Guide: About exposure (3/3)


Status
Not open for further replies.

zoossh

Senior Member
Nov 29, 2005
8,725
0
36
Singapore
Please do not post in this thread in order to maintain a clean and efficient platform.
Status: Updated till 2009 Jan 03



About Exposure

1/3: Nature and distribution
2/3: Inputs and Dependent features
3/3: Judgement and Control

After understanding the various components that lead to the formation of exposure of an image, this chapter basically covers the need to understand the methodology (not the sense of aesthetics) of how to judge and review if your exposure is insufficient, adequate or excessive. Invariably, the judgement made on the 1st shot will lead to further adjustment in the control of exposure by controlling the reference (metering and exposure compensation), the input factors (filters and flash) and the output data (post processing, including HDR).



Summary: Judgement and Control

1. Judgement of exposure
1.1 Overview: Judgement avenues for correct exposure
1.2 Making fast estimate of correct exposure
1.3 Deciding exposure by experience and photography rules
1.4 Deciding exposure by metering
1.5 Feedback avenues of exposure
1.6 Reviewing by eventual output
1.7 Reviewing by LCD screen
1.8 Reviewing by histogram

2. Adjustment of exposure
2.1 Adjustment of exposure
2.2 Exposure compensation
2.3 Why do I need to do exposure compensation?
2.4 Exposure/Exposure compensation indicator
2.5 Setting desired exposure compensation and checking the set value
2.6 How should I go about adjusting the exposure compensation
2.7 Exposure compensation in a dark scene
2.8 Exposure compensation in a bright scene
2.9 Auto-bracketing function

3. TTL Metering
3.1 What is TTL?
3.2 Use of metering in different exposure modes
3.3 Concept of exposure shift and sampling area
3.4 Metering modes based on sampling area
3.5 Spot metering for fine control

4. Control and adjustment of exposure outside the camera
4.1 What is available to control exposure inside and outside of the camera?
4.2 Flash & studio equipments
4.3 The concept of strong light intensity situations
4.4 Exposure controlling filters
4.5 Neutral Density Filters
4.6 Polarizer Filters

5. Regional balancing of exposure
5.1 Handling frames of wide tonal discrepancy
5.2 Filling in softened light: bounce cards and diffusers
5.3 Graduated Neutral Density (GND) filter
5.4 HDR techniques
5.5 Good HDR versus controversial HDR application
5.6 Poor attributes or adverse effects in non manual HDR rendering

.
 

Last edited:
1. Judgement and adjustment of exposure


1.1 Overview: Judgement avenues for correct exposure

Appropriate exposure should be based on

1. emphasis and selection of the important area to show up details
2. judgement of how bright the scene should be
3. correct the given metered value from the camera accordingly from the 1st two points judged by you.


Correct exposure means making best use of your sensor's latitude to expose your composition in sufficient intensity of the important areas, such that the details in it are preserved, coupling with the hue and saturation. As the sensor does not cope well with fine differences of data in dark shadows and strong highlights, the data recorded in this area becomes uniformly absolute, i.e. all zeros and ones, not with fine recorded differences like 0.001, 0.003, 0.002 or 0.997, 0.994, 0.998. These areas of complete shadows and highlights are flat, uniform and featureless and are called "blown out" usually (perhaps there are more correct jargons), and in any post processing, its neighbouring pixels with some details will change accordingly to the post processing, but these areas are unchanged, seemingly like an opaque masking tape pasted onto a piece of glass - they are blocked and shows no changes of light that shines through the glass.

If your sensor cannot cope with that latitude, you need to cover the important parts with appropriate exposure and leave less important parts to fade into these flat areas, which means areas expected to look like real shadows and real blown highlights, for example, you will not see details in the sun, and you may not see details in really dark shadows.

Judging the exposure in the first portion, means judging the area to be appropriately exposed, as that forms the basis of metering areas which is to be explained later.

For that matter after the area, whether we thus talking about the main area to be appropriately exposed, or if the tonal contrast does not differ that much with the whole picture within the range of appropriate exposure, then it sets the next question - how about of the shutter duration and aperture size and at what ISO should one set to give a good picture of spot on exposure?

Big question, it is. This is the main concept of what is considered "correct" exposure.

This depends on what you intend to capture and to reflect. This means that a bright scene that normally looks like 13EV to you, should be given 13EV of light through the camera, and a dark scene that normally looks like 8EV to you, should be given 8EV of light. Exactly because of highly varying lighting situation and different desires to adjust it to suit a mood, that makes the control of exposure interesting and hard to pin onto a set of standard settings that newbies like to ask.

You can leave everything to auto - full automatic exposure mode (auto), or scenes mode where you tell the camera what themes it is for the camera to predict the lighting condition. The computerised system is some kind of artificial intelligence but is not going to know your intention, and it arbitrarily set appropriate exposure at the industrial standard of 18% grey, followed by tons of research into matrix metering to try to predict your shooting pattern and varies the amount of exposure value that it will set as the calculated metered value. For example, it look at a scene that looks like a nightscene, thinks it should be dark and give the metered value at 8EV, and look at another scene that looks like a bright day scene, thinks it should be bright and give a metered value at 13EV. Exactly how these researchers can devise such a system is unthinkable to an end user like me, but it certainly is not infallible - in fact, photographers who yearn for better control to his desire, more has to be done, mostly on top of the programming already present in the camera.



1.2 Making fast estimate of correct exposure

Sometimes the process of photography today is so automated that one doesn't think of what is going behind the process. The number of permutations of settings that lead to a certain exposure value, or the range of exposure value that can be used for a certain frame can be very varied - but becos metering in the camera body already brought you so close to the appropriate exposure, all you need to know is to get the adjustment in.

The process of fine tuning the exposure drops into 3 steps
1. making fast estimate of exposure (by experience/rules, or by metering)
2. reviewing of results
3. fine adjustments to preceding settings



1.3 Deciding on exposure by experience and photography rules

This basically goes back to the film user days. When one uses manual mode and ignore the metering bar, and simply judge his ISO/shutter duration/aperture size based on the situation he is in and with the eyes he see of the frame, that is basically exposure by experience. There are certain rules that people used to apply based on the theme and time of day, which I would not be covering at this point of time.



1.4 Deciding on exposure by metering

Metering refers to the measurement of light intensity (amount of light within frame per time duration). Inbuilt metering ability within a camera however do more than just measuring the light intensity in the frame or selected part of the frame. It also make judgement on what is the required amount of exposure for a picture and help you do calculations of the ISO/shutter duration/aperture size to reach that exposure, i.e. the metered value.

Metering offers a fast way to the calculated optimal exposure where the factors of exposure can be based on.

Remember earlier in the light equation?
frame light intensity x aperture size x shutter duration x ISO = exposure on media

Hence, metering does the following based on the equation above
1. measures the light intensity from the frame/part of the frame (metering modes will be discussed later)
2. judge/show what the exposure on the media should be, and thereby
3. shows what permutations ISO/shutter duration/aperture size will add up to that value.

frame light intensity (measured) x aperture size x shutter duration x ISO = exposure on media (determined by chosen metering mode)

The settings for ISO/shutter duration/aperture size (in blue) are thus what you can decide either automatically or semi-selectively or all-manually for a frame processed by this metering function that determines the 1st 2 steps.

From dpreview on metering.

.
 

1. Judgement and adjustment of exposure


1.5 Feedback avenues of exposure

One has to decide on his own - how much he thinks that the calculated metering is giving an accurate result or how much it differs from what he want. Such a feedback mechanism are variable from estimation down to more quantitative referencing.

There is a few way to assess the adequacy of exposure, namely from
1. the eventual output on print/computer screen
2. LCD post-capture view
3. histogram



1.6 Feedback avenues of exposure - reviewing by eventual output

One thing that used to be done in the film days that is hardly done today, is by experience and instinct. One may subsequently know from his eyes of what he sees, that a certain range of shutter duration and aperture size and a certain amount of resultant EV would give what he wants, without looking at the metered value given by the camera body. There will be a certain part of it that comes with golden rules taught for specific lighting situation but eventually, whether the picture works or not, is judged from the previous output in print, so that adjustment to the settings are made to subsequent takes. Of cos, an additional factor that comes in is the lab factor if the lab does additional processing.

Most of us may still use a part of this method, by a rough estimation from the eventual output. When you always see that your eventual output is darker in your screen and on the printout, you may want to make a positive exposure compensation to your camera.



1.7 Feedback avenues of exposure - reviewing by LCD screen

One quick way to check for satisfactory pictures is of cos to look at the LCD. However, the LCD view may be misleading, although it is the most convienient.

In order to benefit from more consistent result, first of all, you should either
1. adopt the habit of viewing it in the same condition, e.g. always viewing under shade or with a LCD shade/hood (some accessories that adds like a protruding hood that can help reduce direct sunlight from hitting onto the LCD causing flare which may affect your judgement of exposure and colors from the LCD), or
2. develop a mental mapping of how much the environment affects your LCD display and subtract accordingly (takes experience/talent, difficult to be accurate)

Problem is that it is not going to be easy to form a discipline of always viewing under the same condition. If you can bother to always find shelter, you might as well apply the method of using histogram which is of varying ease/difficulty in different camera models. If using a LCD hood, a possibility is to add extra shade by using the hand to cover it further; this is just a postulation and suggestion as i do not use a LCD hood, i merely use my hand to act as a LCD hood which is of some but limited use.

I also do the estimation of gauging how the environment affects the LCD display, but there is no hard science and I also do not find it very accurate, but for lazy bums like me, that is the easiest method.

The next thing to do is to adjust the brightness of the screen to be closer to what you would see on your calibrated monitor. I put -1 for my nikon D50, but you may have to make your own judgement for your own camera. Once you are fixed on a LCD brightness level, leave it alone, and start shooting. Shoot once, look at the LCD, too dark, estimate the difference, roll in EV+1 and shoot again until you have corrected to a satisfactory EV compensation setting, and then leave the EV setting alone in the same lighting environment, then when you went into another lighting condition, then change into another EV again. I usually roll up to plus-minus 1.7EV based on rough estimation.

In short, you can make LCD viewing better, but you cannot make LCD viewing much more reliable. If you wanted more reliability and consistency, you need something that is more quantitative and not affected by flare in the environment and that is by histogram.



1.8 Feedback avenues of exposure - reviewing by histogram

With regards to what is a histogram, please refer to Page 9. Exposure. 1.9 Exposure distribution (histogram) and composition.

Everything about the overall effect of exposure can be visualised from the chart - the proportion of shadow, midtones, highlight, hence even if you do not use it in practice, it is a good thing to know to understand certain concepts.

One can check the histogram on the LCD, which depends on models, can be set as a menu option to show on preview or review. From there, you can see if you have too much to the left (too much shadows) or to the right (too much highlights) and see if that is your desired or undesired exposure distribution. Some models allow review pictures to show up with blinking highlight areas, which my nikon D50 shows.

Understanding of how the histogram changes with the settings and post processing procedures allows one to understand the actual processes of exposure changes. This may require some deeper understanding, which I would leave it alone first.

.
 

2. Adjustment of exposure


2.1 Adjustment of exposure

This is done depending on your shooting mode based on either some preset instructions to how the camera does metering and automatic exposure settings, or direct instructions to the camera on the exposure settings.

Supposedly in non-full-manual mode, the metering done in whichever metering mode, would give an amount of exposure to meet and whichever exposure mode, be it automatic/scene/aperture or duration priority, one just have to do exposure compensation accordingly via the function of EV (exposure value) adjustment, depending on your camera body model.

And in full manual mode, metering is again automatically applied with a reference for you to see and up to you to follow. Just rocking your exposure settings dial to and fro for the shutter duration and aperture size, exposure adjust and changes simply based on whichever setting you set for the camera.



2.2 Exposure compensation

Exposure calculated from the metering can be further adjusted on exposure compensation from an programmed mode, aperture priority or shutter priority, or by directly and freely adjusting the aperture size or the shutter duration in the manual mode which tells you by how much does your settings differ from the expected amount of exposure from the above modes, e.g. spot v.s. matrix metering. and it is up to you how you will alter to bring that exposure nearer or farther away from the calculated value.

The article by Ken Rockwell covers exposure, with issues on exposure compensation and auto ISO.

Exposure compensation refers to a function in most if not all DSLRs, where a preset constant adjustment is made to the metered exposure. This means that if you set exposure compensation as -2 stops or EV-2, when you switch to spot metering with the camera deciding that all your scenes would have the frame intensity x aperture size x shutter duration x ISO adding up to 9 exposure values, all your shots will however be adjusted to EV(9-2)=EV7, which means that that 2 stops difference can be distributed by having a lower ISO for better noise control, a faster shutter duration to avoid handshake or a smaller aperture size to gain more depth of field. Likewise if you used matrix metering with the camera expecting EV11 in one scene and EV8 in another scene, if you set the exposure compensation as -1, you will have EV10 and EV7.

Page 12 of the Photography Notes by sulhan illustrate it in both text and pictures.



2.3 Why do I need to do exposure compensation? The midtone 18% gray principle

The principle of using exposure compensation by our own manual judgement is that the camera is incapable of telling what the object is in real life and thus you have to tell the camera so.

Anything that is brighter in intensity in the frame in real life can be dark in the eventual picture if the shutter duration is short or bright if the shutter duration is long. Remember our eyes sees things in perpetual continuity like a video but a camera captures collective light in pictures over a variable duration. And our eyes recognise a bright object such as the moon as bright while the circular shadow as dark, but to the camera, it is just a circle and its eventual tonal intensity is variable depending on the proportion its stands in the frame and how the overall exposure set will affect it.

The camera body judge the exposure by using the principle of 18% gray. Why 18%? That is believed that what we want to see in most average scenes an average intensity to show most of the picture as midtones, thereby enabling details to be seen. From 0% as black to 100% as white, there is a gray scale in between. 18% is quoted as the midtone gray where, hence whichever frame intensity, ISO, aperture size and shutter duration gives, the camera will complete the picture taking when enough light intensity reaches 18%. As ISO, aperture size and shutter duration are all prefixed by you or automatic, all the camera need to do is meter the reflected light intensity and do the calculation to complete the equation. No matter how you set the semi-automatic exposure modes, such as aperture or shutter priority, or auto ISO, they will all add up to 18% gray. In manual mode, the camera will still meter for 18% gray, and show on the exposure indicator in the viewfinder of what is your setting's exposure in reference to the metered value.

This is just a figure to most of us as a concept of the limits of this midtone principle, except to people who uses grey cards to calibrate spot metering. Thom Hogan mentioned that it should be about 12-13%.

The use of metering based on presumed average midtone however results in erroneous exposure values for scenes that is supposedly very bright or very dark but corrected by the camera body mistakenly as overexposure or underexposure.

The camera tends to neutralise everything to midgrey while our eyes can tell what should be bright and what should be dark. As such, we should increase exposure compensation when we shoot overall bright scene as the camera tends to turn overall bright scene to midgrey, and we should decrease exposure compensation when we shoot overall dark scene as the camera tends to turn overall dark scene to midgrey. What we are doing is to reverse the over-correction of the camera to what is more life like.



2.4 Exposure/Exposure compensation indicator

There is commonly a horizontal bar with interval markings centred around the zero reference point, which can be seen through the optical viewfinder. The bar represents exposure values, with the reference point at zero. This zero reference point will be the exposure value that the camera body decides to be the neutral and optimal value, after metering with your selected metering mode or scene mode.

It usually points to the 18% grey value, but do take note that 18% grey only refered to the sampling area, e.g. in spot metering, and may not be 18% grey for the whole picture. Also each camera body has its own algorithms and may alter it further. For more about metering, please read further below under the metering sections.

When no exposure value compensation is set, an arrow will point towards the zero reference point and remained unchanged. Any EV compensation will be demarcated by the shift of the arrow. If you set the EV to be EV-1.3 or EV+1, the arrow will move left and right from the zero reference point to its respective demarcated value, as shown in Sulhan's diagram in his photography notes.



2.5 Setting desired exposure compensation and checking the set value

EV compensation is commonly denoted by this sign "+/-". Please refer to your respective manual for your camera body for specific instructions.

For Nikon D50, one set the EV compensation by pressing the EV compensation button "+/-" on the top panel controls and rotating the rear dial in Nikon D50. As one rotate the rear dial, the corresponding value will be shown on the top panel. Alternatives may be found in the example given by Sulhan's photography notes, such as a specific dial for exposure compensation, or on the LCD menu interface.

The intervals for EV compensation comes commonly in 1/3 or 1/2 stops, depending on your preference.

To checking what exposure compensation is already set, there are two methods. One is where one set the EV compensation, which may appear on the top panel or the LCD menu, depending on the model. Another is by viewing through the optical viewfinder and looking at the exposure indicator.

.
 

2. Adjustment of exposure


2.6 How should I go about adjusting the exposure compensation

The degree of exposure compensation adjustment thus relied on how much we think we need to correct the given value by the camera body to more closely resemble the actual exposure value of the scene or the desired exposure value of the scene. This is especially when the scene is predominantly black or white, leading to a balanced but flat 18% grey setting which is not what we want.

The application of exposure compensation is based on the understanding on highlight and shadows and how much they need to be let loose or retained. For important details, we need to keep them. But to balance off, we will have to maintain the shadows as dark and the highlights as bright, becos the camera have a natural tendency to push an overall dark composition to be a more neutral less dark midgrey and the overall bright composition to be a more neutral less bright midgrey.

When do we go positive or negative?

In general, if the scene is supposedly very bright, the camera body will try to underexpose it up to 18% grey, and we should thus do positive exposure compensation to reverse the camera body's underexposure. If the scene is supposedly very dark, the camera body will try to overexpose it down to 18% grey, and we should thus do negative exposure compensation to reverse the camera body's overexposure. Alternatively that may be achievable by using the spot metering correctly.



2.7 Exposure compensation in a dark scene

For example, in a scene of the moon, the sky is overall dark while the moon is bright.

Our eyes can keep the sky appropriately dark (EV5) and a small moon appropriately bright (EV10), giving an overall dark exposure e.g. EV6.
For the camera however, it cannot recognise that this small circle in a frame is supposed to be a moon in a dark sky. It will thus try to reach for example an overall exposure of EV8 which is 6+2. This means that the sky would be EV7 from 5+2 and the moon EV12 from 10+2. While it would be nice to have a brighter than usual sky giving some cloud details, we are at risk of losing details in the moon to highlight blow-out. An exposure compensation of maybe EV-1 would give an overall EV7 (from 8-1) instead, giving the sky an EV6 (from 7-1) and the moon EV11 (from 12-1). This means that the sky would be a stop brighter than usual but a stop darker than the metered value, hence giving some cloud details but still remained relatively overall dark so that the night mood is not lost, at the same time, the moon is now kept at a relatively lower exposure so that the details in the highlight will not get blown out.



2.8 Exposure compensation in a bright scene

With the same principle, exposure compensation is applied on the overall bright scene, which is most notably seen in the shooting of snow scenes. Snow makes up the large background and gives an overall bright scene. If left to itself, it may appear white when really bright. A little less exposure, it will gives some hue and some definitions to contour as the slightly darker portions start to contrast against the really bright portions, which is needed when the snowland is not flat but is undulating with contours. But too much correction to midgrey by the camera will cause the whole scene to be underexposed - it will cause the snowy background to look really gloomy as if it is shaded by dense storm-like overcast which breaks the mood of a fun skiiing holiday, and the original midtones details inside the frame will become like shadows and start to lose details. As such, you need to apply the principle to do positive exposure compensation to push back the brightness of the snow which the camera tried to over-correct to overall midtone with underexposure.

How much to adjust depends on your desired mood and also on the proportion of background to subjects and their relative tonal intensity. It is not a hard and fast rule, but at least whether to go positive or negative exposure compensation - that is relatively derivable.

This picture below is from actionman with his permission. Note that the snow is the background but have sufficient midtones subject in large enough proportion. Hence the need to do positive exposure compensation need not be a lot like +1 to +1.7. The settings are 1/1000, F9, ISO 320, Matrix meetering, +1/3 EV. Any over-ardent positive exposure compensation may start to make the snow too white and lose details too.

From actionman,
77116817.jpg


Note that the snow white portion on the bottom left under glare is already very bright and cannot go too far further up. Sufficent exposure compensation however allow the snow in the shadow is to a bit brighter and still retained details.

.
 

2. Adjustment of exposure


2.9 Auto-bracketing function

discussion

auto bracketing simply means pre-setting a plus and minus exposure compensation on successive shutter release.
if you set a + 0.7 EV value, and if you pressed a continuous burst mode of 3 photos or pressed the shutter release 3 times for 3 single shots one after the other, you will have the set exposure, set exposure + 0.7EV, set exposure -0.7EV.

it helps in retaining the best quality for each varied exposure in each photo and letting you decide what is best in the end, but the other aspects determined by shutter duration/aperture will be changed. also if you take 2 photos only and switch to another frame for the next, forgeting to off the bracketing, the 3rd photo may be affected by bracketing. also it increases your storage by 3 times.

.
 

3. TTL metering


3.1 What is TTL?

This metering is often taken from the light that enters through the lens (TTL). This metering is always there, as long as the DSLR is on. Other form of more complex calculation of metering, e.g. using incident light meters, is not required for newbies who are not practising strict studio lighting control.



3.2 Use of metering in different exposure modes

Basically, it is either you follow the given exposure factors automated by the camera or you make adjustment to it. The only situation you do not used it is when you are shooting at manual exposure mode, and you entered your ISO/shutter duration/aperture size without regard of what appears on the metering bar.

The metering bar looks like this, and is usually situated at the bottom of the optical viewfinder. The image came from Rain and Peter, with their website on DIY wedding photography.

It can be a fully automated process of which all 3 values are chosen by the camera to reach that exposure, or it can be a semi-automated process as explained below or to complement a manual process.

When your PASM mode is pushed to automatic or those scene modes, exposure is take care of by the camera, so is metering. When the focusing is done, metering is also accomplished, and with the amount of exposure needed and known, the camera will then adjust the ISO/shutter duration/aperture size automatically to match that amount of exposure.

Given a choice to input any of the three settings (ISO/shutter duration/aperture size) or make addition/subtraction to the total amount of exposure given by that metered value, that is semi-automatic. The control over the three settings (ISO/shutter duration/aperture size) gives you the control over features that are related to them, ie. noise, motion, depth of field. When you make adjustment of the exposure, it is usually done through a fixed preset setting via exposure compensation function based on what you feel to be appropriate or from feedbacks seen on the LCD or the eventual printout/screening on monitor.

In a manual setting, the metered value continues to be given as a midline reference. The camera will not automatically give you the value of ISO/shutter duration/aperture size should you enter any one of the setting. However, by any values you enter for the ISO/shutter duration/aperture size, it will show you on an exposure bar how far it differs from the midline reference, and you can then adjust the values till it reaches what you would expect to differ from that midline reference.

Let's say the light intensity from your frame is 5x of normal, and the optimal exposure calculated by the camera is 1000 units, then the camera body will know that there is a need to come out with an exposure of 5x 200units. If you set an ISO of 100, you will then require the shutter duration x the aperture size to be able to give that 2 unit of light from whatever combinations the camera body can give. Supposed that the light intensity of the frame changes, or the ISO changes, the amount of exposure required from the shutter duration and aperture size will changed accordingly, all together giving 1000 units, which is the calculated requirement.



3.3 Concept of exposure shift and sampling area

Imagine the crucial parts of the picture to be the heart, the fairly important area to be the lungs and least important area to be the fat and muscle wall just outside of the lungs. Metering is to centralise a shield over the heart and most if not all parts of the lungs, considering that this shield is not enough to cover the whole body. Hence the fat and muscle wall to the left and to the right are vulnerable to a bullet shot unshielded by the shield. Sometimes, this shield is larger and is enough to cover either the left or the right or a little of both, but not totally both - and this is where you think about if you want the left or the right to be preserved while the other side to be sacrificed.

As the camera sensor does not have the same degree of tonal range our eyes can see or adapt to, there is a higher likelihood that strong dark shadowy tones to our eyes will appear to be indistinguishable single palette dark tone or absolute black with not much discernable details, or strong bright highlight colors to our eyes will appear to be indistinguishable single palette light tone or absolute white with not much discernable details. This metering is to set the midtone to be exposed just the right amount while minimising the loss to very dark or very bright areas.

However, the way we see photographs are not exactly the same we see with our eyes. We would often expect to see those things we think are important to be sharper, more vibrant and contrasting, whereas less important parts can be in shadows. Hence, by selecting areas that we perceived as important, that will be appropriately given an exposure that a certain aperture size and shutter duration while the areas that are perceived to be unimportant but is in the frame, would be allowed to be in shadows or highlight. In short, it means that the sampling areas selected represent priority areas where exposure is optimal, whereas the non-sampled area will follow whatever exposure is determined for the sampled area.

For example, if we divide a frame into 3 parts, left (shadow), centre (midtone), right (highlight), and that we have the important to be sampled area in the centre, we can have various situations. But eventually the camera decides that it wants the centre to be at an optimal value of 4, where it will become just the right amount of midtone.

if the exposure of a frame is -1, 3, 7, then the outcome may become 0, 4, 8
if another frame have the exposure of -2, 3, 8, then the outcome may become -1, 4, 9
if another frame have the exposure of -2, 5, 8, then the outcome may become -3, 4, 7

Hence, you must understand that although you want optimum exposure for the centre (the important sampled area), the differences between the centre with the left and right remained at the same intervals. Hence there is a possibility that the two extremes may be pushed to very black shadows and very bright highlights of which both can lose details at the extremes. Now it is your duty to take note of that two extremes and make decision whether to tolerate that, or to balance/sacrifice the other two portions.

In general, we sampled the midtone, so that we dun have too extreme a highlight or too extreme a shadow. Most of our subjects that we require optimal details to be shown are in the midtones, such as a face.

Variants can however to made to force more shadow (low key) or more highlight (high key), although strictly speaking both effects refers to a predominant shadow or predominant highlight but with a small proportion but well distributed range of normal tones within.

The low key effect is what we often know as the silhouette effect. We meter a tone between the highlight and the midtone, e.g. the sky near the setting sun, so that now this brighter tone becomes like the midtone, shifting itself to the midtone, and the original midtones and shadows shifting by the same magnitude to become darker, thus becoming a silhouette. We do not meter at the sun directly becos it is really bright, everything else including the sky may become too dark.

The high key effect is however less often used. In this situation, we meter a tone darker than the original midtone and pushing this darker tone to become a midtone. As a result, everything brightens up, and the original midtone and the highlights becomes even brighter. This can be used for example in portraits. If the face is slightly dark, and if you meter at the face against a fair colored background, the metering will result in a calculated exposure that will show the face as correctly exposed midtone with the background turning bright and white.

.
 

3. TTL metering


3.4 Metering modes based on sampling area

After you have understood how metering is sampled from a selected area and applied on the exposure settings which affects ALL parts of the whole picture, here comes the metering mode which allow you to set how you want that sampling to be done.

The important areas where exposure will be based on and calculations made to give the right exposure on that area in higher priority would be the metering areas. This sampling area can be chosen to be
1. spot: on a small object or area at the respective metering points where your exposure is taken from
2. centre weighted: on the larger area in the centre of frame wherever you direct at
3. average: mutiple standard locations in the framing would be sampled from
4. matrix / evaluating (canon terminology) / multi-pattern (nikon terminology): variable areas with an automatic selection of metering points, based on what the camera perceived from, e.g. a form and lighting condition that resembles a portraits, the area taken for the metering would be from mainly the face.

Option 1 to 3 are manual metering modes allows you to sample from any area in the frame (or even outside the frame) - the only difference between them are the size of this sampling area, with the spot metering being the most accurate to the average metering being the least accurate.

Option 4 is an automatic metering mode. The composition will be "intelligently" recognised by the camera body who will then decide on how much exposure should this kind of composition receive.


3.5 Spot metering for fine control

Spot metering is formed on the basis of measurement from a very small area of the frame, and thus is more accurate in identifying a specific midtone desired by the photographer to represent correct exposure of that midtone and other midtones that is similar to it.

The basis of such metering is to allocate fine control by the user. Given a larger area of metering, unless the larger area sampled is as uniform as the spot in terms of variation and range of tones, there will always be some highlight or some shadow in that larger area, in unpredictable variable amounts, that will deviate from the correct exposure of that wanted midtone.

Let's say from a tone of 1-10, a photographer identified an orange on a tree as the desired midtone of 5. In accurate spot metering, the metering is squarely done on the orange, so if the original composition of shadow, midtone and highlight is 2, 4, 8 in terms of tonal intensity, the shutter duration and aperture size will result in the eventual exposure giving 3, 5, 9. But if a centre-weighted metering is applied, assuming that within the area of sampling, half is midtone of the orange and half is the dark shadowy area with dark green leaves, the resultant exposure will attempt to make both the midtone and the shadows in that metered area as close to the desired tone of 5 as possible, in an averaging way. This means that it will be (2+4)/2=3 instead of the original 4. So intead of moving 4 (orange) to 5 (desired tone), the exposure will move 3 (orange and leaves) to 5 (desired tone), and hence the composition will results in the composition of 4, 6, 10 instead of 3, 5, 9.

This effectively means introduction of dark areas into the metering area, will make the camera body attempt to give the shadow area a midtone value, hence brightening up the shadows, slightly exposing the midtone, and blowing off the highlights. This is against the user's wish to correctly exposing the midtone orange as midtone, and keeping the shadows dark and the highlights details kept.

With spot metering, the interference illustrated above would be minimised.

The area of spot/partial metering for some of the entry level models are
Nikon D50 - about 2.5% of frame in the active focus area
Canon 400D - about 9.0% of viewfinder in the centre
but do note that the frame and the viewfinder is not the same although close to each other, the viewfinder approximately >90% coverage of the frame. upon some basic searching, i gave up and realise that specifications are not always given in fine details, and are not usually standardised.

A good way of standardisation in setting the accuracy of spot metering area (the smaller, the more accurate) is to set the area of the spot circle (not the focusing circle) as a percentage of the area of the frame, although the industrial standard is to set the area of the spot circle as a percentage of the area of the viewfinder, which is often closed to but not identical to the area of the frame. The specifications given in physical size in mm is not useful because it is affected by the sensor size. 1% in a larger sensor is larger in physical size than 1% in a smaller sensor, so they ain't comparable.

More reading can be found here.

.
 

4. Control and adjustment of exposure outside the camera


4.1 What is available to control exposure inside and outside of the camera?

Inside the camera,
1. focal length affects on the frame and the amount of frame light intensity
2. metering mode and metering sampling area gives a recommended value that affects your decision over exposure or directly guides it
3. exposure settings - ISO, aperture size, shutter duration

Outside the camera,
1. your standing position and distance from objects affect the amount of frame light intensity
2. where you point at, whether you include light sources in the frame, near the frame or far from the frame, affect the amount of frame light intensity
3. adding light with your own light sources, be it lamps, torches, flashgun or build in flash
4. subtracting light with filters while achieving other effects.
5. using post processing to alter eventual exposure on medium

Composition and light intensity related to that will be covered elsewhere.

.
 

4. Control and adjustment of exposure outside the camera


4.2 Flash & studio equipments

I remained a newbie level because i do not use the inbuilt flash and got an external flash that i never got to learnt well. The art and technique of using external flash with the necessary softening in some cases opens up multiple entirely new avenue to low light situations that range from indoor family, concert, sports, product, wedding, night scenaries to night urbanscape, and that is what I'm still missing out like other newbies. As such there is no further discussion here, but make a note that i still feel this is an important chapter, just that i feel not adequate to provide any summary despite of reading on it. But i'll add a few links as things goes...

flash unit
guide number
Some common problems with direct flash and its solutions: As described from this good article written by Chuck McKern from Vivid Light Photography

.
 

4. Control and adjustment of exposure outside the camera


4.3 The concept of strong light intensity situations

All the while people talks about shooting in low light intensity. Given 2 years of self-exploration, I realised a concept that is often conveyed but never really spelled out as a whole - shooting in strong light intensity.

Combating strong light gives another range of problems, and the solution seems to lie in what i call the exposure controlling filters. Whenever someone ask about what is the essential filters, it seems always to be these set of exposure controlling filters which directly affects exposure settings in the process of shooting and is involved in the capturing of details which hence cannot be replaced by post processing.

Let's first look at the various factors and see how we can or how we cannot combat strong light to avoid over exposure using these factors.

1. Overall exposure setting: By setting negative exposure compensation, light intensity applied will naturally be deducted from the metered value. Or those familiar with metering can use spot metering on a bright spot such as the brightest light spot or the sun (could hurt your eyes) so that the metering will naturally be deducted to midtone the bright source. These settings are still all guided by and within the limits of the individual exposure setting limitations - the ISO, aperture size, shutter duration. You can similarly command control by pushing to the limits of the respective factors.

2. ISO: setting the lowest ISO is always what everyone hopes to do with to ease any noises, and very often we already hit the limit of the lowest ISO, which is often ISO 200, some 100 or even 50. But well, there is little variability and often this is already applied as a base setting when you went out on a sunny day. Just dun forget to reduce it from higher ISO that you set from an earlier time, such as the previous night. The inability to view ISO on the viewfinder is a major reason why everyone forgets to change it.

3. Aperture size: You can set the smallest aperture size to reduce light. However practically one would not like to sacrifice his depth of field choice to lower light. Setting too small an aperture first causes the sweet spot to be lost and secondly limits one's wish for a narrow depth of field.

4. Shutter duration: You can set the fastest shutter duration to reduce light, which is often the case anyway. Motion blur applied in bright day light is not impossible, neither is it not explorable - but it is just quite uncommon. Also when using fill light in bright light (to avoid ugly shadows when shooting macro), the flashgun or built-in light will slow down the shutter speed and prolonged the shutter duration so that it can synchronise with the camera body, say about 1/250s, thereby limiting the use of fast shutter duration in reducing light.

5. Post processing: One can always darkened selectively or en-bloc - but the problem is nothing can be done to the blown highlights. By darkening the highlights, an ugly transition will occur. A possibility is to burnt or unlightened these blown areas, or to clone details over it - but is it worth the tedious efforts that will still yield poor results?

6. Framing: You can increase the focal length which will telescope into a smaller field of area, hence decreasing light, but hardly anyone will do that becos changing focal length changes how something looks totally, whereas exposure may not. By framing away from lights is a possibility, and by avoiding including the sun within or near your frame, that is one of the most effective though not the best way to reduce light, as long as that does not change your composition too much. With or without filters, the inclusion and exclusion of the sun changes many stops of light in the frame.

7. Exposure controlling filters: The last not the least. It is cumblesome, but is your saviour in combating strong light. Let's say you do not want to sacrifice your composition or framing, do not want to sacrifice your wide aperture, have already applied the lowest ISO, the fastest shutter speed and have already excluded the sun, and do not want ugly blown highlights in post processing - which is actually what most of us do in shooting, what else do we have in our arsenal? Decision to be made - bear with the cumblesome filtering and some loss of optical quality, or sacrifice the exposure, or sacrifice your aperture choice or composition. Think about all these factors to see what would you do in your preference.

.
 

4.4 Exposure controlling filters


There is a few filters that are often confused for example in this thread.

Basically, there are 3 filters here that is often confused as they all cut down light, though in different manners. This is especially important for landscape photography, where the sky becomes an enormous lighted background that can overwhelm the exposure and need to be cut down and controlled accordingly.

1. Linear or circular polariser (CPL): polarise and block off reflections from non-metallic surface to a certain extent, saturate colors, some reduction of light
2. ND/ variable ND: uniformly reduces light by a uniformly translucent grey filter (not just reflection, but all light)
3. GND: reduces light in an area only by a filter that is half clear and half graduated (with a gradient) grey

By right the above is not supposed to introduce non existent colors, but they can have defects and introduced a tinge.
on the contrary, there are graduated colored filters which are supposed to introduce non existent colors.

Usage of such filters require understanding of exposure well.



4.5 Neutral Density Filters

Abbreviated ND filters, usually refering specifically to non-graduated solid full length neutral density filters. Their main function is to cut down light in general, used it in bright day light or in situations where there are plenty of reflected light, such as waterfall. It is usually used where the aperture size cannot be reduced beyond the minimum limit, or if less frequently if a larger aperture is required (can't really think of one in this situation: large aperture, long exposure).

Another situation is used in macro of say white flowers in outdoor bright day light - you want wide aperture (increase light) and the fill in light from flashgun to avoid shadows (increase light) while the shutter duration cannot be too short as they are limited by the camera's maximum sync speed (failure to reduce light), very few options is left. ISO already the lowest. Focal length? Nope, macro lens are mostly prime and you already push it to the longest focal length to macro the flower. Post processing with ugly graduations? Seems to go against the shooting of crisp and beautifully smooth flowers.

And as mentioned above, waterfall. Most people travel deep into the woods for most waterfalls, where there is no lodging. And even if it is a large open waterfall, rarely would there be lodging allowed to avoid disrupting the scenaries. To get in and get out, you can't stay there till too late. Light is often strong. And waterfall is typically a subject matter that people do not freeze but rather wanted a silky effect, which puts the typical use of ND in. The main problem in Singapore is that we dun have a scenic waterfall that gets people to invest into an ND filter for that purpose. It can be used for coastal shots for the waves of the rocks, but usually such photography are done in the sunrise and sunset period where there is low light rather than overwhelming daylight.

Neutral refers no color cast, which means that any colors will be only darkened without additional colors added. This is only true for high grade Singh Ray and Lee's filters. Hitech and Cokin filters have a reddish to maroon tinge while Tianya filters have a slightly greenish tinge.

The strength of the density is of the same nomenclature as that of the graduated neutral density filters. I like to use the number of stops cut down as my nomenclature, i.e. 1 stop, 2 stop, 3 stop, 4 stop filters. Other nomenclature corresponding to the same concept includes that by the amount of light intensity, i.e, filter factor ND2, ND4, ND8, ND16 (B+W and Hoya) or in another sequence of filter grades 0.3, 0.6, 0.9, 1.2 (Tiffen and Lee). 4 stop filters are uncommon to find and rarely used.

As described by jcryan55, ND 400/800 filters are "rare filters where many UK landscape photographers like to use. It drops the light to 9 stops allowing u to shoot moving thing like water, waves to become like white mist."

.
 

4.6 Polarizer Filters

Unlike neutral density filter which cut down all light by a fixed amount all the time, polariser only filter out the erractic glare component that is reflected in a haphazard manner. This component of light will produce a washed out veil like effect and reduce contrast, making colors look duller. Polariser works by allowing only a certain direction of light rays to enter and remove the vector component that is not in that direction.

More can be read here from the Luminous landscape and Digital Photography For What It's Worth

In DSLRs, we used circular polarisers (CPL), not linear polarisers, but as a newbie not going back to film, you need not know the differences at the moment.

The purpose is to
1. darken a saturated sky
2. saturate weak colors
3. remove reflections

The result of 1 and 2 naturally increases color contrast. In the darkening process, it also help to reduce blown highlight and bring out details, which in a way increases detail contrast. The removal of reflections also may increase contrast - it depends on whether the reflection itself have more contrast or the details underneath have more contrast.

For point 1, it not just darkens the sky, but darkens everything in general, although the differences is more so noted in the area with a glare component. In a way, it can also function more or less like a neutral density filter, if the sky covers a major proportion. The degree that CPL cut down light can be 1-3 stop off.


Should I always put it on? Can it be used as a protective filter instead of UV filter?
No. CPL are not meant to be put on as protective filter as it cuts down light by stops.

There are occasions where the reflection is desired, in which you do not want to use the CPL. Also the cutting down of light and reduction of clarity means that most of the time if there is no indication of benefits and yet you used the CPL routinely, you may be getting a reduced image quality even if you set the CPL to the minimum filtering angle.

It can of cos protect the front elements, but for reasons that is stated above, you would not want to leave if on the lens at all time. You can of cos take off the UV filter and add on the CPL filter when you need to use the CPL, but that increases the risk of dirtying the front element. Do it if you are careful and comfortable.


Myths to be cleared.
1. CPL does not create a blue sky, unless there is already a faint blue. CPL is not a colored filter and thus does not change the color of the sky. Also it can only cut down exposure by 1-3 stops, so if it is really bright white, it will remained white even with a CPL.
2. CPL only reduces reflection, and may not totally eliminates it. It also does not work on reflection from metallic surfaces. The easiest to see such changes in reflection is on tranquil water surfaces.


How to use a CPL
To use the CPL, look through the viewfinder and look into the sky, while rotating the CPL till you see differences between maximum brightness and maximum darkening. This should occurs with a difference of 90 degrees. It may produce visible differences but it may not, and it could be either of one's eyes is not attentive or that there is actually no significant glare component in which a CPL does not make a perceptible difference. In the latter case, you can choose to take the CPL off since it does not add to the photo but instead decreases the clarity. night86mare suggest looking at the clouds to see if there is a difference in the clouds detail contrast, as said "the further away the skies are from the direction of the sun, the more pronounced the effect. Note the divisions between the blue and the clouds." This basically relies on the distance and angle from the sun, which is explained later.


CPL and the wide angle

Wide angle pictures invariably means that there is framing of a wide part of the sky (which is why a polariser is added to darken the sky).

Assuming that the sun is outside the frame, on the left or the right out of the frame, the part of the sky nearer to the sun would be brighter, and the part of the sky that is further away from the sun would be less bright. In addition to that, the glare from the sun that falls onto the filter where the part of the sky that is nearer the sun shines through, is more parallel to the axis of the lens, and thus less polarisation occurs, leaving the bright portion bright. The part of the sky that is further away from the sun, has a glare that shines more at an angle to the filter, cutting down on the glare component effectively, causing this darker component to be even darker.

Usually without the polariser, the difference of brightness is not that apparent and looks natural. With the polariser on a wide angle lens showing a wide angle frame of the sky, the uneven polarisation is maximise by the wide angle, and added to the difference of brightness and the difference in the glare component, thus looks unnatural.

The luminous landscape suggest that with a lens wider than about 28mm (in 35mm film terms) the sky will be unevenly polarized. This translates to a focal length of about 18mm or below using a crop factor of about 1.5x in some of the modern DSLRs. The luminous landscape however do believes that if you like the effect, go ahead and use it. My personal opinion is initially that it is best avoided but acceptable, but i'm reserved when i saw this picture offered by fellow forumer, shuttergraphy.

DSC5661.jpg

In this case, the sun is on the right, likely within the frame.



Options and recommendations

1. B+W CPL
2. Kenko Pro1D
3. Marumi DHG series
4. Hoya HMC

Cokin rectangular holder always have a slot-in circular-shaped CPL option that has a corrugated edge to provide friction to stay on the holder, but i find it difficult to slot in and does not recommend this.

.
 

5. Regional balancing of exposure


5.1 Handling frames of wide tonal discrepancy

What tonal range and dynamic range means have been explained in C1.4

The inability for the current digital technology of the sensor to handle high dynamic range (from very dark to very white tones) lead to problems where difficult lighting situation arises with wide tonal contrast between dark and bright tones.

This requires additional effort to overcome. A few known methods to balance up the tones include

1. avoiding such situations
- framing to avoid excessively bright sky or light source if possible
- requires bothersome efforts to deal with exposure, framing and maybe focus point separately
- this is recommended for newbies as a starting point
- wait for the right season/time for light to be softer and even, and even still, go for the well lit rather than the back lit scenes until you are comfortable in handling it.

2. compensate by reducing exposure intensity of the bright areas
- using graduated neutral density filters
- require use of rectangular filters using an external cokin or lee filter holder, which may impede use of lens hood, lens cap or other round filters.
- problems of such filters getting dirty upon heavy duty usage and frequent change of front elements
- as mentioned in the GND section, this thread can be read. And singh-ray showing up with a good illustration of it.
- this method is mostly adopted by landscape shooters who does not need to deal with very irregular horizon and in situations where HDR is not feasible (e.g. no tripod to ensure identical composition).

3. compensate by increasing exposure intensity of the dark areas
- using selective fill flash
- require skills in flash techniques
- extra weight of bringing portable flash
- other tools may include diffuser and reflector
- dependent on the distance and power of flash
- unless with professional skills, flash may overpower ambient light and gives a dull artificial picture with dark background and without mood.
- this method is mostly adopted by professional outdoor wedding photographers and outdoor portrait, which requires a very strong understanding and skill of the photographer.

4. combining and overlapping different set of tonal range
- using multiple takes with different EV and digital post processing (which we called "HDR" aka high dynamic range as a synonymous name of the process to the results)
- require tripod to take two identical composition of different tonal range
- subjects must not be moving during the different takes, hence not suitable for moving subjects such as human beings, animals and structures affected by wind.
- require a sensitive eye to avoid overdoing the HDR and causing unnatural dynamic range with halos.
- this method is adopted often for architecture with difficult sky lighting situation where neither flash nor GND is optimal.

5. getting a sensor with high dynamic range and setting the option of RAW format and high dynamic range, e.g. Fujifilm superCCD



5.2 Filling in softened light: bounce cards and diffusers

Direct point source from a flash may present with harsh white light that replaces the ambient lighting and cause unnatural looking lighting with dark shadows.

Quoted from Photojournalism, "the point of softening light is to scatter the light rays in various directions. This allows light to ease into areas which would otherwise remain underexposed (dark) if directional light is the primary source. In effect, it reduces the impact of highlight areas while filling the shadow areas. As such, it will typically create a reduction in the effective power and exposure of the flash (make appropriate GN deductions)."

There might be many different varieties of devices which softens the light, but in general either makes use of reflected light from a less shiny surface, eg. a white acrylic card, or diffused light through a translucent material, e.g. a wax paper or plastic milk bottle material. check this out from abetterbouncecard.

Such devices is one of the DIYable photographic material. Threads include (1), (2).
And discussion are as such here.

.
 

5.3.1 Graduated Neutral Density (GND) filter

Although less commonly used and cumbersome to use, graduated neutral density (GND) filters is one of the more important filter that i often used.
Here is a thread when I first started from advices from fellow forumers.

The main function is to balance out high contrasting exposure, without the need of a tripod for HDR. Contrary to the use of colored filters, graduated neutral density filters does not introduced a non-existent colors (unless it carries a color cast as a defect), but brings out the colors that is washed out by over-exposure of a bright day.

Graduation refers to a gradient of density. Because a rounded filter with a gradient cut off can only rotate around the midpoint, GND filters are naturally made as rectangulars in a vertical format, i.e. taller than wide, allowing a sliding in one direction (up and down) to control the position and alignment to the horizon in the frame. It can also be tilted to any other angles (360 degrees rotation) as when necessary, but usually not required.

Here is a link to Cokin's filters on how the rectangular holder works. Cokin P series holders comes in 3 slot or 1 slot versions. For the 3 slot version, the inner two slots are intended for the filters while the outer slot can be used in addition as a slot for a coupling ring. Cokin P series holders takes rectangular filters that is 84mm wide, and it has another slot in its rear that is 82mm wide and will take adaptors that can step down to the thread size of your lens. In contrary to the usual stepping rings for filters that has screw threads on both surface, this adaptors can only step up and has a male thread on its rear, while the front surface has no thread. Stacking of GND filters like stacking of any other filters can worsen vignetting. In the case of Cokin P series holder, even a single slot holder can introduce vignette at about 18 to 20mm focal length at 35mm equivalents (i.e. 12 to 14mm on a Nikon 1.5x crop factor). Other size format are available in Cokin's product line but are not commonly used as far as to my knowledge.

The strength of the density is of the same nomenclature as that of the neutral density filter (read above), although it is of variable effect on the exposure in the case of GND where the adjustment of the filter will alter the amount of density covering the frame.

Do note that the strength of the density in a GND filter is a variable. The strength of the filter are often given in the number of stops, but a 2 stop filter does not reduced the overall exposure by 2 stop. First of all, it refers to the approximate light intensity reduction only in the denser part of the filter, and secondly, the filter can be slide up and down, thereby varying the area of density and light intensity reduction.

There may be some variation in the curve of the gradient over length of filter in different brands, but i have no way of measuring it, but some impression. There is also description of the curve of the gradient at the edge between less dense and clear area, known as hard edge and soft edge. A soft edge will enhance a subtle blending of tone around an undulating horizon line, whereas a hard edge or reverse hard edge is meant for very straight horizon, such as the open sea, but one has to align accurately as hard edge has less tolerance of error. A reverse hard edge is used typically for low sunset shots, where the most dense portion start from the edge with the clear area and becomes less dense towards the top, as the sun is often at the horizon.

Singh Ray, being the most expensive GND filters, does produce top notch quality but did not spare me the degradation of heavy use. Due to cumbersome usage and the lack of ease of fast and minimal handling, it can still get dirty and scratched easily and cannot be covered by a lens hood. What they have done in return is at least to give a good website on GND as over here.


HDR v.s. GND

Read more into HDR below.

GND and HDR serves different purposes in different situation. a tripod, absence of undesired subject motion between takes and user expertise are required in HDR. GND serves in situation where lack of a tripod or presence of subject motion after a take makes HDR less favorable.

GND enable retention of details at controlled exposure. it can be simulated on ps but not to the same quality as details are not captured at optimal exposure and degrades with processing.

.
 

5.3.2 Graduated Neutral Density (GND) filter: Metering issue

There is usually a clear component meant for the foreground where you want the midtone exposure to be optimised, so that is where you should meter if your metering mode is centre-weighted or spot and the overtly bright portion on top, usually the sky, would be darkened accordingly to the strength of the graduation of the density. This basically means if you do metered according to your subject interest without the use of GND filter, you will get a correctly exposed subject and a potentially overexposed sky. And if you add a GND filter without covering the subject interest in the dense portion of the filter, you will still get a correctly exposed subject and suppression of the exposure of the sky according to the strength of the GND. As the sky is not within the sampled area of metering, the camera will not compensate in metering and any suppression of the intensity in the sky would be full fledged. Whether the sky is adequately and appropriately suppressed of overexposure, it all depends on how bright is the sky and how strong is the GND in terms of density in terms of number of stops. If your sky is not too bright, and you use a 3 stop GND, the sky will look too dark. If your sky is very bright, and you use a 1 stop GND, the sky may still be very bright and only looks slightly darkened.

In most cases, subject interest in focus are usually in the clear area or in the less dense area, so adding a GND filter without adjustment of metering, usually does not affect the shutter duration much as the metering is based on the spot of sampling, giving a pre-capture setting of aperture size and shutter duration which will fulfil the same exposure for that spot with or without the GND filter. If the metered area is covered by the less dense area of the GND filter, shutter duration would be slightly longer. Although the metered area has little change of exposure amount with or without the GND filter if metered through the clear area, that is only true for the metered area. Overall exposure is still reduced, largely from the area covered by the denser area of the GND filter.

If metered through clear area,
Light intensity of metered area (same) x shutter duration (same) x aperture size (same) x ISO (same) = exposure of metered area (unchanged)
Light intensity of overall area (lower) x shutter duration (same) x aperture size (same) x ISO (same) = exposure of overall area (lower).

However, if metering is averaging, or matrix (also called multi-pattern), then the denser area of the GND filter will be taken into consideration of the metering. The camera body will sense that certain areas are darker and will try to compensate for it. This basically means if you do metered according to the whole frame without the use of GND filter, you will get a potentially underexposed foreground and a potentially overexposed sky. And if you add a GND filter, you will still get a slightly better exposure of the foreground and a slightly suppressed exposure of the sky. In any situations where the darkening of the sky drops below too much below the midtone, the camera will increase the exposure to compensate for that drop of intensity. There is thus some decrease in the strength of the GND filter in the brighter area. For the overall frame intensity, matrix metering already suppressed the overexposed sky by reducing shutter duration (or aperture size or ISO depending on your setting), so any additional suppression by the GND filter will also be less pronounced.

Using arbitrary numbers to represent the strength of light intensity with midtone at 10,

With a 2 stop filter and spot metering through clear area
Foreground +10 becomes +10
Sky +40 (too bright) becomes +10 (corrected)

With a 2 stop filter and matrix metering
Foreground +0 (too dark) becomes +0 becomes +6 (slightly more dark)
Sky +30 (still too bright) becomes +7 becomes +13 (slightly more bright)

Common question: metering when using GND

change to spot/centre weighted metering and make sure the spot you meter through is through the clear portion. if you do spot and centre weighted metering through the grey portion, you may get partial to full compensation, resulting instead of blowing of highlights in the clear portion. Partial effect will occur when using averaging and matrix metering.



5.3.3 Graduated Neutral Density (GND) filter: Market availability

To date, such rectangular holders known to me consist of only two brands - cokin and lee. If it didn't interest even those bootleggers to steal the designs, it must have been that people using such filters must been a rarity and i might even be the next rarity to be using it beyond tripod-held landscape shots, noting that it is simply so cumbersome and excludes the routine use of a lens hood and lens cap. This rarity might be the reason why so little choices we have and why there is so little thought to deal with the design flaws.

I seriously think one should invent a GND filter holder without vignetting problem, and function with suitable holder hood and holder cap, and allow insertion without dirtying it. Maybe the filter should have an opaque portion that allows the thumbs and fingers to hold it firmly without smuding the filter portion.

But anyway, for Cokin holder, there is a number of sizes available. The p-series holder is more commonly used, at least in this part of the world, and has a number of supporting brands, from Tianya (HK), Cokin (France), Hitech (USA) to Singh Ray (USA) to support people of various budget, which hence means that one can upgrade or downgrade their filters while using the same holder. There is however slight differences to the specification or the variation of the factory quality control, and some brand may have slight difficulty in slotting in, i.e. too tight, but can be solved by sanding the edges.

Lee's filter, i think, is 100mm wide.

Both Lee's and Cokin's filters and holders are available at Cathay Photo in Singapore. The other main agents seems not to hold it, but i may be wrong. Singh Ray and Hitech needs to be imported. Tianya filters is brought in by one of the advertiser in this forum and I will link here when their customer service issues are ironed out.



5.3.4 Graduated Neutral Density (GND) filter: Handling and maintanence issue

All filters are prone to dirt, scratches and smudge, no matter how cheap or how expensive they are. Well, Singh Ray has better optical quality and that may gives the initial impression that it is more resistance to dirtying, but it is not in time. Not everyone has the same problem of dirtying it or various degree of problem with it. If you are the type that uses the filter only twice during your 2 weeks overseas trip, it is not likely they will get dirty if well kept during most of the time. How frequent you use them or how comfortable you change the front element determines all.

Some people who are afraid of vignetting will hand hold the filter in front of the lens without using the filter holder, thereby cutting down the frontal protrusion. That has a higher risk of dirtying the filter because you hold the filter for a longer duration. If one holds it by the peripheral flat surfaces, it can only be at the top or the bottom, and as we need the dense part to remain clean for use, it is slightly better to hold it at the clear side as that part is less frequently used. BUT of cos, it doesn't make sense if one hold the surfaces when he can hold the edges to reduce smudging the surfaces with our fingers with natural sebaceous contents, and it is more naturally that one holds it at the width 82mm rather than the length of 100mm unless one's hand is really big. Handholding the filter at the width however means that the fingers are closer to the lens's circumference and a little hand shake can bring the fingers transiently into the field of view. Handholding the filter at the length is safer but more awkward especially when one has to handhold the camera at the same time. It is thus more suitable with tripod shots.

I use the slot, but frequently handling or bad keeping still dirty them. i haven't find a way to keep them well but easy for access, although at the moment, it is best left in its plastic hard casing held in a journalist jacket. The plastic casing offers a firm container but remains fixed in the pocket that does not need much counter-resistance to pull out the filter and can be taken out with one hand. Trouble is how many of us is willing to wear that conspicuous jacket when we travel? If one use it solely for landscape on tripod and have all time for fiddling, it should be ok though.

.
 

5.3.5 Graduated Neutral Density (GND) filter: Vignetting issue

That skeleton (filter holder) that holds the filter is almost like a human pelvic bone that protrudes a certain distance from the front of the lens front thread. While a lens hood is designed specifically for each lens, enabling proper shading from flare but not causing a vignette, a generic GND filter holder that is applied on all lenses does not seem to have that consideration addressed.

I did not use Lee as yet but heard that it is bigger and causes less vignetting, but by how much, i do not know.

As mentioned above, for cokin holder, when applied at 10mm without any UV filters on, it will start vignetting at about 12mm on a 1.5x crop factor camera and at about 14mm if an ordinary UV filter is applied to protect the lens. One can choose to remove the UV filter to gain the 2mm focal length wider, but such attempts need to be taken with care because of increased risk of dirtying the front lens element.



5.3.6 Graduated Neutral Density (GND) filter: Color Cast issue

With the extra cost of Lee and Singh Ray, understandably they offer the best quality and neutrality from color cast. I have tried the p series size for Singh Ray, Hitech, Cokin and Tianya, of which the hitech filter having the strongest reddish cast. The end results ranges from anything from blue to magenta to pink depending on the natural colors of the sky or hidden color behind the glaring white, as well as the way you applied the white balance.

I personally find that such color cast works for skies with a warm color temperature or works against a blue sky with a colder color temperature. That is however based purely on my sense of color combinations for my works. I find that it is more natural for a color to fade into another similar hue that is complementary and not contrasting, something i have learnt since drawing with crayons when in kindergarden and applied on textbooks when highlighting text. Violet goes with pink. Orange goes with yellow. Red goes with magenta etc. Even if blue mix with yellow or with pink in the sky, they would supposedly follow similar color temperature, for there is no single blue shade, ranging from very warm to very cold. This is one reason i find some of the HDR attempts very unappealing for the laws of tones and hues are disrupted beyond aesthetics.

Hence, color cast is something one have to consider in handling in photoshop or to face what your filter gives you and just accept it. I have doubts that one who bothers to use a GND filter is the type of people who will just leave what the picture is and dun bother to perfect them after going through so much trouble to mold the tones, but that shall not stop you from doing what you want.

Some people uses blue tac to stick their filters onto the lens to avoid using handholding or the filter holder. I have no attempt with that and have no faith in that. I do not know if blue tac leaves a smudge, or if it will drop. Do considering googling if you are keen on the idea. It may work.

I find that since tianya offers not much worse than cokin and hitech in quality (i can't tell the difference, they basically all degrades quality when dirty) and has better color neutrality at least compared to the expensive but disappointing hitech, might as well get tianya filters.



5.3.7 Graduated Neutral Density (GND) filter: General advice

If one is determined to get the best quality for landscape and dun mind the hassle of using this cumbersome tool (the use of tripod itself is one cumbersome thing anyway, and including the hassle to maintain the cleaniness of the filters) - Singh Ray. This advice works along for the same for those getting lens and bodies - how sure you are in getting the best and how likely one is likely to stick to using it.

Alternatively, Lee filter and its holder is bigger in size, but causes less vignetting but is slightly cheaper than Singh Ray, probably equal or close to Singh Ray in quality, but then Singh Ray filters with cokin holder offers downward and inter-compatibility. One can bring different brands of filters, and to use them at various situations. You can use the cheaper one when it is really dusty and the expensive one when you have all the time on earth to ensure you dun dirty it.

If you ain't sure, get tianya first and try. When u see that you are really frustrated with the color cast but with careful handling your filter is still relatively clean with a few outings, then consider if u want to switch. If you hate the color cast and dirties your filters easily, e.g. within 3 weeks, then reconsider if you are willing to pay S$150 for a filter that just last you for 3 weeks. Maybe you should train yourself until you can maintain your cheaper filters clean enough for a substantial good time, e.g. 2 years and above, then it is probably then worthy to buy these filters. The exception is unless you are really very rich or is earning big bucks from the photos you take, but i guess those who earns big bucks from their photos are probably more familiar with these tools than i am and definitely won't be reading this guide from a newbie.

.
 

5.4 HDR techniques

As described above, HDR is an alternative.

It is a specific post processing technique that makes sures of overlapping but a different set of tonal range on different exposures, and mapped together to provide a wider latitude that our sensor or the old film cannot reproduced. Details are as described above. I haven't really got into this field as my travel photography still need to keep a certain element of reality. HDR pictures typically give a feeling of being in another world, and given the power of mastery digital manipulation, the degree of realism can be altered with control. The important note is watch the degree of color harmony and not to allow grossly unnatural hues to show, as well as controlling the level of exposure composition without introducing halos.

We have a forumer here who have so far shown many works of mastery over this technique. With his permission, I have kindly selected one of my favorite to share with newbies what a good HDR picture should look like.

From Hazmee, and his thread
IMG_6938.jpg


Although HDR pictures does not exactly produces what our eyes would see or feel familiar with, it is still important not to overthrow certain rules of nature and introduce a sense of disharmony. For example, as above, the tones in the water are in general usually darker than the sky, and it would always be softer than the sky depending on the shutter duration. It is also natural for the water to give an extra tinge of whatever colors it have, rather than to have a neutral white balance.

The ability of HDR to give a wider latitude allows dark shades and highlight gradients to show, hence we can delineate steps under the pavilion and see a soft glow within the clouds rimming the sunset, as shown above. It gives a silhouette with details and a highlight with details - which our sensor cannot cope with usually, especially with highlight. The ability to introduce graded tones in the highlight allows HDR to have an adge.

Often during photo enhancement with HDR, there is also control of the tonal composition, which can derives studiolight-like effects. Realism can still be retained with a different touch, which is better seen in his picture #5 in the above thread.

From Hazmee, and his thread
IMG_6259.jpg


Basically there are natural range of tones. Cover up the sky and look at the foreground. Then cover up the foreground to look at the sky. They are both natural looking on their own, but when brought together, it demonstrates a different contrasting lighting condition and introduce drama.

One of the HDR gallery recommended by eikin is here, but i think hazmee did a better job in determining supposedly dark regions.

.
 

5.5 Good HDR versus controversial HDR application

The reason why I make a differentiation between the two is as follows.

Good HDR, just like good sports capture, good portraits and good landscape, are almost immediately and universally regarded as good, with non-controversial attributes of its strengths that is often more objective than subjective. How good or how much one likes it, is of cos subjective, but there is generally no disagreement about whether a great work is good or bad. It will be hard pressed to find someone who says Steve McCurry's works is medicine for me and poison for him. Ansel Adam's works is not overwhelming appealing to me, yet i dun think i will say it is bad or that one man's poison is another man's medicine. Many things are of cos in the intermediates, but the discrepancy in opinions of the really good ones and really bad ones are usually relatively small, between very good or good, or between very bad or bad.

Controversial HDR work, is of cos, controversial. It basically falls into a cult group who believe in averaging all highlights and shadows in the frame into pure midtones and creating dramatic and strong coloring. The strong colors can represent either personal expression like that of the Fauvist art movement or an over-application to create vibrant appealing colors, whereas the averaging of tones appears to be a high tendency on automated HDR software but adopted by supporters as desired. This of cos draws strong opinions from conventional viewers versus a new generation of photographers, some of whom have done great works in non-HDR before, while some have ventured and indulged into HDR rendering without foundation on composition of forms.

I came across a wikipedia article on HDR lighting for video games and computer generated movies, which lay down good objective principles on what good HDR is. As quoted from the above mentioned article, good HDR photographs share the following attribute as good HDR computer graphics.


Preservation of detail in large contrast differences

One of the primary features of HDR is that both dark and bright areas of a scene can be accurately represented. Without HDR (sometimes called low dynamic range, or LDR, in comparison), areas that are too dark are clipped to black and areas that are too bright are clipped to white. These are represented by the hardware as a floating point value of 0.0 and 1.0 for pure black and pure white, respectively.

Graphics processor company nVIDIA summarizes one of HDRR's features in three points:
Bright things can be really bright
Dark things can be really dark
And details can be seen in both



Accurate preservation of light

Without HDRR, the sun and most lights are clipped to 100% (1.0 in the framebuffer). When this light is reflected the result must then be less than or equal to 1, since the reflected value is calculated by multiplying the original value by the surface reflectiveness, usually in the range 0 to 1. This gives the impression that the scene is dull or bland. However, using HDRR, the light produced by the sun and other lights can be represented with appropriately high values, exceeding the 1.0 clamping limit in the frame buffer, with the sun possibly being stored as high as 60000. When the light from them is reflected it will remain relatively high (even for very poor reflectors), which will be clipped to white or properly tonemapped when rendered. Likewise when light passes through a transparent material, the light that passes through has a lower brightness than when the light entered.


My conclusion of good HDR

Therefore, as above mentioned by what gives good HDR, they are supposed to have good tonal range from shadows to highlights with retention of details in both shadows and highlights, and not to make good tonal range into a single narrow range of midtones. And secondly, good HDR should gives a comfortable simulation of lighted surroundings, and not disrupt expected behaviours of light. A picture with no sense of light direction will lack depth. With similar discussion of this topic, someone posted a link to a good example by Feijão-Mestre, of HDR rendering to an indoor lighting giving a comfortable feel of alteration of light through the tinted window.



5.6 Poor attributes or adverse effects in non manual HDR rendering

Many of these attributes are deeper dimensions of judgement, balance, extent and suitability.

1. Colors and tones are part of composition, and not just lines and forms. An artificial rendering of colors and tones to individual taste may result in disruption to the sense of composition. The photographer may himself over-indulges in the rendering of texture and fail to recognise or appreciate the failure of composition.
2. Loss of sense of depth or direction of light due to unnatural alteration of colors and tones. This creates problems beyond being not realistic but visually uninspiring.
3. Conflicting color balance. This may occur with selective area processing in non-HDR related techniques as well. It basically mean having obviously erroneous mixture of color balance in relationship to space. Take for example, an apple is red and its color balance may change with the surrouding light. Imagine it is a beach with warm saffron (orangish) lighting, the apple is put in the middle of the beach and is not under any shadow, so it should be red with a saffron tinge. However, if there is erroneous color rendering, and it appears with a blue tinge, it will conflict with the spatial relationship of expected tones and color.
4. Unnatural hue or saturation. This may occur with over-processing in non-HDR related techniques as well.
5. Halo. A seemingly common feature in some automated HDR programmes but also common in abrupt margin processing in non-HDR related techniques as well.

Night86mare has shown some pictures with regards to his opinions on Point 4 and Point 5.

.
 

Status
Not open for further replies.