Resizing

I am trying to learn resizing and have the following issues. First, I unchecked the Resample box and entered various values but the picture remained the same size after I clicked OK.
It did change size when I rechecked the box although I understand the former is called a non-detrimental change.
When or why under what circumstances do you resize via pixels, resolution or actual size? Also, is blowing up or enlarging a photo always detrimental to its quality? Thank you.
Louis

louis, either I’m misunderstanding the question (always a likely possibility) or you’re getting what you’re seeing on the monitor confused with the actual size of the image as measured in pixels. Most often they have no relationship to one another. I got my "clue" from the sentence in which you said you enter various values under Image>Resize>Image Size, but that the picture looked the same after you clicked OK. Yes, it most probably would.

To change what you’re seeing on the screen, you have several options. You can use the Zoom tool (in the toolbar), you can use the Zoom Out/Zoom In feature in the menu (View>Zoom Out, etc.), you can use keyboard commands (Control/Command + or -), or you can use the one I just recently discovered, which is the Navigator – found under Window>Navigator. The last one can be kept in your pallet well or on the desktop.

To change the size of an image in a way that impacts the print size is a whole different ball game. I suggest you take a look at this link, where actual image resolution is described. After you’ve read it – assuming my first assumption about your question was correct – feel free to repost and we’ll give you more information about the parts you’re still finding confusing.

<http://www.scantips.com/basics2c.html>

Beth
I read that web site and didn’t fully understand it. I wish it had been more step by step, ABC. I was looking to see the photo change size on the screen and understand what you are saying about how it would print as opposed to how it appears on the monitor.
Why is it though that with one step, as I described, I saw no change on my monitor, where with the other there was a size difference.
Also, to the other part of the question:
When or why under what circumstances do you resize via pixels, resolution or actual size? Also, is blowing up or enlarging a photo always detrimental to its quality? Thanks.
Louis
BTW, do you work for Adobe?

I can’t give you a reason for why you saw a change on the monitor when you performed the one experiment. It could have been coincidence, but it’s kind of hard to say without "being there"!

The one thing you do need to remember is that you can never rely on a relationship in size between what you see on the monitor and what you would see when printed. So, I covered the ways to change your monitor view yesterday. Now we’ll move to pixels. And, in response to part of your question – pixels ARE the resolution of an image, so, for practical purposes of this discussion, whenever you resize an image you are also changing the physical size and the resolution.

Have you ever worked with graph paper? If so, you know that you can buy it in various sizes – some of it is scaled so that each set of 8 little squares across and down is equal to one square inch, or a total of 64 squares. Some of it is scaled so that each set of 4 little squares across and down is equal to one square inch, or a total of 16 squares. So, if you graph something out using the scale of one square = one foot on the paper that’s 8 squares per inch, your drawing will appear much smaller than if you graph the same image using the scale of one square = one foot on the paper that is 4 squares per inch. This same concept applies to pixels contained within an image, but your pixels will be much smaller than the squares on the paper.

I should also say that there are some different schools of thought on how critical it is to use a specific method to resize an image as well as different schools of thought on the relationship between the resolution (in pixels) of an image and the quality of the print. I’m going to designate myself as being from the "old school" on this for right now, because I think it’s important to understand the basic concept first. At some point, you might decide to use a different method for resizing or learn that you can get prints you’re happy with at a resolution above or below what I’m going to use as the target. That’ll be fine. Most of us do that, but we do that after we understand how all of this works so we know the limitations.

Let’s use my camera as an example. It’s a 3.2mp. When I take a picture with the camera set to the highest resolution and lowest compression, the image that downloads to my computer comes in at 180 pixels per "running" inch, and the physical size if I were to print it just like that would be roughly 11.3 inches wide by 8.5 inches high, or 2,048 pixels across and 1,536 pixels high. Physically, that picture is too big for me to print out on my printer. I usually like prints to be about 4 X 6 inches. If I were to just take 4 X 6 inches out of my original image, though, I’d lose a whole lot of it. So, I resize it by increasing the resolution without resampling. I’ve found if I go to Image>Resize>Image Size and change the resolution to 330 pixels per inch, I wind up with an image that’s just a tiny bit over 4 X 6, but it still has the same 2,048 pixels across and 1,536 pixels down that it had originally. I haven’t sacrificed one pixel; I’ve just compacted them, so that instead of having 32,400 pixels in each square inch of my picture, I have 108,900 pixels in each square inch. In theory, my image will be sharper and clearer at that higher resolution, because I’ve got more digital information crammed into each square inch.

The "old school" likes to maintain an image resolution somewhere in the range of 150ppi to 300ppi for printing. I often will print my 4 X 6 at that 330ppi, though, just because it’s not too far over the range and it saves me from having to resample. If I want a picture that’s 5 X 7, I resample to 290 pixels per inch. I can actually go as low as 240ppi and still get a nice print, and sometimes that’s what I wind up with if I have an image that has a lot of background in it that I want to crop out. Personally, I always try to do most of my resizing through adjusting the resolution so I make use of all of the original pixels I get from my camera. In some cases, I need to resample.

Resamping: You can resample upwards or you can resample downwards.

Resampling upwards can cause a loss of clarity, because when you tell Elements to resample to create a larger image, it "makes up" more pixels to fill in. Actually, you can get by with doing that quite nicely in a lot of cases, but the results will vary depending on how far you try to push, the kind of image you are working with, and the method you use for resampling (which I won’t go into right now.) If you have an image with a lot of fine detail, the results of "making up" those pixels might be much more noticeable than using the same technique on an image without as much detail.

Resampling downwards isn’t as risky. When Elements does that, it tosses out unneeded pixels. Using my camera image as an example again, if I were to take that 4 X 6, 330ppi image and ask Elements to resample down to 300, I probably wouldn’t notice any loss of clarity because it’s still going to have 90,000 little pixels in each square inch of space.

I’ve been trying to type this while keeping an eye on a couple of other things, so I’ll just hope it makes sense! Some people have more trouble grasping this concept than others, but once you "get it" it makes perfect sense and can help you get prints you’re very happy with.

Digest this and then repost. I’ve gotta trundle off and do something else right now, but I’ll check in again later.

And, no, I don’t work for Adobe. I’m very happily unemployed, thank you! 🙂

Louis,

Since Beth isn’t here this morning, I’ll take the liberty of answering this for you. She does not work for Adobe, she is just a pleasant and knowledgable person who fields questions here on the forum.

Your image did not change size on the monitor when Resample was Unchecked in the Resize box because this does not change the number of pixels in an image. Your monitor only understands pixels. Think of it as a grid of squares (pixels). For instance: for x number of pixels it displays as 3" on monitor when you have the zoom factor at 25% and the resolution at 200. If you resize to 300 res, (resample unchecked) there are still x number of pixels, so it will still display at 3" at 25% zoom factor. (the only way this image will appear differently on your monitor is if you change the zoom factor)

When Resample/Constrain is checked, you are CHANGING the number of pixels in an image, either upwards or downwards. More space is occupied on the monitor by your image if you have more pixels and vice versa.

Resizing by changing the number of pixels is generally a bad idea, though less destructive somewhat if downsampling. This is because you are either fabricating pixels not present in the original capture or just leaving some out. Resizing by resolution controls the print out size. Resolution can merely be assigned a number. Generally people believe that a res. between 200-300 is a good value to use to get a decent quality printout (printing at say 72 res. will give you a pixelated, blocky look). When you resize, you’ll want to keep the aspect ratio of the image intact. Changing one dimension will have the program automatically fill in the other dimension and the resolution. Physical printed size is inversely proportional to resolution. The larger the size you request, the lower the resolution. So type in the size of one side you would like to have it be when printed but look at the res., if it is below 200 your image just doesn’t have enough pixels to be printed with good quality at that size.

Beth,

Whoops, so sorry Beth, I didn’t mean to step on your toes. 🙂

Nancy

Not a problem, Nancy! We were consistent with the concept, so maybe having it presented in two slightly different ways will be helpful. 🙂

Beth,

Your graph paper analagy was a brainstorm!

Thanks, Nancy! I’ve been waiting a long time to get that one worked in. 🙂

By the way, louis, you said you wanted step by step instructions. In order to do that in a way that would be meaningful to you, post some information about the images that come from your camera. By that I mean, open a brand new (or never before edited image) in Elements, and then go to Image>Resize>Image Size and post the dimensions in terms of width and height in pixels, width and height in inches, and the resolution that’s shown. Once we have that, we can give you something you can do on your own computer that will show you the process from initial opening to getting a print.

You are both very helpful. I will try to work with what I have been given and see what other challenges that leads to.
The photos I have been working with have been downloaded from Phillip Andrews book, Adobe Photoshop Elements 2.0. It seems to be the most’basic’ book on the subject I could find.
Don’t know if the for Dummies series has one on this program, but even though when it comes to this I am one, I don’t like their books or series that I have experienced.
One of the photos I need to work with is at my Web site. Please do not consider this a blatant promotion. It is not intended as such.
You can, however, put a face to these posts. My Native American name is Eagle Warrior so you are not confused when you go there. The photo I need to downsize for another application (without distorting it as I have done) is at www.nativeamericansmusic.com. Then click Artists Bio.
Wanishi.*
Louis
* "Thank you" in the language of my Lenape Tribe.

I downloaded the image from the website. What do you need to do with it? If you want the whole picture, but just smaller, do this:

First, do a Save As and change it to either TIFF or PSD so you can work and resave without losing any quality. Then, assuming you want the entire image, but you just want it smaller, go to Image>Resize>Image Size and (with resampling NOT checked) gradually INCREASE the resolution until you see the document size of the photo – as shown on the resizing page not on your monitor – change to the size you want. For example, right now it’s roughly 5 X 7 (a hair short of that.) If you simply increase the resolution from 150ppi to 300ppi, you’ll change the size of the image itself to about 2.5 X 3.5. If that’s still too big, increase the resolution more. Each time you make the resolution number bigger, you will be compacting the pixels and causing the document size to become smaller.

If 2.5 X 3.5 is too small, REDUCE the resolution some and stretch the pixels back out. For example, at 250ppi, it would become about 3 X 4. Try to get in the habit of doing as much resizing as you can through manipulating the resolution of your image. Once you get it close, you could then crop out a bit of the length and width to create an even sized picture.

And remember not to expect to see a change on your monitor! If you do, that’s an accident and not something you can rely on having happen every time!

I wish I had more of an intuitive idea and feel for this area. I won’t even say it’s like Chinese to me because I speak that language somewhat and found it a lot easier than understanding graphics programs.
I am struggling through learning the program even from Andrews seemingly simplistic book and your helpful input. Your last post is one I will need to read and re-read several times before the light goes on.
I notice though that nowhere do you mention what to me would be simply resizing in inches. In other words, I want X photo to be
4 X 6 instead of 3 X 5 or 8 X 10?
Also, the photos I am working with (from Andrews book) are all JPEG. Should printed photos be TIFF or PSD (is the latter a photoshop extension)? And those used on a Web site, JPEG? Is that pretty much a criteria?
Louis

… In other words, I want X photo to be > 4 X 6 instead of 3 X 5 or 8 X 10?\

You can use resize, set to inches if you like. Uncheck "resample" and set the new dimension on one side, the other side will adjust automatically, in proportion ("constrain proportions" will be locked). The ppi will adjust proportionally, no pixels are lost. As print size goes up, ppi goes down. As print size goes down, ppi goes up. All using same number of orginal pixels.

Note however, in your example, if you had an 8×10, it cannot become a 4×6 OR a 3×5 without cropping or distorting, as these are three different aspect ratio sizes.

8×10 = 4.8 x 6 inches
8×10 = 4 x 5 inches

That’s simply math.

Also, the photos I am working with (from Andrews book) are all JPEG.

Stay away from JPEG whenever possible. You should never do multiple saves of JPEG. As you as you begin editing, save as TIFF or PSD. Publishers like TIFF fine. Except for the necessary evil of starting with a JPEG from some source (usually digicam) and perhaps outputting JPEG as a final product version for website use or screen viewing via email, there’s really no reason to ever use JPEG at all. It’s a lossy format, and the more often you open a JPEG, change it and resave, the lossier (crummier) it gets.

Mac

Louis,

Convert your images to .psd (the Photoshop format) before editing them. This format preserves the layers construction (which you will use when you get a foothold on the program) and is a lossless format. Jpg images will degrade with repeated opening and saving. So make your changes in .psd and store in psd. If you want to email or put one on a website, make a copy and use the Save For Web feature. This will put the image in the compressed jpg format and drastically reduce the file size. Store your originals in a separate folder. Always make a copy of one to edit it. In doing so, your original images are preserved untouched as though they were your negatives.