Skip to main content

Options for implementing image and video transcoding

When it comes to transcoding media, you have two main options you can choose from in terms of how to implement it: using a premade solution, or doing it yourself. There are merits to either approach, but don't try the DIY approach unless you know what you're doing.

Option 1: Using a premade solution

Some cloud providers offer a built-in image transcoding solution (for example, as part of their CDN offering). While this is convenient, it also has a lot of pitfalls and downsides that need to be considered.

Pros
  • No coding effort required to implement
  • Won't require any active maintenance from your developers or yourself
  • Doesn't present any issues if you decide to move to another provider later on
Cons
  • Likely to cost more and be less efficient at compression
  • Some resizing tools don't allow you to specify important parameters, like chroma subsampling, and do not let you choose an encoder to use (which can matter a lot for e.g. JPGs)
  • Usually involves lazy 'on-demand' resizing, which isn't always desirable

Option 2: Doing it yourself

Implementing image and video transcoding yourself isn't easy, but can be well worth the pain provided it's done correctly.

Pros
  • Allows much better control and customisation of how the media is transcoded
  • Lets you resize proactively
  • Will generally result in better filesizes and quality, reducing your CDN costs
  • Will cost less to resize a given number of images compared to a premade solution
Cons
  • A lot more effort than using a pre-built solution, both implementing it and maintaining it
  • More effort involved if you migrate to another provider, since you'll have to move your code over as well, which may need some tailoring for the new provider's infrastructure/APIs
  • Isn't going to perform well unless you use the correct libraries
  • Requires knowledge of media formats and encoders for good results
Back to top