High zoom lenses preferred by lousy photographers

Photography equipments are full of trade-offs, and amongst them, cost, optical quality, speed (aperture). When you have a cheap lightweight zoom lens, you're giving up other things: amount of light it outputs for the film, image quality, and variety of pictures.

Amount of light:

For a given focal length, a fast lens (I'm not talking about focus speed, but "how much light it can send to the film/sensor") is usually larger and heavier because it needs more optics to take in light, and to output more light. Fast lenses today range between f/1.2 to f/2.0. Almost all the professionals I know prefer at least f/2.8 or less. Each f-stop is the difference between twice or half the light output (e.g. going from f/1.4->2.0 is one stop, and 2.0-2.8 is another stop). A f/1.4 lens is twice as bright as a f/2.0 lens, and four times as bright as a f/2.8 lens. Given the same focal length a f/1.4 lens will almost always be heavier and bigger than a f/2.0, and much much heavier and bigger than a f/2.8. Not surprisingly the cost of lenses go up exponentially as the f-stop number decreases only slightly. For day time pictures, none of this will matter because there is so much light that using f/5.6 to f/8 to f/11 will suffice. But for night time, or for fast action, you'll need to open up the aperture to allow more light. Let's say it is sunset and you need to shoot your kid's soccer game (approx EV 10). You have the following choices (given that you want at most 400 ISO for a 11x16 picture):

• ISO 400, f/5.6, 1/100 sec
• ISO 400, f/2.8, 1/500 sec

The two exposures above are almost equivalent, but you really need faster than 1/100 sec for sports. Let's say you'll accept 1/500 sec f/2.8 for a sharp-enough of sports motion (f/2.8 allows FOUR TIMES more light than f/5.6). Since most cheap zooms at the longest length START at f/5.6, you won't be able to take the shot, and even if you try to take the shot 1/100 sec will result in blurred motion. Another option is to boost up to ISO 1600 (400->800->1600 is 2 stops or FOUR TIMES more sensitive film), but now you're increasing grains that is not acceptable for 11x16 picture). In this situation the cheap zoom is too difficult to use due to motion blur, and almost useless at night time even when flash is used.


Zoom image quality:

Cheap plastic zoom lenses (aka "kit-lens") usually produce very soft pictures. Even at f/5.6, they tend to be too soft. Compared to a professional f/2.8 lens, at f/2.8 it can yield sharp pictures usable on magazines, and when they go to f/5.6, the quality of the pro lens will surpass the consumer lens by leaps and bounds.

Frequently cheap plastic zoom lenses exhibit a lot of undesirable traits such as vignette, chromatic aberration, etc. The bokeh on cheap lenses tend to be very busy, and ugly. There are those new "super convenient zoom" lenses out there with 11X or more ratio (e.g. the 18-200mm zoom). These may be convenient to use, but their image quality is very soft, the colors are subdued, the bokeh is horrible, and people are now just discovering that these super-zooms take horrible pictures; you'll see a bunch of them used, marked down, on eBay.

In short, you may end up with an affordable long zoom, but you may not end up using it much because you'll be disappointed with the images it produces. Everything is a trade-off of cost, zoom, and image quality.


Substitute for cheap plastic zoom lenses:

The target audience for cheap plastic zoom lenses are beginners, or people who simply don't know any better.

For people who just started in photography, prime lenses (no zoom) are great substitutes because they cost just as much as cheap zoom lenses, but more importantly their image quality rivals that of professional zooms that cost thousands of dollars. They are usually lighter or have bigger aperture. Equally important, they are great pedagogic tools for students to learn about exposure and composition. Prime lenses force the beginner photographer to WALK AROUND to get the composition they want. The act of simply walking around will yield a wide variety of unique shots, and people tend to like variety over pictures that look the same. Variety is the spice of life. You know some people sit in the same location and take hundreds of shots? They will have boring looking pictures; their pictures are very 1 dimensional because they lack varieties.

Moral of the story:

Camera engineering is all about trade-offs. A lens with lots zoom is sacrificing image quality. Camera companies want you to think that you're getting a lot for your money, when in fact they're also taking away other things. Don't fall for the zoom numbers. They mean very little.

High megapixel cameras preferred by lousy photographers

Photography equipments are full of trade-offs, and amongst them, cost, optical quality, speed (aperture). When camera companies pack many megapixel in a little sensor, the sensor will gain an undesirable trait: noise.


Megapixel race:

A lot of beginners think megapixel is a desirable attribute. There's a brand that starts with "C" that is very good at marketing to consumers; they tend to produce plastic cameras with high megapixels bundled with cheap plastic zooms (e.g. 55-250mm IS). Don't get me wrong, I use this brand, and I love the 55-250mm IS as a lightweight substitute for the 70-200mm f/2.8 monster. But I digress...

High megapixels are great if you need to print big sized pictures (16x20"), or if you tend to crop frequently. For most people, this is irrelevant because they only print up to 4x6, or upload to the internet at 1600x1200 (only 2 megapixel). A 12 megapixel APS-C (cropped sensor DSLR) is way more than enough to print a nice 12x16 wedding album. Heck even a 8 megapixel APS-C is more than adequate.

Thanks to marketing, common people now want lots and lots of megapixels (kind of how the layman cares about a CPU megahertz but doesn't care about cache+RAM). Company "C" loves to pack lots and lots of megapixels to consumers. But here's a catch. Noise. The moment you pack a lot of megapixels (and given the same sensor and processing technology), more noise occurs. The noise becomes much more apparent at high ISO, which makes it difficult to use at night. Technology is about trade-offs; nothing is free.

Let me ask you this. Given the following two choices (without using flash), what would you buy?

1. 15 megapixel that looks great when shot during the day and you can print up to 16x20. Good grains up to ISO 800.
   
2. 8 megapixel that looks great when shot during the day but you can only print up to 8x10. Good grains up to ISO 3200.
   

It really depends on your style. For me personally I shoot at night-time frequently. I love shooting at night not because it's easy, but because it's extremely difficult as it is very demanding on the capabilities of the camera sensors and the lens, and the skill of the person behind. Despite years of experience I still make crappy pictures at night, and occasionally good ones; I do it because it's tough, and because it's always an educational and fun learning process. Most people can take technically good pictures (proper exposure) during bright day time using full auto-mode, but very few can shoot well at night. Personally I go for the second option. In fact, since I never print 16x20 and almost all of my pictures are internet sized (1600x1200 or less, or 2 megapixels), choice 2 is more than enough for me. In fact, if I can pick 5 megapixel with super high ISO (3200-256000 ISO) for night time, I'll take it over a 15 megapixel sensor with only 800 usable ISO.



Newest low-end Brand A vs. Brand B at 100 ISO. The two classes are quite equivalent, and look great at ISO 100.



Top Brand X 15MP vs. Bottom Brand Y 12MP, ISO 3200. Pictures saved as GIF instead of JPG since JPG has a way of "smudging" over noise. Note the blue "chroma" noise for Brand X, despite having more megapixel.

Wasted megapixels:

The other interesting thing about camera companies these days is they pair up a high megapixel camera with a cheap kit lens. What consumer end up is having a sensor with resolving power much higher than that of the lens, so the extra megapixels end up being wasted. Pairing up a high megapixel camera with a kit lens is an engineering mismatch, but apparently, it is brilliant from the marketing perspective.

For example if the lens can barely project 15lpmm (lines per milimeter) but the sensor can record up to 30lpmm, then the sensor will still only record 15lpmm from the lens. With this combination, it is no different than buying a sensor that has 1/2 the megapixels, then upsizing it. By doing so, no extra details will be recorded, and the only thing a consumer will end up is having a bigger image and a bigger file, and an underutilized camera.

One similar analogy I can think of is a car company that pairs up a super expensive Y-rated sports tire with a car that barely has 50 horsepower. The Y-rated tire can go up to 186mph. But seriously, I don't know any sportscar with 50 horsepower that can go up to even 100mph. It's wasteful and a plain mismatch to put such a high quality tire on a slow car. However, marketing knows better, and marketing knows that consumers will buy based on meaningless numbers and slogans alone (megahertz? megapixels? 16X CDROM speed? MMX! IT'S A PENTIUM INSIDE!)


Moral of the story:

Camera engineering is all about trade-offs. A camera with high megapixel is sacrificing the ability to shoot well at night. Nothing is free. Don't fall for the megapixel numbers. They mean very little by themselves. One needs to evaluate a camera system as a whole.


--------------------------
High ISO shoot-out comparisons:

Nikon D5000 (12.1MP) vs. Canon T1i (15 MP)
http://www.cameralabs.com/reviews/Nikon_D5000/noise.shtml

Nikon D700 (12 MP) vs. Canon 5d mk 2 (21MP)
http://www.xtremephotography.ca/Canon-5D2-Vs-Nikon-D700/

Nicer body vs nicer lens?

A typical conversation starter with another photographer is "What kind of camera do you have?"

Many people take pride in their $2000-$4000 camera body. More often than not however, I see people with a really nice prosumer $2000 camera fitted with a $300 subpar zoom kit-lens. The best analogy to this I can think of is buying a super fast Pentium Quad Core fitted with only 512MB RAM running Windows Vista; this is huge component mismatch. Likewise, when you fit a nice camera with a cheap lens, the limiting factor to achieving high image quality is the lens; you're not utilizing the camera to its full potential; most of the pixels will be recording imperfections from the lens. I suspect people use cheap lenses on expensive bodies for the following reasons:


Cheap lenses are lighter. People using cheap lenses don't have to worry about damage/lost.

Yes, these are good reasons to use cheap lenses. If you need something light or compact, by all means buy consumer level (plastic) lenses because they're lighter, and the cost of losing or breaking them or losing them is nearly none compared to owning pro lenses.


Cheap lenses are good enough on a $2000 body.

Noooo! The lens takes in light as input and outputs light on the film, or in these days, a digital sensor. The camera is simply a box that records what the lens sees. If your lens has imperfections, those imperfections will be recorded in the box (film, or digital signals->files). Cheap lenses are full of imperfections, and with a high megapixel camera you will record all the imperfections.

One may argue that in the digital age, you can correct some imperfections using software. This is true for simple to correct imperfections such as vignette. Distortions can be corrected as well, and to some extent chromatic aberrations. However, there are other aspects that are difficult to correct with a badly designed or badly built lens, and in some cases, a corrective procedure may not be desirable or even possible. Below are a list of problems caused by cheap lenses, and the corrective procedure (if any):

• lenses with not enough resolution; the camera film/sensor captures more lines per millimeter (lpmm) than your lens can project sharply. In another word, your lens is too "soft" and can't resolve as many lines as your sensor can. This is as bad as up-sizing an image-- you're not improving image quality. Corrective procedures: downsize your picture (waste resolution) or resharpen (but add noise) or both. Also try stopping down (f5.6-f8) while adding shutter (blur) or ISO (noise)
  
• lenses with certain tint. Corrective procedures: turn it into B/W. Play with tinting.
  
• lenses with badly looking bokeh. Corrective procedures: use Alien Skin Bokeh and Photoshop and spend 15-30 minutes per picture... pain.
  
• slow lenses (with high f-stops) which causes your camera to auto-set the ISO too high (or worse, or that causes your camera shutter to drag). Corrective procedure: not much.
  

A cheap lens+expensive body will produce pictures with very visible imperfections that the lens produce. On the other hand, an expensive lens+cheap body will produce magazine quality pictures.

Money is better spent on the body than a lens?

Noooo! A financial metaphor to a camera equipments is that the camera body is like a computer, and the lens is like a monitor. Both will depreciate in time, but the computer will depreciate much faster than the monitor. Like a computer, a digital camera is equipped with the latest bells and whistles and will depreciate 50% in 2-3 years. Next year, there will be a better camera; with higher ISO, more mega-pixel, more FPS for the same price or even less than your camera current one. Moore's Law rules in the digital world.

On the other hand, pro lenses (Nikon gold-rim and Canon L red-rim lenses) do not depreciate as fast. After the 60s and 70s, glass and optics innovations and breakthroughs haven't been as life-changing as microprocessors have been. Sure there were defining moments in integrated lens technology like auto focus, IS/VR, flouride coating, nanocoating, so on so forth, but in general, the field of optics is a very mature field compared to electronics. The laws of optics doesn't change every year like computer does, good glass also retain their values well. Case in point, let's say 10 years ago you were interested in the 80-200mm focal range (and you don't care about size/weight). You could shell out money for a consumer plastic version Nikkor 80-200mm f/4.5-5.6 for $200, or a professional 80-200 f/2.8D for $1200. The professional version is over 1 stop advantage at 80mm, and a whopping 2 stops advantage at 200mm. It's also a lot sharper. Let's say the $1000 price difference is too significant and you end up getting the consumer [plastic] version. After 10 years, you'll have lost many potentially good shots. For example, let's say on stage, a shot that would suffice with a professional 200mm lens with f/2.8 1/200 sec at 800 ISO, now you'll need to use it at f/5.6 at 1/50 sec at 800 ISO on a consumer version. When aperture is open wide on the consumer lens, the image quality is very soft, and the increased in time exposure will usually result in too much hand shake and/or motion.

With this same example, you may be surprised that a pro lens doesn't cost much more than a plastic lens simply because a pro lens will not depreciate much. The true cost of ownership is really the cost of purchase minus the cost of the lens (minus 0-8% transaction fees) when you sell it. Professionals buy and sell all the time. Going back to our examples, after 10 years, you can sell the consumer 80-200mm lens for about $70 on eBay today. On the other hand you can the professional 80-200mm for about $900 on eBay (2008 price). The consumer lens doesn't retain its value well, but the pro lens does. In the end, your final "true cost of ownership" in this example is as follows:

1. Consumer Nikkor 80-200mm f/4.5-5.6 initial cost: $200. Selling price: $70. Cost of using the lens for 10 years: $140. High depreciation.
2. Professional Nikkor 80-200mm f/2.8D initial cost: $1200. Selling price: $900. Cost of using the lens for 10 years: $300. Low depreciation.

In the end, your true cost of ownership is $300(the pro) vs. $140(plastic lens). In another word, if you picked the professional version you'll have spent only $300 in 10 years, or $10/year that produced amazing looking, magazine ready pictures. On the other hand if you had picked the consumer version you'll have spent $140 in 10 years, or $14/year that produced many lousy shots. You may even end up so discouraged with the consumer lens and not use the lens at a later point.

----------------------
In short, a good quality lens will always retain its values for a long time. It has been that way for decades, and will be that way for as long as the laws of optical physics hold. On the other hand, prices of fancy electronic camera bodies and cheap plastic-feel lenses will drop as fast as computers. They're horrible "investments". A common problem I see today is that people buy nice bodies fitted with a cheap lens. This combination has the worst of both worlds: 1) a cheap lens will not resolve details as well as the body, so megapixels are wasted 2) an expensive body will depreciate 50% in 2-3 years.

A better combination is to fit a super cheap body with an expensive pro level lens; you'll get magazine ready photos, and you'll save more money in the long term because the lens will retain its values well.

You know how photographers often ask each other "What kind of camera do you have?" They really should ask "What kind of lenses do you use?"

--------------------------
Notes:
1) Cost of lens prices based on observations from historic prices from eBay and CraigsList, and cross referenced with Ken Rockwell's website.

2) Cost does *not* take inflation into account

3) More datapoint backfill in 2009:
June 15th 2009 marked the 10 years anniversary of the the first professional Nikon DSLR - the Nikon D1: 2.7 megapixels were selling for $5850 back in 1999 (in 2009 the D1 sells for less than $ 200 on eBay). In 10 years, the price dropped to 3.4% of its original value.

Multiple cameras? Sync up your clock!!!

Every decent photographer organizes photos by grouping them, comparing similar pictures, picking out good pictures, and deleting bad pictures. This has been true even in the age of film photography.

Then came the digital age. People machine gun their cameras because there is no longer a notion of cost to each picture. Storage is cheap so just shoot a lot and hope to get a few that looks right. Unfortunately, machine gunning has a cost: time. Now, people are spending more time than ever in front of their computer, organizing tons of badly taken, machine gunned photos.

I just came back from an event where I took 600 pictures with 3 different cameras. Yes there are times I'm also guilty of machine gunning my camera, but at least I throw out bad ones. Usually I can sort through hundreds of pictures within an hour. My workflow to get the best 20-30 best pictures entail the followings: 1) grouping similar photos or photos 2) giving rankings to each photo (judging from technical merits, emotional merits, and uniqueness) 3) picking out the ones with the highest rankings 4) enhancing the ones that have high rankings

Digital cameras make step 1-2 very painful because it's so easy to machine gun your camera. You have the pleasure of taking tons of bad pictures, but end up with hundreds of photos (instead of just 2-3 rolls of film per event in the old days) that you need to stare at again. You end up looking through them, thinking "why the hell did I take this lousy picture?" Hours and hours in front of the computer will pass by, and by the time you're done, you're thinking "That took a long time and it was very very painful. Digital sucks."

Well, digital sucks even more if you have hundreds of pictures that have different time stamps that make sorting and grouping them impossible. Case in point, I just went through about 600 pictures taken with 3 different digital cameras. Unfortunately, all three had a slightly different internal time stamp, so when I tried to group them, the sequence of pictures overlapped one another. I actually had to go through them one by one to group them manually for comparison. I was too lazy to write a script that extracted EXIF and modified EXIF to match the time stamps, and in retrospect that was the right thing to do. Anyways, what could have taken just an hour or two, ended up to be almost a day of work. Having wrong time stamps really messed up the work flow.

Moral of the story #1: it really pays to spend a minute to sync up your digital camera clock.

Moral of the story #2: everyone is going to run into this situation at some point anyways, so find a script/program that can shift/sync EXIF time stamps.

Back to basics: M.A.S.P. and M.A.Tv.P.

Before we talk about why setting your camera mode to manual is preferable in some situations, let's first go through a very brief history of how some of the fancy features of a modern camera came about. During the process, we'll explore how to make great pictures by not using fancy new features.

The beginner photographer from 1900 to 1950s have only 3 big variables to think about in terms of exposure: aperture, speed, and ISO. Let's call it the "Big-Three." These three fundamental settings need to account for how much light is available (e.g. EV level). Life is easy to think about in the context of the Big-Three:

1. aperture: exactly how much depth of field do I want or have?
2. speed: if I handhold my camera, what is the minimum speed I can set to before there is hand-shaking blur?
3. ISO: low ISO yields good grains but now you may have hand shake. High ISO yields course grains but at least your pictures may be sharp. What exactly is a good cut-off line between little hand shake and little grain?

For each lighting condition (EV value), you have several ways to configure the Big-Three to match the EV. The usual procedure is as follows:

1. estimate or measure (via a primitive Selenium light meter) how much EV is available.
   
2. configure the Big-Three to compensate for EV, using experience or exposure charts and slide rules. It's actually quite easy-- if you passed DMV exams, you can memorize the chart just as easily. There are charts/rules you can download from the internet. You can also scroll down to the end of this blog to download them.
   

With your newly downloaded chart, let's start with an example with our good 'ol 200 ISO film. It is sunny outside, bright sun, hard shadows. This is EV15 according to the chart. According to the slide rule, for EV15 you have the following options:

• f/16 1/250 sec <-- sunny 16 rule
• f/11 1/500 sec
• f/8 1/1000 sec
• f/5.6 1/2000 sec
• f/4 1/4000 sec

All of these settings are equivalent in terms of exposure (how much light will reach into the camera). At f/16 and 1/250 sec, you can use a focal length of 250mm without worrying about blur (based on the reciprocal rule focal length rule and speed rule), and many objects will be in sharp focus because the depth of field is deep. At f/4 and 1/4000 sec, any motion will be frozen in time while the background will have a decent portrait-like blur to it. Regardless of the setting, you can feel comfortable "getting the shot" reliably.

By the way there is this Sunny-16 rule of thumb old timers like to talk about: "When it is nice and sunny outside, the proper speed at f/16 is the reciprocal of the speed of the film." So when it is sunny outside using 200 ISO, the proper speed at f/16 is approx 1/200 sec (we're off by a bit, but +/- 1 EV is not a big deal). Likewise, when using 400 ISO, the proper speed is approx 1/400 sec.

In general, when there is ample lighting, and lots of choices in the Big-Three, pictures turn out great. But when there is very little light, photographers have very little choices; the challenging lighting conditions often exceed that of the capabilities of their equipments. Let's take an example of EV5 (indoor night home, or lit by campfires or bonfires). At EV5 with 200 ISO on a 50mm lens, here are the choices:

• f/1.4 1/30 sec <-- can only focus one eye
• f/2 1/15 sec
• f/2.8 1/8 sec
• f/4 1/4 sec <-- lots of hand shake blur

In these two spectrums, there are two equally bad choices: only one eye in focus, or getting a lot of hand shake blur. Anywhere in between, you'll still get a crappy picture. In this lighting condition, the photographer simply cannot make good shots reliably due to hand shake and limited depth of field. He/she can use a tripod and ask the subjects to stay very still, or use a much higher ISO film (increase grain), or use a flash, or a combination of. In short, limited lighting will limit the Big-Three choices, drastically. Night shots are challenging to even the most seasoned photographers.

A good photographer will immediately know the limitations of his equipments in particular environments and reconfigure the equipments for different lighting conditions in a nick of time. A beginner photographer will try his best and fiddle with the equipments and eventually figures out that the lighting condition exceeds the capabilities of his equipments... usually when the precious Kodak moment has already passed!

In the 1950s, cheap, small consumer light meters came out. People no longer had to guesstimate the EV level, but still had to configure the Big Three manually. Later on, some manufacturers integrated light meters into the speed dial of the camera (Leica MC/MR-4), such that moving the speed dial also moved the slide rule that told you what aperture you should use. Finally, you manually move the aperture based on that slide rule.

Throughout most of the 40-50s most cameras didn't have a built in light meter. In the late 50s the legendary Nikon F came out (later on with F mount Photomics), and slowly convinced rangefinder users to switch to SLR. The Nikon F with Photomic had a built in light-meter such that setting the Big-Three on the camera also changed the view finder's meter for proper exposure. With this light-meter/body integration, there is no longer a need for the traditional process of 1) measuring EV, and 2) configuring the Big-Three separately. You can do 1 and 2 in the camera, in one step. But you still had to adjust the Big-Three.

The 70s and 80s saw much more electronic integrations between light meters and camera controls. First, S/Tv priority came out. No longer did people have to set all Big-Three. A photographer only needs to set the the aperture and the ISO, and allow the lightmeter+analog electronics figure out how much time should be used to expose the film. The Big-Three could now be called the Big-Two. With this new S/Tv priority mode, photojournalists who need to shoot in fast changing conditions where they didn't have time to to configure the settings, can now make shots, fast. Now, they can remove themselves from technical knowledge of cameras, to other important aspects of photography such as timing, composition, angle of view, interpretation, etc.

The late 70s and 80s saw the advent of A mode. Photographers could manually specify how much time to expose to the film, and the electronics figure out what aperture to set the lens to. New lenses had to be made such that the body could change the lens aperture electronically.

Soon afterward both A and S/Tv mode come out, the P mode (program mode) also came out. A photographer would first set the ISO, and the electronics do a best effort guess on time and aperture. It's almost dummy mode. To make things easier (or more complex, depending on your point of view) companies like Canon came out with special modes like Sports/Portrait/Night Scene, etc on cheap plastic Rebel bodies. They marketed using photos of Andre Agassi taken with a $300 consumer Rebel body fitted with a super $5000 white professional telephoto lens (very misleading). They marketed the Rebel as the camera that anyone could use; just look at this amazing picture of Andre Agassi! Buy this Rebel body! Proud sponsor of Wimbledon. Fine print: taken with a cheap Rebel fitted with a $5000 white professional lens.

Nowadays, people take their SLRs and just set it to A(auto) or to one of the modes, like Portrait, Sports, Scenery, etc. They have absolutely no idea whether their cameras can take the picture that they want given the lighting condition they have. In general, they'll try to take the shot, look at the LCD, and be disappointed with the results. Then next year, they want a slightly better camera because they think it'll take 10X better picture. They do this without any understanding of exposures, and waste money along the way. So this is what I say to all the beginners who rely on auto mode-- STOP USING AUTO MODE!!! Instead, set it to fully manual for a day. Tape over your wheels and only set the Big-Three. Yes it'll be extremely painful to have to think while trying to get the shot, but it'll be a great learning experience. I promise that after that one day, you'll gain more knowledge and appreciation for exposures, and you'll make better pictures for the rest of your life. I recommend you download the exposure chart, and think in terms of 1) lighting conditions and 2) what type of Big-Three setting you need for the style of photograph you want to take.

After that one day, go ahead and set the mode to whatever you want. With a fresh understanding of your equipments and their limitations, you'll make better pictures. Here is my rule of thumb for setting the modes:

• AUTO: a mode I set to when I give the camera to my mom.
• P: I'll use this mode only when I need convenience, and I am mostly outdoors (EV11), and moving between different locations (indoors, outdoors). I also set to AUTO ISO.
• A: a mode I set to when I specifically need the depth of field (people), and have ample light. Appropriate for slow/still things.
• Tv/S: a mode I set to for sports or water or clouds moving and such. I need 1/500 to 1/2000sec to capture fast moving objects.

The only difference between film and digital is that you can set the ISO easily on a digital camera. I started doing the followings when I went digital:

• M+auto ISO: I set this when I'm indoors (EV5 or below) with no flash. I will only set 2 variables (aperture/speed) and let the computer pick the ISO. When I'm indoors, I will use a prime lens (f/1.4), and I absolutely need a speed of 1/60 sec or faster. So I only set the aperture and speed and let the computer pick the rest. For example if only EV5 is available to shoot people using a 50mm lens, I have very limited choices because I have limited light. I start at f/2.0 1/60sec (computer will auto choose 800 ISO) for EV5. If f/2.0 is too shallow, I may reconfigure to f/2.8 1/60sec (computer auto chooses 1600 ISO). If that's too grainy, I pick f/1.4 1/60sec (computer auto chooses 400ISO).
• M+manual ISO: I set this when I'm in one place and need very consistent exposures. No matter how fancy your matrix metering is, it is still reflective metering, which although is wondering is easily fooled frequently, and I often end up with 10 indoor picture that all have slightly different exposure that I need to recalibrate afterward. In this situation, it's better to measure your EV level once (gray card or incident metering) and be done with it. Sometimes I also use spot-metering to see the light level on the stage, and keep the setting the same. When your exposures are correctly set AND consistent, you don't need to play with post processing as much. If you get it right once, you never have to worry about it again. This is the ultimate mode when shooting in indoor studios, when you really know what you're doing.

Your camera has a bunch of other modes that really is just a combination of M/A/S/P and other settings:

• Portrait: the computer will set as wide of aperture as possible (blur background), and some cameras will give more warm tone to it
• Landscape: the computer will set the lens to a sweet spot (usually f/5.6 DX or f/8-11 FX), depending on brand and lens and other factors. Shutter will expect to be slower, but it's landscape, so it's ok.
• Sports: the computer tends to set wide aperture, medium/higher ISO to capture fast moving objects. Depending on model may add more sharpening in the JPG
• Night scene: the computer will do a slow-sync, flash the foreground with slower shutter speed for the ambient lights

-------------------------------
Andrew L Lawn's exposure meter is very pretty to print and make. I love it
http://squit.co.uk/photo/downloads.html

Interesting reading, on the side of being overly verbose.
http://www.fredparker.com/ultexp1.htm