Dan's Data letters #176
Publication date: 25 October 2006.Last modified 03-Dec-2011.
Deep focus
Seeing as the tip of your pinky finger knows more about photography than I do, could you please impress upon me some of your vast photograph-taking knowledge?
I'm just getting into photography. I took the attached shot and quite like it...
...but I'm frustrated by it's relatively small depth of focus. This was taken with a Nikon Coolpix S1 (using its macro mode), which I realise is no photographer's tool, but if I were to buy or borrow something more expensive, could I improve such shots? Or am I doomed to taking 2 or more shots with different focal points and Photoshopping them together?
Daryl
Answer:
I was about to tell you "no", but in the particular case of the S1, maybe you actually could get rather better results
from a different camera.
I'm going to have to ramble on a bit here to explain why this is. I will commence by patronising you with a description of how depth of field works.
Depth of field depends on the focal length of the lens (which is how far away from the film/sensor/whatever a pinhole would have to be to throw an image the same size as that thrown by the lens) and aperture size. The shorter the focal length of the lens - the actual focal length, not some fiddled-with "35mm-equivalent" focal length - and the smaller the hole in the aperture plate/iris/whatever that limits the amount of light that can get into the camera, the greater will be the depth of field.
Here's a neat-o calculator.
The Coolpix S1, like a many super-tiny pocket-cams, has no manual controls at all. There's a manual white balance, but that's it. Cameras with no manual controls make it difficult to deliberately set the aperture small. They're likely to do so if you're shooting a brightly lit scene - because a smaller aperture means less light gets into the camera - but there's no way for you to tell them that you're happy with a slower shutter speed in return for the smallest possible aperture.
The next step up from this in the manual control department is shutter priority and aperture priority, where the camera sets its exposure automatically but you get to tell it what shutter speed or aperture to use. Lots of "consumer" cameras have these modes, and they're all you need for a lot of the photography that most people, even professionals, do.
The next step up from that is a proper full manual exposure mode, where you can set your own shutter speed and aperture.
If you've got a camera with aperture priority and/or full manual exposure control, you can tell it to use its smallest aperture (which means you set it to the largest "f number" it supports) when you want maximum depth of field.
Consumer digital cameras generally offer quite a lot of depth of field even at their larger aperture settings, because they have very small sensors.
To get a reasonable field of view from a camera with a very small sensor, you need a lens with a very short focal length. And a short focal length, as I mentioned above, is one of the ingredients of a wide field of view.
The S1 has a "1/2.5 inch" sensor. This number appears to not make much sense (one divided by two point five inches? That's 0.4 inches, right? But 2.5 inches is 63.5 millimetres - one divided by 63.5mm is 0.016mm!) because it doesn't make much sense; the weird fractional notation is because that's the way old TV camera tubes were specified. But, to be fair, if you do the calculation in inches, that actually does tell you the diagonal measurement of the sensor chip - 0.4 inches, in this case.
Anyway, this sensor is not the smallest you'll find in a consumer camera, but it's pretty small.
Do you remember the old "110" cartridge film format? It was cheap and simple, and thus popular before digicams were available, but was too small to ever give sharp results with prints much bigger than your basic six-by-four snapshot. 110 film had teeny little negatives compared with 35mm, but they were still 17 by 13mm in size.
Your S1's sensor, which is not the smallest used in digicams, is only 7.2 by 5.8mm.
So, in order for the camera to have a reasonably wide view of the world, its lens needs a very short focal length, and will in turn give much more depth of field than, say, a 35mm camera lens that has the same field of view.
Lots of depth of field is bad when you want "selective focus", the ability to get your subject in focus with the background (and maybe some visible foreground as well) out of focus. People shooting movies on camcorders make fantabulous contraptions to reduce their depth of field. But when you want a deep in-focus area, consumer cameras are great. They're definitely better, for this, than professional and semi-professional digital SLRs, which have much larger sensors (around the 25-by-17mm "APS" mark for cheaper DSLRs, full 36-by-24mm 35mm-frame sized for the expensive ones).
There's not a lot of difference between the depth of field capabilities of almost all consumer and prosumer digicams. Some have bigger sensors than others, some have longer lenses, but there's really not a lot in it. There's practically never any point in changing from one digicam to another, if all you need is more depth of field.
So, as I said, I was about to tell you that there was no point trying another camera. Until I looked at the S1's actual specs.
I'm not at all sure, you see, that the S1 actually has adjustable aperture. Its main "aperture" control appears to be a neutral density ("ND", clear grey) filter that it can snap into the image path when it needs to reduce the amount of light getting in. That's not actually an aperture change at all - there's no smaller hole letting light through the lens. It's just grey sunglasses, if you will. An ND filter doesn't do anything to depth of field.
The base lens spec for the S1 is as follows:
"F=5.8-17.4mm (35mm equivalent 35-105mm), f/3.0-f/5.4."
What that means is that the actual focal length of the zoom lens is 5.8mm at full wide angle, and 17.4mm at full zoom, and that this gives you the same field of view range as a 35 to 105mm focal length lens on a camera that has a "sensor" the size of 35mm film.
The f/3.0-f/5.4 part tells you the maximum f-numbers the lens can deliver at its widest and narrowest zoom settings. Now, a normal camera that can change its aperture will be able to stop down those apertures to f/6-to-f/8, or something; those apertures are still on the large side by 35mm standards, but with the very short real focal length of a tiny-sensored camera's lens, they give quite a lot of depth of field.
The specs section of the S1 manual (PDF) says, under "Aperture Range", "Electronically-controlled ND-filter selection, f/3.0 and f/8.5 (W)".
If it's saying that the lens actually can be stopped down to f/8.5 at full wide angle, then that's on par for other small-sensored cameras and should give lots of depth of field.
If your sample picture really is the best you can manage, though, then I'd surmise that the spec means that when the ND filter's in place, the lens lets in as little light as it would if it were stopped down to f/8.5. But it actually still has an f/3.0 aperture, and the depth of field that goes with it.
At that aperture, then if you're focussing a metre away, your depth of field will be something like 70cm to 1.75 metres - though it depends on where you draw the "too blurry" line - for small prints, the effective depth of field will be better.
If the camera can really manage f/8.5 at full wide angle, then focussing a metre away will give you sharp results from about 45cm to infinity. If this were actually the case, though, then the camera would presumably have done it in the nice bright daylight picture you attached. You could still easily see the entire depth of field when you focussed really close, but I doubt you'd have a problem with it.
So I really think that what you've got there is a fixed-aperture camera with a grey filter it can use when it has to.
And so, finally, yes, you could indeed get better depth of field from a camera that has an actual adjustable aperture.
Out of curiosity, I downloaded the manuals for the Coolpix S2 (PDF here; the S2 is very much like the S1) and S4 (which has a bigger lens assembly, with one of those swivel arrangements that used to be a Nikon trademark, and 10X optical zoom - PDF here).
The S2 manual says the same thing about aperture as the S1 manual does, but the S4 manual says "Electronically-controlled aperture and ND-filter selection", giving a list of F-numbers up to f/13.6. So it would appear to have the same ND filter doodad as its smaller cousins, plus a proper adjustable aperture. I bet the f/13.6 setting, which is extraordinarily high for a 1/2.5"-sensor camera like this, is just Nikon's weaselly way of saying that that's how much light gets into the camera when you use the next setting down, f/6.8, along with the ND filter.
Even f/6.8 with a sensor that small and the lens at its 6.3mm full wide angle setting, though, will give lots of depth of field. So will any other small-sensor camera with a similar real minimum aperture.
I didn't know before now, though, that there were big-brand proper autofocus zoom-lens cameras out there that didn't actually have any real aperture adjustment at all. You learn something every day, I guess.
Buttload of batteries
Hi Daniel, first just letting you know something fun, you answered a question for me in the past - the second one on this page.
Anyway, I came to my senses and bought some nice 60cm dishes... Got signal just fine after a few months of acquiring all the stuff, and on the night I had managed to get it all up and running, my brother came home to tell me that we can now get ADSL via Telia, argh, Murphy!!!
Anyway, this isn't why I'm writing to you. I've got a rather large UPS...
...that I got as a freebie. It's a Riello Power Dialog 600 6000VA 4.2KW. Since it was made in 1998, I have been impressed that the batteries have lived this far, but a few days ago reality caught up. One of the batteries had failed completely (was cracked open) and the others look bad at their contact terminals... Some blackish goo like substance on them...
Anyway, there are 22 12v 7.2Ah batteries in it, called FIAMM-GS Model FG20722 (PDF).
Now, at Elfa.se I can get a pretty good deal with similar batteries (part #69-532-28), since I need so many I get a pretty good discount - only 137SEK each ($US18.74 with today's exchange rate). If I want the correct terminal size too, I have to add 34SEK for the part #69-532-30. I have used these batteries in the past in an APC Smart-UPS 700 and they work great.
The thing is that I fixed an APC UPS I also got for free a few months ago, a Smart-UPS 1500. I used elfa.se #69-479-07 batteries with that. What's so special about these batteries, bar the about 48% cheaper price compared to APC originals, is that they come with a 10 year instead of 3 to 5 year rated life, according to the Eurobat guide.
So what I am wondering is if you know of, or can find, 7.2AH batteries with this long design life. While the original batteries where also only rated for 3-5 years, they still held up for almost 9. So I think these batteries with a 10 year promised life would work really well, if I now could find them in the correct size and ampere/voltage rating.
Note that I can't just hook up any high amperage battery to this UPS, because they are wired to get 310VDC (at least that is what the management software tells me).
The batteries would also need to be located in Sweden, or at least Scandinavia, since the weight of 22 7.2Ah batteries would be about 60kg - else I'd have to pay an arm and 4 legs for shipping :P
Per Hansson, editor @ Techspot.com
Answer:
Sorry, but I don't have enough experience in the field to make any very useful recommendations. I've bought a fair
few cheap Chinese SLAs over the years (they last about as long as you'd expect them to...), but I haven't sampled
enough good ones to have much of a clue.
Apparently Panasonic and Power-Sonic (no relation) are both good; I dare say the Italian FIAMM-GS batteries are fine too, but I'm only guessing.
Note that when batteries last a long time in a UPS, that always means the UPS's charger is not mistreating them - not charging too hard, not overcharging, not allowing them to run all the way down. Even cheap SLAs can last surprisingly well in a UPS with good charge circuitry. Likewise, the best batteries in the world will be beaten to death in a year or so by a crappy charger.
And, of course, if the batteries don't have to work very hard then they're likely to last longer - that's the second essential part of the recipe for really long battery life. If the batteries have to deliver a full cycle every few days because you're running your computers in Baghdad or something, then they'll wear out much faster.
And then there's the environment. The batteries will be happier if they're kept cool and not bashed around.
Top quality batteries will deal with all of these factors better than cheap Chinese no-brand ones, but a "five year" battery will probably not last for two years if it's cycled three times a week, charged roughly, and run too hot.
So the quoted "life" ratings are all very well, but there's a reason why people who say they expect their batteries to last for five years do not actually guarantee that they will.
Accordingly: Do feel free to buy brand name batteries rather than Chinese cheapies, but don't obsess over it, because there are a lot of variables here and most of them currently seem to be on your side anyway.
The 310-volt thing is plausible enough - 310V is about 22 times 14V, the and 14V is a realistic peak value for fully charged "12V" lead acid batteries.
(This also, of course, means that this battery bank is capable of killing yo' ass very dead indeed. You don't need gloves if you're working on a 24V bank; you do if you're working on 310V. Thick gloves. And possibly a hand behind your back.)
I bet you could run that UPS just fine from any twenty-two 12V lead acid batteries, though. Your floor joists might or might not appreciate the weight of 22 car batteries, of course, and the charger might be mean to 22 SLAs if they're smaller than the standard ones. If you get yourself a deal on 22 12V batteries that aren't stupidly large or unacceptably small, though, then you could hook 'em up and have them work just fine. It probably wouldn't be a good looking setup, though, even by my standards.
