EOSHD testing finds Canon EOS R5 overheating to be fake

[UncleBobsPhotography]

44 minutes ago, Stanly said:

@Andrew Reid Sorry for re-posting, but you probably missed it – R5 does record 4K HQ via HDMI for 4 hours, not pixel binned.

He got problems when he turned the screen on. I wonder how internal recording works with the screen off using an external monitor.

To be honest, I'm just jealous my order isn't arriving any time soon so I can do my own testing.

[Lux Shots]

14 hours ago, horshack said:

There aren't any other cameras on the market that can do 4K full-oversampling from a 45MP full-frame sensor, so the fact that other cameras can achieve it in lesser configurations (smaller sensors or fewer megapixels) is not all that instructive. Same reason why the R5 can get warmer even just at idle - more pixels, more sampling, more data movement. There are also different levels of competency in design and manufacturing of large-scale ASICs, same as there is in software. One design may achieve a computational load more efficiently than others, which translates to less heat. This can be seen in AMD vs Intel with their different process technology and thermal characteristics.

I'm certainly speculating as I go along. That's what engineers do. They continuously evaluate the information as it becomes available and tailor their theories based on that changing information. The alternative is to start with a presumptive theory and then stick with that theory in spite of contradicting information. I try to avoid that when possible.

DIGIC is responsible for the full image/video processing pipeline once the data is ingested from the sensor. It's a multi-core ASIC with different computation units designed for each of those roles. As to the GPU scenario, actually yes, when the CPU heats up it heats up the air around it, which increases the ambient temp within that region. If that heat isn't conducted away it can heat up the area around the GPU as well, and vice versa. The operating temperatures of these chips is based on the ambient temperature around them - the higher the ambient, the quicker the chip will reach its tMAX. Based on previous Canon designs the EXIF temperature is on a chip well removed from the sensor and DIGIC complex, which would explain why it plateaus well before DIGIC does. We'd have to determine the precise location of that chip to understand how its temp corresponds to the sensor/DIGIC. The thermal photos posted online show both hot spots and cooler spots.

As an engineer myself, I hear your arguments. Ask yourself this as you try to calculate plausible thermal resistance transformations. Isn't this just too damn complicated?

I have never had such a difficult time with thermal management in any design I've done.

Artificial firmware limitations are what's the culprit. Of course, if Canon does fix it in firmware, they will say it's a bug, which is a definite lie.

[mechanicalEYE]

Yes, he got 4 hours but he definitely had to have the screen off.  The dummy battery doesn't play much of a role beyond allowing recording past the LP battery life.

I've tried all his his test, and the R5 screen seemed to be the difference in whether I got the overheat warning or not.

23 minutes ago, UncleBobsPhotography said:

He got problems when he turned the screen on. I wonder how internal recording works with the screen off using an external monitor.

To be honest, I'm just jealous my order isn't arriving any time soon so I can do my own testing.

External recording... GREAT! Obviously prores makes things a bit better dealing with the codecs of the R5...

My only issue here is, I WANT TO KEEP THE CAMERA SIZED AS IT IS out the box! Am I asking too much? LOL Thats o no go if Canon doesnt make absolute sense of these time limitations + recovery times.

 

 

 

[no_connection]

2 hours ago, Snowfun said:

Not sure that is the issue. It’s a question of how much heat a certain component(s) generates and, possibly more critically, how the software interprets that to initiate a shutdown. 
The latter could well be a combination of temperature, increase in temperature and time. Very little to do with the total heat generated by virtue of the camera’s power use.

While it in the short term matters where the heat generated in regards to thermal mass and how fast say the sensor heats up, but you are still looking at the outer shell to dissipate the heat. And it does not really matter too much in the long run weather it's the sensor, processor or a lightbulb hidden inside. Point is I want to know if 8k RAW draws 20W while pixel binned 4k draws 10W since that will indicate how sustainable a certain mode is.

Thermal transfer from sensor to air is ofc important but I don't buy the argument where you rely on thermal mass and count on something to take time to reach critical temperature effectively overdrive it beyond what it can sustain.

We need to look at what is coming in to set correct expectations for what we can get out.

[UncleBobsPhotography]

35 minutes ago, no_connection said:

Point is I want to know if 8k RAW draws 20W while pixel binned 4k draws 10W since that will indicate how sustainable a certain mode is.

This might be obvious, but this is rather easy to figure out. 2 ways:

1) Use a dummy battery and attach an amperemeter

2) Avoid overheating by putting the camera in a freezer and count how much you can record on a full battery

[Video Hummus]

1 hour ago, UncleBobsPhotography said:

He got problems when he turned the screen on. I wonder how internal recording works with the screen off using an external monitor.

Something like this:

if (HDMI_SIGNAL_OUT && (LCD_SCR_STATUS == OFF)) {
   CRIPPLE = false; {
else{ CRIPPLE = true; }

[Andrew Reid]

Anyone spot the problem with the charts like this?

[no_connection]

48 minutes ago, UncleBobsPhotography said:

This might be obvious, but this is rather easy to figure out. 2 ways:

1) Use a dummy battery and attach an amperemeter

Yes it is both obvious and easy which is why I asked/wanted/wondered why noone had done it yet.

[Andrew Reid]

The recording time charts don't take into account the time spent in live-view or menus.

So at some point during the first 40 minutes of the day, you will be looking at a 5 min limit in 4HQ instead... and then 0 mins.

Unless you work like this:

Shoot a 30 min continuous shot. Power down camera for 2 hours.

Come back and shoot another 30 mins. Power down for 2 hours.

Nobody works like this.

[Emanuel]

This is disgusting. These guys are crazy. And unprofessional.

[Stronz]

Whenever someone defends the R5 because of "45MP" remind them that the R6 is just 20MP and has the same issues.

[mkabi]

Not defending Canon, but suppose they created the same camera without 8K and 4K/120p....

Everything else will be available with longer record times and no overheating.

Then they secretly passed on the code to MagicLantern to unlock 8K & 4K/120... would that have been better....???? Obviously, Magic Lantern would unlock all the secrets including the overheating issue.

[Andrew Reid]

11 minutes ago, mkabi said:

Not defending Canon, but suppose they created the same camera without 8K and 4K/120p....

Everything else will be available with longer record times and no overheating.

Then they secretly passed on the code to MagicLantern to unlock 8K & 4K/120... would that have been better....???? Obviously, Magic Lantern would unlock all the secrets including the overheating issue.

Erm. How about they just give us what is advertised?

And tell the truth.

[John Matthews]

7 minutes ago, mkabi said:

Not defending Canon, but suppose they created the same camera without 8K and 4K/120p....

Everything else will be available with longer record times and no overheating.

Then they secretly passed on the code to MagicLantern to unlock 8K & 4K/120... would that have been better....???? Obviously, Magic Lantern would unlock all the secrets including the overheating issue.

Personally, I'd feel like I just bought a hotrod of a car that can only drive 5 miles.

[Andrew Reid]

Canon could have released a EOS R5 to compete against Sony A7R IV... for stills.

It has the slightly soft 4K pixel binned mode like the Sony does, but in 10bit.

That's not too shabby.

Then, release an EOS R5 C, with all the video bells and whistles enabled and properly functional.

For some strange reason they decided to go full beans with the 8K RAW and 4k 120p only to not let us use them! Marketing tick boxes only!

Why bother engineering these features if they are not going to be usable by the intended customers?

And how in any way would a mirrorless camera cannibalise sales of a $15,000 full frame C500?

[Video Hummus]

41 minutes ago, Andrew Reid said:

Canon could have released a EOS R5 to compete against Sony A7R IV... for stills.

It has the slightly soft 4K pixel binned mode like the Sony does, but in 10bit.

That's not too shabby.

Then, release an EOS R5 C, with all the video bells and whistles enabled and properly functional.

For some strange reason they decided to go full beans with the 8K RAW and 4k 120p only to not let us use them! Marketing tick boxes only!

Why bother engineering these features if they are not going to be usable by the intended customers?

And how in any way would a mirrorless camera cannibalise sales of a $15,000 full frame C500?

This is dead on.

And people would have paid $1000 extra for the this “R5 C” just because it’s a Canon and had 8K RAW and 4K120p with the autofocus and clog3.

[currensheldon]

55 minutes ago, Andrew Reid said:

Canon could have released a EOS R5 to compete against Sony A7R IV... for stills.

It has the slightly soft 4K pixel binned mode like the Sony does, but in 10bit.

That's not too shabby.

Then, release an EOS R5 C, with all the video bells and whistles enabled and properly functional.

For some strange reason they decided to go full beans with the 8K RAW and 4k 120p only to not let us use them! Marketing tick boxes only!

Why bother engineering these features if they are not going to be usable by the intended customers?

And how in any way would a mirrorless camera cannibalise sales of a $15,000 full frame C500?

It's really true. Or just charge an extra $1000 and put in a fan and have unlimited recording times. People would have bought it. I would have bought it and then probably would have bought a Canon cinema camera (the ones they are trying to protect) to go along with it once the RF versions are released.

Instead, I think I'll stick with Panasonic for awhile. 

 

[independent]

Overheating warnings without shooting anything is a sloppy implementation or a bug. It’s indefensible, and I think this will definitely have to be addressed in the firmware update. 

As far as any other changes, I can deal with the continuous recording limits to protect the cinema lines, but the recovery times need to be greatly shortened. 

How about this for a compromise?

Canon revamps the algorithm to better reflect real use case of their R5 target demographic: content creators and indie filmmakers. For most, the average duration of takes ranges from seconds to only a few minutes, so Canon should offer a more usable ratio of shooting time to recovery time...

Unlimited 30 second takes?
30 second cool down for 2 min takes?
1 minute cool down for 5 minute takes? 
5 min cool down for 10 min takes?

[arourke]

Have people tested removing the camera battery to see if it resets the timer? You may also need to remove the battery and insert a different memory card in case reset times are stored in a local file.

Thai may not work since some electronics have an internal battery for BIOS settings and/or an internal memory.

[noone]

Does shooting in C log or not make any difference to overheating (10 bit in log VS 8 bit otherwise)?