Want more people to comment on your photos? There's an easy way to increase online engagement: Simply apply a digital photo filter on your snapshot, according to a study published on Wednesday by Yahoo Labs.

Three Yahoo researchers teamed up with a Georgia Tech professor to look at a dataset of 7.6 million photos posted on Flickr. The team got access to all the photos' public metadata, which showed whether the photo had a filter applied to it or not, as well as how many people had viewed it.

Turns out, if you apply a filter, your photo on Flickr is 21 percent more likely to get views. Filtered photos also elicit more engagement and are 45 percent more likely to receive comments.

However, filtering your photos isn't the only factor affecting whether people look at your pics. In terms of views and comments, whether your snapshot has Kelvin on it matters less than how many followers you have on Flickr, naturally.

But what filter should you use? Not all filters are created equal. According to the study, some filters drove views and comments better than others. Filters that increase contrast or correct exposure usually increase the amount of engagement. You also want to opt for filters that make your photos warmer because cooler color effects are less engaging.

"Specifically, we find that filters that impose warm color temperature, boost contrast and increase exposure are more likely to be noticed," the study notes.

All of the photos studied were uploaded to Flickr by a mobile phone: 4.1 million of them from the Instagram app and the remainder from Flickr's own app. Although over half the photos studied were also posted on Instagram, Yahoo's study didn't have access to Instagram's page-view data -- so the research doesn't say you'll get more likes on Instagram if you use, say, X Pro II. There's also the chance that people who are more likely to use filters simply take better photos, which could also account for the increased engagement.

But if you want more eyeballs and discussion on your duckface selfies, you should probably try a filter. It worked for millions of photos uploaded to Flickr.