cgardner
Offline
• • • • •
Upload & Sell: Off
|
erichard wrote:
hobbyist vs. Hobby-ist
I have a question. I haven't kept up with this forum, but has anyone challenged the principle assumption of using a white towel each and every time? .... I think it'd just be better to look for clipping in areas where you expect no clipping... One technique that I use, that I've not seen written about elsewhere, is to actually find the exposure where critical clipping disappears on the LCD and then up the exposure one notch (1/3 to 1/2 stop).
The White Towel thing is consistently trolled and mocked but never challenged on it's technical merit or proven not to work.
It has it's genesis back when I was doing Zone System B&W. Adams, who was both craftsperson and artist always strived for a full range of detail in his photos similar to what might be seen as someone in person scanned and their pupils adjusted to highlight and shadow. Then once one managed the technical side so the photo fools the viewer's brain that a 2D reproduction is real you put something interesting in the photo.
My next job was shooting weddings for the photographer how popularized the two flash approach used by most wedding photograhers in the 70s and 80s because by using fill to lift the shadows and then overlapping key light it is quite simple indoors — no brainer simple — to record detail everywhere in the groom's black suit and bride's white dress all at the same time with color/print film which has a shorter range.
Capturing the image on the negative got fit the overall scene range to the final product, a print. Why was it important to record the full range as seen by eye? The simple fact photos look better when highlights aren't blown out or too gray and shadows have detail. That was possible with B&W without using flash, but not with color film because it had half the range. You would better understand this had you shot and developed your own B&W and color prints as I have.
With film photography the process didn't end at capture. With both B&W and color we routinely "post processed" by dodging and burning the midtones to alter them under the enlarger, a process similar to making shadow puppets on a wall with your hand. The paddle and cupped hand dodge and burn icons in Photoshop are a homage to those old manual techniques.
The requirements of a print made with a digital camera aren't any different from that of one made with film. They look more real when the white paper base with no ink is used to reproduce the brilliant sparkle of catchlight specular reflections and there is a full range of detail down into the shadows that match what the eye sees in person, not just what the digital sensor is able to capture. Flash is only necessary when the sensor can't match the scene range. You don't for example need flash on an overcast day or in open shade if no direct sunlit is visible in the shot. The used of two flashes in a key / fill arrangement changes the scene range to fit the sensor.
Why fit scene to sensor? That's what is required to get a full range of detail on the print that matches what was seen by eye so the print looks "normal" not over- or under-exposed.
Because the range of a sensor is shorter than outdoor lighting most of the time it is impossible to match scene to sensor most of the time, unless one uses flash or bracketed exposures and HDR post processing blending of the exposures.
Learning to control the technical aspects of reproduction is like driving: experience and practice increase understanding of how a car works and how to control it at will.
When you start out clueless about how the pedals and wheel react you are likely to wind up in a ditch because you don't know how to control the car. Once you learn how to drive a car and control it with precision there is nothing to prevent you from driving the car off into the ditch if you feel like it (i.e. intentionally blow highlights or block shadows).
When you master the skill of driving a car you will gain the ability to instinctively react to situations, like the back-end of the car sliding out on a slippery curve or a deer running out in front of the car.
At an even higher technical skill level one is able to use subtle feedback from the car to allow it negotiate the curve sideways in a four-wheel drift with front and back end kissing the two sides of the road. I used to love to drift my 68' BWM 2002 around the traffic circles in Washington.
The latter is what fitting scene to sensor is — exposing the scene so shadows and highlights just kiss both sides of the histogram.
All images, including a black cat on a coal pile, have critical highlight detail. On a black cat and coal most of the clues come from where the 255 specular reflections glint off the tops of the hair shafts in the fur and angular high points of the lumps of coal. The cat has white teeth and eyes. Not 255 white mind you. But to reproduce the scene and make EVERYTHING wind up looking real you would want the catchlights to be 255 and the teeth to be around 245-250.
The same is true with white on white. In that situation a very narrow range of tones define shape on objects such as the 3D shape of the gently curved hood of a white car. So its is even more important to control exposure so most of the car is rendered at around a 235-250 eye dropper range and on the higher parts of the surface 255 catchlights are seen.
With a white car or bride's dress its easy to get them rendered optimally by first raising exposure until the solid whites are clipping per the camera warning THEN BACKING OFF 1/3 stop. If you do that when the file is opened for editing you will find the specular highlights are clipping at 255 but the parts of the dress or car where you saw detail by eye are in the 235-250 range. The clipping warning is like a radar detector that prevents you from getting caught speeding.
What you perhaps don't appreciate yet is that the same control of highlight exposure is needed in the shot of the black cat in the coal bin. The 3D shape depends on the 255 specular highlights and ideally you want the teeth and eyes in the 235-250 range just below that. The problem when shooting is being able to see when that is happening. Unlike the white car or dress there is no big white object in the scene so you don't see the highlights very well or at all in either the clipping warning or the histogram bar height. That's the type of situation where I would "throw in the towel".
With the white towel sitting next to the black cat I can see when the towel, which has about the same range of tones as the whites of the eyes and teeth, clips in the playback. I then do the same thing as with the white dress—raise exposure until I see the whites start to clip, then back down 1/3 stop. In the image when opened I will find 255 specular reflections and correctly exposed teeth.
Using a white towel as an exposure aid is no different in concept than the practice years ago before spot meters of putting a gray card or the palm of the hand out in the same lighting as a scene and taking a meter reading off of it. Due to the properties of the neg/print process and how meters worked that was that was needed to produce a full range B&W print. With color exposing to the gray card would make it easier for the lab to get the faces looking normal, but wouldn't ensure there was detail in the sky or deep shadows. The same is true of digital. Assuming you own a gray card if you filled the viewfinder with it in the same light as the scene then used the same shutter/aperture/ISO to shoot the scene in your photo you would find a perfectly exposed mid-tones, but if scene exceeded the sensor range exposing for the middle will blow highlights and lose shadow detail.
When a scene does exceed sensor range and you don't have a way to change the scene lighting and contrast you are forced to decided whether to expose for the faces, if there are any in the shot, the sun lit highlights or the shadow detail. It's like moving and having more stuff than will fit on the truck and deciding to what to leave on the curb. If you actually try what you've read about — over exposing highlights by 1/3 stop — you will leave the highlight detail of your image on the curb.
Blowing highlights can work fine, if there is no important detail in the highlights. In the case of the black cat if the cat is small in the photo the difference between clipping the eyes and teeth vs. rendering them correctly probably will never be noticed by the viewer of the photo. But if it were a close-up portrait of dear little Midnight for the cat's owner you'd need to find some way to record both the teeth and the black parts of the cat accurately with detail at the same time. To do that you would need to learn to use flash to fit the range of the cat to the sensor. Once you learn how to do that you can not only shoot a decent close up photo of a black cat you will be able to photograph a black and white cat sitting together, both with life like detail?
How to you gain that skill? Practice controlling the scene range with a target the one I use when setting my lights in the studio:
http://super.nova.org/TP/CardTargetHisto.jpg
Turn on just the fill. Raise fill until you see detail in the folds of the black towel in the playback.
Then turn on the key light.
Raise the key light's power until you see the white towel start to clip, then back off until the clipping warning disappear.
What you will find in the file is a full range of tone. Now put anything in same spot and shoot it with the same lighting. It will have the same full range of tone and perfect exposure.
You will need to do this a few times before you'll be able to accurately judge the RAW shadows and highlights based on the camera feedback. But once you do you'll have the skills to record a full range of tone anytime at will. That will also give you the skills to opt to clip the highlights or shadows by a precise amount, like drifting the car with one wheel up in the air, not because it's the best way to control the car in a drift, but because you can.
When you can do all that please come back and challenge the white towel thing based on your experience and suggest what you think is an easier approach, especially if using speedlights that can't be hand metered accurately.
Canon flash is like my Mazda MX-5. I could have bought a rag top with a 5-speed manual transmission. I previously own a manual sports car and commuted on motorcycles for over 30 years so I knew the pros and cons of manual control. I opted to instead by an automatic with a power hard top. Not an automatic like my first car, a '68 Dodge Dart, but one with a 6 speed automatic with F1 style paddle shifters. I figured if not using a clutch worked for F1 it would work OK for me. A test drive of both manual and automatic versions confirmed that. I can up and down shift at will the same way as with a clutch, only faster because I now do it with my fingers without either hand leaving the wheel. Hard vs. rag top was for convenience and security - it goes up and down in 12 sec. and I don't worry about someone cutting the top to break into it.
After getting the car I waited about a month for an overcast day to photograph it because I knew from experience that the low contrast of the light would allow me to fit the entire range to sensor without flash and create the uniform highlights needed to render the 3D shape of shiny objects. I used the old trick of wetting down the pavement to create darker reflections on the side of the car. I didn't use any flash for the shot because again experience told me it wasn't the best tool for the job. Natural light and the patience to wait for just the right natural light was. 
http://super.nova.org/TP/MX-5_01.jpg
How did I learn to drive? Not like most of my peers in high school. My junior/senior years in high school I worked for the school's driver ed program setting up the cones, gassing the cars, driving all over metro Chicago swapping the loaners at dealers. We had a dozen cars ranging from a stick shift VW bug, Dodge Challenger, Pontiac LeMans, and Buick Electra 225. My schedule got fixed so I had no afternoon classes so at noon I'd grab a car and take off. When the boss wasn't around I'd practice weaving through the cones as fast as I could without hitting them and did donuts and drifts in the snow. I once got the VW completely airborne over a railroad crossing. It was a ton of fun, taught me to drive quite well and paid for my diving and camera gear.
|
