DPReview: Review of the Canon EOS R5

[dtaylor]

I have not seen anything conclusive showing that "having the camera on reduces record time". There have been a couple forum posters, but not a single reviewer that has demonstrated this that I've been able to find. It's certainly not been my experience to-date.

I know Armando mentioned it, and I thought both DPReview and the Northrups mentioned it. Whether you experience it or not will likely depend on ambient temperatures and sunlight exposure. Also possibly on settings like overheat control.

Also, I'm not sure what part of your background gives you the expertise in consumer products/engineering to suggest it would have been "minor to fix" this. We know "why" this is happening, and it's the fact that it's the sealed design of a 5D-class body and small size (combined with everything else that's generating more heat). By my estimation, solving this problem would require fundamental changes to the camera's design one way or another.

The competition doesn't have this problem. You could say "but they don't have 8k or 4k120" except this happens with oversampled 4k. And if you say "but they're not reading 45mp", the R6 has the same problem.

I have a relevant background but it doesn't matter whether I'm right or wrong about how "easy" the fix would have been in the design stages. (Admittedly I don't have a unit to tear down to try to prove/disprove my own statement.) What matters is that the competition is able to deal with this and do so in small, sealed bodies. (Well...in Sony's case we could debate the sealing. Never the less...)

 

[davidhfe]

The comments about non-HQ 4k on the R5 vs. "regular" 4k on the R6 intrigued me. Has anyone done an analysis comparing the quality between the two qualitatively?

I have an R5, and am still inclined to use non-HQ because it lacks the heat limitations, which is more relevant to my use (and still better than FHD by a country mile). But it would be interesting to know which one is truly putting out better video when comparing them directly.

I've seen side by side crops posted to YouTube. The HQ footage has significantly more detail, less aliasing, less noise.

When looking at real world footage, it's a bit harder. There were some shots where the HQ footage was clear, e.g. Chris' hair was almost distractingly sharp. Just outstanding quality. The standard 4K still looked good, and wasn't jarring when intercut. If they were identically framed, I think you'd notice a cut if even at 1080 YouTube. I did watch another set of clips that were at IPB though and at that point I started losing which was which. I also thought Armando's 4K-std shots were outstanding but those were professionally lit, and even 1080 likely would have looked great.

The 4K HQ is the best 4K you're going to see short of an ARRI or RED in terms of sharpness. It's really stunning. Others have better DR (A7SIII is like a full stop better practically) and are much better suited for 60/120, which don't have a HQ mode.

The DPR line sums it up for me:

If the quality difference between 4K HQ and standard 4K capture were not so dramatic, this would bother me less. However, once you start viewing and editing the gorgeous 4K HQ footage, it makes it that much harder to go back to inferior line skipped 4K, and that’s a type of disappointment I don’t want to be dealing with on a shoot.

 

[dtaylor]

The comments about non-HQ 4k on the R5 vs. "regular" 4k on the R6 intrigued me. Has anyone done an analysis comparing the quality between the two qualitatively?

I have an R5, and am still inclined to use non-HQ because it lacks the heat limitations, which is more relevant to my use (and still better than FHD by a country mile). But it would be interesting to know which one is truly putting out better video when comparing them directly.

Tony did side by side magnified comparisons in his videos. The HQ mode is definitely better. That said, I'm torn on the value of pixel peeping video. To give an example: if you compare the M6 II and the X-T3/4 in DPReview's video comparison tool, which magnifies the view, the M6 II looks horrible. But if you download 4k samples from someone like DSI Pictures on YouTube they both look great.

So there may not be that much practical difference between 4k and HQ 4k when viewed as intended on a 4k monitor. Stills can be printed very large and then inspected at close distances. Movies are not like that. Motion both adds to your impression of detail and smears away flaws you might otherwise notice in a print. Regular 4k might be just fine for most uses.

I'd still have an issue with the thermal limitations in large part because I love 4k60 slowed down.

 

[koenkooi]

Once again for those in the back of the class: the problem is not the record time limits. Those are actually decent. The problem is heat retention. Just turning the camera on eats into the record time limits. Shooting stills can drop them to zero. And even if you get the full time from a cold start, you then have to wait hours for recovery. And this is true for over sampled 4k30, not just for 4k120 or 8k.

Several competitors have over sampled 4k30 without this problem. I don't know why it becomes a religious/political issue to just admit Canon got the thermals wrong in this design. If you don't care about video then great, the R5 is an amazing stills camera. Love it, enjoy it. It's still frustrating to see Canon have a potential game-winning home run that stumbles on something that would have been so minor to fix in early development.

I'm having flashbacks to the lack of 24p discussions.

The comments about non-HQ 4k on the R5 vs. "regular" 4k on the R6 intrigued me. Has anyone done an analysis comparing the quality between the two qualitatively?

I have an R5, and am still inclined to use non-HQ because it lacks the heat limitations, which is more relevant to my use (and still better than FHD by a country mile). But it would be interesting to know which one is truly putting out better video when comparing them directly.

This one has all the 8k and 4k options, but only in a short, unscientific test:

 

[WhatDoesMStandsFor]

Once again for those in the back of the class: the problem is not the record time limits. Those are actually decent. The problem is heat retention. Just turning the camera on eats into the record time limits. Shooting stills can drop them to zero. And even if you get the full time from a cold start, you then have to wait hours for recovery. And this is true for over sampled 4k30, not just for 4k120 or 8k.

Yes, but I've also saw the article on No Film School were they shot +1500 photos befores started recording and managed to get close to 20 minutes before the camera shot down due to heat. Look at the 8K 4:2:2 10bit outdoor test.

Too Hot To Handle? The Canon R5 Heat Check

Testing the Canon R5 overheating in the real world.

nofilmschool.com

My guess is that the Overheat Control feature needs to be turned on all the time if you're going to shoot both stills and video.

 

[Chris.Chapterten]

I've seen side by side crops posted to YouTube. The HQ footage has significantly more detail, less aliasing, less noise.

When looking at real world footage, it's a bit harder. There were some shots where the HQ footage was clear, e.g. Chris' hair was almost distractingly sharp. Just outstanding quality. The standard 4K still looked good, and wasn't jarring when intercut. If they were identically framed, I think you'd notice a cut if even at 1080 YouTube. I did watch another set of clips that were at IPB though and at that point I started losing which was which. I also thought Armando's 4K-std shots were outstanding but those were professionally lit, and even 1080 likely would have looked great.

The 4K HQ is the best 4K you're going to see short of an ARRI or RED in terms of sharpness. It's really stunning. Others have better DR (A7SIII is like a full stop better practically) and are much better suited for 60/120, which don't have a HQ mode.

The DPR line sums it up for me:

The 4K HQ footage does look amazing. Much better than the standard 4K. I would really like to see how the 1.6x crop oversampled 4K footage compares to the HQ mode. I would rather deal with the crop and keep a higher quality personally.

 

[tron]

As much as we wish there is no camera that is perfect for anything (despite that there are more general purpose cameras like the 5 series).

 

[davidhfe]

I have a relevant background but it doesn't matter whether I'm right or wrong about how "easy" the fix would have been in the design stages. (Admittedly I don't have a unit to tear down to try to prove/disprove my own statement.) What matters is that the competition is able to deal with this and do so in small, sealed bodies. (Well...in Sony's case we could debate the sealing. Never the less...)

The competition does deal with it, but those bodies also have compromises:
- Z7, A7R: No 4K-HFR or HQ modes.
- A7S: Limited resolution
- SH1: Active cooling, larger body
- Sigma fp: Large passive radiator, No HFR
- Fuji, GHR: Small sensor, fuji still has thermal problems
- R5: Huge set of options, the most interesting ones have significant thermal limits that make them unsuitable for some types of pro work.
- R6: I have nothing positive to add

All of the above are very reasonable engineering tradeoffs. I don't think it's fair to say the competition is "able to deal with it" though.

 

[Chris.Chapterten]

Yes, but I've also saw the article on No Film School were they shot +1500 photos befores started recording and managed to get close to 20 minutes before the camera shot down due to heat. Look at the 8K 4:2:2 10bit outdoor test.

Too Hot To Handle? The Canon R5 Heat Check

Testing the Canon R5 overheating in the real world.

nofilmschool.com

My guess is that the Overheat Control feature needs to be turned on all the time if you're going to shoot both stills and video.

That's a really strange test. After overheating occurred they left the camera powered on with the display timer set to 30 minutes so it was constantly running. I would like to see the same tests done again with the camera powered off when cooling down...

 

[dtaylor]

All of the above are very reasonable engineering tradeoffs. I don't think it's fair to say the competition is "able to deal with it" though.

You left out Z6, A73, S1...probably forgetting something else. Do they have all the modes of the R5? Nope. Do they recover faster after shooting over sampled 4k? They certainly appear to, though DPReview's testing looks better than some of the other reports on the R5/R6.

fuji still has thermal problems

Not really as they recover quickly.

 

[twoheadedboy]

Canon EOS R5/R6 ProAV TV <- This link, time stamp 55:20.

Also, Mehdia was the only saving grace for pure honesty from Canon on overheating I saw in the very early days of pre-orders.

Thanks for this. What I took from what she said is that the temp is going up from non-photo/video ops, but what I didn't hear is a direct correlation with "first video recording length" or a limited ability to shoot stills. What that suggests - to me - is that the 1st video length is the 1st video length (+/= a small amount in modes limited to less than 29:59), but in a situation where the video was not started "cold", the recovery/refresh time is going to be longer, particularly in a hot ambient environment, because of the additional heat unrelated to the video recorded. Are you seeing anything that suggests that someone who's had the camera in standby and/or taken some stills cannot get 20 minutes of 8k raw in a "room temperature" environment?

 

[twoheadedboy]

I've seen side by side crops posted to YouTube. The HQ footage has significantly more detail, less aliasing, less noise.

When looking at real world footage, it's a bit harder. There were some shots where the HQ footage was clear, e.g. Chris' hair was almost distractingly sharp. Just outstanding quality. The standard 4K still looked good, and wasn't jarring when intercut. If they were identically framed, I think you'd notice a cut if even at 1080 YouTube. I did watch another set of clips that were at IPB though and at that point I started losing which was which. I also thought Armando's 4K-std shots were outstanding but those were professionally lit, and even 1080 likely would have looked great.

The 4K HQ is the best 4K you're going to see short of an ARRI or RED in terms of sharpness. It's really stunning. Others have better DR (A7SIII is like a full stop better practically) and are much better suited for 60/120, which don't have a HQ mode.

The DPR line sums it up for me:

Not HQ. I understand HQ is better. What I want to know is R5 "Non-HQ" 4k 30p UHD vs. R6 "standard" 4k 30p UHD.

 

[Kit.]

Not 120.

Can you please show me your calculations?

No, surface area exposed to airflow is precisely why airflow would efficiently remove the heat if it was the sensor overheating.

No, it's precisely why it wouldn't.

Why do you think there are extended surface heatsinks attached to CPUs, and not just fans directly blowing to CPU surfaces?

Is that a problem for the Z6, Z7, A7 series, A7r series, or XT series?

What does it have to do with your theory that it's "DIGIC processor was [not] properly heat piped"?

 

[Deleted member 384473]

Thanks for this. What I took from what she said is that the temp is going up from non-photo/video ops, but what I didn't hear is a direct correlation with "first video recording length" or a limited ability to shoot stills. What that suggests - to me - is that the 1st video length is the 1st video length (+/= a small amount in modes limited to less than 29:59), but in a situation where the video was not started "cold", the recovery/refresh time is going to be longer, particularly in a hot ambient environment, because of the additional heat unrelated to the video recorded. Are you seeing anything that suggests that someone who's had the camera in standby and/or taken some stills cannot get 20 minutes of 8k raw in a "room temperature" environment?

Yes. Having the camera turned ON in standby mode will increase internal temperature.

 

[twoheadedboy]

The competition doesn't have this problem. You could say "but they don't have 8k or 4k120" except this happens with oversampled 4k. And if you say "but they're not reading 45mp", the R6 has the same problem.

Is there any SLR-format camera other than the forthcoming Sony a7S III that features a video mode greater than 4k 30p which isn't limited by heat (but is limited relating to stills, file formats, card formats, and other areas)?

 

[twoheadedboy]

Yes. Having the camera turned ON in standby mode will increase internal temperature.

That didn't answer my question.

 

[Deleted member 384473]

That didn't answer my question.

When you start the camera into standby mode the clock begins to tick and you lose the full potential of estimated time provided by Canon and other testers before overheating. Might not get a full 20 minutes if you've been composing a shot and etc for a few minutes with the camera on standby as one would - White balance, exposing, and so on. Cool down should be about the same after a overheat shutdown no matter how you got there.

 

[dtaylor]

Not HQ. I understand HQ is better. What I want to know is R5 "Non-HQ" 4k 30p UHD vs. R6 "standard" 4k 30p UHD.

Tony compared them in his video. R6 is naturally better than Non-HQ 4k R5.

 

[Kit.]

I'm frustrated. I've held off on purchasing a more video capable body hoping that Canon would nail it with the R5. They've come so close. Absent the thermal issues this is an amazing camera across the board.

I don't want another brand for video. I want to be on FF Canon.

Isn't Canon yet to release its video-oriented RF body later this year?

 

[dtaylor]

Can you please show me your calculations?

I just love it when someone thinks needless nitpicking will prove something. The recovery times do not improve with airflow over the sensor. Unless the laws of thermodynamics fail inside the RF mount, it's not the sensor. And I don't need "calculations" to prove it.

No, it's precisely why it wouldn't.

Why do you think there are extended surface heatsinks attached to CPUs, and not just fans directly blowing to CPU surfaces?

What does this have to do with your original analogy, or with the statement: 'If it was the sensor overheating blowing air on it with the shutter open would have reduced recovery times."?

What does it have to do with your theory that it's "DIGIC processor was [not] properly heat piped"?

It was in reply to your belief that if the R5 shed heat more quickly you would need gloves to handle it.