Camera Overheating Tests Are Dumb
Вставка
- Опубліковано 10 чер 2024
- There's a major problem with overheating tests on cameras.
👕 Banana Shirt: bit.ly/49jEDkA
📼 Banana Review: • The TRUTH About Banana...
🛒 Gear I Use to Make Videos: bit.ly/3RlRmgx
🔌 Undone Cables: www.geraldundone.com/merchandise
🟣 Gerald Undone Merch: bit.ly/2ViBkcV
=============================
Table of Contents:
=============================
0:00 - The Plan
0:57 - The Problem
2:14 - Disclosure
2:53 - Results & Setup
7:29 - Theories
9:37 - Conclusion
=============================
Gear Used to Make This Video:
=============================
Camera: geni.us/lfNKufl
Lens: geni.us/wVTM
Microphone: geni.us/I40Dn
Recorder: geni.us/ZLyb
Key Light: geni.us/rGFf00
Complete Gear List: kit.co/GeraldUndone/youtube-v...
=============================
Best Deals for Camera Gear
=============================
Gear Focus Marketplace: geni.us/gearfocusGU
Amazon Top Deals: amzn.to/3d9tdT4
B&H Deal Zone: bhpho.to/2SCDrnj
=============================
Music:
=============================
I use Artlist for my background music needs. Use this link to get two extra months when signing up: bit.ly/2TV5sYT
=============================
Follow Me:
=============================
Twitter: / geraldundone
Instagram: / geraldundone
Discord: / discord
=============================
Patreon & PayPal:
=============================
If you wish to support the channel monetarily, you can use my Patreon page here: / geraldundone
Or directly via PayPal here: www.paypal.me/geraldundone
This is appreciated, but unnecessary, and no content will ever be behind a paywall.
=============================
Affiliate Links:
=============================
Some of the links in my video descriptions are affiliate links, which means at no extra cost to you, I will make a small commission if you click them and make a qualifying purchase. If you have a different purchase in mind, you can also use these storewide links below.
🛒 Amazon: geni.us/1m1G32
🛒 B&H Photo: bhpho.to/2MYRKBE
=============================
Recommended Channel Tools:
=============================
🚩 If you're a fellow UA-camr, I highly recommend you try TubeBuddy. It's helped my channel immensely: www.tubebuddy.com/pricing?a=U...
=============================
#overheating #sonyalpha #camerareview - Наука та технологія
FAQs:
1. Yes, I also tested with a dummy battery. So did Philip Bloom. Results still varied wildly.
2. Yes, I practiced proper heat-control methods. E.g. screen out, threshold set to HIGH, etc.
3. Everything was identical. Same battery, same cord, same card, same placement, same target, everything that I could control was the same. Exactly the same.
4. For those that misunderstood the conclusion: I'm merely suggesting that less confidence be placed in overheating test results as we don't know enough to explain what's causing the shutdowns. I thought it was ambient temp previously, but clearly something else is at play. And if it's more than just ambient temperature, a controlled heat chamber might still be insufficient to provide meaningful data since we know too little about what actually triggers an overheating shutdown.
Is it possible that the sensors are just cheap and unreliable? (The cameras obviously have a temp. sensor within to tell them WHEN to turn off...) - My car thermometer, for example, will tell us it's 40 degrees when it's actually 31... and then on other days it says 20 degrees when it's 25.
@@lamontmcleod2 Interesting. To check that theory, someone could have a thermal cam set up to monitor the rear plane behind the screen to gauge the cutoff temperatures.
Sounds software related.
If it overheats and you reset the camera and restart your computer, does it work?
A thermal camera would definitely be interesting to see more info. Especially WHERE the heat source may be. No guarantee it’s from the area we would assume.
I was just going to ask about the dummy battery. I have mentioned to lot of people how using one extends life, but apparently still just insanity.
Yep, I had my ZV-E1 live stream perfectly for 100 minutes, so I was super confident the next time I went to live stream… stopped 17 minutes in.
I bought the ulanzi fan for it. When I tested it it worked for 4-5 hours until I switched it off, tried it on a zoom meeting and it only worked for 40-50 mins 🤷♀️
What's your settings? 422? 420? 8/10 bit? how warm is the room? connected to a monitor? thanks! Curious to do a podcast for 60-80 minutes but livestream apparently draws more power - also, dummy battery
it’s temperature dependent and if the camera is on during that time of month. Just get a fan and be done with it
@@5k4nkHun7_42 I bought one too! It is still very inconsistent for me also. The fan has helped because It has never stopped 17 minutes in again but it also seems like 40ish minutes is the potential overheat point for me even with the fan. I’ve also been using a dummy battery.
@@jaymills1720 when it lasted 100 minutes then 17 minutes the room was 67 F and I had a fan pointed up toward the back of the cam to help with air flow. I was in 10 bit 4:2:0. No monitor plugged into HDMI. Since then I have used a Ulanzi fan and a dummy battery and have been able to double that minimum overheat time, but it is still unpredictable.
Would be nice if they added the option to display the internal temp reading so you'd at least have an idea of how quickly it might reach shut down temps in any given situation
Some camera fans display a temperature reading from the back, which is nice
I had the same experience with my ZV-E10. I didn't do 30 runs, I just did 1 per setup. My findings are:
1. Screen flipped out makes the camera run longer.
2. Plugged-in power does not overheat, battery power overheats.
I used my IR thermometer to find the camera's hot spot, and I could predict when it would be close to overheating. The hot spot wasn't the sensor or the screen or the battery, it was something internal under the menu button.
I think the reason the cameras are overheating at variable durations is because the battery is charging/discharging while connected via USB. The test you showed at 4K24 that ran for 2h 19m was the best result probably because the battery was just discharging the whole time. You should try re-running a few tests using a dummy battery as the power source. I'll bet you'll see a dramatic improvement in run time.
It's because they made them too small and told us they are video cameras
@@navpreetsingh8156Nah, it's because batteries get hot when they charge
Yes charging while recording makes a huge difference. I tested this myself with a thermal camera and 4K60 was no issue with the a7cII but it quickly became an issue when USB-PD was plugged in. I have not tested if it makes a difference when the battery is full or empty AND charging
The issue with dummy batteries is that they can zap your camera and Sony doesn’t provide a first party dummy battery that plugs in straight into the wall. A lot of us aren’t willing to take that chance.
@@Soundwave857You can’t run USB streaming when the battery is empty. They slowly lose battery when connected to USB and streaming. On my A7C R, the 4K30 streaming stopped after 12 or so hours and it wasn’t because of overheating. It was because the battery slowly drained to 0%.
There was a live stream from Panasonic's Sean Robinson, sitting in Austin, Texas, on a nice vantage point, in the sun during summer, at around 35-40°C (100F). Streaming was done with the GH5-II. He showed what "-10° to 40°C - no overheating" means and that Panasonic makes sure that cameras with this specification meet the mark. After almost an hour it was obvious that the camera could go on but Sean ended the stream a bit earlier than usual.
Yeah Sony suck
what about s5? Gh5 just has M43 sensor.
In Thailand, which got 25C-30C in 'aircon to the max' room, or 40C+ outdoor. There is one UA-cam guy whose content is to setup people's live stream. Hundreds or even thousands of stream setup he had made, he always use Panasonic.
@@ThanhThanh-it1pmSensor size is overrated.
This stream was made when the GH5-II was new (Sept. 21) and it was made to demonstrate the new streaming features of the GH5. The original S5 has no fan, so it would most likely get into overheating in the sun in Texas.
I'm not sure if there is any FF camera without a fan that can be used without recording limit at high quality 4k (oversampling, no line skipping). I'm also not sure if this is needed for a UA-cam stream. We are talking about low quality 8bit. A good action cam or smartphone could deliver the same quality if you find one that doesn't overheat.
@@Ulrich.Bierwisch GH5 has M43. it's even bigger than some sony camera. It has fan. sensor is kind of big chip. M43 is 1/4 FF size.the video is about FF sony camera.. we should compair something similar must compair with FF camera like S5 S1H R6 R5.
it's a trade off, bigger size or overheat
My laptop overheated watching this in 4K so I had to switch to 240p.
I've seen the same variance in my experience with my A7IV and A7S2 (not just in webcam mode, this also applies while recording long takes/streaming using the HDMI output).
In the end I've just stuck to a couple of best practices, and since then I've never had any overheating issues.
- Even the smallest bit of moving air will make a gigantic difference vs completely static air. Even a small and slow fan near the camera, where you can barely feel any air movement, will make a big difference in my experience.
- Opening up / flipping out the screen whenever possible so air can move along the back instead of trapping the heat.
- The biggest contributor for me: using a dummy battery. The battery is often a large contributor to the generated heat, so this moves that heat to outside the camera and replaces it with a cold part that can even soak up some more heat from nearby components instead (or if your camera allows it, you could power it through usb and leave the battery door open). On top of the heat benefit you can then also use much larger external batteries or AC power for infinite runtime.
Ive left the a7sii on and pushing video through HMDI for up to 3 days!!!! the a7iv has been a nightmare with inconstancies sometimes only lasting 1.5 hours
What dummy battery brand do you use? I'm a bit leery to try them as one review I read said they tested the output voltage at 8.4v (it should be 7.2v).
@@Digitalfiendscom I don't really remember what brand mine are, I just check before first use if the pins correspond to the right ones when adapting straight from a larger NP-F battery (voltage on those are the same so no conversion needed), or that the voltage is correct when using AC or any conversion.
I don't think there's anything wrong with the unit you were looking at though. If the "nominal voltage" of the battery for your camera is reported to be 7.2V, it will range from 8.4V (fully charged) to somewhere slightly above 6V (empty). If the dummy battery provides 8.4V, your camera will show that the battery is 100% charged.
The way I calculate those values: a standard lithium cell goes from around 4.2V to around 3V with a nominal voltage of around 3.6V (different cells can have slightly different specs, like 3.7V nominal or 4.3V max). If a battery has multiple cells in series, every value will be a multiple of that. In this case the nominal voltage of the battery is reported to be 7.2V, which is 2 * 3.6V meaning two cells in series. That is why each value is doubled (thus the range is 6V to 8.4V).
I worked on a multicam shoot in January and due to a snow storm LensRentals couldn't get our camera package out to us in time for the first few days of the shoot...not ideal
We were filming in a very small town and had to source some cameras from a local colleges film class, they were very generous and let us borrow a bunch of Canon R6 cameras
I ran some tests in the hotel room to make sure we wouldn't run into overheating issues, I managed to have a camera fill a card without overheating and one drain the battery..no overheating
Long record times , no issues
We get onto set and set them all up on tripods to frame up all the various shots (indoors , air-conditioned convention center) and before we even start shooting you could see the record times on all of the cameras slowly drop lower and lower
By the time we were close to shooting the cameras said they could only record for 10 minute intervals..that's worse than shooting on a 5DmkII...
They were getting hot before we ever even hit record
Screen flipped out, oem batteries, screen brightness down
And mind you, this room was chill, some of the cast/crew were wearing sweaters
I use a cheap 30$ HDMI capture dongle (plug-n-play, no drivers or software required) and a dummy battery when I stream with my camera. I know the video quality is not like super sharp oversampled 4K when you capture the live view of a camera but it's still *MUCH* better than using a webcam and it never overheats. Also, a lot of camera have settings for the video quality output when plugged in HDMI.
OVERHEATING! Everything overheats on me. Not just live streaming but recording in general. 4K 30P on the A7SIII, A7IV, ZV-E10, ZV-E1... All of em! I ended up biting the bullet and going with an FX3 and FX30 to put an end to it.
Really? Might depend on the environment are you shooting inside or outside? I use the zve10 in my studio for long podcast shoots with dummy batteries i never had them overheating during a session but i do keep the room A/C on
Theory: the temp sensors that trigger the overheat protection aren't really measuring the overall "warmth" of the camera, they are instead measuring one, or more likely several, critical components prone to overheating, i.e. processing chips. These can run indefinitely (for some cameras) when the only thing the camera is doing is streaming the video over USB, but if one or two other tasks start up in the background, it will cause a chip to exceed the thermal dissipation it can handle, and it will quickly trigger the protection. So the discrepancy in your tests may then have less to do with the ambient temp of the room than some other function of the camera, or a buggy interaction with the host computer. For example, if the camera is connected to wireless (I know some modern cameras are starting to do this, not sure about the ones you tested) it might be phoning home, looking for updates, etc. Or the algorithm that monitors the locally available wireless networks or nearby bluetooth devices might have some associated periodic task or whatever. Or the camera might just have some unrelated internal periodic task that triggers at unpredictable times. Or something happens on the host computer that triggers an invisible background task reaction in the camera, etc. It certainly sounds like something a firmware update could address. Others have also mentioned that battery charging/discharging creates warmth, and although I assume these cameras were all fully charged during these tests and running on wall power, that doesn't mean there isn't some battery-related stuff happening in the background that might generate warmth or trigger tasks. And one final shot in the dark: very slight/imperceptible variation in how a camera is latched into a tripod might dramatically alter how much thermal coupling between the camera and tripod occurs. So if you had the stomach for more tests, I would try running the cameras with no battery installed, in "airplane mode" if such a thing exists for cameras, with no memory card installed, streaming to a minimal linux installation, and not mounted to anything, just to rule out as much stuff as possible, and see if the results become more consistent. One can also be paranoid and wonder if manufacturers have started detecting when overheat testing is underway (e.g. static unchanging scene, minimal audio fluctuations) in order to push things a bit farther than they otherwise would.
Theory. Sony suck
@@coin777 physics sucks
Electrical Engineer here, I work with embedded systems and I sometimes see erroneous readings out of the temperature sensor. If the embedded software doesn't handle those edge cases proprely, it may consider it as a high temperature and decide to shutdown. Those glitches are random and could explain why your results are random. Also, most of Sony camera may share a common drivers across all their camera and this could explain why the overheat randomness is accross all Sony cameras
I still don't understand why we have in the FX30 the "Auto Power OFF Temp." with Standard and High options in the menu if it has a built in fan in the body... 🤦♂
I have a feeling that if the time is so variable it’s either:
A. The camera is getting charged by the computer and it’s throwing the thermals all over the place
B. The usb streaming requires an chip on the circuit board that isn’t frequently used outside of that use case, so it’s thermals are wacky. Either not monitored super well, not well integrated into the thermal management in the camera, or something similar
so this works for laptops also. the heat dissipation is better after a warming period! when everything is starting slightly warmer, the dissipation starts off better, and for some reason that makes the test different. But hey, they ALL get hot at some point, yes, they do, but it seems to actually matters if it STARTS hot in the first place... For some reason, if it starts off hot, it will, given the same environment, perform better than if it starts cold. I noticed the same thing with my laptop. Exporting a huge project took considerably less if the laptop started the export hot. Same environment, same ultimate cooling performance and general temps, it just performed better when starting hot ...
Potato Jet's older camera overheat tests also support this 'hypothesis'
Great tests and investigation! Love the rant at the end!
I am a little confused - is the takeaway that we should purchase these cameras and not obsess over the heating issue for indoor environments or that we should worry and not purchase them if we record longer than 30 minutes?
My personal experience with Sony is that there are only a few models that are actually great with heat management:
- FX3, A7S III (same sensor)
- A7R V and A7C R (same sensor)
- FX30 (adequate cooling)
Those are the only cameras I’d consider if extreme runtimes are desired.
Both
The takeaway is that overheating testing is very difficult because the results can vary wildly. Because of that, you also can't trust these cameras when doing long recordings/streams because one time they may work fine for hours and the next time they may shut down after 20 minutes with all variables seemingly equal. The point is that you can use this info to make your own decisions. I did similar testing because I found it useful to know, and I decided to use the smaller cameras daily for situations where I don't need 100% reliability, and I rent larger cameras when I do need maximum reliability for long recordings/streams.
Panasonic S5II: "Hold my cold beer"
Honestly everyone is sleeping on Panasonic S range, they are solid!
Cannot wait for S1 MK2.
@@Spencer_Whiteno one is sleeping we just don't want the Crop on 4k 60
@coin777 Almost every full frame camera crops at 4K60 including the A7IV and A7CII. Only cameras at considerably higher price range don't. So either your lucky enough to own a very expensive camera or are being hypocritical 🤷
Just buy a quality lens that's just a little bit wider for 4k60 etc. crop. It's not that hard to adapt.
@@coin777 i have a feeling that even without crop, people wouldn’t give panny a chance because it’s so foreign. Crop 4k60 is mostly no big deal since most use for b-roll/slow motion.
You have to be lucky to buy a camera without overheating problems. I used the Sony A7IV to record a documentary in the coastal area of my country and the average temperature of 25 to 35 degrees C. We recorded interviews of between 40 minutes to 1 and a half hours directly under the sun and in some cases at midday and never We had overheating problems. But I know people who have the Sony A7IV and in 10 minutes they get hot recording in an air-conditioned place.
I've heard of people getting drastically different overheating results from the ZV-E1. I would love to see one of these done for Canon to see if it's also inconsistent.
go do it yourself 😅and publish a video 😁
I think its really cool that Sony sent you a camera to test for a stress test on a known issue. I assume Sony Cameras and Sony/Discovery are managed by very different people.
I think the variable people might be missing is airflow? Varying amounts of airflow around the camera can help it dissipate heat at different rates.
100% this. Even the slightest difference in airflow over the other side of the room will make a huge difference in the way the heat dissipates. Basically all of the heat dissipation is through convection (apart from a tiny amount of conduction into a tripod stand)… and the tiniest air movements can completely alter how much heat convection is taking away from the camera!
Thank you for the work you've put into these tests. I can imagine it's been a lot of work, but luckily for us you've got that obsessive curiosity gene :)
1080p is more than enough for video calls or whatever. For 4K just get a camlink
Camlink doesn’t fix the overheating. In my (admittedly limited) experience, it’s the 30p in 4k.
@@cabellofatto camlink works by capturing the hdmi output. It's the same as connecting your camera to an external monitor. Haven't seen any camera overheat when doing that so I think it's different
The ZV-E1 is advertised as 4k unlimited recording
@@cabellofatto When you use the webcam mode the camera is encoding the video in the body to send it over USB. When using a Cam Link you're just sending out an HDMI signal and the Cam Link does the processing - so there's quite a difference in performance because of this. The one thing that could still be a factor or causing overheating issues is if you're powering via battery and not using a dummy battery - dummy batteries won't generate heat as they power the camera so can be very useful in this use case.
@@Steggy interesting. My zv-e1 would still overheat with a dummy battery running through the camlink in 4k30 after ~30 mins. That said, Sony had me factory reset the camera, and after that, I started getting longer runtimes, so *shrug*
This might sound strange. I hope that those weird times for sony zv-e1 are result of the bug in the software. I’m using ulanzi cooling fan and sometimes the camera still works at 45C and sometimes shuts down at 37C.
No vortex ending? Awww.
Thanks for this test on tests. As someone who never streams or films interviews, I've never taken overheating tests seriously. I understand those that do take it seriously if it does affect their work. But, you've now shown that overheating tests in and of themselves don't provide information that is useful or reliable due to the drastic variations in results.
I'm so glad I subscribed to your channel. You always come through with the most useful information on video production, etc.
A question; Does it heat up faster/slower if there more or less motion being captured? What about lens cap on vs bright lights over exposed.I am wondering if the amount of pixels changing to the cpu for compression are affecting it?
You need to take the lens cap off to capture video!
@@navpreetsingh8156 Obviously, but with all black video, there is almost nothing to compress.
it’s probably something to do with the SOC having inconsistent heat and stuff… my guess is that the inconsistency of the heating performance is what let the manufacturers to reduce the maximum heating temp before turning off so that the customer experience was more consistent
Check the lumix camera. Maybe they are more consistent?
yes. GH5, GH6, S5II and S5IIX are more reliable than sony mirrorless cameras.
yeah, but not many cameras has a physical cooling, so it'll be interesting to test Lumix cameras against themselves
The level of dedication you invest in these test videos is absolutely mind blowing!! Shout out to what you are doing for all the filmmakers/producer's community!! 🙌
Even when there is no new cameras release, you take your time to educate us with your much detailed tests!! You are raw!!! 🙏
Yup, I hate doing these tests for very frustrating & unexplainable inconsistency...
One of the major reasons that these cameras overheat in the first place is that the thermal design is pretty nonexistent, just take a look at the Canon R5 in where the main heat spreader is sandwiched between the cpu and the power supply board and the heat has nowhere to go, add a small copper plate that contacts the heat spreader with the back plate and no more overheating issues
Yup. Unlike Panasonic who had a fanless full frame camera that didn't overheat in 4K (the original S5), both Canon and Sony are yet to discover copper
The battery compartment has a temperature sensor, could the batteries be the issue as they may discharge at different temps? Sometimes ill take a battery out after a hour and they are hot
I've had to do my fair share of overheating tests in the past - though my goal was to simply answer "can this camera go longer than 4 hours connected to a Cam Link" - but even then you're right, ambent temperatures can play a big role. What I'd be curious about though is where your thermometer is in your testing room because in my office my computer area will be the hotspot so depending on how much my laptop wants to be a radiator at that time will cause that corner of the room to go up in temp by a whole lot while the overall room temp might experience a smaller change in degrees. A solution to that is to keep the camera far away from the computer in testing so IDK if that's your type of setup or not.
A big factor that most Cam Link users run into who experience overheating is they will plug in a USB cable for power while keeping the normal battery in. This is a big generator of heat and even when just passing through the HDMI live view to a Cam Link - the heat generated from the battery can be enough to cause overheating. Meanwhile dummy batteries don't warm up like this so you get a marked improvement in performance. So even for Sony's webcam mode I would suggest using a dummy battery just as a safeguard. It's also just convenient to not have to worry about any slow battery drain from the usage either - dummy batteries never die.
Also, not to shill here - but I think that if the user described at the end, if the live streaming feature is really the determining factor between 2 models (Let's say I'm deciding between ZV-E1 vs. A7CR) rather than spend another $1k for the extra time in webcam mode, it'd be more cost effective to just go for the ZV-E1 and just use a Cam Link to handle the webcam aspect of it (IMO)
valid points indeed
You've already met some guys who have the exact setup you're looking for, for the exact purpose of investigating this sort of stuff.
Hey, @LinusTechTips I think you've got a[nother] job for the Labs guys!
Thanks for sharing! Your results show that it's needed to attach a thermal couple to the camera and record its temperature over the duration of the overheating test.
From your description, it seems that the camera temps stabilizes pretty close to the overheating threshold, therefore the time to reach it can be from a few minutes to days.
My theory is something something happening with heat sensors, and the sensors might be similar if not the same in all those models. Remember Canon R5 situation, when it "overheated" upon release pretty quickly but after firmware update they kind of fixed it? So my guess is the heat sensors could potentially go crazy because of some small variables, so in the reality we can't truthfully rely on those. I also really interested in infrared temp check of the same camera when it's running for 8 hours while being ok, and "overheating" for like 30 minutes. If there's no change in 'internal' temp, it's a major issue since the longevity of the cameras can potentially drastically harm because of those potentially false overheating indicators.
Ambient temperature isn’t the only issue. Being in direct sun can cause a camera to overheat quickly too. My former Sony A1 would overheat in minutes in direct sun, but draping a white towel over it would extend recording time substantially. Now some cameras that are properly engineered to handle heat don’t have that issue, even in direct sunlight.
We experienced a similar problem at our studio that primarily does long form podcasts. We used to run a6300s but they would need sit idle in 1080p as the 4k idle would cause them to overheat. We swapped to FX30s for the built in fans and they never seem to overheat now in 4k30 idle. Our A7iii center camera still does if it is left on for 3+ hours. I have not experienced such large fluctuations in duration though before it overheated. A really good thing to hear though as it's not something people think about often.
This is interesting.
I find that my body temperature can cause my R6 to overheat faster. If I'm in a situation where I'm holding the camera close to my body or on my chest, it tends to overheat a lot faster. That thing really wants a fan on it.
I suspect the reason maybe even more weird. For example, on my a7m3 it start heating as soon as I occasionly switch to HDR mode (default on PP10). Even if I turn to other mode, it still gettinig really hot at left side of body. No recording or streaming, just playing with camera keeping in arms and pointing to objects like trying to focus. My theory is it some suboptimal firmware algorithm that kick in when HDR used (I guess, more aggressive NR) and it did not properly shut down and keep shredding bits even when gamma mode changed.
Hello Gerald! My experience is that the camera is still affected by the codec used, which means there is variance in H / HS, and which mode is selected. perhaps take the A6700 as its the most sensative and test various codecs?
A couple things I didn’t hear you say, was the screen open off the back in the same amount each time? And did you have an external temperature sensor on the cameras to determine if the camera was overheating or the sensor was incorrectly determining the camera was overheating?
The charging factor might play some role but it seems this is a reflection of some other random process running within the camera that is competing for CPU with this streaming / mirroring process. If that process is running at the time you start the live stream, the two compete for CPU and keep the temp elevated until it shuts down. It could be a function of some other menu mode that was invoked just prior to starting the stream. If the physical camera CAN run for 8 hours, the core process and hardware support it clearly CAN operate efficiently without thermal problems. The "camera OS" has some other problem the maker needs to fix.
I wonder at this point whether the camera's decision to oversample or pixel-bin is random, based on the initial conditions when streaming starts 🤔
did you try putting a computer fan on the a6700? I have an a6500 and when i need to do a long recording i put a 12cm computer fan behind it and it never get the temperature warning, tried in a 12 hours recording and it just worked fine, 12 hours in one file with the camera plugged on a usb charger and the fan plugged in a 12v power supply.
I use $10 HDMIUSB capture device from Amazon, and my Sony A7IV never even shown high temp sign ... I used it for some online interviews (job interviews), so it was quite important to not overheat :D. Longest was ~2h - zero problems. Oh and powered via USB, so guess additional heat source.
I think this issue isn't that tests are dumb -- it's that the cameras aren't engineered in a way that delivers relatively stable, consistent performance (and on top of that, doesn't give enough feedback about its current status). If you know the camera will overheat "sometimes" in situations where you need it to function reliably, it really doesn't matter if it's one time in five or one time in 10. It's not an option for your work.
With a battery adapter I've streamed for 12 hours at a time at 4k 30 with the GH5 at several events and never had an issue.
stop flexing! 😅
@@Ra-Hul-K Bahaha. I know I need to upgrade. shooting with the same camera for events 8 years now.
Switched from Canon to the ZV-E1 - can confirm my camera randomly overheats in no consistency
I've streamed with it once over an hour
Another time it made it about 40 minutes then overheated
Recording normal videos - shoot & stop - sometimes lasts over an hour, other times it's lasted 30 mins then overheats
All of mine is off battery though
Overheating is a design flaw but most people have come to accept and even defend it because "it's not a video camera" :)
The only thing I can recommend is to do more tests and plot them on a graph. I'd also do tests with the temperature maybe at 19C, as well as 25C. It may just be that it's on the fringe of overheating, but because the camera temp + room temp can fluctuate, including possible drafts, the point of overheating is basically tilting on the edge of a cliff.
When we think about the derivative (think acceleration, it is the derivative of velocity) of heat, something coming to a temperature is logarithmic. It will *start* heating up quickly, but as it approaches it's target temperature, it takes much longer to transition. Do *not* think of a heater or oven. Think of a computer coming to temp under a load test, or an object sitting in a room and coming to temperature.
It may be that the overheating point that is set in the camera's firmware is just barely on the edge of it's max physical/actual temperature when in that mode, so little fluctuations can cause it to randomly hit that temperature limit.
Besides that, there could be a dynamic nature to whatever encoding/processing it is doing. It could depend on what it is recording. Similar to how CRF rate control can dynamically change the bitrate based on complexity, and the encoder requires more processing when it is a complex/fast scene.
Just to add my experience with the a6700: I’ve consistently gotten an hour and a half out of it in this mode before it overheats. ~75f, battery fully charged, display open, battery door closed.
It’s never overheated on me with a dummy battery, even after five hours. Everything else equal to above except “USB Charging” off; don’t know if that matters though.
First gen Ulanzi fan on low also lets you run indefinitely, even without the dummy battery, in my experience.
Interesting video Gerald. When I tested the webcam function on the X100 VI, it only output 1080 60p, no other resolution or frame rate was supported. It worked fine in UA-cam Live, but I wondered did you manage to get it to actually output 4k over USB?
It said it was doing 4K. I never thought to check thoroughly, but the menu said 4K, the frame size in OBS was full UHD screen. I'm not sure. Maybe it's a connection thing? On the Sony cameras if you don't use a high enough USB spec, they limit the output resolution. It could also be a mac vs PC thing. I'd have to dig into it further. The manual doesn't mention a resolution unfortunately.
@@geraldundone it was weird, regardless of what I set the camera to in the menus, it was outputting 1080 60p, when I connected it to my Mac. Still worked well though. If OBS on yours was reporting a UHD input though, I would assume it was 4k it was receiving.
Thank you for your informative and entertaining videos! I really love them, they scratch both my photography and my science itch and it's the best!
I'm mindblown by this video. Thank you and congratulations for succeeding in making videos always so interesting 🙏
Seems like a great reason to visit LTT, again, and do some deep tests! More 'pointless' videos, please.
I did a 4 hour shoot last weekend and my A7Cii never overheated. It did however, feel almost too hot to touch by the end
You should try the Sony FX30 that has a cooling fan built inside, i don't think that will overheat at all, you should test that one..
Can the FX30 do 4K 30FPS over usb?
@@tenpoundfortytwoonly 4K 15fps. The 6700 does 4K 30 over webcam
Gerald, appreciate the huge effort in trying to do this testing!! Does the Fuji X100VI output 4k30 in UVC Webcam mode?
It does, yes.
Good job! I don't trust these hybrid cameras for long shooting. 😅
I made a clip on fan back for my a7cII and used a usb-c thermal camera from InfiRay to monitor the temps. It didn't cool much in overall temps but with the fan on i was able to record 4K60 with usb PD unlimited. Without the fan it always shut off after ~90minutes pretty much all the time. I don't exactly remember it but overheating warning always came on at ~46°c and at just over 51°c it turned off. I should test it again
I think PotatoJet got some kind of egg incubator to test overheating. So he can setup desired temp. which is controlled via thermostat.
I like his incubator tests because it illustrates something of a "worst case scenario", essentially if I know some camera won't overheat in the incubator at 100 degrees, it sure isn't going to overheat when I'm shooting some talking head stuff at 70 degrees.
That’s crazy… maybe a wine refrigerator or something could help keep it at some kind of room temperature… but even some refrigerators fluctuate in temperature as their compressors turns on and off.
Great job on the video Gerald! Thanks for making it.
Thanks for this throughly made tests that can show us that we could go slow on those test results
I love how visible the containment of your final expletive was throughout the video. 😂 But now we know and knowing is half the battle.
more like knowing half is half the battle
I work in a lab, and whenever we have such problems, my first question is to see the actual setup. Often one can get an idea what is going wrong by just seeing the setup and how the experiment is conducted. But having said that, even if we figure out how to measure those things consistently, that info won’t have any value for the customer in the real world.
AMEN - AMEN. Buy the camera to fit your needs. If you need to record for hours on end in 4K or 8K you need a high end camera that is made for that purpose. That $1,500 Sony/Cannon/Nikon camera is NOT designed to run for hours. The same goes for GoPro. Just go buy a camcorder to record your kids baseball game.
Thanks for the hard earned info ❤
I do a 2 hour livestream every week with a Fuji X-S10 (different channel). Gives me a warning after less than an hour but works fine for the full 2 hours, except in the summer when I have to turn the lcd display off otherwise it powers down before the 2 hours is up.
did you do all the tests with a fake battery? or charged the battery of the devices while the live-streaming function? i see the problem.
very eye opening results. Gerald is the best.
THANK YOU!! I've been screaming this in ZV-E1 all year
I used A7RIV and later A1 for home office conferences but I was using an HDMI cable connected to a PCIe capture card running 4k60, rather than internally encoded through USB. Some conferences lasted for nearly the full day, skipping 2 hours during the lunch time. Neither camera has ever overheated. I always have the screen stretched out for better cooling.
I'm wondering if the speed of your connection to whatever you're streaming could be a factor. Like if there were some additional fraction of a second delay down stream to the computer could that be causing the cameras bot buffer the data leaving the camera? I've had cameras that in the past have overheated when using slow SD cards that ran fine for hours when using fast SD cards and my assumption was that the extra buffering between the image generation and the writing to the card was resulting in the camera's processors running harder/getting warmer. You're talking about live streaming so in a perfect would there really shouldn't be any buffering in the camera, but in reality there is going to be some bottle neck (miliseconds) between the camera and the computer and then between the computer and the internet. I'm not an expert but it's conceivable (although I don't see why it would work this way really) a bottle neck on your internet connection could translate all the way back up stream to the camera's processors.
Thanks for doing this, Gerald!! The overheating really does vary stream to stream. My Ulanzi fan has been working great for my Sony a6700 and ZV-E1 so far. No overheat warnings after 1.5 hrs 4k30, but who knows - might be totally different next time 🤣
my man Mark Bennett 🖖
Great video, thank you for explaining all of this to us! Many of us that are asking for overheating tests are from places like Texas, and we’re going to be taking a black camera outside to film for hours in the hot sun if we’re doing weddings and things like that.
I’ve heard some UA-camrs even kind of berate people for asking for overheating information, but if we are purchasing a workhorse camera, and it completely becomes a useless brick (like the R5 when it overheats) it could cost a lot of money in botched work. That said, what this video shows is that we may not even be able to get that information reliably 🤷🏼
I found this video amusing because I think it may be the most frustrated I’ve ever seen you on camera 😂
Thank you for always producing what we need to know!
I don’t know if I’ve ever gotten more than 45 minutes using the a7iv as a webcam over USB. Even with a dummy battery.
I always have it set on 1080 (30p). Is the oversampling the issue there?
It must be related to the double slit experiment somehow. Like, “cameras only overheat when you need them to work”. These tests are like watching the photons with one eye closed. 😂
I had this experience on the ZV-E1, 7.5 hours before I got bored in a 26c ambient temperature room, then in a 21c room a few months later it overheated in 45 minutes.... why??
Thanks for sharing it. ✌️
You sure the fuji webcam on latest cameras is 4k? Cause my xpro3 and xt4 only do 1080p and it's really bad quality.
Stick it in the oven? 😅 This is one of the main reasons I still opt for Panasonic over Sony, the active cooling is a real weight off when shooting long videos. I expect other brands will follow the active cooling route.
Where can I get that t-shirt?
BTW, My S5IIX never overheated on my while "streaming" but the HDMI lag is too big for my taste so I tent do use my A7IV for calls.
It's always funny to see "professional' IT companies on calls which look like they are streaming from the dark basement vs my YT studio look with natural bokeh :D
Because they are professional it companies not OF creators
@@coin777 I'm no OF creator either, but got few cameras and nice light. And TBH, they just need 1 small light to look better on teams/zoom calls even with their shitty web cameras. But they tend to lit their faces with monitor corpse like glow :D
@@JoATTech I think you got the joke, but still responded seriously, amaright?
@@RomanNuss-pj3od Yep :D
But jokes aside, I cannot grasp why CTO of 2000 ppl account joins the call on video and looks like he's a zombie calling from his crypt.
Just had my a7r5 overheat on me several times on my shoot yesterday. Recording internally and also sending feed to atomos. Ended up dropping down to 4K and just recording on the atomos. ugh.
I’m curious as to what your setup was. Was it powered internally or externally? Also, did you follow Gerald’s recommendation to pull the screen away? Even when I do these things, I generally only get about an hour to an hour and a half max of continuous 8k on my Alpha 1
@@feederbrian9457 First I forgot to pull the screen out...I had internal battery and was powering via usb. I swear the house I was recording in was like 80 degrees (My glasses fogged up when I went inside and I can't remember a single time that's happened to me before in Seattle Washington, lol). I think I'll just use 4K moving forward. I don't need the extra pixels and that way I can get 10bit 422 instead of 420 anyway.
I always use HDMI out for video at 4k60fps and USB for charging. Never had the A7IV overheat for stream (up to 8 hours) but know someone who has.
I tried the ZV-E1 and it overheated twice within 12 minutes of a stream.
I now have the A7CII and tested it over 7 hours with no overheating but was warm to touch.
What about with a fan attached?
Canon EOS webcam utility also allows plug and play webcam features. It even supports super old cameras like my Canon 600D/T3i. Simply plug and play webcam via USB
😂😂😂 thank you Gerald! This was very informative! Luckily my A7S3 lets me live stream and Google meet/Zoom with no issues. But then again I’m only doing it at 1080p 30fps 😜.
could it have corrolation to the brightness or no brightness of a specific number of pixels in some weird way the hardware reacts while oversampling and or binning. Im just brainstorming trying to come up with ideas where there could be varibility other then ambient.
I had my zv-e1 left on for 2 days after I concluded a podcast/live stream and the unit wouldn’t turn back on. I shipped it back to Sony under warranty and they said the motherboard fried. They replaced it thankfully but there is an issue with over heating. I had it connected via HDMI way 1080 30 fps and it had that result
This lines up with my experience using Sony A series cameras for long shoots. For streaming, it’s hard for me to recommend a camera without a fan and HDMI capture. I lean Panasonic but the FX and A7S lines are solid too
Honestly feel bad for so many streamers being convinced ZV and cheaper A series cameras were good for streaming. Should be a big up front disclaimer on (video centric) camera reviews IMO
One alternative thinking: I don’t care how LONG it COULD last, I care how SHORT it MIGHT last. The range in results is absurd, but it suggests that a bad day could result in blowing a stream or a shoot
I saw reviews of various Sony cameras and overheating issues. It significantly impacted my decision to get a Sony FX3 as I do primarily video work of live events and can’t call a stop for the camera to cool down. The Sony ZV-E1, FX30, and the A7 series were all considered, all had vague reports of overheating. I went with the safe, although more expensive route. It also means I still need to pickup an A7 of some sort for stills since I’m now investing in Sony glass. Overall I’m happy with the FX3, just ranting a bit.😀
I wish there were consistent/repeatable tests to feel out the limits. Can it do 2 hours without overheating in a typical indoor space? Does it only have issues in higher temp environments? What about in direct sunlight? I appreciate the rant on this subject. It needs more attention. Thank you
I just use the Elgato HDMI to USB adapter for my A7C. Longest stream at 1080p60fps was just over 6 hours, with simultaneous recording. Not even remotely overheating.
I know it's not 4K, but is this an outlier?!
People flamed Dan Watson years ago for his overheating tests in Florida, but he had a point. We shot a wedding last Summer in Charlotte, NC and the A7SIII overheated outdoors shooting 4K24p while in outdoors during the ceremony. No warning or anything. We had the FX6 running as well, which saved the ceremony footage. Sony has improved tremendously over the last 8 years, but there are still concerns in some situations.
Well that kinda sucks, was planning a usb livestream with a few of those cameras. Not in 4K but did want to use the reframe function. Seems like this wil not be a very reliable setup.
Did you use the exact same power source and SD card in all tests?
Have you considered trying Ulanzi's fan that can attach to the back of some of the Sony cameras? Wonder if it would give you more consistent results without some of the wild swings in run time.
I did it with my Fuji X-S20 in a torture test (35 degrees celsius, camera body under full sun), the Ulanzi fun solved overheating issues - in fact, the camera just got to the 2nd heat warning (the red one), but never shutdown until the SD Card was filled. Without the fan, the camera shutdown arounf 50 minutes, if I remember correclty.
Some other users reported that the generic Aliexpress fans gave the same results than Ulanzi's.
But as said, was a torture test. In real world, never used the fan.
@@MarcioK doesn't the Fuji X-S20 have an optional official Fujifilm fan accessory?
@@Ra-Hul-K It have, and as far as I know, it works very well, and is more practical than 3rd party fans. But is US$199,00, the Ulanzi is $39,95.
In my opinion, if you need the fan for professional use, for the price of the X-S20+ Fujifilm fan, for $350 more you buy the X-H2, a much superior camera...
This is pure comedy! 😂 Great video!