 it would suit my needs for several years.

I still use the 20D as backup to my 50D. I bought $2,500 of studio lights not because I needed to make a living but because I was curious. I switched to Canon flash from Vivitars to take advantage of the camera's advance features. Not all the time with ETTL flash or HHS, but to have it available for when I found it more convenient to use.

Had I been a professional photographer I would have made different gear decisions. But I'm just a hobbyist and my #1 priority for gear isn't whether it will meet the needs of my clients as yours are, rather whether it is within my self imposed "toy" budget and convenient to use. I tend to be frugal and have different goals than yours, one of which was to save and be able to retire at age 50. I could have, but hung in another five years because the money was very good and the job still interesting

Yes I am a hobbyist, giving advice to other hobbyist, but finding myself constantly defending and trying to explain my approach and methods which is largely based on professional work experience in reproduction. I'm just a hobbyist, sort of like how an ex-PGA tour pro who did ever made past the Nationwide or European tour is a golfer. I still enjoy the game and others find they can benefit from my experience and advice. Not professionals mind you, just fellow hobbyists who enjoy the game but are struggling to keep the ball in the fairway.

The advice give is the same Ansel Adams gave me via his books, Monte Zucker gave me via hands on training and daily C&C, reinforced by 30 years of photographic reproduction experience:

PHOTOS LOOK MORE REAL WHEN THEY HAVE A FULL RANGE OF TONE SEEN BY EYE

The origin of the "White Towel" method was a result of switching to Canon flash and using it in ETTL mode. I didn't meter with manual speed lights because I worked systematically by distance. With ETTL one must rely entirely on camera feedback to correctly exposed highlights.

In ETTL mode you don't need to worry too much about the shadows in the flash lit foreground because the lighting ratio controls shadow tone. One simply needs to know what ratio will put sufficient detail in shadows to met THEIR criteria for acceptable exposure, then adjust FEC until the highlights have detail.

Last time I checked the RULES OF PHOTOGRAPHY handbook it was filled with blank pages. So I can choose to make my shadows devoid of detail, the same as seen by eye, or even lighter than seen by eye. When not shooting range tests I do the latter. With experimentations I found by starting from a ratio of A:B = 1:2 with my approach the resulting file will usually have lighter than seen by eye (i.e. normal looking shadow). In other works I set the ratio on the flash so the flash creates a scene range, in the foreground at focus distance, which is LESS THAN SENSOR RANGE WHEN HIGHLIGHTS ARE EXPOSED ACCURATELY.

That no different than my goal had been with film. But the problem I found with flash shots with very small highlights was: 1) the large metering zones didn't accurately measure them and adjust the flash, and 2) the highlights were also so small they did not register on the histogram.

I didn't have the same problem in situations where the background was dark and the highlights were relatively large because when I saw the right side of the histogram running of in the first shot as in this overexposed frame...

http://super.nova.org/TP/ClippingHighlights.jpg

I could see, react and adjust exposure lower. Here's the ex F-18 fighter jock
my oft seen towel shot without the towel SOOC with the Levels histogram superimposed...

http://super.nova.org/TP/TowellessGary.jpg

See the problem? The histogram is useless for judging exposure in the highlights when they aren't big enough to register.

My common sense solution? Add larger highlight area to the photo for a test shot AS A PROXY FOR THE HIGHLIGHTS THE CAMERA COULD NOT DISPLAY...

http://super.nova.org/TP/TowelGary.jpg

See the difference? The towel when accurately rendered creates eyedropper readings in the 230 - 254 range when exposure is adjusted just below clipping, which in a dark scene will create an easy to see hump. Adjust exposure until the hump just kisses the right edge and simple as pie, perfectly render highlights towel — AND THE FACE NEXT TO IT.

Mind you this was soon after getting my 20D when I hadn't experimented with the clipping warning.

When I started using the clipping warning in addition to the histogram in the playback, exposing per the histogram towel spike, I noticed that when I got the highlights dialed in via the histogram the clipping warning also disappeared. The RAW file would be a bit under in the highlights, but by the time I made the JPGs they would be back similar to what the clipping warning in the camera shows.

The advantage of the clipping warning vs. histogram was that it shows WHERE highlights are clipping. So if the white towel was in front of a white background the I could tell via the clipping warning when adjusting foreground and background lighting when each was clipping INDEPENDENTLY.

That gave me the ability to adjust lighting like this without a meter, based entirely on camera feedback. Not because I was anti-meter on principle, but because I COULD NOT METER ETTL FLASH OR MANUAL FLASH WITH MY CANON FLASHES.

http://super.nova.org/TP/WhiteBGTowelCard2.jpg

Do I do it perfectly per UK's exposure rules? Apparently not, but it was for me a very practical solution that turned the histogram and clipping warning into a powerful diagnostic aid when setting exposure and ratio.

I did not, until recently, include a black towel in exposure shots. I would look at the separation of the border and black patch on the target, and the 3D knobs on the stand and clamps holding the target. The problem with flat targets, besides glare, is the don't show how light will hit 3D faces. I don't aim for a numerical match, I adjust shadows until they look as seen by eye perceptually are not running off the side of the histogram on the left.

The problem reading shadows accurately via the histogram is the same as the highlights. On a light background small ones don't register. My solution? Make them bigger in the test shot by adding a 3D object which is as dark as any detail in the scene. Anything black will work.

The fact I have several other objects on my stand when setting lights has doesn't mean I use them for setting lights. It a bit like haven't both a thermometer and barometer next to each other on the wall. If you want to know water is in a liquid state you'd use the thermometer to see if the temp is between 0 and 100°C not the barometer.

What is not blurred on the target UK had a hissy fit about is all that is pertinent to how I use the towels, histogram and clipping warning...

http://super.nova.org/TP/CardTargetHistoBlur.jpg

When using speediights I first adjust white towel by pushing it to the point of clipping using the brightest light hitting it, which in that shot is the hair light. Could I do that with a flat target? No, which is why I uses a white towel. After setting highlights I check the detail is seen in the folds of the black towel and adjust fill as need. The HHS done in 2010 didn't have a towel. My method has evolved with time.

With studio lights where I can control the lights independently I have all the lights on but below the point of correct exposure so there isn't a huge difference in spill fill as I adjust them. I finalize the fill first based on the tone in the detail in the towel normally placing them at Zone 1 for a "normal" contrast or lighter (Zone 2) for a "softer" look, then accent to Zone 9 below clipping, the key light until it looks in balance (Zone 8) then finally the background lighting if I;m using any.

Keeping black and white towels between the point of clipping is like keeping water in its liquid state, not freezing anywhere, not boiling off either. It is not necessary to have a thermometer to know if water is between freezing and boiling, or know the exact temperatures that occurs (which varies with elevation in any case) you just need to look at the water. If it isn't solid and isn't bubbling and steaming its liquid.

A basic tenet of process control is that you can only control precisely what you can measure with accuracy. That's all the towel thing is about — making the histogram and clipping warning more effective and useful. I you have exposure parameters that are different than mine, the same method would be useful to you, you would simply need to use different tone targets. For example if for some reason you wanted a .45 density object to clip you'd want to use an object with reflective density .45 with the clipping warning.

In concept adjusting exposure per the end points with towels is not any different that measuring it off the middle of the range by centering the histogram with a 12% card. But the way I look at the problem if your nose itches (highlights) or your toes itch (shadows) to you solve the problem by scratching your belly in-between?

While a professional open minded enough to try my approach might find it useful in situations where exposure must be set out of necessity by histogram and clipping warning (e.g. the incident meter dies) it mainly geared towards the use of speedlights that can't be metered as a way for hobbyists starting out can use to both control exposure and better understand the cause and effect of fitting scene to sensor and how doing that improve the appearance of the image content. It will not make their photos any more creative, just more competent.

I trust if they get to the point of see a target patch is 20 points low they will also have the skill level to remedy it if they find it is a problem. That the broader point of the exercise.
Jan 25, 2012 at 11:36 AM
1       2       3       …       5      
6
       7       8       end
FM Forums | Lighting & Studio Techniques | Join Upload & Sell

1       2       3       …       5      
6
       7       8       end
    
 

Welcome back
Log in to your account
Reset passwordNo account? Register