dlakelan

If I'm reading the wikipedia article on axial chromatic abberation correctly, then using an 81A/81B plus white balance should improve things since the most violet portion of the spectrum will be much more heavily affected than the mid-range blues, so you should be able to recover good whitebalance. It's the equivalent of "blue blocker sunglasses" for your camera :-)

Can you post an example image?

 

You can definitely put optical filters in front of your lens, but if the color is visible you will necessarily affect the visible spectrum. If you try to enhance the affected colors via whitebalance in post processing, the fringe will just come back with more noise...

 

If the cause is actually CCD blooming due to high levels of a blueish color in the lighting conditions, your 81A/81B filter can squash that and the post-processing whitebalance would be fine. This is actually one of the first questions I had when I went digital: is it worth it to use color warming optical filters to prevent overexposure in teh blue channel when taking photos in deep blue skylit conditions?

 

I think the answer with a modern DSLR is to simply underexpose so that the blue channel doesn't clip, and then boost the other colors via whitebalance/post processing since you have 10 or more stops of latitude in the RAW file.

Can you get the effect you want with just the center autofocus point (which is the most accurate point) and focus-recompose when you have a single subject?

 

I doubt you can get the effect you want at f/2 with two or more people in the frame just because their eyes are unlikely to both be in the plane of focus. It's not a technique for use with groups unless you are doing careful studio work perhaps.

 

If you are setting up to shoot a lot of head-shoulders portraits, dial your autofocus sensor to be the one closest to the eyes when you frame a shot in vertical orientation, and then shoot in that orientation without adjusting the autofocus point. the goal is to focus-recompose with a minimum of movement (so as to not disturb the plane of focus).

 

try your shots on a tripod with a willing test subject and see what it takes to get what you want at different apertures.

 

This seems to be a technique issue, and if you want to master the technique it will take careful experimentation. shoot 3 or 4 shots with various settings and a test subject, and then review them... choose new settings/technique and try again... until you're getting consistent results.

J. Harrington: You got an excellent result using your technique, but I recommend trying a spot meter off the water, expose 2 to 3 stops *over* that exposure.

 

Metering off the water will put the water at neutral gray, hence expose MORE than that to put it near full brightness (just like in snow).The water will be bright, but should not be clipped (check the histogram and tone down the exposure if necessary to prevent clipping). This will give you optimal exposure for preserving information in the shadows. Then in post processing you can tone down the water and pull up the shadows with less noise in the shadow regions. As long as you are not clipping the histogram, the water can be toned down with no loss of information in the highlights.

GIMP is working on integrating GEGL which will do all color manipulation internally in 32 bit floating point. Their latest release has GEGL as an optional feature. Based on that, I expect GIMP to be usable for 16 bit RAW manipulation within a year.

 

For photography I use digikam for organizing photos. dcraw for converting raw to editable formats, and GIMP/imagemagick for image manipulation of editable formats. Imagemagick does THE BEST job of downsizing photos when used with the "-filter lanczos" technique

 

lcms does color management and GIMP and digikam can both use it. There is even some software for doing screen profiling with hardware (argyll I think). imagemagick can apply output icc profiles for printing.

 

I believe you can do pretty much everything in Linux, so go for it and try it out.

Also check your metering mode, sometimes if you have a camera in spot metering mode the exposure will vary wildly depending on where you point the camera. If you point the metering spot at a dark object, everything will be overexposed, if you point it at a light object everything is underexposed.

 

This happens because the camera meter assumes whatever the meter is pointed at is a neutral grey tone... When you use a center weighted averaging or evaluative metering mode the meter is less sensitive because it looks at more of the picture...

 

Don't know much about your camera or what metering modes are available, but check to see what metering mode you're using and then read the manual about the various ones.

The things you can't get from a P&S that you do get from a DSLR:

 

1) fast interactive focus

 

2) lower noise at high ISO

 

3) shallower depth of field

 

4) A broader range of focal lengths

 

5) Quality optical viewfinder

 

All of these are controlled by fundamental physical aspects of the DSLR (such as methods of construction, size of

sensor, geometry of focusing)

 

You can get good photos with a P&S, if you get tired of carrying your DSLR around, by all means, go grab your A75

or even one of the newer A series P&S cameras, and use that for a while.

Robert:

 

DPI and PPI are different things. The D means "dots". a printer doesn't put down variable color dots, it puts down either a dot of a certain color, or no dot... You need to average many of these dots together to form a pixel which is the P in PPI.

 

So basically, the printer forms a pixel from many dots. More DPI means more color gradations within the pixel (ie. you're averaging over more dots)... So high DPI on the printer is always good... it means better color resolution... whereas higher PPI in the image is good because it means better spatial resolution (ie. ability to see the difference between things that are close to each other on the page.. like grains of sand in your image or something like that)

 

Some professional printers that print onto photo sensitive paper actually use a different system, where the intensity of the laser paints the pixels, not a set of little dots... on those printers they really are printing pixels, so there's no DPI just PPI.

 

Hope that helps.

Tips:

 

1) Shoot from 1 hr before sunrise to 2 hours after sunrise, or similarly around sunset

 

2) The higher your ISO setting the faster your shutter speed, and therefore less problem with motion blur. BUT

also higher ISO means more noise in the images (grain). Try out different ISO settings just at home in your house

on some still-life test photos, and see what the highest ISO you find acceptable is.

 

3) Color balance is important too. The three images in a row have different color balances. The middle one is

warmer. Try setting your white balance to "sunny" or to "shady" and see which one gives a nicer result.

 

4) Blurred backgrounds: to blur out the background more you need a small f number (which is a larger aperture).

f/2.8 is good but smaller numbers are even better. Depending on what lens you have, your lens may not have such

small f/numbers available. If you're mainly shooting pictures of people in the rainforest, an inexpensive, high

quality lens with a large maximum aperture (small f/number) is the Canon 50mm f/1.8. Large apertures (small

f/numbers) also give you more light to work with.

The simple equation is pixels = inches * (pixels/inch)...

 

When the image is in the computer, it's just a set of pixels, there is no notion of inches...

 

Your image has a certain number of pixels across its width, (or height) so if you want to print it at a certain

width (or height) in inches, your pixels per inch is just (pixels across image) / (inches across print).

 

Suppose you want 300 ppi and you have a 1000 pixel across image... you can solve for W the width of the image

 

1000 pixels / W inches = 300 ppi, solve for W... W = 1000 pixels / 300 ppi ~ 3.33 inches

 

so you can't adjust the inches and the resolution at the same time... they're linked by the mathematical equation.

 

Better thing is to set the width that you want, and calculate whether your ppi is acceptable. Suppose you want 6

inches width...

 

1000 pixels / 6 inches ~ 167 ppi is that enough for you? if it is, print it... otherwise make it smaller...

 

suppose you just want the largest image that is acceptable to you... choose the smallest resolution that is

acceptable, perhaps 150 ppi... then the size is determined from the equation...

 

(numbers above are just illustrative... not meant to be a real example for your case)

Penny,

 

Whisper thin depth of field can be a good thing, giving a soft look to some portions of the face...

 

On the other hand, if you want more depth of field, it is easy to shoot at a smaller aperture (larger f/number) the one that is listed on the lens is just the largest aperture available (smallest number) most lenses will go to f/16 so anything between f/1.7 and f/16 is available to you, if you have enough light... and each stop will increase the depth of field.

Honestly, I think it would be worthwhile to look at an inexpensive Canon P&S digital, like the A590 IS. You will progress much faster if you don't have to wait and pay a lot of money for film to find out how well your images are turning out.

 

The Canon A line of P&S has all the manual controls (aperture, shutter speed, ISO even a not that great manual focus mode) even in the lower cost range. There are much better P&S cameras, but they cost more, and you're better off saving the money to upgrade to a DSLR eventually.

 

The cost of taking say 200 photos (a typical shoot for a weekend photography "lesson") in 35 mm is maybe $50 to $75... which is around half the cost of buying the P&S. In my opinion the P&S will give better quality digital images than consumer grade scans of negatives or slides. And it will show you things like histograms of the image immediately after you've taken the picture... which will help you learn... The cost of shooting and storing the 200 photos on the P&S is a few pennies.

 

For a beginner, the low cost per shot, fast feedback, and direct access to the digital files are all really good reasons to go digital as soon as possible. Once you've learned the basics with manual controls on a digital P&S, you can move directly to a DSLR and enjoy all the wonderful (but more expensive) step-up features.

 

Film has its place, but these days, it's more an "experts" things (mainly due to the cost, and the reduced availability of good film processing).

Not that familiar with the Alpha lens line, but for a 1.5x crop body, the standard prime lenses for portraits are 50 and 85 mm.

 

For a given framing, the out of focus background is a function of the absolute focal length (not the crop factor equivalent). So you can really blow the background out of focus with long lenses like 100mm or more, but you have to be very far away from the subject...

 

f/2.0 or better is best for low light performance and out of focus backgrounds. f/2.8 is acceptable as well, especially on longer end. Anything with a smaller aperture (larger f/number) will not give you a good out of focus background.

 

if you can get a minolta 50mm f/1.7 in good condition, for $130 +- or so that would be the way to start IMHO.

 

One place to look for used stuff is http://www.keh.com

 

I have the tamron 17-50 mm f/2.8 and I think it's a good general purpose lens that includes portraiture at the long end. very sharp, not as good low light performance or extreme out of focus backgrounds as 50 f/1.8 on my canon but decent.

 

It looks like you might want to consider the Tamron 28-75 mm f/2.8 as well, relatively inexpensive, and covers the portrait and normal range. Bob Atkins has a favorable review on his site:

 

http://bobatkins.com/photography/reviews/tamron_28_75.html

I like the crop...

 

As for unsharp mask... this depends on what size you're unsharping... for this exact image (ie web sized) I tried:

 

radius=1.0

amount = 0.3

threshold = 10 (in GIMP, should be similar to PS I think)

 

I got the following, which I think is pretty good<div></div>

I like the stock up on flash cards method, but perhaps consider stopping into photo shops where they can D/L to DVD before you ship the flash card home... I'd hate to have only one copy of my images and have it lost or damaged in the mail...

 

depends on where you're going though.

seems like shallow depth of field, partial desaturation, and sharpening combined with some kind of scrim lighting placed relatively close to cause the light to fall off quickly... using a strobe so the exposure time is very fast.

 

that's my guess, but I'm not a lighting expert. I think this effect is mostly in the lighting. Doesn't hurt to have a highly textured model too :-)

<div></div>

Save to JPEGs or TIFFs and merge in photoshop. What you're doing is very valid, the RAW file has 12+ bits of data per channel, the final JPEG will have only 8 bits... so your RAW has valid additional data and you can pull it out... it's called dodging and burning in the darkroom, HDR in the computer...

Whitebalance setting is exactly what you need, yes. If you shot RAW you can go back and do it when you process the RAW file. If you shoot JPEG you have to get the WB correctly in-camera. Best way is to shoot a grey card custom white balance. Most people here will recommend RAW for this reason. I still recommend shooting a grey card when using RAW, because it gives you a known-good measurement of the whitebalance that doesn't rely on your eye/monitor/room light combo

A few were soft, and some had a distinct blue cast due to the shade, or were underexposed. I liked the B&W ones best. I think the most compelling shot is the closest one in B&W... The catchlights in the eyes from the tree are great in all of them.

 

Most important things to look out for: focus on the eyes, best ones are where the baby is engaged with the camera

Light falls off as the square of the distance between the source and the subject (for a point source), which means if you double the distance you lose 2 stops of light, if you multiply the distance by 1.4 you lose 1 stop of light. If you're over-exposing at f/11 on low power output, you need to back the lights up (or use softboxes/umbrellas etc to cut the light).

 

Shooting through (or off of) an umbrella also loses some light. Measure the distance from the flash to the subject, and multiply that distance by 1.4 (approx sqrt(2)) and move the light back that much, you should drop your exposure by 1 stop. If you don't like that, multiply the new distance by 1.4... lose another stop.

 

I would stick my camera on 1/60 s at f/8, crank down the power to minimum, use only one light at 45 degrees, and back up the light until I had a good exposure via the histogram, then fine tune using the aperture setting.

I agree with Mark and Chris. The upshot is that at 1/200 the frame is fully open for a small amount of time... Any faster, and the frame is not fully open for any amount of time. (it's a partial frame slit throughout the exposure).

 

Best bet is 1/125 or 1/90 or something like that. This prevents the possibility of the shutter not quite syncing and leaving a shadow on the top of the frame.

 

the flash only lasts 1/1000 second or so. So there is no motion blur even at 1/60 or longer... the only issue with the shutter is if you leave it open long enough the room light starts to contribute enough that your exposure is no longer primarily from the flash. This is called "dragging the shutter" but indoors in low lighting, you'd have to leave it open for 1/2 second or 1/8 second or something like that.

Rick. Even with your method, it pays to shoot the grey card to get a good consistent reference for the white balance. Even if you don't tell the camera to use it... but then telling the camera to use it is only two buttons away, and you can always ignore the camera balance in Lightroom...

 

Definitely doesn't pay to try to set the white balance for each shot though, if that's what you thought I meant. Just once per each lighting situation, same thing you do, only with the grey card for the reference.

Donald. I don't think so. Flash sync speed = maximum speed you can go where the full frame is exposed all at once. Faster shutter speeds happen by moving a slit across the image. So only a portion of the frame will be lit by the flash... 1/200 is fastest speed for flash photo on this camera.

whoops. I wrote $1.35 / TB but I meant $135 / TB... but it was just a typo, the rest of the math is still

correct... so it's still $0.34 for 1.5 GB of "decent" images.

 

I run all Linux, and I do use unison for synchronizing, but it's not possible to get several TB on my laptop,

that size drive doesn't exist yet. Still I'm not close to filling my laptop yet.

 

I don't feel like I was blazing away. This shoot was 2 sessions 2 hours each, with several changes of clothing.

so about 100 images an hour... of those about 5 were "perfect" for what the client wanted (several character

looks), and about 75 or 80 were usable for various additional purposes (modeling, advertising, etc).

 

I think the advise to not blaze away, and to edit ruthlessly is key. It was too easy to say "i'll do it later"

when JPEG file sizes were so manageable... larger RAW files will probably improve my photos in more ways than

just better dynamic range and high resolution... they'll force me to finally take the time to edit carefully :-)