Tip of The Year

[Lisa Nikodym]

I just went fully digital early this year, so I've been learning a *lot*, but the one technical trick I figured out on my own and haven't seen anyone else mention:
A quick but very effective way to convert color RAW images to B&W in ACR, much better than using PS's channel mixer IMO, is to set Saturation to zero and then use the two white balance sliders to effectively change the mix of color channels to get the best mix for contrast, detail, etc.  This has only two controls instead of the three in the channel mixer (the third in ACR is Brightness), and it keeps the brightness much more constant than tweaking the channel mixer, making it easier to try various combinations without the change in brightness throwing you off.

A more general tip...
My spouse has come up with a theory for why he thinks I get so many good photos when we travel, despite the fact that it contradicts the way many pro photographers work...
Instead of picking one or two places and devoting the entire day (or week) to waiting for the right light, we move at high speed when we travel, hiking and/or driving to see as much as possible in our limited time (a side effect of having a hyperactive spouse!).  In that one "perfect" spot (usually one already discovered by hordes of photographers) my image may only be 75% as good as the images of those who've waited around, but for each of those images I have about thirty images of places those other guys never got to, and of those thirty there's probably a few in unexpected places where I hit the timing right and that are *better* than the one I would have gotten by just waiting around in that one spot, with the added advantage that they are of more unusual places.  It's all statistics.
I'm not claiming this theory is necessarily True with a capital T (it's hard to prove one way or the other), but it's a potentially useful alternate approach...  Any comments?

Of course, now I've reread Didger's comment, and it seems to be more or less the opposite of mine.  As Didger has said, "To each his own."  I'd be interested in other people's opinions on the subject.

Lisa

[boku]

Tip #1: Keep photography a hobby - it is WAY more rewarding that way.

Tip #2: Embrace the fact that I can work better with images of the great outdoors than people and products. (Which of course, leads to tip #1).

Tip #3: Regardless of motive, L glass feels good to the touch (and looks good to the gathering crowd).

Tip #4: I have a long way to go to develop my skills, but never before has access to the resources been so plentiful.

Tip #5: Tripod, Mirror Lockup, DOF Preview, Histogram, Cable Release.

Tip #6: Crop to suit. Crop for drama.

Tip #7: Avoid irritating forum threads.

Tip #8: Less BS, more photography.

Tip #9: It's never too late to make a comeback. You can shoot until you die.

Tip #10: Get good gear even if it means getting less gear.

[RobertJ]

Mr. Jonathan Wienke taught me this:

"If you want a camera, just buy it."

Christmas is coming soon!

T-1000

[Ray]

My favorite tip:  Using the TSE lenses shifted one direction in combination with the camera shifted the opposite direction to generate three images that can be then be stitched into a PERFECTLY SEAMLESS image.  

Jack,
You've got my attention, but can you elaborate on that technique please. Whenever I've tried to combine camera shift with lens shift it seemed to bugger up the stitching process (for nearby stuff in the foreground, that is).

With the 20D, from one extreme lens shift to the other, I get very close to two horizontal frames with a narrow overlap that's usually sufficient for perfect stitching.

With camera vertical it's necessary to take 3 frames, each with a substantial overlap, resulting in a stitched image of about 2 1/2 frames.

If I were to shift the camera in the opposite direction to the lens shift, as you suggest, it would merely increase the overlap in horizontal mode and reduce the over all width of the panorama. In vertical mode, it would increase an already over generous overlap as well as reducing the width of the panorama.

Do you mean, shift the camera in the same direction as the lens shift?

[BryanHansel]

Jack - Great tip.  Now I just need a lens that will do that.  Almost motivation to buy one right now, but I'm waiting to see how much I have to budget for a D2x.

didger - So true on the backpacking tip.  Is it a backpacking trip or is it a photo trip?  I've learned my lesson carrying a 8 or so pound tripod around the Smokies for about a 100 miles.  Not to mention the 70-200 2.8 heavy heavy lens, that I used for maybe 30 shots.    

These are the things that I learned this year:

Tip One: Leave a high paying stressful job to move to the middle of nowhere with no job and not knowing what is just around the corner.

Tip Two: On long distance expeditions make sure you know your partner well enough before you head off into the woods.  Make sure to do many many shake down trips before heading out with them.

Tip Three: Don't paddle 52 miles in a day ever again.

Tip Four: Using ND Grads with digital cameras is fun and it provides a slightly different look than digitally producing the effect. Plus, it takes less time.

Tip Five: Digital is a heck of a lot more fun than film.  Buy a digital camera.

Tip Six: Shameless plug.  I wrote a personal essay about three things that I've learned in my life that have helped me manage other people while helping them to grow.  If you want to read it here's the link: http://www.nessmuking.com/gettingby.htm

[Ray]

The advantage of three images, is the center image involves the center sweet-spot of the lens and IMO make the overall image better.

.....these resulting three images will line up PERFECTLY in PS on layers, and you can easily mask out the joints.  Moreover, any close or far objects also line up perfectly.  I practiced this procedure on a group of telephone poles with lots of wires.  Works like a champ  

Jack

Okay! Got it. Thanks. I guess I didn't follow this because I rarely have trouble stitching TS-E images. When I do, because of grass and the like in the very near foreground which can get slightly out of register in the overlap area, producing a noticeable double image effect, I've always been able to fix it by cropping off unnecessary overlap.

A great feature of Panavue's Image Assembler is its capacity to stitch different sized images as wll as images with a very narrow overlap. The aprox. 7% overlap of 2 horizontal images that Andy mentions is never a problem as long as there's detail there for the flags. The 3 vertical images (or horizontal images stitched vertically) with a substantial overlap are the ones that occasionally give trouble, but only before cropping. That's my tip   .

However, I'll check out the idea of making the central horizontal image the main one, which with Image Assembler would involve tacking on a half image on each side of the central image, with minimum overlap.

[gryffyn]

Shoot more and don't forget the sunscreen.

[jwarthman]

YOU can do a perfect job of assembling the images in PS -- better than any phot-assembler program, and once you get the hang of it it takes only a few minutes.

Nevertheless, I do need to practise assembling images in PS because IA has a resulting file size limitations of around 400MB. It would be a nightmare stitching 60x20D images in PS individually. But merging 3 or 4 groups, prestitched in IA, should be easily manageable.

Actually, the good stitching programs do *lots more* than just line up the pixels, such as you might do with Photoshop. I'm not familiar with Image Assembler, but I suspect it's similar to PTMac. (Although PTMac doesn't have that file size limitation.)

And, to keep on-topic, I would say that trying out Panorama Tools (and its Mac front-end, PTMac) is my best tip of the year! Windows users can also use Panorama Tools with a variety of front-end software such as PTGui.

PTMac lets you mark "control points" in your images for stitching. But you also tell PTMac about the lens you were using, so it knows about the field of view. It uses this information to compute how to "project" your images onto the surface of a sphere, that represents how the actual scene should have appeared (imagine yourself at the center of a sphere, looking out at the world).

You then tell PTMac what "projection" to use, e.g. rectilinear if you're making a flat print, or cylindrical if you're making a Quicktime VR panorama. The software projects the image from thesphere outward, onto the flat "paper" beyond. This is exactly the same as a map maker projecting a globe onto a flat surface, e.g. Mercator projections.

With all this power, PTMac can also take out lens distortion (e.g. barrel distortion) and chromatic aberrations.

Once you're satisfied with your settings, you tell PTMac to create the panorama, and it goes to work - sometimes for quite a long time. The good thing is, it determines the correct mathematical transformation, and applies it only once to each pixel, thus minimizing any possible image degradation that might come from repeatedly changing your pixels.

In any case, the results are quite remarkable - and much more realistic than simply tacking images together in Photoshop. You might like to take a look at some links.

A Luminous Landscape Discussion:

     http://www.luminous-landscape.com/tutorial...panoramas.shtml

Here’s a great tutorial on the use of Panorama Tools and related software:

    http://epaperpress.com/pano/

And here is a good explanation of “projections”.

    http://epaperpress.com/pano/proIntro.html

I hope you take the time to more fully explore panorama-creating with one of the best tools out there.

Enjoy!

-- Jim

[DiaAzul]

We are rapidly reaching the end of the year and I thought, just for fun, that it would be nice for everyone to share the one top tip or technique that they learned this year. Target is to get at least 20 tips from all the posters on the forum and perhaps Michael can compile them all up, with his own collected thoughts and wisdom, to give us a top 10 tips & techniques to improve our photography in 2005 article. All tips and techniques welcome - even if your key to succesful photography is to that you wear your undergarments inside out.

So, to kick off...

During a photojournalism course this year, it was impressed upon me the need in all situations to take at least one landscape and one portrait picture of each situation - primarily because some newspapers need portrait, some landscape to fill the hole on the page.

I find this quite relevant to landscape and nature photography as it forces a rethink each time I am composing a picture in a given situation. All to often in the past the easy option has been to select landscape and stick with a panaromic view and everything on the horizon. However, turning to portait forces me to consider how to use foreground elements more effectively and get a balance between the ratios of foreground, background and sky elements. This is now a key technique in forcing me to 'work the scene' in order to get better pictures.

Merry Christmas to everyone and happy picture taking in the new year.

[Jack Flesher]

Well, here's a combination forum effectiveness tip and slightly early New Years resolution:
Keep messages short and simple and not very subtle in the hope that folks will actually read and understand before replying.

Great lesson Didger!  

My favorite tip:  Using the TSE lenses shifted one direction in combination with the camera shifted the opposite direction to generate three images that can be then be stitched into a PERFECTLY SEAMLESS image.  The stitch is done manually in PS -- a background layer plus three image layers and one mask -- and takes less than 3 minutes once you get the hang of it.

[didger]

but it's a potentially useful alternate approach...  Any comments?

Well, it's obviously just the ticket for your situation, since you previously mentioned a husband that's hard to keep tied to one place very long anyway.

I try to have my cake and eat it too by getting out so much that I can see a huge variety of places that hardly anyone ever gets to and cover each place in minute detail and various times of the day and year.  All ga gotta do is just go out practically every day and work all the light hours of the day for months on end, years on end.  No problem, uh except that I've already got thousands of Sierra images and little Photoshopping on them so far.


The B&W conversion trick sounds really cool and if I hadn't already bought and fallen in love with the ImageFactory converter (PS filter) I'd surely try your trick.  I don't care much for the PS channel mixer conversion either, but the ImageFactory filter is something else; a really quick intuitive powerhouse; that's another tip, I suppose, but Tweedledummer gets the credit.

Of course, now I've reread Didger's comment, and it seems to be more or less the opposite of mine.

No, no.  My tip was for backpackers, NOT for ladies traveling with their impatient husbands and in any case having previously declared not even be into car camping, what to say of backpacking.  I go for a combination of the "shoot a lot of frames" and "minute detail, waiting for the light etc" Ansel Adams approach, but few people can manage that much time shooting, so, to each his own.  It's all good if we're having fun and we don't push whatever works for us on others.

[didger]

Avoid irritating forum threads.

Production or consumption.

[kaelaria]

Practice makes more perfect.

Try different or new techniques before getting to your destination - you'll have more time for composition rather than trial and error to get a 1/2 decent shot.

[Jonathan Wienke]

Don't eat yellow snow.

Never store pizza and photographs in the same envelope.

Milk is not archival, especially if exposed to the sun.

Get the best quality gear you can afford, and learn it inside and out.

[GordonMcGregor]

A photoshop one:

Channels = selections = masks = alpha channels = 8 bit images

You can switch back and forth between any of these and treat each just like the others.  Selections can be made automatically from channels.  Masks can be made straight from the original image without any wanding or selection tools. Complex horizon selections take a single click.  Selections from channels can be painted to suit.   You can sharpen masks, grow them, shrink them, use filters on selections, uses curves, levels on masks and so on.

Probably the simplest, most significant, useful, powerful thing I've managed to avoid understanding about photoshop for many years until about a year ago.

[Jonathan Wienke]

As another aside, I find the shadow/highlight tool almost unusable for any decent starting exposure.  You can pull a whole lot of image data around using it, but the results seldom seem worth looking at afterwards.

Would that include this image?



Or this one?

[GordonMcGregor]

A photoshop one:

Channels = selections = masks = alpha channels = 8 bit images

With Photoshop CS, none of the things you mention require 8-bit mode. It's one of the best reasons to upgrade if you haven't. That and the shadow/highlight tool and the updated Camera RAW.

What you say about CS being good is true.  The value and point I was making is almost completely unrelated to bit depth though.

Perhaps more usefully and to avoid meandering away from the value of the point, it would be better stated as

channels = selections = masks = greyscale images

and leave it at that.

As another aside, I find the shadow/highlight tool almost unusable for any decent starting exposure.  You can pull a whole lot of image data around using it, but the results seldom seem worth looking at afterwards.

[Jack Flesher]

My favorite tip:  Using the TSE lenses shifted one direction in combination with the camera shifted the opposite direction to generate three images that can be then be stitched into a PERFECTLY SEAMLESS image.  

Ray,
What Jack is describing is intended to keep the lens position stationary and effectively "shift" the camera and sensor across the fixed image it projects in three steps.  This eliminates the possibility of parallax making the stitching of the foreground a problem.  (Some LF cameras allow you to move the film carrier like that.)
  On the 20D you do only need two exposures with the ~ 7% overlap.  You can also move the camera as Jack suggests but for the horizon (for example) you still only need the two shots, but the image of the blade of grass an inch from the lens doesn't move (the way it would if you moved the lens position).

Andy

Andy's got it.

The advantage of three images, is the center image involves the center sweet-spot of the lens and IMO make the overall image better.

the techniques is simple.  Since all three TSE lenses shift a maximum of 11mm in each direction whether you are oriented horizontally or vertically, you simply shift the camera in the opposite direction by 11mm for each frame.  My camera has an L bracket on it and I have made to marks, 11mm and 22mm in from the edges of the baracket.  So I start with lens full left, camera full right and take frame 1.  Shift the lens right to the center position, and shift my camera left 11mm and take frame 2 (the middle frame).  Then I shift the lens full right and the camera another 11mm left and take frame 3.  

these resulting three images will line up PERFECTLY in PS on layers, and you can easily mask out the joints.  Moreover, any close or far objects also line up perfectly.  I practiced this procedure on a group of telephone poles with lots of wires.  Works like a champ  

Jack

[Pete]

One of my favorite and most fun techniques learned this year?  Gotta be playing around with rear-curtain synch, dragging the shutter, and throwing in a bit of twist on the zoom.  Great for party pics and weddings.  Happy Holidays and Happy New Year to All!

[Graham Welland]

Get yourself a small portable digicam that you can easily carry about your person anywhere and ALWAYS take it with you.

If I had a $ for every wonderous sights I've seen through the years but didn't capture for the sake of carrying a camera ....