Canon EOS R5 records 4 hours of 4KHQ 30p to an external recorder, with a couple of simple tweaks

[twoheadedboy]

So I just have to carry my cards in portable fridge, joke to the side. I see another test thjat can be done. Film without cfexpress card in skipped line 4k and then when needed put the cfxpress for higher frame-rates, after doing them take it out and continue to film in ordinary 4k.

You don't need to do that. You can always film line-skipped, even when you have overheated from a different mode.

 

[genriquez]

Ok, I will do the dumb question.

Anyone tried with just a SD card?

DPreview updated their review and got a few more minutes with recording to just the SD card at room temperature. What a plot twist.

 

[Respinder]

Thankfully I don't pay attention to that guy. So many other people out there with a positive outlook on life and with actual talent. Who needs that kind of negativity on a daily basis?

It gets better - now Mr Reid posts a story about 2 min of video footage versus the 4 hours reported on here. His website seems to have gone the tabloid route! Sensationalize everything! ;P

 

[genriquez]

It gets better - now Mr Reid posts a story about 2 min of video footage versus the 4 hours reported on here. His website seems to have gone the tabloid route! Sensationalize everything! ;P

To be fair he did include a blurb about the 4 hours with external. But he wasn't enthusiastic about it and still wants a recall.

 

[ahsanford]

Ok, I will do the dumb question.

Anyone tried with just a SD card?

Umm, doesn't SD lock you out of all the better R5 recording modes?

- A

 

[Jonathan Thill]

Umm, doesn't SD lock you out of all the better R5 recording modes?

- A

just 8k alli, raw and 4k120 oh and I think 4k60 but have to check my camera

 

[Bdbtoys]

Umm, doesn't SD lock you out of all the better R5 recording modes?

- A

From the manual... can't use SD on just these modes... 8K Raw, 8K ALL-I, 4K 120, 4K 60 ALL-I

 

[dcm]

CFast cards in my 1DX2 are sometimes pretty warm when I remove them right after shooting. Not much of a surprise here given the amount of bandwidth being sent to them. What was surprising was how hot they would get when attached to my Mac via a SANdisk USB 2 reader. They continued to heat up after I was done using them and no data was being transferred. Hot enough to burn my fingers if I touched the metal part. I learned to eject them immediately when the transfer was done.

No surprise that CFExpress might have a similar issue. Both speed and the number of parallel channels to the chip will increase heat. I think you’ll find that the available recording times are somewhat related to how fast you are pumping data to the card which allows some modes to record longer than others. There may not be a 1-1 correspondence since the digic-x, battery, etc. are also contributing factors. It could be a driver/software issue or the controller on the card causing it to heat up even when not in use. I haven’t studied the protocol.

Most modern processors throttle down when they reach their maximum temperature. When my 6-core Processor in my Mac Mini hits 100C/212F under a heavy load from a machine learning algorithm, it begins to throttle the clock speed down to reduce the temperature to 100C even with external fans cooling the aluminum cover. I don’t even have to be using all 6 cores to cause this to occur. In my application this means it just takes longer for the experiment to complete. If I’m processing video it slows down the frame rate I see. In the camera it might mean the digic-x can no longer process the data stream fast enough if it similarly must throttle its performance to prevent overheating. It’s not a simple on/off switch that some here believe it is.

This is just a part of modern computer architectures which I’m sure includes digic-x and can affect the memory cards as well, both CFExpress and high speed SD. The reason that processor clock rates no longer get faster is the amount of power they required which generated excessive amounts of heat. You don’t see many above 4GHz these days. What you do see is more cores at the same clock rate. Data centers now spend as much to cool their systems as to run them, locating them near ocean or northern climates in some cases to reduce costs.

 

[pcho]

Why can’t they have an adapter to have the cards fitted externally and for camera to still work with memory card door opened. This way the camera is not as hot and internal recording my go longer as well.

 

[Quarkcharmed]

There is still card activity and battery activity unless external recording with external power

I'm not 100% sure what's going on there, but if you record externaly and don't write to the cards, there should be no CPU power spent on the cards anyway. So I guessed a purely physical reason why cards removal might be helping.

 

[usern4cr]

He got a wonderful increase in record time to 3 to 4 hours for 4K highquality.
BUT, he pulled 2 things out of the body: the battery and the mem cards, and added external power to power it and a "dummy battery" - which I assume lets the external power get to the body.

To say you're getting the extra time just because of taking the cards out is silly, as that would infer that taking the battery out had no affect. When we all know that if the battery was providing power for 4 hours (if it could) when you're not hooked up to wall power (almost 100% of the time for most casual shooters) then it would produce heat in the battery itself as it discharges. And that heat, coupled with the tight seals of the R5, has nowhere to go but to heat up the camera. In fact, it's possible that taking the battery out had *more* of a cooling effect than taking the memory cards outs. He needs to test it with just the mem cards out, and then with just the battery taken out, and let us know what happens.

So? Test it without cards only, then test without battery only, take 2 aspirin and call us in the morning!

 

[Otara]

Battery grip have any impact? Cant see anything on it, might not do much.

Edit: Wonder if the hack if delaying Canons ability to respond, the timing of those guys is pretty horrible for Canon from a ransom perspective.

 

[Quarkcharmed]

CFexpress cards get hot. They need and have heatsink.That is why they have an on-die thermal sensor and SMART reading info from the card is available to the device firmware, similar to info from your computer hard drive.

This theory can be tested by removing the Cfexpress card only. I wonder why would the system read thermal data from the card when not writing to it. If it's the case, it can be a software bug or oversight.

 

[David - Sydney]

So just leave the card door open and clip heat sink fins on the end of the cards. Problem solved!

I just checked my R5 and also Gordon Undone confirmed that opening the card doors turns off the camera. As soon as the door closes, it starts back up. Makes sense in some ways but it clearly doesn't allow card swaps a la A7Siii

 

[Otara]

This also might be rather obvious but has anyone tried alternating CFexpress cards?

 

[David - Sydney]

A little bit more on temperature management of CFexpress cards.
As this article indicates: “In general, we find techniques like throttling, which may be employed to reduce SSD temperature, to be effective at reducing the failure rate of SSDs. We also find that SSD temperature is correlated with the power used to transmit data across the PCIe bus, which can potentially be used as a proxy for temperature in the absence of SSD temperature sensors.”
In other words, at the end of the day, it may be the CFexpress card itself that employs some kinds of power throttling to reduce its temperature, and cause the 8K/4K read/write unsustainable. The difficulty is that different cards may have different heat tolerance and management policies.
Perhaps someone can experiment with different CFexpress cards and report the results here.
For the R5 itself, it would have been difficult to have policies to manage overall temperature for all current and yet to appear CFexpress cards besides the other heat generating elements so a kind of collective throttling (i.e. setting a safe time limit that guarantee the card and the camera will not burn out) was perhaps the only option. Good news is that the R5 is not hot to touch (unlike some other cameras such as XT4, for example) therefore, PCB heat is reasonably managed.
My guess is that, as the newer CFexpress cards with heat sensor communicate their temperature to the camera, better heat management mechanisms and policies will become available via firmware updates.

This may explain why people with the CFe Sandisk 128GB card say that they have had no issues using it but Canon says that it is not approved (1DXiii or R5). Not many people have multiple CFe cards from different manufacturers - yet

 

[David - Sydney]

This may also be the reason why Canon doesn't allow duplicate video recording to both cards as both cards heating up contributes to less video time.

 

[Jonathan Thill]

He got a wonderful increase in record time to 3 to 4 hours for 4K highquality.
BUT, he pulled 2 things out of the body: the battery and the mem cards, and added external power to power it and a "dummy battery" - which I assume lets the external power get to the body.

To say you're getting the extra time just because of taking the cards out is silly, as that would infer that taking the battery out had no affect. When we all know that if the battery was providing power for 4 hours (if it could) when you're not hooked up to wall power (almost 100% of the time for most casual shooters) then it would produce heat in the battery itself as it discharges. And that heat, coupled with the tight seals of the R5, has nowhere to go but to heat up the camera. In fact, it's possible that taking the battery out had *more* of a cooling effect than taking the memory cards outs. He needs to test it with just the mem cards out, and then with just the battery taken out, and let us know what happens.

So? Test it without cards only, then test without battery only, take 2 aspirin and call us in the morning!

He repeated the test with the included Battery and it ran till the battery was exhausted.

 

[Respinder]

To be fair he did include a blurb about the 4 hours with external. But he wasn't enthusiastic about it and still wants a recall.

That's the definition of a tabloid: sensationalize the headline but bury the real facts into the story.

 

[usern4cr]

He repeated the test with the included Battery and it ran till the battery was exhausted.

He specifically said he took the battery out and replaced it with a dummy battery. Where did he say he used a real battery (instead of the dummy battery) and if it went to the end of the battery then how long did he say that was?

The 3-4 hour test & claims were for no memory cards and no battery and with a dummy battery inserted so the camera would run. Hence you have no battery to heat up since it's now just an electrical pass thru of external power with no resistance relative to the camera.

To be clear: I think this is GREAT news for R5 video users. But please, we need an identical follow up test without just cards, and another without just the battery. We've already got one without both. Then we'll know what's really making the difference.