In my recent post on diffraction blur and lens sharpness, I started to take a closer look at factors to be considered when wanting to make full use of a high resolution sensor such as the one in my new Canon R5 with 45 megapixels. In this post I would like to continue and have a look at exposure times and how they may have to be adapted. Higher resolution sensors with smaller pixels are obviously more prone to shake from hand-held shooting, so I’d assume that the old rule of using an exposure time of 1/focal length may not hold true anymore and shorter exposure times are required for full sharpness on pixel-level. However on the plus side, the R5 comes with in-body image stabilization (IBIS) and I also wanted to test whether Canon’s claim of five stops of image stabilization can hold up and help delivering sharp images on the high resolution sensor.

Disclaimer 1: hand-held testing is of course personal and your own results may vary depending on how steady your own hands are.

Disclaimer 2: again, in this post we will indulge in pixel-peeping to assess sharpness of images coming from a high-resolution sensor. For reasonably-sized screens and prints, viewed from normal viewing distances, the requirements when it comes to resolution and sharpness are much lower than the level we look at here. Only when cropping images or producing very detailed large prints, the additional resolution and higher demands to shooting technique come into play.

To test the sharpness delivered by different exposure times I have done eight test series covering all combinations of:

• IBIS disabled and enabled,
  
• Two lenses: the Canon EF 35mm F1.4L and the RF 85mm F1.2L and
  
• Two distances: minimum focus distance and a focus distance of about five meters.

Each test series was shot as follows:

• Images were taken in increments of 2/3 of a stop of exposure time.
• Five images were taken for each exposure time, to account for variations due to the hand-held nature of the test.
• With a few noted exceptions, each lens was set to the aperture at which it delivered its sharpest result in my previous test on lens sharpness, which is F5.6 for the 35mm and F2.8 for the 85mm.
  
• Electronic shutter or EFCS was used to avoid any shutter shock interfering with the test results.
  
• For the minimum focus distance, ISO was variable to adjust for the different exposure times, for the longer focus distance a variable light source was used and ISO had to be only slightly adjusted in some cases.
• For comparison, the images were cropped and the crop scaled up by 300%.

Test Results without IBIS

To get a good baseline, let’s start by looking at the results without IBIS.

When looking at the results from the 35mm, it shows that the short distance is a bit more demanding than the long distance. At the short distance 1/30 or 1/50 are not sufficient to produce consistently sharp images on pixel-level with the 45 megapixel sensor, though at least the results from the 1/50 are usable for most applications. Also at 1/80 there is still noticeable blur in some images and even at 1/125 a slight blurriness remains visible. Only at 1/200 the results are sharp across the range, with only a very slight hint of increased sharpness at 1/320. Compared to that, at the longer distance, results at 1/30 and 1/50 are noticeably sharper than at the short distance and already usable at 1/30. Sharpness gradually improves till 1/80 and after that shows only marginal gains.

For the 85mm, the difference between the two distances is not as significant. At the short distance results from 1/100 are noticeably blurry but could still be usable, while 1/160 starts to produce results that should be usable for most applications. When looking for maximum sharpness on pixel-level, on the close distance stopping down to 1/400 is necessary, while on the longer distance 1/250 is sufficient, two thirds of a stop less.

These results show that shorter focus distances demand short exposure times, and longer focal lengths demand shorter exposure times, not only in absolute numbers but also relative to the old rule of 1/focal length. At 35mm, 1/focal length produces usable results, while at 85mm stopping down by a full stop is necessary. When aiming at making full use of the high-resolution sensor, going down two stops from 1/focal length produces good results on both lenses, after that no relevant increase of sharpness can be observed.

Below you can find the test results - since even the enlarged image is scaled depending on your screen size, you may want to download the full resolution to get a complete picture: 35mm short distance, 35mm long distance, 85mm short distance, 85mm long distance.

35mm, short distance, without IBIS - click to enlarge

35 mm, long distance, without IBIS - click to enlarge

85mm, short distance, without IBIS - click to enlarge

85 mm, long distance, without IBIS - click to enlarge

Test Results with IBIS

Now, let’s have a look how the IBIS can improve our results and allow us to achieve maximum sharpness with longer exposure times.

Looking at 35mm, there is again a very notable difference between the short and the long distance. At the short distance, results at 1/50 deliver similar sharpness as 1/200 without IBIS, so only a two stop improvement. Five stops down, at 0.5 seconds, results are unusable. Compared to that, at the long distance 0.5 seconds delivers very good results, that only slightly improve with shorter exposure times. So there clearly seems to be a difference in IBIS performance depending on subject distance, and only with subjects being further away the IBIS delivers the full advertised five stops.

With the 85mm at the short distance at 1/13 results start to be usable, comparable to 1/100 without IBIS, and at 1/50 results are comparable to the 1/400 without IBIS, with only a slight further improvement at 1/80 and not much to be gained beyond that. That is a three stop improvement over the results without IBIS. At the long distance, the IBIS again easily delivers five stops of improvement, with the results at 0.3 seconds being at least as good as at 1/100 without IBIS, if not equal to 1/160.

To sum this up, while already without IBIS shorter distances were more demanding, the IBIS further amplifies that by delivering its full advertised potential only at the longer distance. This is nothing Canon or any of the other vendors talks about, but is not totally unexpected since a lot of reviewers have already noted the decrease of IBIS and IS performance in macro and close-up applications. Still, even two to three stops improvement at short distances are a very welcome help and in low-light situations might make the difference between getting a shot or not. The five stops delivered at further distances really amazed me and very significantly expand the low light capabilities of the camera compared to ones without IBIS.

Below you can again see the test results - or if you prefer, download them in full resolution here: 35mm short distance, 35mm long distance, 85mm short distance, 85mm long distance.

35mm, short distance, with IBIS - click to enlarge

35 mm, long distance, with IBIS - click to enlarge

85mm, short distance, with IBIS - click to enlarge

85 mm, long distance, with IBIS - click to enlarge

Conclusion

So, what conclusion to draw from this test?

It is not easy to derive general rules from the results, with focal length, subject distance and variations in IBIS performance to be considered.

If you want to make full use of a high resolution sensor such as the one from the Canon R5, without IBIS add two stops to the old rule of exposing for 1/focal length of a second. For subjects further away, we might get away with a little less. But for most applications we will not need that level of sharpness on a 45 megapixel sensor. So if the situation demands it, 1/focal length will deliver usable results on shorter focal lengths, while on the longer focal lengths adding a stop would be good in all situations except emergencies. Add in the IBIS and you can reduce your exposure times by two stops at the very closest focus distances and shorter focal lengths, and gradually increase this up to five stops for subjects further away and longer focal lengths.

While these rules will deliver good results in most situations, if you are maneuvering on the limits of available light, it is worthwhile to know and have tested your own equipment and how it behaves in different situations. If possible, check your results constantly while shooting to make sure results are as expected and take multiple shots to later choose the sharpest. And in the end, if the situation allows a sturdy tripod is what virtually guarantees sharpest results.

As always, I hope you found this post useful and feel free to contact me if you have any questions or comments.