TTL and manual

Learning goal

You will be able to decide, by looking at the scene, whether to leave the flash's power to the camera or fix it yourself. You will explain how TTL measures, through the lens, the light coming back from the subject after a pre-flash and chooses the power of the main burst, and how, because this measurement rests on the same middle-gray assumption as the light meter, it goes wrong with dark and light clothes and with dark and light skin. You will correct TTL's decision with flash exposure compensation; with manual flash you will set the power in fractions and use the fact that each halving is one stop, and when the distance changes you will set the power again according to the inverse square law. You will compare the recycle time that shortens at low power, and the trade-off between speed and consistency in the two methods, through a moving event and a fixed portrait setup.

The question

A folk dance group is getting ready for its end-of-year show at the district's cultural center, and they have asked you to shoot the dancers backstage. Half of them wear dark navy, almost black vests; the other half wear bright white shirts. You bounce the flash off the ceiling and start shooting with the camera's automatic flash setting. When you shoot a dancer in a navy vest up close, their face comes out much brighter than you expected, almost washed out. When you shoot their friend in a white shirt from the same spot right after, this time the face is dark and murky. The only difference between them is the clothes. The flash is the same, the distance is the same, the ceiling is the same.

Who should decide the power of the flash? If the camera makes a decision it considers right in every frame, why do the results come out so different, and what is the price of taking this decision out of its hands?

Context

In the bouncing lesson, as you turned the flash head, the path of the light lengthened and the surface changed, but you never touched the power setting. What made this possible was an automatic system calculating how much light the flash would give in each burst. This lesson tells how that system works, where it goes wrong, and when fixing the power by hand is better.

Choosing between the two methods isn't a sign of skill or a matter of superiority. Experienced photographers use both; which one they choose depends on how much what they are shooting moves and how much the frames need to resemble each other.

Simple explanation

How does TTL choose the power?

TTL is the abbreviation for through the lens in English, and it names through-the-lens flash metering: the flash metering method where the flash's power is decided automatically by measuring, through the lens, the light coming back from the subject. When you press the shutter, the flash fires twice. A pre-flash is a small measuring burst of known power fired just before the main exposure; the camera measures the part of this burst that comes back from the subject and enters through the lens, and calculates how strong the main burst needs to be. The time between the two bursts is so short that most people see a single light.

The great strength of this method is that it adjusts itself as distance and surface change. If the subject comes closer, more light comes back from the pre-flash and the main burst is turned down; if you turn the flash to a high ceiling, less light comes back and the burst grows stronger. At a moving event, in a wedding hall or in a crowded backstage area, you don't need to do a calculation for every frame.

But TTL shares the same basic assumption as the camera's light meter: it accepts that the area it measures is middle gray (Lesson 9.1). If a dark vest fills the frame, little light comes back from the pre-flash, the camera thinks the subject is dark and raises the power, and the face is overexposed. A white shirt reflects a lot of light, the camera turns the power down, and the face is underexposed. The difference between the two dancers backstage comes from here.

How is TTL's decision corrected?

Flash exposure compensation is the setting that raises or lowers the power TTL calculated by as many stops as you want. It is the flash-specific counterpart of exposure compensation (Lesson 9.3): it doesn't erase the camera's decision, it only shifts it by a certain amount. If the face comes out bright on a person in dark clothes, you set flash exposure compensation to minus; if the face goes dark on someone in white clothes, to plus.

This correction also applies to skin. On a dark-skinned face filling the frame, TTL sees that little light comes back and raises the power; the skin can look lighter than it really is, grayish, and robbed of its depth. A light-skinned face filling the frame, on the other hand, can lead the power to be turned down because it reflects a lot of light, and the skin to come out murky and grayed. If there are different skin tones in the same frame, TTL decides by the average of them all and may not show anyone exactly right. The solution is to enlarge each face in the test frame and adjust the compensation for the face; the goal isn't for the skin to be light or dark, but to stay faithful to its real color.

TTL has one more problem: because every frame is metered anew, small changes in the framing change the result too. If the person turns a little, if a white wall enters the frame from behind, if they pick up a black bag, the power changes as well. When you put a series of frames side by side, you see that the brightnesses don't match each other.

When is manual flash better?

Manual flash is using the flash with a power that you, not the camera, decide, and that stays the same in every burst. The power is set in fractions: full power, half, a quarter, an eighth, and so on down. Each halving halves the light, that is, it is a one-stop decrease. Going from full power to an eighth is three stops.

The strength of manual flash is consistency. The color of the clothes, the tone of the skin, the wall behind change nothing; the flash gives the same light in every frame. As long as the distance between the subject and the flash stays the same, every frame comes out at the same brightness. That is why manual is generally chosen for a fixed portrait setup, product shooting or a series of frames that need to resemble each other.

Its price, on the other hand, is dependence on distance. As the subject moves closer and farther, the light falling on the face changes too (Lesson 18.2): if the distance doubles you lose two stops and need to quadruple the power. At a moving event, setting the power again at every step is hard. The practical way to find a starting point with manual flash is to shoot a test frame, enlarge the face, look at the histogram and correct the power by one stop at a time. Many photographers first shoot a test with TTL and note the power it found, then switch to manual and start from that value.

Recycle time is the time it takes for the flash to be ready to fire again after a burst. At full power this time grows longer, and in fast shooting one after another some frames can come out without flash or with too little light. At low power the flash recharges faster. When working in manual, not using more power than you need both shortens the recycle time and makes the batteries last longer.

Visual evidence

When you put your first frames from backstage side by side, you see the difference clearly: the faces of the dancers in navy vests are washed out and bright, the faces of those in white shirts dark. A dark-skinned dancer's face, with the effect of the navy vest too, has come out much lighter and grayer than it is; their light-skinned friend has dropped to a murky gray tone next to the white shirt. You set flash exposure compensation to minus for those in navy vests and to plus for those in white shirts; the faces return to their own colors. But in the crowded backstage area, the result still changes a little depending on who turns where.

Then the group calls you to a corner to shoot a portrait of each dancer one by one. You put the flash in a fixed place and mark the spot where the dancers will stand on the floor with tape. With a test frame you find the power at a quarter and switch to manual. Now the one in the navy vest and the one in the white shirt get the same light; all twenty portraits are at the same brightness, and when lined up side by side they stand like a whole. When a dancer steps back one step from the mark, you see the face go dark and call them back to the tape.

Common mistakes

Summary

TTL measures, through the lens, the light coming back from the subject after a pre-flash and chooses the power of the main burst itself; because it adapts as distance and surface change, it is fast at moving events, but because of the middle-gray assumption it goes wrong with dark and light clothes and with dark and light skin, and can give inconsistent results from frame to frame. Flash exposure compensation shifts TTL's decision stop by stop; manual flash, on the other hand, fixes the power in fractions, each halving is one stop, and as long as the subject stays in place every frame comes out the same. Where speed is needed TTL is chosen, where consistency is needed manual; in both, what checks the result is the face enlarged in the test frame.

Mini assignment

Do the following tries with a flash that can work in TTL and manual.

1. Shoot the same person with TTL, first in dark clothes, then in light clothes, from the same distance; compare the brightness of the faces.
2. Shoot again, setting flash exposure compensation one stop to minus in the dark-clothes frame and one stop to plus in the light-clothes frame.
3. Switch to manual and shoot four times with the power at full, half, a quarter and an eighth; see in the histogram how much darker each frame is than the one before.
4. At manual power, send the person to twice the distance, repeat the frame, then catch the old brightness by quadrupling the power.

Check yourself

• Can you say in advance in which direction TTL goes wrong in front of dark or light clothes?
• Can you turn manual power fractions into stops?
• On different skin tones, did you enlarge the test frame and adjust the compensation for the face?

Glossary

Mini quiz

1. By what does TTL decide the power of the main burst?

• A) The lens's focal length
• B) The camera's white balance
• C) The memory card's write speed
• D) The measurement of the part of the pre-flash that comes back from the subject and enters through the lens

2. If a dark navy vest fills the frame, how does the face shot with TTL come out, and why?

• A) Exposed correctly, because TTL tells the clothes apart
• B) Overexposed, because when little light comes back the camera raises the power
• C) Underexposed, because dark clothes reflect light
• D) Its colors shift, because the vest paints the light

3. In manual flash, lowering the power from full to a quarter is a decrease of how many stops?

• A) 2 stops
• B) 4 stops
• C) 1 stop
• D) 3 stops

4. You will shoot portraits of the group's twenty dancers that resemble each other, in a fixed corner. Which way gives the most consistent result?

• A) Metering every frame separately with TTL
• B) Turning off the flash and raising the ISO
• C) Finding the power with a test frame in manual flash and marking the spot where the dancers will stand
• D) Shooting each dancer in another corner of the backstage area, changing the spot

Answers

1. D — TTL measures the returning light of the pre-flash through the lens and calculates the main power from it.
2. B — With the middle-gray assumption it takes a frame that returns little light for dark and raises the power; the face is overexposed.
3. A — From full power to half is one stop, from half to a quarter one more; two stops in total.
4. C — Manual power doesn't change with clothes or skin; if the distance is fixed, every portrait comes out the same.

Summary card

• TTL: meters with a pre-flash, chooses the power itself; it is fast and adapts to distance.
• TTL assumes middle gray: it raises the power for dark clothes and dark skin, turns it down for light.
• Flash exposure compensation shifts TTL's decision stop by stop.
• Manual: the power is fixed, each halving is one stop; if the distance changes, set the power again.
• If there is movement TTL, if consistency is needed manual; in both, check by enlarging the face.

Bridge to the next lesson

In the portraits in the corner backstage the faces were flawless, but the stage lights and colorful curtains behind could hardly be seen. When you slowed the shutter speed, the face stayed the same and the stage behind lit up; when you changed the aperture, both the face and the background changed together. It was as if two different lights were being recorded in the same frame at the same moment. Why are there two exposures in one photo? The next lesson tells how flash and ambient light are controlled separately, and which of the two the aperture and the shutter speed manage.