Notes - Feb 15, 2016
Ok, I have decided to learn as much as I can about Nikon’s CLS (Creative Lighting System). Right now, the only thing I can see is creative is how much it confuses in how it exposes images under various conditions. Since I mostly use flashes off-camera, what CLS does while the flash is on-camera is less interesting to me. Unfortunately, off-camera flash is much more complex and unpredictable.
Sources (so far):
http://nikonclspracticalguide.blogspot.com/2008/01/nikon-flash-two-separate-metering.html
Any of Tom Hogan’s stuff, including his D810 e-book.
Random notes so far.
1. Nikon has termed their lighting technology iTTL. Camera menus generally refer to TTL. There is also CLS. I’m not sure there is any difference between these.
2. Cameras have two difference exposure systems - one for ambient (non-flash), one for flash. These systems can work independently or collaboratively.
3. On-camera, TTL flash modes are either TTL or TTLBL. TTL uses only the flash exposure system and exposes for the the “main subject”. It does this through center-weighting the exposure so that whatever is in the center of the frame is correctly exposed. TTLBL uses both the ambient and flash exposure systems. It takes readings from both to balance the exposure from the ambient background to the main subject. Presumably, the combination is an exposure that averages to 18% gray as it would be for a regular ambient exposure.
4. When the flash is off-camera, controlled by the CLS commander, it’s not clear how the exposure is determined (TTL or TTLBL). The procedure is that the camera commands each flash group to transmit short “pre-flashes” which enable the camera to determine the correct power setting for each flash to illuminate the scene. It then sets the appropriate power level into each flash and takes the picture. With multiple external flashes, placed randomly, it’s hard to understand how the camera could get it right in terms of exposure.
What appears to be the case is that, when the flash is off-camera, it exposes for the entire scene (18% gray).
So, time for some experiments.
Head shot. Light primarily determined by flash.
This example is where both the subject and the background are primarily illuminated by the flash. The ambient exposure is set so that very little light reaches from the ambient.
Here is a shot, without flash, aperutre priority, f/4, 1/20 sec. ISO 100
This is typical of an exposure against a light background. using an incident light meter, I exposed for the meter reading:
1/13 sec, f/4, ISO 100
This require 1/2 stop additional exposure.
All subsequent images, with flash were f/4, 1/60sec, ISO 100.
Next is an image taken with a SB910, on camera, TTLFP, Aperture priority, f/4
Next was the same settings except TTLBLFP.
Notice that the last picture is darker by ~1stop from the TTL image. Since the entire image is illuminated by flash, I’m speculating that the difference is due to the exposure with TTL is center wieighted whereas the exposure with TTLBL is for the whole scene, balanced to 18% gray. The histogram is quite similar to the first image, taken without flash and matrix metered. Similar results were seen independent of the position of the flash head or if the diffuser was on or off.
The flash was now taken off camera and TTL triggered, right next to the camera.
Notice that the image is dark with a similar histogram to the on-camera flash, TTLBL. I’m guessing this is because, off camera, the flash is set for an exposure of the whole scene, not center weighted like TTL.
If this is the case, it explains why, under these circumstances (head shot with a close-by, light background) it is necessary to add ~1 stop of flash comp to get a correctly exposed image. This adjustment is shown below.
As a further experiment, I added the Pocket Wizard remote triggers (Flex, TTL) and took a shot with the flash in the same position as above but with no flash comp.
Mysteriously, the exposure looks more correct, even without compensation. Don’t know why and will investigate later.
Conclusion.
1. On camera flash. In this circumstance, there is a difference in exposure between TTL and TTLBL. The first exposes for “main subject” i.e center-weighted. The second exposes to balance the whole scene (18% gray). Because this scene has a light background, the main subject is under-exposed by ~1stop.
2. Off camera flash. It appears that the camera exposes for the entire scene and, again, results in under-exposure of the main subject. I guess this makes sense since the camera doesn’t know where the flash is so it can’t make any assumptions about where the main subject is. Nevertheless, this is something to keep in mind when taking head shots with a light background. Ideally this should be done with manual exposure. However you run into the problem of pre-flashes screwing up the flash meter. Interestingly, this does not seem to be a problem when using the Pocket Wizards. However, requiring a Flex for each flash is a bit expensive.
3. Confused why the pocket Wizards give a different result than CLS. Will need to investigate further.
Notes. February 27, 2016
Next I wanted to try CLS when the background was bright and distant, thus not affected by the flash. Our flat in San Francisco, overlooking the Bay, on a bright day, was the ideal setting.
First image is full-auto. Aperture priority, no exposure or flash compensation, off-camera flash (to the right), triggered by the on-camera flash in commander mode.
The camera has chosen to make the backgound dominate, thus the subject is slightly under-exposed relative to the background. In the next image, I dialed back the camera exposure (-1EV) leaving flash comp at 0EV..
The background drops in brightness as expected. However, the subject stays the same brighness suggesting that the flash exposure is solely determined by the flash itself. The next image has +1EV dialed into the flash exposure. Camera exposure comp remains at -1EV
Here the background remains the same, as expected. The subject increases in brightness by the expected +1EV.
Lessons learned.
1. In full auto mode, the camera makes a good stab at getting exposure and flash/camera balance correct.
2. With this scene, camera exposure controls the brightness of the background only.
3. Flash controls the brighness of the image only.
4. Without resorting to a light meter, the process is to start with zero comp and add camera comp and flash comp until the desired effect is achieved. probably quite a quick process.
Just for fun, I also took a shot, manual exposure to see how “FP” mode worked.
Seemed to work just fine but I needed to set ISO at 800 to get the flash to illuminate the subject properly.
Mastering the use of Gel Filters in Flash.
In the Joe McNally workshop I noticed he would create an effect where the background was deep blue but the skin tones, lit by flash, in the foreground were just right. He did this by setting the camera wihite balance to “Tungsten” (incandescent, ~3200K) and then gelling the flash with an orange filter to neutralize the blue effect for the portrait. I liked this idea and wanted to experiment with gels.
First up, the SB-900 has a neat feature where if you insert a (nikon approved) filter, and the camera is in “Flash” or “Auto” WB, then the camera adjusts the WB in order to make any subjects soully lit by the flash + filter, have the correct WB. I experimented by taking a picture of a colorchecker card in an unlit room illuminated only by flash. In “Auto” mode, the camera almost perfectly compensated for the orange filter (SZ-2TN) (Gray - 102,105,110 versus 105,106,110 without the filter) but required +1 stop exposure comp
Next up, I set the WB to “incandecent” and took another picture. This time the image was a little warm. Not quite balanced. I found by adjusting the WB to the following, I was able to achieve almost perfect WB:
Incandescent B2.00 G1.00 (gray - 98,100,99)
Note that the setting B0.00 G0.00 (Default) was slightly warm and pink (gray - 108,103,100)
Here are the results of my analysis:
Auto FP High Speed Sync Explained
One of the most confusing aspects of the Nikon flash system is this thing called 'Auto FP High Speed Sync'. This blog will hopefully clear up the mystery about this mode.
Before you can understand FP High Speed Sync, you have to understand what FP means and how the Normal Flash Sync works.
FP stands for Focal Plane and it refers to the type of shutter used in most modern DSLR cameras. A focal plane shutter is actually two precisely timed curtains positioned between the lens and the sensor that can either block light from hitting the sensor or allow light to hit the sensor. The reason there are two shutter curtains is to be able to get much higher effective shutter speeds.
It is important to understand is that these curtains open and close in exactly the same amount of time. So the the shutter speed is set by timing between the start of the first curtain opening and the start of the second curtain closing.
Notice that the entire sensor will be open to the light at every shutter speed up to the speed of the curtain movement itself. This is the maximum normal Flash Sync Speed. To say this another way; at all speeds up to the maximum normal Flash Sync Speed, the first curtain completely opens before the second curtain begins to close. At any shutter speed higher than this, the second curtain will begin closing before the first curtain gets fully open, thus never exposing the entire sensor at any one time. At really high shutter speeds, this results in very narrow 'slit' of light that travels across the sensor.
These two curtains travel vertically across the opening from the top to the bottom, and this in itself causes some strange effects when using high shutter speeds (small slit). If you shoot something that moves horizontally really fast like a race car from the side, you can sometimes see that the wheels seem to lean forward a bit because the top part of the wheel was exposed after the bottom part and the top moved forward a little bit as the slit moved from bottom to top. Of course, the image is inverted on the focal plane, exposing from the bottom to the top of the image, and that's why race car wheels lean forward even though the shutter that moves from top to bottom. Also, this effect only occurs if you don't pan with the car. You can totally eliminate the forward leaning effect by panning.
As I mentioned, a focal plane shutter mechanism moves the curtains at a very precise speed. This speed is determined during manufacturing of the mechanism and is governed mostly by how recently the shutter was designed. In older 35mm cameras, this speed was 1/60th second, but with time, shutters got faster and faster, and in the new D300 this speed is 1/320th sec. And those race car wheels lean much further forward with an older 1/60th shutter than with the D300 1/320th shutter.
Once you understand how the shutter works, you can begin to see what needs to be done to synchronize the flash. Depending on the design of the flash, the length of a full power flash will vary. In an SB800, the maximum flash lasts about 1/1050th sec.
The key point is that the flash, firing at 1/1050 sec is much faster than the curtains which move at 1/250th sec on the D200 (or 1/320th sec on the D300), so to synchronize the flash it must fire at some point after the first curtain has fully opened, but before the second curtain begins to close, so the sensor is completely exposed to the light from the flash coming through the lens. You may also see that you have a choice of when you fire the flash; either right after the shutter opens (front curtain sync) or right before the shutter closes (rear curtain sync). The timing of the flash has profound effects on the light trails when using very long shutter speeds. It will place the light trails either in front of the subject when using front curtain sync and behind the subject when using rear curtain sync. Rear curtain sync usually looks more natural for light trails.
Up to this point I have been only discussing the normal flash sync. However, Nikon has added a mode called 'Auto FP High Speed Sync'. You select this mode in the menu on the camera. In this mode, the flash duration is stretched so that it fires continuously for the full curtain travel time, ie, 1/250th sec on the D200. So, the flash pulse that normally takes 1/1050th sec is now stretched across 1/250th sec. The official name for this operation uses the word 'Auto' in front of 'FP High Speed Sync', because in this mode the flash will sync normally and fire normally below the flash sync speed, but it switches automatically into High Speed Sync (stretched flash) when the shutter speed is set higher than Normal Flash Sync Speed.
This all sounds like a pancea until you find out that stretching the flash pulse and using higher shutter speeds causes the power of the flash to be reduced from what it normally is. In fact, the times where you would really like to be able to use a mode like this, like fast action sports in bright daylight, you usually can't, because the high speed flash sync mode is not powerful enough.
In fact, the power becomes less and less as you increase shutter speed, because the slit gets narrower and narrower. So, in this mode the flash is dependent on aperture and shutter speed, and if the shutter speed is increased to 1/8000th sec, the power gets so low that a subject would have to be less than four feet away.
However, when stopping action is not the goal, and a really high shutter speed is not required, then FP Sync can be very useful. In fact, Auto FP is excellent for shooting portraits in bright daylight. Then, you can use a shutter speed well above the normal sync of 1/250th coupled with a wide aperture to blur the background which greatly helps isolate the subject. I typically use camera A mode, ISO 100, FP Sync, f/ 2.8 and 1/1000th in bright shade, and I get a flash range of about 10-15 feet. If I am in really bright light, like on the beach, the shutter has to be around 1/1600th, and this reduces the range to about 8-10 feet which is still reasonable for portraits.
If a greater distance than 10 feet is needed in bright daylight, there is no choice except to use regular flash sync and accept the requisite smaller aperture. This happens frequently when shooting a party outdoors, where I typically shoot in camera S mode, regular flash sync, ISO 100, 1/250th, and f/7, and I get acceptable fill out to 20 to 30 feet or so.
POSTED BY RUSS MACDONALD AT 1:54 PM
Note: I own a D200 but not a D300. They are very similar but not identical, so I am using the D300 manual to help me with this.
Here are the steps:
1. With both the camera and flash OFF, mount the flash on the camera hot shoe.
2. Turn on both the flash and the camera (order doesn't matter).
3. Go to menu e1 on the camera and select 1/250 s (Auto FP) or 1/320th (Auto FP). Hit 'enter' to save it.
4. 'Wake up' the flash if it is in STBY by half-pressing the shutter button. Then, on the flash press the 'Mode' button several times to cycle through all the modes. You should see 'TTL FP' and maybe 'TTL BL FP' among the modes. If 'TTL BL FP' is not shown, you probably have your camera in 'Spot' metering mode. Switch it to 'Matrix' to allow BL. Set the flash in 'TTL FP BL' mode for fill flash testing.
5. On the camera select Manual mode and verify you can set the shutter at speeds above 1/250, all the way up to 1/8000.
Now, you should be able to select camera A mode and shoot in bright light with fill flash.
However, the higher the shutter speed you select, the less flash power you will have.
Hope that helps,
Russ