Across the board distinctive elements of each filter have been compromised. Filters that were washed out are now more contrasty. Filters that were contrasty are now more washed out. They’ve all drifted towards the same look.
I did feel like Instagram 2.0’s filters had gotten weaker, but couldn’t be bothered to check, and didn’t really want to believe they had screwed it up so much. Or maybe I didn’t want to find a confirmation because I knew it would piss me off and make me want to stop using the app. Look at the comparisons: basically, the original image’s colors appear in the final image much more than they used to in the most extreme filters.
I’m pretty sure I understand why: when I programmed my own photography app for iPhone, I did have to make choices between speed and filter intensity and richness. Some things can be achieved with GPU-accelerated functions, others can not.
But the problem is, Instagram made the wrong choice. They compromised their best asset, distinctive filters, to solve two problems that nobody cared about: low picture resolution and having to wait for the effects to be applied. Guess what: Instagram wouldn’t have become the success it has if that really bothered the public so much. (Yes, people are impatient, but Instagram was already quite fast. As for the resolution, Joe User definitely doesn’t give a damn that his pictures are now 2000-pixel wide.) Instagram was far from broken, and they tried to fix it.
Oh, and they made the interface worse, too — but I’ve already talked about that, and there’s no need for side-by-side comparisons to realize it. By the way, the article includes a quote from Instagram confirming that, no, you can not change the gradient on the tilt-shift feature (and you probably still won’t in the future, but at least it looks like they’re going to make it softer).