It's no secret that we believe our image processing and delivery is one of the quickest tools for developers on the market. Quite often we get asked 'but how quick is Libpixel?' and 'how do you guys compare against your competitors?'. Well we decided it was time to run a test, first starting with Imgix. Firstly we’d just like to give a quick nod to the team at Imgix. They’ve built a great product to help developers with their image processing needs. Both Imgix and Libpixel provide a similar set of features. Libpixel has some differences from Imgix that might make it a better alternative for you or your development team.
The Test
To make this as impartial as possible we picked an image hosted and originating from a 3rd party stock platform. We took the image below (originally 11MB but compressed for this post) and converted it into a 320x320 thumbnail. By not applying any additional image manipulation effects (blur, saturation or brightness) we just wanted to concentrate on the resized image and it's speed of delivery. Both Libpixel & Imgix were using the US region for processing, with delivery being transferred back to Finland.
The results
Imgix returned the processed image in 0.95 seconds, with Libpixel processing at 0.52. That is a significant difference, Libpixel being nearly 2x faster. One of the main factors involved in how quick we can result the processed image lies with how we've positioned our processing locations.
Libpixel has coverage on all continents and this ensures that we can locally process and optimize delivery, for example if your users are visiting from Germany, they'll view images that are processed from one of three possible locations in Frankfurt. Libpixel has over 90 edge locations globally which ensures immediate delivery.
Whether you're running a global web service or just working on a hobby project, speed matters. Libpixel is here to support you with your image infrastructure needs.
