Chief Alchemist - Business Consulting For The 21st Century Via A Holistic & Intelligent Approach
Share RSS 2.0 feed for Sign up for the mailing list. Follow Chief Alchemist on Twitter. 'Like' the Chief Alchemist's page on Facebook. See what Chief Alchemist has been Q&A'ing on Quora. Chief Alchemist bookmarks & highlights on Follow the Chief Alchemist on Chief Alchemist channel on Chief Alchemist on Flickr. Mark 'Chief Alchemist' Simchock on LinkedIn. Free Initial Consultation. Email Chief Alchemist. Phone Chief Alchemist.
  • Mark ‘Chief Alchemist’ Simchock
  • 'Email me.Email => ca .at. ChiefAlchemist .dot. com
  • 'Phone me.Alchemy United => 732 997-0028
  • Free initial consultation.Free => Initial Consultation
  • Please be sure to subscribe to your communication channels of choice.
  • Click To Close => The small green (consultation), red (email) or blue (phone) icons in the top upper right.
Business Consulting For The 21st Century Via A Holistic & Intelligent Approach

How to best upsize an image in Photoshop

FYI => “Tip of the Month: Bigger: Not Always Better (But Sometimes Unavoidable)” via iStockphoto email newsletter.

Note: I couldn’t find the article on the iStockphoto site so I resorted to copy & pasting it in full below. Here’s the link to the iStockphoto website if you’re not familiar with it already:

During more than a few episodes of crime shows like CSI, you can observe “experts” take a grainy, low-res photo, press a few buttons and magically create a crisp blown-up version that clearly shows who the bad guy is.

This is, of course, patently ridiculous.

The magic “clean it up” button has no basis in reality. If you upscale an image you will always lose quality. This is because when you upscale an image, either you’re increasing the size of each pixel or you’re adding new pixels that your computer more or less has to guess at. If you need a larger image, your best bet is to buy an one that’s already big enough. Unfortunately, that’s not always an option, so let’s take a look at how to minimize the damage when you have to upscale.

In Photoshop, click on “Image > Image Size” in the menu. Under the “Document Size” section, you can modify the size of your image by inputting new values into the width and height boxes. For best results, increase by a percentage instead of using exact dimensions. Also, it’s generally believed you can optimize the results by increasing the image size through multiple smaller increments instead of a single large one. If you need to boost the image to 150% of the original size, try doing it 10% at a time (or even 1% — create an Action in Photoshop to speed the process along). This way you can monitor the quality loss as the image grows, and determine exactly when enough is enough.

Enlarging an image in PhotoshopIt’s important to make sure your image resampling is set to “Bicubic Smoother,” which is more clearly marked in the latest version of Photoshop as being “best for enlargement.”

If you’re planning on printing the image when you’re done, the professional standard resolution is 300 ppi. However, you can usually get away with 240 ppi, and if your project doesn’t need perfectly crisp images upon close inspection, you can get it down as far as 200 ppi. (If you’re just printing personal photos to stick to your fridge, 140 ppi might even be enough for you.) This cheating can score you precious size for your upscale.

You can try to hide problems that do arise in the boosted image through a combination or sharpening and blurring filters, though there’s no hard and fast rules as to what works, so it will require some trial and error on your part.

Finally, if you’re enlarging the image for large-format printing, keep in mind that if the poster or banner is being viewed from a distance, a lot of the defects (especially blurriness) won’t be readily visible. Basically, if there’s some physical distance between the image and the viewer, the upscaling is a lot more forgiving.

Hopefully this will help you get the most out of upscaling. Again, it’ll never be perfect. It’s just not possible. At least not until those CSI “clean it up” machines become a reality.

Actually, the first paragraph isn’t entirely accurate. There was an article on a couple months back that discussed an algorithm that was proving to be highly effective at interpreting “low-res” data.

“Fill in the Blanks: Using Math to Turn Lo-Res Datasets Into Hi-Res Samples” by Jordan Ellenberg

Interesting stuff, eh?

Will not be published. Required.
Please include http://