Fringing

Part of this course is also designed to make you think about the limitations of your equipment. An issue that cropped up with my first assignment was with fringing, and I know this can be an issue with my lens/camera. However I have never really tested to see at what point it becomes worse, at what point can I no longer 'rescue it' in post production, does the ISO have an effect on the extent of the fringing?

Bearing all of this in mind I set out to do some testing. The first lens I tested was my Canon 50mm 1.8 MkII prime lens. The shots were not of anything terribly interesting; my son had a rugby training session so I set myself up to take some shots across the field, the trees against the sky,cricket nets and sight screens to provide contrast. Parents were asked if they objected to the photographs being taken, although not intended to be directly of their children they were running across the view; none had any issues.



Test Scene





Using a tripod, so that the scene remained the same, many frames were taken starting at ISO 100 @ f22 - f2. It was a good exercise to undertake as I also discovered that the lens is not as sharp at f22 and that the fastest shutter speed on the camera is 1/4000. However the sharpness could be down to the slower shutter speed and despite being on a tripod it was a slightly breezy day.

When I hit the maximum shutter speed I did not take anymore shots as they were a little overexposed. I could have used this as a chance to see what was recoverable with exposure but felt that would just confuse my already fried brain trying to note apertures and ISO's!

With regards to fringing, the wider the aperture the more this occurs and the higher the ISO the less apparent it becomes. To examine the images I enlarged to 400% (max enlargement in ACR) and looked closely across the frame. When the fringing became first apparent it was only on one small area, an upright white post, which was easily recoverable using defringe in ACR. As the aperture became wider fringing became more noticeable in other high contrast areas, but even so was still recoverable in ACR. At f4 and below at lower ISO, it became more difficult and at f2.8 and below the results were not so good on any.

Below are the comparison shots I took. With some it would appear that I get a colour halo, rather that fringing, when there are contrasting bright colours. On the ISO 200 @ f16 a child has run across the posts which appears fine, but his yellow shirt clearly has an 'edge' to it. Same with the ISO 200 @ f5.6. The only area fringing appears in the contrast on the rugby shirt strips.

In conclusion the 50mm lens seems to have a slight issue with fringing. At a rough guide, with the lighting conditions on the day I was testing, this occurs mainly when the aperture is at its widest and really starts to appear at f8 at ISO 100/200 and f4 at ISO 400. ACR does a great job of recovering fringing with Defringe until the really wide apertures are used and the fringing/pixels are at a very soft focus. It could be that you could use other desaturation methods, reducing the magenta levels etc but I did not experiment with that. Bearing this in mind I'll now know that shooting at the widest end will run the risk of fringing when shooting areas of high contrast and avoid if possible.

The same set up was then used using my 18-55 kit lens.

I shot at focal length 18mm, 24mm, 35mm and 55mm throughout the ISO range. Lots of frames to inspect and far too many to create comparison images as above. However I did note my observations.It seems across the board this lens has fringing no matter what you do! Admittedly it is not so bad and then gets worse. Anyone feel rich enough to buy me a nice new lens? Or a camera? Or both? Results coming as soon as I've had a break from my PC.....