How do I downsize???

[KobraCarry]

I've been working on a photo, using some filters, etc and now when I try to upload it online my service says the file is too big. How can I shrink it? Thanks

[Some Idiot]

Are you using photoshop?

[KobraCarry]

Elements 8

[srwatters]

Two ways to shrink a file, reduce the number of pixels (re-size) or reduce the jpg quality. If you're saving at quality 9 or above, reducing to 8 will yield some reduction in size with only a small loss in quality. Remember the more complex the image, the larger the file size will be because the jpg compression algorithm looks for common values and reduces their quantity. Simple plain backgrounds will often compress very well.

[KobraCarry]

Thanks. Also, if I'm trying to convert to a certain size, say 8x10 etc, how do I know what is the recommended amount of pixles per inch, etc. Sometimes when I try to convert or change size and I put 8 inches by 10 inches, the photo just shrinks so small you can't even see it.

[srwatters]

PPI or more correctly in printing, DPI, will vary as the physical size of the image increases. Typical DPI for 4x6 is 300 or more. I like to hold that number for up to 8x10 or perhaps even 11x14. Larger size prints are typically hung on a wall and viewed from a further distance. I've made 15x20 and 18x24 prints at 240 or even 150 DPI in a pinch. I'm sure there's a reference out there somewhere that will be more exacting, but those numbers I've given should give you decent results.

[Flea77]

Instead of relying on rules of thumb what you should do is contact the person or lab that is making the print for you and ask them what DPI/PPI they recommend. I have had labs run between 240 and 360 so guessing just is not a very good idea. Once you know the correct DPI/PPI, put that in your resize first, then scale it down to 8x10. That way to get the best possible print.

Allan

[srwatters]

Rather than rely on what your lab says, I'd also suggest considering the content of the image. The printing process will interpolate your file based on the device that is rendering the image. If your image is a soft portrait, interpolating up to some arbitrary DPI might render odd artifacts in the edges. If the content is edge rich, then softening might be the result. Blindly following a lab will not always give you the best possible results.

[Flea77]

While in theory Scott you are correct, he specifically stated an 8x10 which even at 360 DPI requires only 10.4MP and according to his profile he is shooting a 7D which is an 18MP camera, no interpolating up would need to take place. He would in fact be reducing the number of pixels. Now if he were shooting an 8MP camera, or even a 10MP camera, that would be an issue. In this case it is not.

Allan

[KobraCarry]

Thanks for all the input. So, if I don't mess with pixle size, but under Image:Resize I change the settings to "inches" and put 8x10, or whatever size I need, why does the image turn to the size of a quarter on the screen?

[KdLaneJr]

Originally Posted by KobraCarry

Thanks for all the input. So, if I don't mess with pixle size, but under Image:Resize I change the settings to "inches" and put 8x10, or whatever size I need, why does the image turn to the size of a quarter on the screen?

Here is what the image size dialogue box looks like in CS5. If you change width, height and resolution to 8 inches, 10 inches and 72 respectively, you should get what you are looking. Double check the units after you enter the data for your new size.

[srwatters]

You need to understand that when resizing your image, if you change the setting to other than pixels, the resolution now comes into play. Measuring units x pixels per unit = total pixels in the image. When you look at the image on your screen, it's the pixels that count. I think you're going about this the wrong way. What I do is basically forget about DPI when I'm in the digital world and focus on pixel count and preserving as much as I can by proper framing when capturing the original subject. Always remember to capture with the final ratio in mind. That means if I'm shooting a 35mm based DSLR, the ratio is 2x3. If I want an 8x10 headshot as the final product, I will allow for cropping in the frame. When shooting my P45 digital back, the ratio is 3x4. That works perfect for the 9x12 images most models need for their books.

When you prepare to print the 8x10 image, the first step is to change the resolution without interpolating the data. That means unchecking the box that says "resample image" and setting the output resolution as needed (i.e. 300). Once you've done that, you will know the size of your image in inches at that resolution. Conversely you can set the inches and see what the resolution will be. From there you can now check the box for resampling and change one of the values (resolution or inches) and Photoshop will resize your image accordingly using the chosen algorithm.

Hope that helps...

[KobraCarry]

That's all awesome info. But I still don't understand why after doing all that, the image I see on the screen is now tiny and doesn't appear usable.

[srwatters]

I hate to be captain obvious, but you are looking at the image in "actual pixels" mode (command or control + 1) right? Please do a screen grab showing the image and contents of the image->image size dialog box and I'll try to help you. Something like this:

[KobraCarry]

Hmmm. I could be Sgt Stupidhead. :) Guess I'll have to check that when I get home! :)