Okidoki... I've just done a semi-scientific test with my R7. Preface: I love this camera and what it gives for the price. Huge upgrade over my 7dii.
That being said, I recently watched a video from Jan Wegener which claimed that the Canon R7 has a shutter shock issue with EFCS mode. Here is the video: https://youtu.be/TBf2SnTbhRA
Now, shutter shock on full Mechanical mode is already well documented. But the general advice is to switch to EFCS and you'll avoid the issue. This has worked on other cameras. As such, I didn't believe that experiencing shutter shock on EFCS mode would be possible on the R7.
However, after performing my tests below, I believe Jan's claims about shutter shock in EFCS mode are substantiated. I didn't think it would be possible, but I'm seeing the result myself.
Here are the details...
Process
I put a book on my dining table with appropriate light for the test. I focused on a word in the middle of a paragraph in One Shot mode, centre point AF, Spot focus. I did a burst of 10-15 shots in EFCS then full Electronic shutter under the following conditions:
- Shutter speed lower than 1/100
- Shutter speed higher (range 1/200+, e.g. 1/640 etc)
- IS on
- IS off
I repeated this process a 2-3 times in each scenario to ensure my results were consistent.
I then reviewed and compared the results of each shutter mode at 10x magnification.
The test was performed with the following lenses, all adapted:
- 10-18mm IS
- 15-85mm IS
- 100-400mm IS ii
- 24mm f2.8 (non IS)
At the start of my test, I did a quick 'control' with the full Mechanical shutter with the 15-85 to verify what shutter shock looked like. I didn't repeat this with the 100-400 or 24mm, since I was primarily testing EFCS vs. Electronic.
(As a sidenote, Mechanical shutter showed evidence of shutter shock below 1/100 throughout the entire shooting burst. Compared to Electronic shutter, the result was noticeable enough that I'd not use Mechanical unless absolutely necessary (e.g. needing circular bokeh with high shutter speeds). Others have already demonstrated this).
On to the results...
Results
For the 10-18mm and 15-85mm, results matched each other. The Electronic shutter yielded perfectly sharp results across the entire range below 1/100.
As Jan noted, however, the EFCS yielded a perfectly sharp result on the first frame only below 1/100. The subsequent exposures in the burst were slightly blurry. I was able to replicate this numerous times, both with IS on and IS off.
The 'blurriness' was slight but noticeable. It was not as bad as the Mechanical shutter, perhaps showing that there is at least some dampening at work. I also haven't noticed a problem in the field; it's really only under these 'test conditions' that I see it, magnified to 10x and focusing on a word on a page. I have tack sharp photos of birds that are part way through a burst on EFCS. However, I'd suspect the first frame of those bursts would be ever so slightly sharper at the pixel level, even if I can't notice it.
Additionally, as expected, the problem went away at higher shutter speeds. I wasn't testing for where that threshold was, but others have suggested about 1/200.
For the 100-400 IS ii and 24mm f2.8 (non IS), results matched each other. There was no evidence of shutter shock on EFCS below 1/100, at any point in the burst range. This was unsurprising for the 24mm since, as others have noted, it seems the presence of IS has a role to play in shutter shock. What surprised me more is the lack of shock with the 100-400 ii (either with IS on or off). This may be down to it being a heavier lens, a more sophisticated IS mechanism, a full metal construction, or something else. Either way, both these lenses were safe to use with EFCS at any shutter speed.
Conclusion
Here's a summary of my findings:
- After the first frame of a burst shot, the 10-18 and 15-85 demonstrate slight shutter shock below 1/100 (at least) in EFCS mode.
- This problem disappears at higher shutter speeds.
- The 100-400 ii and 24 f2.8 do not present the same problem.
These findings lead me to believe that the R7's EFCS mode has a slight problem with adapted EF lenses of lower build quality that also have IS.
It should be noted that the 15-85 is a heavy-ish lens, so it's probably not weight that causes the problem. Furthermore, the 10-18 is not an overly old lens, so age probably isn't the issue either. From my perspective, it could be inferior IS implementation or something in the lens construction (e.g. plastic vs metal mount).
Given conclusion (1) and (2) above, I'd suggest using EFCS at shutter speeds faster than 1/200 and Electronic shutter at lower speeds. I've programmed this to a button on my R7 for quick switching (I use the "Silent Shutter" option to activate this; please let me know if anyone has a better idea!).
If you're not a pixel peeper, you may well be fine simply accepting conclusion (1) above. I'm 50/50 on it personally, though I'm encouraged that I haven't noticed this problem out in the field. Time will tell. In any case, I'm plenty happy with the R7 for my use as a hobbyist. It's a lot of camera for $1500 USD.
Next steps
If possible, it'd be good for others to pursue the following questions:
- Does this problem apply to the new RF-s lenses, which may be of similar build quality to the EF-s lenses I tested?
- Can we demonstrate or rule out the problem with any other adapted EF lenses? Can we determine a pattern?
- What is the exact shutter speed threshold where EFCS yields photos of the same sharpness across the entire burst range (under perfect conditions, of course)?
It should be possible for Canon to take a step with this, as well. One simple fix would be a menu item that auto-switches shutter mode based on shutter speed. This would at least remove the need to think about the problem in normal use. Hopefully they can provide this in a future firmware update; I'll be contacting them to suggest it, and would encourage others to as well.
Please feel free to ask any questions about my findings here. I haven't yet loaded the photos from my test onto the computer, but could supply them if needed.
