A Simple Trick to Kill Image Noise in Your Photos

Chris Gampat The Phoblographer Sigma dp and 50mm f1.4 product images first impressions (6 of 12)ISO 64001-40 sec at f - 4.0

We’re going to start this post off by saying that at this point in the year 2015 and this stage in digital photography, image noise isn’t a major problem unless you’re printing. But for most folks who shoot and upload to the web, image noise isn’t a major issue. It can easily be fixed in post-production with the push or a slider, you can get back incredible amounts of image details, preserve your colors, and you can also choose to shoot at a lower ISO and push the files because shadow detail recovery is that damned good.

If you’re a pixel peeper that lives and dies by looking at your images at 100%, you’re living in an archaic age that doesn’t really exist anymore. For Pete’s sake, folks are shooting ad campaigns with an iPhone.

Chris Gampat The Phoblographer Fujifilm 27mm f2.8 first impressions (2 of 18)ISO 64001-210 sec at f - 2.8

And now to our simple trick: there is a simple way to lessen the look of image noise in your photo. We’re not talking about color noise–instead this has to do with luminance. The easy way to get rid of it or at least lessen its effects is to overexpose the image. The more light hitting the sensor tends to overexpose areas that then make the image look like it’s not as noisy as it really is. We recommend overexposing by a full stop or so if possible, and then you have the option of pulling the images in post-production

Try doing this through use of the aperture or shutter speed. Of course, we completely understand that this won’t work in every situation but it can surely work in lots of them. In fact, as an example note the image in this post: it was shot at ISO 6400 but it doesn’t show a lot of noise. Even when you look at it at 100%, you’ll notice that the image noise isn’t that terrible.

At the time of publishing this post, this will work with most cameras released within the past four years up to ISO 6400. But if you go beyond this, then you run the chance of introducing color noise and that becomes much more complicated to deal with.