You NEED to tune your DSLR autofocus!

Evidence

Reikan FoCal has been available for over 11 years now, and in that time we’ve had millions of results uploaded. We process these results and build information about the typical performance of cameras and lenses which we give back to users and the community, both for kit comparison in FoCal Pro, but also in our completely free Lens and Camera Performance site, FoCal IQ.

It’s from FoCal IQ that we’ll take a look at the evidence to seal the argument for needing to calibrate your DSLR autofocus.

Camera View

Let’s jump straight in with a scary chart – here are the adjustment values required by the percentage of our D850 users for all lenses from a total set of a little under 100,000 results:

A chart showing the focus adjustment required for a Nikon D850
Focus adjustment required for the Nikon D850 across all lenses

An AF Fine-tune of 0 (in the middle of the chart) would mean that no adjustment was required, and this accounts for just under 5% of users. To put it another way, 19 out of 20 users would get some benefit from calibration!

Now, we tend to say that an adjustment of -1 or +1 won’t make a noticeable difference to your image sharpness (although it will tip you slightly in favour of sharper shots), and a difference of above 4 units results in clearly sharper images, so I’ve overlaid some coloured bars to the chart:

D850 data highlighted with Minimal (green), Noticeable (Yellow) and Significant (Red) adjustments highlighted
D850 data highlighted with Minimal (green), Noticeable (Yellow) and Significant (Red) adjustments highlighted

All cameras in the RED area – that’s about 60% of all D850s – would produce significantly sharper images after calibration.

And here’s the same image for the Canon EOS 5D Mark IV, again from around 100,000 test results:

Focus adjustment required for the Canon EOS 5D Mark IV across all lenses
Focus adjustment required for the Canon EOS 5D Mark IV across all lenses

Things seem slightly better in Canon-land – probably as unlike the Nikon F-mount, the Canon EF mount has never supported screwdriver lenses, which are particularly poor at focusing. But we’re still looking at around 40% of all 5D Mark IV producing significantly sharper images after calibration.

The raw information for these charts is available directly from FoCal IQ here (choose the calibration link under each camera), which also includes the comparison with the Nikon Z9 (F-mount lenses) to show how Mirrorless cameras are much improved… but not perfect!

Lens View

Instead of looking at the results for a specific camera with all lenses, let’s look at a specific lens with all compatible cameras.

24-70mm f/2.8 Lenses

Here are the results for the Nikon and Canon 24-70mm f/2.8 lenses:

Nikon 24-70mm f/2.8: approx 73% needs an adjustment of 5 or more units
Nikon 24-70mm f/2.8: approx 73% needs an adjustment of 5 or more units
Canon 24-70mm f/2.8: approx 30% needs an adjustment of 5 or more units
Canon 24-70mm f/2.8: approx 30% needs an adjustment of 5 or more units

There’s quite a noticeable difference between the two, with over 70% of the Nikon lenses getting a good boost in image quality after focus adjustment, compared to around 30% of Canon users. Before you think you don’t need to calibrate the Canon lens, remember that:

  • 3 in 10 of these lenses will show a significant improvement
  • Approximately 7 in 10 of these lenses will show a noticeable improvement
  • This is the newer “Mark II” version of the lens – the original Canon EF 24-70 is much more likely to require calibration

Source data at FoCal IQ – Canon and Nikon 24-70mm f/2.8 Comparison.

70-200mm f/2.8 Lenses

Here are the results for a couple of comparable first-party 70-200mm f/2.8 lenses:

Nikon 70-200/2.8 chart of focus adjustment required
Nikon 70-200mm f/2.8: approx 80% needs an adjustment of 5 or more units
Canon 70-200/2.8 chart of focus adjustment required showing significant front and back focus at focal lengths
Canon 70-200mm f/2.8: approx 70% needs an adjustment of 5 or more units

The 70-200mm f/2.8 variants show even more need to calibrate – the performance of 8 out of 10 Nikon and 7 out of 10 Canon lenses would improve significantly with calibration.

There are also some interesting things shown in these charts.

  • The Nikon 70-200mm f/2.8 has a tendency to back-focus at all focal lengths (the results are pushed towards negative numbers, meaning the adjustment brings the focus point nearer to the camera)
  • The Canon 70-200mm f/2.8 seems to back-focus at 70mm, but front-focus at 200mm. Luckily, most Canon cameras allow tuning of both Wide and Telephoto ends of zoom lenses, so you can fix this issue cleanly with FoCal.

Source data at FoCal IQ – Canon and Nikon 70-200mm f/2.8 Comparison.

35mm f/1.4

And here, we’ll step away from zoom lenses and take a look at a fast prime – the 35mm f/1.4:

Nikon 35mm f/1.4 chart of focus adjustment required
Nikon 35mmm f/1.4: approx 40% needs an adjustment of 5 or more units
Canon 35mm f/1.4 chart of focus adjustment required
Canon 35mmm f/1.4: approx 32% needs an adjustment of 5 or more units

Again, we can clearly see that you need to check your wide-aperture primes with 30-40% of lenses across the two manufacturers improving significantly, and around 70% getting at least some visible improvement.

Source data at FoCal IQ – Canon and Nikon 35mm f/1.4 Comparison.

Last one – how to save a LOT of money

For the last lens, I just wanted to give a practical example.

Here’s the focus adjustment required for the very popular Canon EF 85mm f/1.2L II lens:

Nikon 85mm f/1.2 chart of focus adjustment required
Adjustment required for the Canon EF 85mm f/1.2L II lens

If your copy was one of the more than 70% that require an adjustment giving a significant improvement, you might as well have got yourself an EF 85mm f/1.8 lens – a saving of over £1,000! Every shot you would take at f/1.2, f/1.4 or f/1.6 would be soft, and potentially a well-tuned copy of the 85mm f/1.8 lens would outperform your very expensive f/1.2L version as the f-number rises above f/1.8.

If you’ve invested in high-quality fast glass, you really owe it to yourself to at least check the autofocus performance with your camera and lens, otherwise you’ve almost certainly wasted your money.

I’ve personally seen a lot of night-and-day performance improvements for even moderate-priced glass – autofocus calibration is like getting an upgrade to your camera and lenses!

Next up

On the last page, we’ll review what’s been said above, just to hammer home the point that you really do need to at least take a good look at the autofocus on your DSLR kit…

One comment on “You NEED to tune your DSLR autofocus!

  • Howard Phillips says:

    Fascinating. I’ve passed this in to a number friends who are professional photographers. Have you conducted tests with your software solution on Leica l, Fuji or Zeiss glass as well?

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *