Photo Size/Format ? callin' all Gurus

[Skunkboy]

This is for someone who is used to a 640x480 cheap digicam so either one will suffice; but, in your opinions, which should be used??

Objective: A photo that's able to be printed on a 2x2 sheet spread (4 pages printed). I realize it will obviously have some quality issues since this isn't a 10mp camera or anything; but, which would be better?

I can take as TIFF and the sample I took turns out a 6.5mb file but the size is only 1792x1200.

At the same time, I could take the highest resolution JPG and get a 2240x1500 image but the filesize was only 93k. The compression quality on it is 94% too.

The sample I took was pretty much in the dark just looking at the floor so it wasn't a "busy" picture but at least something to go by.

Now... TIFF is uncompressed so no loss; but, pixel size is often the better bargain. I realize this is then only 3.3mp and going to a 2x2 spread or maybe even bigger is a stretch but from even the 2.1mp setting of the TIFF is better than the 640x480. Then it's also the difference between Vivitar and Kodak.

I can't really screw this one up; so, I figured I would pull from the friends I've made here.

[Skunkboy]

I'm surprised this appears to have been lost in the suffle... there's a few in particular that I figured would comment - hmm - must be vacation time for everybody (I sure wouldn't want to be on vacation right now... too warm to ski but too cold to enjoy the outdoors)

[Mighty]

I'm not sure what the question is. Are you trying to decide between two different cameras? If so, a comparison of pixel resolution should probably not be your primary consideration. There are many other factors that determine what camera would be best for your situation. If pixel count were the only factor then the choice would be easy.

What are you going to use the camera for? You mention printing 2x2 images per page, but you don't describe what they're pictures of, who they're for, what kind of printer it will be printed on, and whether resolution really is the most important thing.

If it's just amateur pics of people, and the audience of the printed pictures is the subject of the pictures, then 2.1 megapixels is probably Good Enough for all but the most discerning. And if someone is unhappy with 2.1 mp, then they'll probably not be very happy with 3.3 mp, either. That's a difference that's noticable when the pics are side-by-side, but it's not a huge difference. Especially if it's going to be printed on a consumer inkjet printer. In which case it would probably be difficult to tell which is which.

Therefore, in this range the resolution is one of the least important considerations. Instead, you should be looking at what other features the cameras have that you'll actually use. What's the lighting going to be like for most of your pictures? If it might be in less-than-optimal lighting situations then you might want the camera that has the largest apeture and thus grabs the most light. Are these going to be action shots or posed shots? If action shots, then you might want the one that has the least shutter lag (ie, quickest response time from when you push the button.)

If you want advice on a camera choice you need to give a lot more info on what you want to use it for and the price range you're looking at. And what specific models you're currently considering.

[Skunkboy]

not two different cameras - same camera - two different options. There's actually far more settings than just what I've listed but those are the two best. Do I dare go JPG and risk the compression loss? If I go TIFF then the dimensions (pixel wise of original) won't be as large. There's only one chance for the photos and I was hoping to get it set for whichever would be best for blowing up to a 2x2 page (4 page total) spread.

[Mighty]

So it's not four pics per page, but rather you're blowing it up to 22x17 inches? Ah, that's a bit different. Hmm.

Chances are, the TIFF image is closer to the actual dimensions of the CCD detector. I'm guessing that it scales that up a little for the JPG, since it's a lossy, anyway. The smoothing inherent in the JPEG algorithm could give the impression of there being more info than was ever really there.

If you google the model of the camera you might be able to find the actual usable dimensions of the CCD, and learn how it's filtered and interpolated (http://www.dansdata.com/digicam.htm). Knowing that info might help you decide which is more realistic.

Why do you only have one shot at this? Are you printing this on a printer that you don't have regular access to? Because my advice is to print a few test shots. They don't have to be the entire 2x2 page output. Just pick a representative section, like something with lots of detail that you're interested in, and print just that section at the appropriate size. So, if you're blowing 1792 pixels up to 20 inches, take 270 pixel section out of the middle and blow that up to 3 inches and see how that looks. You could get four such tests onto one page, allowing you to test different images or different pieces of the same image.

There are too many variables to make an informed guess at what will look best. It might be that the JPEG smoothing will help when it's blown up that much. Alternatively, depending on how aggressively it's compressed that much magnification may exacerbate the artifacts. While I'd normally say the example you gave is a lot of compression, as you said, your test image isn't representative, so it's impossible to say at this point. Also, the printing technology could make a difference. A really good quality high res printer might make the pixel boundaries stand out really strongly from the TIFF image.

If it were me, and I didn't have an opportunity to test, then I'd choose TIFF, then scale it up in Irfanview or a paint program to guarantee a high quality rescaling algorithm. But my guess is that neither is going to look very good when blown up that much, unless viewed from a pretty good distance, like at least five meters.

[Skunkboy]

The reason it's one chance is because it's a chance for decent pictures of my brother who I barely ever see.

I've only got 2 32mb cards, a 20, and a 16 but am going out to buy a 128 (maybe 256 depending on cost) just for this chance.

Camera at hand - Kodak DC290 I'll probably have the pictures taken both ways - all depends on what size card I get my hands on. The 290 is 2.1mp / 3.3 interpolated. Interpolated I can get the 2200 out of it but I was always under the understanding that TIFF would be a better image. So... bigger but lossy or smaller but uncompressed.

[Mighty]

It depends on how aggressive the compression is. JPEG isn't a fixed compression ratio. At minimum compression it uses an LZW-type compression that gives about a 2:1 ratio with no loss. It can be dialed up to outrageous levels. I haven't gone searching yet to see how aggressive the DC290 is set for. My guess is it's cranked up medium-high.

Again, I recommend taking test pics of other people before your brother shows up and do some test prints. That'll give you a better idea of what will work better with your particular equipment.

But don't expect miracles.

[Mighty]

I found a review of the DC290 (http://www.dpreview.com/reviews/kodakdc290/page8.asp). There, he has pictures of the same subject in different modes. He shows an example of a High res TIFF vs an Ultra res JPG. Because the camera is interpolating based on the raw data it can produce a better Ultra res image than taking the final TIFF and blowing it up in a paint program. (If you read that earlier link (http://www.dansdata.com/digicam.htm) you'll see that all cameras are interpolating even at their "natural" resolution.)

So, it looks like you probably want to go ahead and use Ultra mode. That will give you the best image to zoom up for high res printing.

I'd still suggest blowing it up at least to 200% in a paint program with bicubic interpolation. Who knows what scaling algorithm your printer driver uses.

Drake

[Skunkboy]

I'm not expecting the best... I realize this isn't any 10mp camera. I did some searching and now need to do some experimenting; but, the nice thing about the DC220,260,290, and it seems like there's another one in there, is that Kodak designed them to where you can install scripts. I found one script that sounded like (not sure yet as I've not tried it) it can adjust the compression the camera uses. I've had a few scripts on both 290's I used in the past to do things like long-term exposure and such; so, I'm hoping to find something that will let me get rid of as much compression as I can. **well, that was replied to based on the bit that was in the e-mail notification... now I noticed there was the second reply**

Yeah, if I can't find something to nearly eliminated the lossy on the 2240x1500 then I was thinking tiff may be the better way to go. hmmm.... looks like it's time to install photoshop on a computer again. I usually get frustrated with it and kill it before it sits on my computer too long. Never really had time with anybody to run me through some of the things I want to do... though I've figured out quite a bit since 1998 or so when I first installed it (I've got a copy of 7 now).

Thanks for the assistance. Perhaps I'll put a pic or two in a link on my site after they're taken.

[Mighty]

For just scaling, cropping and gamma correction, I use www.irfanview.com. It's simple, and it comes up much faster than a paint program, even on fast machines.

[Skunkboy]

http://dc290.skunkboy.net there's a script that will go through and shoot with every possible setting. I did so - aiming at a few things with detail plus my pcmcia CF adapter which had some color to it. I'll probably go with the biggest tiff. I'll look into irfanview too :)