Macworld Forums: Blow Up 3 proves reliable for photo enlargements - Macworld Forums

Post your comments for Blow Up 3 proves reliable for photo enlargements here

Hmmm, no Aperture plug-in? That's a shame, although using it as a standalone app is no great hardship.

How does it compare to PhotoShop's scaling? I would be disappointed to learn that the $600 PhotoShop uses inferior interpolation methods.

Don't expect an Aperture plug-in. Apple and Adobe have been at war for years!

If you're blowing up an image to poster size, the artefacts sometimes won't matter, because no one's going to be viewing it with their nose to the image -- just like looking at those massive billboards, where print resolution just doesn't matter.

I wonder how often you can say "Blow UP" be for echelon picks it up and sends men in black to your front door?

jrdvt, on 05 December 2011 - 08:00 AM, said:

Be disappointed. All the add-on plugins or standalone programs use use proprietary algorithmic methods whose patented approaches and math were developed in the last 20 years. It isn't just doubling the pixels like Photoshop does. You can do a 20% blowup in Photoshop that looks decent. These other programs can double the size or in the case of posters more than that.

This is a perennial problem for bitmap images and photos. Know what you are doing ahead of time. That tiny insect in a photo cannot be cropped and blown up to be an entire photo. That scan of a medal needs to be done at 2400dpi minimum to make it the size of a small plate. That printed logo on a business card cannot be salvaged and increased to dominate a sheet of paper. If taking a photo learn to frame and take multiple shots. If creating a painting, start with a large size. If creating a logo that must scale to several sizes, use a vector program.

jrdvt, on 05 December 2011 - 08:00 AM, said:

Incorrect. Photoshop’s Bicubic interpolation was updated (Smoother and Sharper) a few versions ago, the interpolation algorithms in Lightroom and ACR are even better yet and more recent (adaptive interpolation on raw, linear encoded data). Test either, with output to print, not screen, then evaluate that output. Then you know what’s ‘better’.