Moderator: E.J. Peiker

All times are UTC-05:00

  
« Previous topic | Next topic »  
Reply to topic  
 First unread post  | 22 posts | 
by Alexandre Vaz on Mon Aug 03, 2020 9:56 am
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
I know this may sound like a stupid question, but please bare with me.

I'm guessing in the last few days we all have heard about some specific cameras overheating while recording high-resolution videos (and taking a long time to cool down).

I know nothing about electronics, but I would like to know what is the major source for heating? Is it the sensor itself or somewhere else?

What do you guys think explains that different cameras (without cooling fans) perform so differently in this regard? Would you guess is mostly sensor technology? Data transference protocols? The ability of different materials to dissipate temperature or even the temperature set to turn the camera of? Something else? 

Thanks in advance

alex
 

by SantaFeJoe on Mon Aug 03, 2020 10:12 am
User avatar
SantaFeJoe
Forum Contributor
Posts: 8622
Joined: 28 Jan 2012
Location: Somewhere Out In The Wilds
There are several factors that can cause overheating. Battery drain, slow memory cards, ambient temps, LCD screens, etc. Just Google “why do cameras overheat” and you will find many answers. My opinion is that high battery drain is a large factor. If you have ever used a really bright EDC flashlight, one of the first things you notice is that it gets pretty hot very quickly at maximum output. Really almost too hot to hold. I’m sure that it is the same for camera batteries within the camera body.

Joe
Learn the rules like a pro, so you can break them like an artist.  -Pablo Picasso
 

by E.J. Peiker on Mon Aug 03, 2020 10:16 am
User avatar
E.J. Peiker
Senior Technical Editor
Posts: 86761
Joined: 16 Aug 2003
Location: Arizona
Member #:00002
It's because of the large amount of image processing required by the CPU inside the camera.  When shooting high rez videos at high frames per second, you are moving and processing insane amounts of data, which means electrons moving which means heat being generated all inside a very small enclosed weatherproof enclosure where it's difficult to get the heat out.  The way out is ever faster and more efficient microcircuits using smaller and smaller features while consuming less energy.  It's the same reason that when your computer is working hard, the fan gets louder and the computer gets hotter.  When it isn't working very hard it gets quieter and cooler.  All a digital camera is is a computer with an image sensor, a lens and one or more displays.

Note that overheating is a non-issue for still photography even at high frame rates.
 

by Alexandre Vaz on Mon Aug 03, 2020 10:41 am
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
E.J. Peiker wrote:It's because of the large amount of image processing required by the CPU inside the camera.  When shooting high rez videos at high frames per second, you are moving and processing insane amounts of data, which means electrons moving which means heat being generated all inside a very small enclosed weatherproof enclosure where it's difficult to get the heat out.  The way out is ever faster and more efficient microcircuits using smaller and smaller features while consuming less energy.  It's the same reason that when your computer is working hard, the fan gets louder and the computer gets hotter.  When it isn't working very hard it gets quieter and cooler.  All a digital camera is is a computer with an image sensor, a lens and one or more displays.

Note that overheating is a non-issue for still photography even at high frame rates.
Thanks for the explanation E.J., but considering what you've said, what explains why similar cameras with identical settings behave so differently in this regard? Would you guess that it's mostly related with software or hardware? 
 

by E.J. Peiker on Mon Aug 03, 2020 12:49 pm
User avatar
E.J. Peiker
Senior Technical Editor
Posts: 86761
Joined: 16 Aug 2003
Location: Arizona
Member #:00002
Alexandre Vaz wrote:
E.J. Peiker wrote:It's because of the large amount of image processing required by the CPU inside the camera.  When shooting high rez videos at high frames per second, you are moving and processing insane amounts of data, which means electrons moving which means heat being generated all inside a very small enclosed weatherproof enclosure where it's difficult to get the heat out.  The way out is ever faster and more efficient microcircuits using smaller and smaller features while consuming less energy.  It's the same reason that when your computer is working hard, the fan gets louder and the computer gets hotter.  When it isn't working very hard it gets quieter and cooler.  All a digital camera is is a computer with an image sensor, a lens and one or more displays.

Note that overheating is a non-issue for still photography even at high frame rates.
Thanks for the explanation E.J., but considering what you've said, what explains why similar cameras with identical settings behave so differently in this regard? Would you guess that it's mostly related with software or hardware? 
It's just physics, the more electrons you try to move in an increment of time, all else equal, the more heat you generate.  SF Joe also makes a good point, higher battery drain causes more heat from the batteries and those too are trapped inside this small weathersealed, unvented housing we call a camera :)
 

by SantaFeJoe on Mon Aug 03, 2020 2:14 pm
User avatar
SantaFeJoe
Forum Contributor
Posts: 8622
Joined: 28 Jan 2012
Location: Somewhere Out In The Wilds
In the third to the last paragraph in this article, it tells you how much more data 8K has to move. 17 times more than 1080 and 4x more than 4K.

https://www.dpreview.com/articles/86532 ... not-enough

Joe
Learn the rules like a pro, so you can break them like an artist.  -Pablo Picasso
 

by Alexandre Vaz on Mon Aug 03, 2020 5:09 pm
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
SantaFeJoe wrote:In the third to the last paragraph in this article, it tells you how much more data 8K has to move. 17 times more than 1080 and 4x more than 4K.

https://www.dpreview.com/articles/86532 ... not-enough

Joe

Thanks, but say both the new Canon cameras and the new Sony, both at 4K (and with similar sizes and materials) apparently deal very differently with the subject. That´s what puzzles me...
 

by E.J. Peiker on Mon Aug 03, 2020 5:14 pm
User avatar
E.J. Peiker
Senior Technical Editor
Posts: 86761
Joined: 16 Aug 2003
Location: Arizona
Member #:00002
Alexandre Vaz wrote:
SantaFeJoe wrote:In the third to the last paragraph in this article, it tells you how much more data 8K has to move. 17 times more than 1080 and 4x more than 4K.

https://www.dpreview.com/articles/86532 ... not-enough

Joe

Thanks, but say both the new Canon cameras and the new Sony, both at 4K (and with similar sizes and materials) apparently deal very differently with the subject. That´s what puzzles me...
12mp direct read sensor vs 20mp sensor whose image needs to be processed down to 4K - a huge difference in the amount of compute power required.  Also, Sony has way more experience with heat management since they suffered these problems in earlier cameras.  Also, Fuji and Canon sensors are very different - Sony is the undisputed world leader in sensor technology and Canon has been playing catch-up for many years.  You don't just catch up in one generation, it often takes 3 or 4 generations to catch up when you've relinquished leadership.

A good example of this is AMD vs. Intel.  It took AMD 15 years to catch-up and in some respects exceed Intel in microprocessor technology.
 

by WJaekel on Tue Aug 04, 2020 10:54 am
User avatar
WJaekel
Forum Contributor
Posts: 663
Joined: 30 Jun 2007
Location: Germany
E.J. Peiker wrote:...Note that overheating is a non-issue for still photography even at high frame rates...
I guess the jury is still out if this is also true for the just released R5 if you shoot stills in a hot enviroment all day long. According to the extensive tests they did at Dpreview the heat is building up and cumulating from the moment you switch on the camera, dive into menus, set AF etc. Probably that's not a problem for isolated still shots such as landscapes. But it remains to be seen, if extensive high frames are not effected either as some reports claim.
That said, I'm shooting stills for 95% with occasional intermediate videos if the scenario, i.e. wildlife behaviour, benefits from it. But the cumulative heat and long recover time rises the question if you regularily will face a situation that the camera does not even allow to record a short clip for 5 min or so in 4K HQ, 4K 120 (not to mention 8K) - or just shuts down very soon, at least,  leaving you with a 30min recover time in which you cannot shoot stills either and thus may be missing all wildlife scenes happening by then. That's really a no go for me and furthermore leads to the question if all the HQ video options are more or less useless for mixed application - at least in hotter enviroments. Though the R5/R6 seem to be really impressive as for IBIS, animal eye tracking etc you'd essentially limited to still shooting (if there's in fact no heat issue in the long run) - or to Low Quality video, at least, also available in the DSLRs. But then all the fancy HQ video options of the R5 are pretty useless and just marketing. I really don't know if Sony in fact is better for heat dissipation as you say. I personally will wait for more unbiased reports in the field before I come to a conclusion to go for the R5 - or stay with my trusty DSLRs. For now, I really hope that animal eye tracking will be upgraded in my 1 Dx III via a frimware update, too.

Wolfgang
 

by Alexandre Vaz on Wed Aug 05, 2020 7:28 am
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
Have you guys seen this one?

https://www.eoshd.com/news/a-round-up-o ... eld-tests/

This suggest the main issue is not sensor or CPU related but mostly something happening during information storage, right?
 

by Scott Fairbairn on Wed Aug 05, 2020 9:54 am
User avatar
Scott Fairbairn
Forum Contributor
Posts: 5131
Joined: 13 Jan 2005
Member #:00437
Alexandre Vaz wrote:Have you guys seen this one?

https://www.eoshd.com/news/a-round-up-o ... eld-tests/

This suggest the main issue is not sensor or CPU related but mostly something happening during information storage, right?


Unless I read that incorrectly, it appears that for video usage, the camera is extremely sensitive to temperature. So much so, that even using it for a minimum number of stills will severely impact the video capturing ability.
As soon as you turn any mirrorless camera on you will get some heating effects, but normally that doesn't impact shooting much. It would appear that the Canon's have very little latitude with temperature.
It's too bad because the cameras seem very good from a stills shooter perspective and the negativity surrounding the video specs is getting most of the press.
 

by SantaFeJoe on Wed Aug 05, 2020 10:47 am
User avatar
SantaFeJoe
Forum Contributor
Posts: 8622
Joined: 28 Jan 2012
Location: Somewhere Out In The Wilds
Here’s a review of the camera from a still photography point of view:

https://petapixel.com/2020/08/05/forget ... ographers/

Joe
Learn the rules like a pro, so you can break them like an artist.  -Pablo Picasso
 

by Alexandre Vaz on Thu Aug 06, 2020 10:04 am
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
This addresses most of my initial questions, but thank you all for contributing

https://www.dpreview.com/articles/86532 ... fic_source
 

by Alexandre Vaz on Mon Aug 10, 2020 12:35 pm
User avatar
Alexandre Vaz
Forum Contributor
Posts: 2051
Joined: 4 Sep 2003
Location: Portugal
More about the subject here

https://www.eoshd.com/news/i-bought-a-c ... solutions/
 

by Ed Cordes on Mon Aug 10, 2020 3:13 pm
User avatar
Ed Cordes
Forum Contributor
Posts: 4874
Joined: 11 Mar 2004
Location: Corning, NY
Member #:00700
Alexandre Vaz wrote:More about the subject here

https://www.eoshd.com/news/i-bought-a-c ... solutions/
Well, I respect and understand that this author is entitled to his perspective.  However, I am not sure how objective or valid the article is.  Adding new apps for the processor to handle seems like this alone will tax the system more. I also don't think the R5 is primarily a "professional video tool".  It is a still photography camera with some high end video capabilities available if you want to add them to your session.  Granted, Canon allowed the marketing of the 8K video and hype to occur.  I think that was a mistake.  However, in the end it is a still photography camera first.  Yes, I might be biased in my view as I only shoot very occasional video, and concentrate 98% on still photography.  This author's call for a class action suit and accusation of unethical behavior is over the top as the camera, as tested by many others, performs as stated in Canon's specs, operations manual, and post introduction comments.
Remember, a little mild insanity keeps us healthy
 

by Coreyhkh on Mon Aug 10, 2020 4:46 pm
User avatar
Coreyhkh
Forum Contributor
Posts: 1090
Joined: 8 Dec 2012
Location: London Ontario
E.J. Peiker wrote:
Alexandre Vaz wrote:
SantaFeJoe wrote:In the third to the last paragraph in this article, it tells you how much more data 8K has to move. 17 times more than 1080 and 4x more than 4K.

https://www.dpreview.com/articles/86532 ... not-enough

Joe

Thanks, but say both the new Canon cameras and the new Sony, both at 4K (and with similar sizes and materials) apparently deal very differently with the subject. That´s what puzzles me...
12mp direct read sensor vs 20mp sensor whose image needs to be processed down to 4K - a huge difference in the amount of compute power required.  Also, Sony has way more experience with heat management since they suffered these problems in earlier cameras.  Also, Fuji and Canon sensors are very different - Sony is the undisputed world leader in sensor technology and Canon has been playing catch-up for many years.  You don't just catch up in one generation, it often takes 3 or 4 generations to catch up when you've relinquished leadership.

A good example of this is AMD vs. Intel.  It took AMD 15 years to catch-up and in some respects exceed Intel in microprocessor technology.
I am pretty sure Canon has cought up with this sensor.... its first class
-------------------------------------
http://www.coreyhayes.net
 

by Jan Wegener on Tue Aug 11, 2020 10:12 am
User avatar
Jan Wegener
Forum Contributor
Posts: 5116
Joined: 29 Mar 2007
Location: Melbourne, Australia
Member #:00917
My R5 overheated while I was filming it with another camera for my Youtube review. I had it on standby for like an hour and was left with 1 minute of video. So that's the biggest issue for me. You are losing the video without even taking any video at all.
I can take 2000 images and still have decent amount of video time left, but 30minutes of standby and I'm down to 5-10 minutes of video.

Besides that the camera is amazing.
 

by E.J. Peiker on Tue Aug 11, 2020 12:34 pm
User avatar
E.J. Peiker
Senior Technical Editor
Posts: 86761
Joined: 16 Aug 2003
Location: Arizona
Member #:00002
Interesting tear down with an eye towards identifying the source of overheating - looks like some better manufacturing could mitigate at least some of the problems:
https://www.eoshd.com/news/chinese-user ... ime-limit/
 

by Jan Wegener on Tue Aug 11, 2020 10:27 pm
User avatar
Jan Wegener
Forum Contributor
Posts: 5116
Joined: 29 Mar 2007
Location: Melbourne, Australia
Member #:00917
https://www.eoshd.com/news/eoshd-testing-finds-canon-eos-r5-overheating-to-be-fake-with-artificial-timers-deployed-to-lock-out-video-mode/
this is, even more, damming, suggesting it's just a software imposed limit, maybe to protect Cinema series?
But why the outrageous recovery times and making you lose video time without even taking video. 
 

by E.J. Peiker on Wed Aug 12, 2020 9:31 am
User avatar
E.J. Peiker
Senior Technical Editor
Posts: 86761
Joined: 16 Aug 2003
Location: Arizona
Member #:00002
Jan Wegener wrote:https://www.eoshd.com/news/eoshd-testing-finds-canon-eos-r5-overheating-to-be-fake-with-artificial-timers-deployed-to-lock-out-video-mode/
this is, even more, damming, suggesting it's just a software imposed limit, maybe to protect Cinema series?
But why the outrageous recovery times and making you lose video time without even taking video. 
In most thermal control systems you build in what is called hysteresis which basically means that you don't turn things back on until the temperature drops well below the maximum set-point.  Without this, imagine a home thermostat, it would basically cycle on and of and on and off rapidly.  Clearly, if it is true that it is software controlled as is the conjecture on several sites, the low hysteresis point was set way too low and it should be a very easy FW fix or even FW hack.

One of my engineering classes in college was on thermal control systems ;)
 

Display posts from previous:  Sort by:  
22 posts | 
  

Powered by phpBB® Forum Software © phpBB Group