It’s time to fill those memory cards. Canon releases firmware v1.8.1 for the Canon EOS R5. 400mp stills are now possible

That one is obvious: to limit the backlash if this feature happens to be unsuccessful.
Precisely the reason why I said:

"The answer I think, is that Canon are using the R5 as a testbed for this feature, rather than risk seeing it flop in a new model. The reaction here has been extremely negative. If they'd waited for the R5ii to launch the feature, and it had received such a negative reaction, it might have severely dented sales of the new camera".
 
  • Like
Reactions: 1 users
Upvote 0
Precisely the reason why I said:

"The answer I think, is that Canon are using the R5 as a testbed for this feature, rather than risk seeing it flop in a new model. The reaction here has been extremely negative. If they'd waited for the R5ii to launch the feature, and it had received such a negative reaction, it might have severely dented sales of the new camera".
In my head I hear Rudy Winston saying “With improvements coming in future firmware updates”.
 
Upvote 0
Precisely the reason why I said:

"The answer I think, is that Canon are using the R5 as a testbed for this feature, rather than risk seeing it flop in a new model. The reaction here has been extremely negative. If they'd waited for the R5ii to launch the feature, and it had received such a negative reaction, it might have severely dented sales of the new camera".
I agree completely. But I also think that, as usual on a gear forum, people had totally unreasonable expectations. As many who have used pixel shift on other brand's cameras have mentioned, uses are extremely limited. Plus, other brands go no higher than about 200 mp. Anyone who has gone up to a higher MP camera (even from 24 to 45) knows that pixel peeping will show you worse results. Yet, that's what many if not most folks are doing to compare the 400 MP images to the 45 MP. I did some experimenting this morning. I am not surprised by my results. 400MP is better than 45 if I use a wide angle lens on distant subjects. Using my 24-105 L, at 24mm, viewed at 100% for the 400MP image and 300% for the 45MP image (same size on screen), 400MP gives me a bit more resolution. At 50mm, 400MP still better resolution. At 105 - 45MP better as the 400MP image now pixelated at 100%. Using a zoom lens, the RF 100-400, 400MP is no better and usually worse at 100%. Conclusion based on very limited shooting (but it makes sense), 400 mp shots are for distant subjects shot with standard and wide angle lenses. From what I recall, these are the types of shots that Olympus users often use pixel shift for - cityscapes mainly. The firmware (not surprisingly) is all the rage on the facebook groups. From what I saw in limited browsing, people shooting distant panoramas were happy, other not so much. Cropping is limited, you probably want to use 400MP if you are not going to crop or crop minimally. Again, this is all based on limited testing and your - or others - results may differ. I used a simple tripod (not heavy-weight) and the 10 second delay (2 seconds will not likely be enough).
 
  • Like
Reactions: 1 user
Upvote 0
In the first gossip news on March 6, there was a statement like this. I guess I and many people like me are stuck with this statement.

"While we don't have the new firmware notes, "we did get some information as to what to expect." As you'll notice, these features are available in the Canon EOS R6 Mark II."

  • Possible Canon EOS R5 Updates Coming
  • Lens breathing correction
  • Pre-shooting buffer
  • Removal of 30 minutes record limit
  • Vehicle tracking AF mode (bikes, trains & planes)
  • Additional tracking for more animals
what about " "we did get some information as to what to expect." If this is some information, where is it? :unsure:
 
  • Like
Reactions: 1 user
Upvote 0
I took 6 more. The shutter was pretty slow, so even on a rubber pad, each of the 6 shots had varying amounts of vibration with the best showing greatly reduced artifacts but still too much. I'll try it on concrete with my heavy tripod and fast shutter tomorrow. I'll use a remote shutter release as well. Even then, vibration may find its way in. I used my 24-70 lens which has no IS. I'll try my 100L tomorrow to see if that is any different with IS off and on.
This morning, I took the same indoor photo but using a 10 sec shutter delay. There were no artifacts. It was pretty obvious last night that they were vibration related.

Just like a person would expect, it takes very careful technique to get a vibration free image at 400MP.
 
  • Like
Reactions: 8 users
Upvote 0
I can’t speak for others who may have been expecting more from this firmware update but for me the only major feature of the upgrade is the 400MP IBIS High res option and it’s not only unimpressive, it’s a bit of a dud.
I bought my R5 knowing the specs, and capabilities at the time I purchased it and so far has performed superbly from purchase, and after several firmware updates. I understand that modern mirrorless cameras are more computers than optical cameras like dSLR's and it's primarily software which can be improved and features added. I myself did not buy my R5 thinking that down the road it will have even more features and capabilities. I'm bewildered by some of the people here wanting extra capabilities beyond what the R5 originally had.

Firmware updates are great, and I'm more interested in updates that make my R5 perform better. The 400mp "feature" is fine I suppose but people seem to have unrealistic expectations when they buy the R5, or any camera really.
 
  • Like
Reactions: 1 users
Upvote 0
Here are some thoughts on expectations for this feature. The high res image is exactly 3 times the linear resolution of the R5 sensor with a few pixels lopped off the height (presumably to avoid edge effects from the stitching algorithm). The base sensor is sitting at 113.8 lp/mm, which puts the high rese image at 341.4 lp/mm. Roger Cicala did some testing of really good lenses at very high resolution and the results suggest that you are not going to get too much happiness out of most of your lenses
and

Roger does not have a fixture to test RF lenses, so he hasn't tested the RF 85 f/1.2, which is arguably the sharpest of the lot, but in chart testing, it doesn't appear to be hugely sharper than the Sigma 135, which is pretty amazing to start with and Roger tested that one.
Now bear in mind, we are talking about resolution that is 100 lp/mm higher than Roger's most extreme test , so the MTF of any lens you have is going to be a very small number at such resoultion. Keep this in mind when evaluating the feature and recogize that your best result may be to scale the image back down to something in the 100 to 200 MP range. That said, My preliminary test suggests that the feature taken by itself works surprisingly well (that is to say the effective MTF of the shifted sensor is pretty darn good). So all who want to know how good their lenses really are can use this feature and a really high res chart to find out :ROFLMAO:. The camera likely will not be a significant limitation in your measurement.

What we don't know is if Canon is using 9 base 45MP images or 9 90MP DPAF images, but the results might suggest the latter.

So bottom line to naysayers is to recgonize that there is no such thing as magic and in the absence of magic, this feature seems to do what you would expect plus a bit more and given that the JPEG output file is on the order of 170 MB, you probably didn't really want a "RAW" file that would likely have to come in the form of a DNG file (e.g. full res RGB) or 2.4 GB per image. Watching LR gasp for air when you click on the JPEG is your first clue. The surprising part is how fast the camera does the processing.
 
Last edited:
  • Love
Reactions: 1 user
Upvote 0
I agree completely. But I also think that, as usual on a gear forum, people had totally unreasonable expectations. As many who have used pixel shift on other brand's cameras have mentioned, uses are extremely limited. Plus, other brands go no higher than about 200 mp. Anyone who has gone up to a higher MP camera (even from 24 to 45) knows that pixel peeping will show you worse results. Yet, that's what many if not most folks are doing to compare the 400 MP images to the 45 MP. I did some experimenting this morning. I am not surprised by my results. 400MP is better than 45 if I use a wide angle lens on distant subjects. Using my 24-105 L, at 24mm, viewed at 100% for the 400MP image and 300% for the 45MP image (same size on screen), 400MP gives me a bit more resolution. At 50mm, 400MP still better resolution. At 105 - 45MP better as the 400MP image now pixelated at 100%. Using a zoom lens, the RF 100-400, 400MP is no better and usually worse at 100%. Conclusion based on very limited shooting (but it makes sense), 400 mp shots are for distant subjects shot with standard and wide angle lenses. From what I recall, these are the types of shots that Olympus users often use pixel shift for - cityscapes mainly. The firmware (not surprisingly) is all the rage on the facebook groups. From what I saw in limited browsing, people shooting distant panoramas were happy, other not so much. Cropping is limited, you probably want to use 400MP if you are not going to crop or crop minimally. Again, this is all based on limited testing and your - or others - results may differ. I used a simple tripod (not heavy-weight) and the 10 second delay (2 seconds will not likely be enough).
I think cityscapes would be a challenge, due to omnipresent vibrations from traffic. Landscapes almost a guaranteed failure due to moving foliage and/or water, not to mention resolution-destroying atmospheric haze and heat haze.

The *only" circumstances I can think of where this feature would be usable (OK I've got a limited imagination, there may be others) would be in a studio, where it could be used for exacting product photography; or in a museum environment, where it could be used for digitising works of art (although the latter are usually handled by very precisely stitching several dozen or even hundreds of images.)

For any kind of "normal" photography where very high resolution is wanted, the 100MP of a Fujifilm GFX should be enough (and will have better tonality), and of course there is always Gigapixel AI, which apparently produces superior results to this Canon pixel-shift implementation.
 
Upvote 0
I bought my R5 knowing the specs, and capabilities at the time I purchased it and so far has performed superbly from purchase, and after several firmware updates. I understand that modern mirrorless cameras are more computers than optical cameras like dSLR's and it's primarily software which can be improved and features added. I myself did not buy my R5 thinking that down the road it will have even more features and capabilities. I'm bewildered by some of the people here wanting extra capabilities beyond what the R5 originally had.

Firmware updates are great, and I'm more interested in updates that make my R5 perform better. The 400mp "feature" is fine I suppose but people seem to have unrealistic expectations when they buy the R5, or any camera really.
I sort of agree, in as much as I didn't consider potential future features when I bought my R5. Like you, I bought it on the basis of the original specification, and on the expectation that firmware upgrades would gradually eliminate the bugs that are inevitable with cameras with such complex firmware, and the near-infinite permutations of settings available via customisation.

Where we differ, is that I think it's absolutely reasonable to also expect *enhancements* to e.g. AF, and to expect a limited number of new *features*. All the other brands add features via firmware, so Canon has little choice to follow suit. It also makes a huge amount of sense from a sales point of view, to add features that will keep a camera competitive with models released a year or two later by other brands.
 
  • Like
Reactions: 1 user
Upvote 0
You are already past the diffraction limited aperture for R5 sensors. Probably (but not necessarily) you see no artifacts because you see no extra detail either.
You're right, I should use 2.8 or larger next time. Was just a quick test to get slow shutter speeds. I do see artifacts with longer focal lengths and moving things like leaves.
 
Upvote 0
I think the possible good news is that by not including the AF advances from the R6 II, we might expect an R5 II sooner rather than later.
The R5 is so damn good that I doubt I’d upgrade to a Mk 2 just for the r6 ii af upgrades. I understand that for some folks it’s different.

So selfishly, knowing that I would skip the generation, I’d prefer it as a firmware upgrade now

Do we even know if the AI hardware in the r5 can run the newer AF routines? Every year has seen a big jump in neural engine performance recently. The hardware just might not be capable. Anyone have any insight?
 
  • Like
Reactions: 1 user
Upvote 0
With initial tests with the new firmware and the pixel shift, have to agree with a lot of the comments here.. underwhelming and very finicky. It shows some potential, but would need maturing.

used my 180mm Macro with manual focus as it's a sharp lens and removes any movement from zooming or focusing.. still got artifacts once at 100%. What I found odd / interesting is that while I had set autofocus to manual on the lens, it still seemed to use the motor for something, not sure what. I only noticed as its older motor and fairly loud so was surprised by it and likely it generated additional movement/vibration I didn't want. Anyone else experience that?

as for the images while I did get some artifacts, when looked at on a 30inch monitor without pixel peeping, the pixel shift image did look sharper than a control image I had without it (old dead leaves on a plant)

It may be possible to get better results with a wider prime.. will keep playing.
 
Upvote 0
For those of you taking test photos to compare 45MP vs the new 400MP jpg, I have a few suggestions:
* Don't use electric light - The old lighting can pulse at 120 (or 100) Hz, and the new LED lighting can have other artifacts that aren't constant. There's a lot to be said about using natural light that is as constant as possible.
* Avoid wind completely.
* Use manual focus only (this may be forced by the new software - I don't know). I'd test with a prime lens instead of a zoom to avoid any possible focus shift.
* Set fixed aperture, speed & ISO so that it can't change between the 9 shots (that may be forced with new software - I don't know).
* Ensure the tripod & what it is on is rock solid (not a wooden floor which might vibrate somewhat)
* Use a time delay and/or remote shutter release cable.
* I'd use a quality test chart, otherwise a rigid subject with much contrast & detail so your results can show how good (or bad) it is.
* To show results, show a small patch of 400M results at 1:1, and the same patch from 45M up-res'd 3x3 bigger with your best post software to compare and mention which is which.
* Finally, get a box of tissues to wipe your tears of joy (or sorrow) when you see what it can (or can't) do! :oops:
 
Last edited:
  • Like
Reactions: 1 users
Upvote 0
For those of you taking test photos to compare 45MP vs the new 400MP jpg, I have one suggestion (in case you hadn't thought of it).
Do NOT do this test with electric lighting (incandescent, compact flouresent, or LED). All electric lighting will have pulse artifacts at 120 or 100Hz which might be the reason that some are seeing artifacts, since the 9 frames are taken at different times.

My only other suggestion is that you photograph a quality resolution test chart (and of course eliminate all vibrations).
Or a 3x3 grid of test charts if you really want to test for 400 MP :ROFLMAO: .
 
  • Haha
Reactions: 1 user
Upvote 0
Did anyone else\'s camera freeze up after installing?

Finally stopped showing the post-update screen with no responsiveness to any control, and now black with no response to any control.

I found it sat at 0% progress for like 2-3 minutes before finally starting to creap up. If you are able to try again, give it time. But it sounds like it's now totally unresponsive?

> Switched from Nikon. Still zooming the wrong way.

LOL. I decided to switch to a new paradigm where turning any control right would bring the meter right: bigger aperture, slower shutter, higher ISO. Even after a couple months I invariably turn a control the wrong way if I haven't adjusted any of these in ten minutes.

I think the old paradigm I was used to was that turning to the right gave you a higher number, which meant meter moves LEFT for aperture and shutter but RIGHT for ISO.
 
  • Like
Reactions: 1 user
Upvote 0