noshir_patel

Thanks for the info Dave... If it holds, it sounds like perfect timing for me.

Thanks for the responses so far.

 

Ok, a follow up for those who have been there in the fall...

 

Do I need reservations for the cheaper motels or will they not be full that time of year? I'm thinking of splitting my time between Jackson Hole and West Yellowstone. I'd like to stay flexible on when I'm where (depending on wheather, etc.), but I don't want to get stuck with high hotel rates. This really needs to be a budget trip...

I am planning on going to Grand Teton the last week of September

(starting the 25th). Any guesses as to what the fall color situation

will be like at that time this year (my hope is to catch the aspen

trees while they're yellow of course)? Anyone know a good source of

information for this sort of thing? Anything I can look for in the

weather reports?

 

Thanks...

Nosh

Canon's new camera is the 10D. Basically an improved D60 with the same resolution and lower price. Affordable and lighter full frame digital won't be here for awhile it seems.

Part of the problem is loose use of terminology. It makes sense to those of us who have done computer graphics work for years, but for others it is confusing to have two different concepts of "dpi".

 

Here's how I would suggest thinking of it. Image resolution is Pixels Per Inch (ppi) and printer resolution is Dots Per Inch (dpi). Now the bit somebody else wrote about many dots being needed to make one arbitrarily colored pixel is basically correct. After a certain point, higher dpi is not so much about making an image sharper or getting more detail as it is creating smoother tones with less of a noticable dither pattern.

 

I can't say what works best on an Epson 2200 (since I don't own one), but for the other Epson "photo" printers there's not much to be gained by sending it more than a 360 ppi image. I'm tempted to say nothing to be gained. Most people wouldn't notice the difference between 240 ppi and 360 ppi (although subtle pixelation can be detected up close).

 

Also, the bit about using even divisors of maximum printer resolution is tempting, and I have to admit that I do it, but I'm pretty sure it doesn't matter. There is really no reason for it to matter (unlike scanning where it matters a great deal due to sensor alignment).

Um, given that a test won't even cost you film...

 

Setup a bird substitue in a tree, take a picture, and muck with flash compensation until you like what you see. Then you'll know.

It really doesn't matter what dpi you save as for internet viewing. The browsers ignore it. All that matters is number of pixels wide by number of pixels tall. A common screen "resolution" is 1024x768. I usually size my pictures on the internet to 500x400. I do tag them as 72dpi, but occasionally I forget and leave them tagged as 4000dpi (which is what my scanner's native resolution is). Theoretically, this means that the image is 1/8 inch wide (pretty darn small), but guess what... Internet explorer doesn't display it any differently...

You can get one reasonably cheaply on bhphotovideo.com. I lost the one that came with my camera and bought another. Just be careful you get the one for the Rebel 2000 specifically. It's a slightly different size for different Canon cameras.

$5000? Microtek 1800f (or 2500f) and it won't cost you nearly that much. I have the very much cheaper Epson 2450 and it is ok.

I don't think the Epson even comes close to an effective dpi of 1800. I think it's more like 1200 (which is actually ok for what I use it for... 4x5). Dedicated film scanner should blow it away. I'd stick to a big name brand like Minolta, Canon, Nikon, or (maybe) HP.

A suggestion...

 

Get the new Kodak Readyload holder (the single sheet variety). It works well with both Kodak Readyloads and Fuji Quickloads (the Fuji holder only works with Fuji Quickloads, so I have heard). I own both holders and think the Kodak is easier to work with. I can't imagine bothering with traditional holders (loading hassles, dust problems, extra bulk, less film in the field). I usually carry the holder and up to two boxes of film (total of 40 sheets) on a photo dayhike.

 

I haven't had any sharpness problems. My scanner captures enough detail to make nice 20x24 prints and a look at the transparencies shows clearly that the scanner is not capturing all the detail that's actually there. You just have to set focus correctly and stop down the correct amount. (I use the method for optimizing sharpness that you can find on www.largeformat.info).

I'd expect the Nikon Coolscan 8000 (and similar dedicated film scanners) to blow the Artixscan 1800f away. (I own the Coolscan 4000 which I use for 35mm, by the way. Don't really see how it could be any better except I'd prefer if it was supplied with a better profile... Wouldn't be an issue if I got a custom profile, I suppose.) I have heard of some banding issues with the 8000, but I don't remember the details. Since I don't shoot MF, I didn't care to research it.

The Artixscan uses an actual 1800 dpi ccd (slight assumption, the 1100 certainly used a true 1000 dpi ccd), where Epson (in the case of the 2450) uses two 1200 dpi ccds and pretends that this is as good as one 2400 dpi ccd. It is not (because the samples overlap way too much). Also, the artixscans have significantly higher dmax than the Epsons (shadow noise is certainly significant on the Epson). We can reasonably expect the new Epson to use 1800 dpi ccds (two of them), but they will not have the dmax of the Artixscan.

 

The Artixscan 2500f does some optical tricks, I believe, to scan smaller than full bed originals at a higher resolution than the actual ccd natively would support. This may partially explain the degradation seen with that scanner. The 1800f doesn't function that way. I'm not saying it will do as good as a dedicated 1800dpi film scanner, but I would expect it to be close.

 

All that said, I only have first hand experience with the Epson 2450, which I think is great for the price. I do believe if I were willing to pay four times as much, I'd get a better scanner. (Maybe not four times better...)

 

If I could have any of the above regardless of price, I suspect the Artixscan 1800f would be my best choice.

I almost hope the rumor of a digital SLR based on the EOS 3 is false, because if it is true I might not be able to stop myself from buying it (and I certainly can't afford it).

Since the Epson uses a two CCD approach instead of one CCD (which I believe the Microtek uses) the Epson's performance is nowhere near true 2400 dpi. This is because the sampled dots overlap each other considerably. I own it and think it's good for the money, but I consider it a 1200 dpi scanner. Everything I've read about the Artixscan 1100 leads me to believe that the 1800f would be far superior to the Epson in both true resolution and dmax. (In fact, I believe even the 1100 would be better despite slightly lower resolution, though there's no reason to buy one now.)

 

In any case, I've used Epson scans of 4x5 transparencies to produce very nice 20x24 prints (preflighted lightjet service from WCI). I do scan at 2400 dpi, not for the resolution, but for the noise reduction you get by downsampling (which shouldn't be necessary with the 1800f). I think MF to 16x20 might be acceptable with the Epson, but it could be marginal.

I would guess there is a slight light leak at the film door. Might make a difference where you store the camera when not in use (back pointed toward a light source...).

 

You could try something like the following. Put in some cheap slide film, advance it several frames with the lens cap on. Leave it for a while in a very bright place, preferably with light hitting the camera from several different angles at the back and sides. Now take one picture (just to make it easy to see where your film was advanced to) and get the film processed (without being cut and mounted) and see where it is fogged. If there is a light leak, this should give you an idea of where it is. By the way, I just made all this up, so while I think it should work, I have no proof... :-)

One of the best ways to use a wide angle lens is to get VERY close to your foreground in order to emphasize it against a landscape (you see this over and over in landscape photography). Like a small group of flowers. Maybe it's not an arms length away, but a finger's length...

Unless you want your images to look pixelated (big rectangular blocks), interpolation is much better than what you are trying to do...

If portraits are your goal, I'd choose the 100 f2 over the 100 f2.8 macro (and I'd choose the 85 f1.8 over both... in fact, I did). f4 limits what you can do. If you can't afford the f2.8 zooms, I say buy primes.

The StaticMaster is very good for this purpose. I use it for every piece of film I scan. Some people like the dust blower air cans. You probably know about Digital ICE, but that's no substitute for getting the dust off in the first place (though it does deal nicely with the very very small particles that the StaticMaster can sometimes miss that show up on 4000 dpi scans).

 

I hope you've done the math for your scanner to see how long it's going to take to scan all that film. Don't believe manufacturers' scan times. Do it yourself, and include time to change film, preview, adjust settings, scan, save, adjust picture post scan, and save again. Everybody starts off "I want to scan it all" but I bet most everyone ends up only scanning the best...

In PS6, use the "save for web" command from the "file" menu. In the resulting dialog, adjust the quality until it tells you a file size close to what you want (I think it displays this under the image). It also provides a preview so you can see how much image degradation you are going to get.

Depends on how much you like the EOS 3. The D30 is more like the Elan II except that it's digital.

Bevel vs straight cut... I used a utility knife to cut my holder. I don't own a mat cutter. Don't worry about rough edges as they will be taped over. Anyway, you can crop them out in Photoshop. By the way, my holder has a slightly larger opening than the Epson one (makes it slightly trickier to attach film, but doesn't crop as much).

 

The art tape sounds like a good idea, but magic tape has left no residue that I can detect. Anyway, I try not to let the tape get in the image area.

I have the 2450 and have recently started using my own holder for film (4x5) instead of the supplied one. I cut a piece of black matboard in the shape of the holder, and lined the opening edges in tape. Now, to attach a piece of film, I tape it to the holder at eight points (corners and edges). Because of the tape lined edges, the tape used to secure the film is easy to remove. It's a hassle, but now I get no newton rings and my film is FLAT. So flat that I can easily combine bracketed exposures with perfect registration in Photoshop (impossible with the supplied holders). Make sure you don't forget to cut out the calibration strip at the top (without this it doesn't work at all).

Wooi, I would expect the 24-85 to be pretty good if stopped down to the sort of aperatures you would normally use for landscape photos. Are you sure it's not a technique thing? I use the 28-135 which I belive is of comparable in quality to your lens and I can get a reasonably good 16x20 by scanning and getting digital prints done.

 

Bear in mind that if you are using hyperfocal distance tables (or depth of field tables), most of those assume you are not going to enlarge past about 6x8. I have found that if I follow these tables without compensating for this that I cannot produce an acceptably sharp enlargement. You might want to either stop down more than the tables say or generate tables for a smaller circle of confusion (which basically means amount of acceptable blur). (I have a Palm program that does the computations for me.)

 

Back to your actual question...

 

I can tell you the 20 f/2.8 is very sharp. I own this lens, but not the 24-70 f/2.8.