Watercooling a Canon R5 to enable unlimited 8K recording | Hacker News Reader

Watercooling a Canon R5 to enable unlimited 8K recording

petapixel.com

This reminds me of a loosely related tale.

In 2008 I was hired by a game studio to build a HW/SW system for synchronizing a Canon EOS 1D Mark III and a series of studio flashes. The idea was to use its ability to take ten pictures a second to take a rapid succession of images with very different lighting. The images were then processed by a series of tools I'd helped develop the previous year.

The timeline was pretty tight, so I had to begin development before I had the camera. I contacted Canon support to confirm that the 10 pics per second was only when given a continuous signal and to ask which speeds it could reach when taking separate pictures via remote trigger. Their support tech insisted it was irrelevant since I couldn't possibly press the button that fast. I tried, to no avail, to explain that it was a computer sending the signal and it could do so (with the hardware I had) a thousand times per second. Reluctantly, he forwarded the question to their technical support, who never answered.

I ended up using a continuous signal and having my program listen to the remote flash connector. It worked splendidly.

I’m curious to see what the visual effect ended up being. Any link to an example?

Sounds like a 3D capture rig than visual effects.

That's correct. I can't say too much about the details. It ended up working really well but wasn't used in production for reasons of internal politics.

The guy who had the original idea for the system, a former classmate of mine, later went on to start his own company, Quixel, based on similar ideas.

It sounds like it might have been involved in a process such as this one: https://mobile.twitter.com/AaronCovrett/status/1125409161808...

If you capture an object with light from various angles you can "re-light" it in post-processing. Drop it into different scenes and have realistic lighting coming from different angles. https://augmentedperception.github.io/therelightables/

Wah this is magic.

Paul Debevec has been working on this steadily since at least 1999. http://www.pauldebevec.com/FiatLux/

Light doesn't exist, it's made up.

I recall something like this was done for a presidential portrait a few years ago.

https://fstoppers.com/bts/how-smithsonian-created-insanely-a...

Sprite Lamp sounds like it may fit the bill as an example, as well:

https://www.youtube.com/watch?v=fy3FvPg01w0

Given the author's other post about a similar technique going on to be used at Quixel, presumably they were inferring normal maps from lighting data, and later metallic, AO, smoothness etc maps for PBR.

> Reluctantly, he forwarded the question to their technical support, who never answered.

Because they don't know off the top of their heads and would have to test it to be sure which is a huge amount of work they don't have to do.

>Because they don't know off the top of their heads and would have to test it to be sure which is a huge amount of work they don't have to do.

And that's called... poor customer service.

Had I been faced with that behavior and had the freedom to switch vendors as OP I would have -- meaning that support behavior isn't without flaw, they lose a certain threshold of customers.

This is just self-aggrandising. No mass production vendor would do R&D for your bizarre use case unless there's clear PR potential or you are ready to pick the tab.

Yeah I used to be in a weird position where most of my time was spent figuring out how to use and abuse a bunch of off-the-shelf things to build out one-offs of whatever crazy crap people could come up with. Lots of hardware and software reverse engineering and integration.

I’d never in a million years expect a company to do my R&D/integration work for me when I’m buying one mass-produced consumer item and using it outside of its intended use-case.

Not a single company is going to recreate your unsupported setup to see what would happen and let you know unless there was a huge payment being made. Since the only reason OP is doing this is to save time, it would make no sense to spend weeks going around to each vendor trying to work out which one to pick.

And often support people give wrong info, you may eventually find a vendor that just says "Sure you can do that" and then you get it and find out you cant.

You can still respond. Even if just to say "that's not a supported scenario and may void your warranty."

They literally did say that it wasn't possible to do what OP is asking using the provided interface. Its not the support persons job to explain what would happen if you hack the device to work differently.

The supplied interface was actually a connector for remote trigger, which is what I used. The only hack was sending the signal from a programmable device rather than a manual button press.

Any manufacturer that’s even remotely serious wouldn’t need to test their own product to find out how it works. Each component in the chain, from trigger button to storage I/O, had presumably been tested and its tolerances documented. Responding to such a query could (should?) be as easy as checking the specs. “In theory this camera can support up to N triggers per second, which is the limit for part X, but you will need your storage to sustain writes at speed S or the buffer will fill up, and by the way this wasn’t tested end-to-end and will void your warranty.”

It’s more likely they just wouldn’t tell for whatever reason. A mass manufacturer probably doesn’t want to give an impression that they encourage a way of using the units not explicitly mentioned in the manual to avoid any complaints/lawsuits. Also, some companies require tech support staff to respond to a certain minimum number of tickets per hour (effectively penalizing personnel for doing anything beyond provided checklists, which consulting part specs definitely would be).

> A mass manufacturer probably doesn’t want to give an impression that they encourage a way of using the units not explicitly mentioned in the manual to avoid any complaints/lawsuits.

That, or they may have a product line which is designed for that usage case, and they don't want to hurt its sales.

In that case they could have linked to that product, right?

maybe you don't want the two distinct customer sets to know about possibly similar products.

I suspect its 1000x easier to just wire an arduino to the trigger and have it pulse rapidly than it is to pour over spec sheets and only end up with a "It looks like it should work"

Are you saying that testing is a huge amount of work, and looking at specs is further 1000x harder?

Tech support staff might not have a test bench at hand at all times (and an Arduino unit out of the blue might’ve required a month to approve and purchase), but they don’t need any equipment to pull specs for a part from internal documentation, which support staff must be very adept at scanning by the nature of their job. I believe this is what tech support more or less for, in a better world.

They probably had an answer at their fingertips without any need for testing, but communicating it was against their/company’s interests.

I doubt a support person from a camera company even has the skills to understand the datasheets. 99.9% of their job is explaining things in the public manual and handling product returns. The only people with the skills to understand it would be the engineers and it would be a waste of time assigning them to investigate this unsupported hack.

I know some tech support folks and consulting docs and specs is more than comfortably within their reach. In fact, it’s what they do at better companies where management does not preclude them from doing their job properly and provides access to relevant data.

It's more likely still that the person answering emails didn't know the answer, and either was unmotivated or unable to put the question in front of someone who did know the answer and then follow up with them to respond.

This is when having contacts inside a company helps. Either be a big enough customer to warrant their attention by default, or make friends with the company rep when she calls or stops by. If you aren't a large enough customer to have either of those options available to you, then you get the attention they feel you are worth as an end user.

Wouldn't it of been easier to just buy the camera and send it back if it didn't work ?

Cost isn't the issue in their comment. Time is. By the time you test it its already your time used.

How exactly would Cannon of tested this for them. It's a odd use case with a custom setup

Title is a bit misleading - while he did attach watercooler to R5 to unlock unlimited 8K, later in the video he achieved the same thing by using custom internal copper heatspreader and small fan attached to bottom of the camera body in 3D printed enclosure - that is much more useful and less "gimmicky" solution.

Even with just custom heatspreader and without fan he was getting 40 minutes of recording, then had to take 5 minutes of break and could record another 40 minutes, as opposed to 25 minutes with hour long breaks before modification.

I think he probably went for watercooling first because it has the promise of silence. A fan on a video rig sounds horrible, and probably sounds horrible too. :)

It's funny how all the pc watercooling awareness is spreading. There are lots of videos about the controversial process of applying thermal paste.

Fan on video rig is a fairly standard solution, used by both Blackmagic and RED cameras.

As far as I know, fans like that are used to run between takes, not during them.

Cinema cameras are also very rarely used very close to an omnidirectional microphone—a slightly audible fan would not be picked up by the external lav and shotgun mics recording the talent / action.

Some people recording with the R5 or other more compact systems might rely on internal mics or a shotgun mounted atop the camera for audio, and that definitely picks up sound emanating from the body, at least a little.

Many lenses' focus motors are loud enough to be annoying in onboard video.

"The promise of silence" is well said. It may be better than tiny fans, but for pcs you basically add pump noise and move the fan's location a bit.

You can obtain the same effect with 12cm+ fans in a well designed case without the risk of leaking.

A good D5 will emit hardly any noise when running at 30-50%, and you can have 140mm fans on 140mm wide rads running very slowly (400-600rpm) and keep everything very cool.

I've done both air only, with high quality fans, and now have. custom loop - the custom loop is far quieter, even at load. And it was fun to do.

PCs also live in cases, and can be moved a non-trivial distance away from the peripherals they drive. Not so much for cameras, whose placement is driven by other factors

Microphones can be moved away from the camera.

Very true, though a bit more difficult for a one-person team if they need to move around.

Well for filming people you can use a wireless lav.

I wonder if the processor (which was the hotspot) couldn't be moved from the rest of the camera though.

Water cooling gives you two things. One, huge thermal capacity for absorbing short bursts of massive heat. Two, the great flexibility of not hanging the radiator directly off of the heat source – which means you can spread the heat across (many) long radiators and they can be cooled by fans running at much lower RPM (that IS quieter than running 1-2 fans at high RPM which is what you get on a tower cooler), you can share the cooling capacity between CPU and GPU, you can have lots more capacity (and without putting stress on PCBs).

I'm not talking about 12 cm fans on the cpu radiator, I'm talking about large fans on the case.

But the liquid cools the CPU/GPU. The bottleneck is there, and you can't overcome it with case fans.

Hell we are now seeing water cooled 3D printer kits.

I think he went with watercooling first to simplify his testing and make sure that his fixes weren't failing due to inadequate cooling power.

Many cinema cameras have fans in them.

The hacked-together watercooling was the experiment that proved that a better heat solution would lead to unlimited filming.

Active cooling in handheld devices is really underrated.

I understand not having it if they're water proof but most aren't.

A big issue is most handheld devices are already packed to the brim with battery filling most of the space left over after the electronics are laid out so there's no space left for meaningful active cooling. There was a teardown of a Red camera on Linus techtips to add watercooling and the contortions required to pack everything in with the cooling was amazing.

Not sure how active cooling plays with image stabilization but my guess would be that not very well - or rather it takes quite some engineering to make it play nicely. Also battery life...

Canon responded to the controversy by pushing out a firmware update that changes how the R5 monitors internal temperatures.

-his R5 suddenly seemingly had no recording time limit

This is the most impressive bit to me. If this were something like a smartphone I would expect the update to remove the ability to reset the timer, not change the whole scheme to be more user-friendly. It's refreshing to see a manufacturer who doesn't respond to hobbyists with outright hostility.

The R5 is a pro-tier camera, and Canon established its market-leading position through good technology and exceptional customer service. They would be throwing away a stellar reputation for little gain if they didn’t try to make the camera work the best that it possibly can within its limitations.

It's an important distinction that is often obscured by the number of products that aren't professional equipment/software, and don't have this kind of attitude, but affix the word "Pro" to their consumer product line for marketing reasons.

"Matt also believes that Canon could have designed the camera to dump the heat toward the bottom of the camera and then sold a separate base attachment for videographers that provides vastly extended recording times."

Seriously, this. I wonder if they could make the area around threaded screw on the bottom for attaching to tripods a heat spreader. These are usually made out of metal already, and have to be structurally sound for the camera to be attached there. If the attached base was a block of aluminum or copper, it could probably suck the heat out of there.

I had the same thought at the first picture. There's already a great big slab of metal visible at the base of the camera with what looks like an easy access path from the processor. If these easy DIY heat sinks catch on, someone's bound to make a longer heat sink, bend the end over in a vise, and tuck it in alongside that existing mass of metal to incorporate it into the heat sink mass. The copper alone doubled the time to overheat, adding that much mass should do substantially more, and almost certainly provide a heat path to tripod mounts.

Whenever I read something like this, I think how despite the rise of the smartphone camera that's good for 99.9% of people, it still seems like there's still a lot of opportunity to improve traditional point and shoot digital cameras. It almost makes you wish that some of these companies (Canon, Sony, etc) put some more research into some sort of skunkworks that could completely reimagine the form of the camera in a way that would greatly benefit the user.

In the end, it comes down to the optics.

The (incredible) tiny sensors in smartphones aren't big enough to get past a certain fidelity. Combined with very small working ranges on the lenses and the limits are pretty hard. Hence the three lens set-ups now on the newer models.

Granted, the optics here are just the coolest things. 'Digital' lenses are some very slick stuff and the engineering is the best you can get.

But in the end, you just need longer working distances and bigger lenses. The optical physics are just the laws you have to deal with, same as Newton had to. No amount of engineering can get past them.

One thing they could do is to put a 'periscope'-like mount on the side, have the sensors slide up and down inside, and slot in a few lenses based on the optics. But the mechanics and likely slow speed of such a device would wear-out very fast and would be very susceptible to drops and dings.

Eh. Modern phone cameras produce a kind of synthesized picture composed by machine learning, based on what you point at. That seems to be the direction of travel. E.g. super-resolution based on multiple samples over time and integrated backwards to eliminate blur.

I honestly want the large camera manufacturers to start tapping into the computational photography realm. That tech, with their amazing sensors and glass, would produce some amazing results.

There was a post here on HN a few months ago where someone went into how the camera manufacturers are going to be challenged by phone manufacturers coming up with better and better algorithms to enhance the quality of pictures that are bumping up against physical limits such as lens size and pixel density.

Edit: Here's [0] the link to the article.

[0] https://photographylife.com/smartphone-vs-camera-industry

It doesn't have to be in the camera. You can do that in post with the raw files.

Generally no, because a RAW file records only the total amount of exposure per "pixel".

Full computational photography requires exposure per pixel over time (so something like a "stacked RAW" needs to be invented) combined with synced data from accelerometer/gyroscope/etc. for deconvolution.

I honestly don't. I wish they'd spend less time developing increasingly bloated and buggy camera software features and more time on simple improvements ( like exposure bracketing using ISO instead of just shutter speed )

It's still a '99% of people' solution.

For the remaining people that really do want/need that last 1%, ML doesn't work. You can't infer data that does not exist no matter how many samples you take. The resolution of every object is set by the optical resolution: a function of the wavelength and the NA of the lens [0].

Aside: super-resolution refers to nanoscopes that break the Abbe' diffraction limit of light [1]. The various set-ups do this in a few very clever ways so the specifics change exactly how well they resolve objects.

[0] https://en.wikipedia.org/wiki/Optical_resolution

[1] https://www.microscopyu.com/techniques/super-resolution/the-... (a good intro to these nanoscopes)

I can't see any major phone manufacturer doing that, although it sounds really cool. Too much room for error in both the manufacturing and user stages.

Agreed. Positioning optical devices takes very tight mechanical control and a good optical feedback mechanism. I don't think they will be out on the consumer market any time soon. But then again, I thought the same about these tiny sensors and the digital lenses, so who knows.

The problem I have with all the lens add-on stuff for phones is that you still have the tiny sensor and now you've complicated the camera that you just have with you all the time. When I want something more than a smartphone, I have an APS-C mirrorless camera that fits in my pocket (with a 40mm equiv. lens). And is pretty compact otherwise for mid zoom ranges.

I doubt they would just dump heat to the threaded base for a few reasons:

1) Heated metal expands, and different heated metals expand at different rates, changing tolerances for threaded mounts and possibly causing thermal locking and/or looseness.

2) Managing heat over time is about diffusing the dangerously high temps (e.g. near boiling) to areas large enough stay safe for human contact while also being exposed to air to shed that heat. A pinch point threaded mount under a rubberized/plastic body will absorb thermal shocks (quick use) but will equalize to very close to the junction temperature of the processor over time, aided only slightly by the other parts of the camera that are poor thermal conductors. Basically, that spot would become a burn risk.

See ASTM C1055 - https://www.astm.org/Standards/C1055.htm

The short version is that you need to keep things under 60C/140F if there's a chance a person could touch it.

Exactly right. It is important to understand why things are designed the way they are before we go about challenging established engineering principles.

https://www.electronics-cooling.com/2016/09/surface-temperat...

When using an Insta360 One X with a monopod for long video shooting, the threaded base gets extremely hot. The rest of the camera does not seem to get hot and I suspected that they were intentionally cooling that way. The accessories tend to be metal as well.

The R5’s whole body is magnesium already.

It's clearly coated in something and is not just magnesium.

I own an EOS R which is built similarly. It’s painted but it’s all metal, aside from the rubbery bits which I think are mounted on plastic over the metal. You can feel it get cold in the winter: https://www.photographyblog.com/news/take_a_look_inside_the_...

>body is magnesium

It's an alloy with aluminum; along with PC GF-30 parts.[0]:

[0]: https://www.canon-europe.com/cameras/eos-r5/specifications/

That’s right, but the vast majority of it is metal.

Yeah but the thermal expansion/contraction between plastic and composite (PC+GF) would not be welcome. Flip note: glass fiber reinforced polyamide(nylon) and polycarbonate have similar strength to aluminum.

Another issue would be use of adhesive that generally dislike heat even more than the electronics do.

I don’t think the composite parts are structural, but I could be wrong. Either way how come the body doesn’t already dissipate heat? What am I missing? Are the thermal pads not connected to it?

At least some such cameras were limited to less than 30 minutes, because EU tariffs classified something that could record for more than 30 minutes as a video camera, which had a higher import duty. Source: https://www.borrowlenses.com/blog/video-recording-limits-in-...

Having the camera overheat seems like a clumsy way to do it, though. The time limit would depend on ambient temperature and manufacturing variance in the heat sink.

This limitation applies to my Sony A6000 too, thankfully there is a software hack to unlock an unlimited recording time.

https://github.com/ma1co/OpenMemories-Tweak

That EU tariff was scrapped some time in around July 2019. The R5 was never sold with such a recording limit (4k 30fps has always been unlimited even with the overheating issues and old firmware). Some new cameras are still being sold with the 30 minute limit so you have to watch out for that if you are looking at another camera.

This is not specific to Canon by the way, even smaller cameras like the newest RX-100s from Sony have limitations on taking high resolution videos because of overheating.

Wonder if you have a link to this. I've tried researching it before, and all I could find was internal discussions of the intention to remove the tariff, no codification into legislation.

Yeah actually did look into it a few days ago and couldn't find a definitive source on the current tarrifs. The best I could find was actually the article to borrowlenses that the commenter linked above which mentions that "This is all supposedly going to get phased out starting in July of 2019".

Here's my previous post about it.

https://news.ycombinator.com/item?id=23436809

The legislation outlining the tariffs in its version from 2006 [0, p. 577] lists the import tariff for a “video camera recorder” at 14%, whereas the version from last year [1, p. 601] lists it as “Free”. This seems to be the result of a WTO agreement [2].

[0] https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=uris...

[1] https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELE...

[2] https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=uriserv%...

Not true, the r5 and r6 come with a 30min hard limit despite tariffs no longer being relevant. This is besides the overheating issue so when it you manage for it to not overheat, it will stop at 29:99

Actual hacker news.

This is really interesting though, in the context of Camera companies like Canon having to adapt to the digital age, and users expectations of what digital devices should be capable of.

Canon is well adapted to digital age. They simply sell much much more expensive professional gear that can do the unlimited 8K recording. If they were to add this feature to R5 out of the box, they could see reduced pro gear sales.

They don't, actually. The R5 is the first, and currently only, Canon camera that records in 8k. They do have a line of cinema cameras, but all max out at 4K currently. Likely because they use the smaller Super 35mm sensor to match the film size film folks are familiar with.

This whole human malware business meant getting decent webcams and Camlinks (HDMI capture devices) extremely difficult and/or expensive for a while.

Logitech C920 were particularly difficult and expensive to get, which long lead times.

So in April, Canon came out with a utility enabling use of many EOS (and three Powershot) cameras as high quality USB webcams.

Despite this being uncommon in consumer video cameras, it's pretty standard to see this in long-exposure astrophotography setups - not to prevent overheating, but to reduce sensor noise. Some setups use Peltier or even cryocoolers to get to very low temperatures, but plenty of commercial CCD modules have water-cooling loops built in, eg https://www.atik-cameras.com/product/atik-11000/

How do you stop condensation on the sensor? Sealed chambers?

That particular model says it uses heater elements to prevent condensation on the outside glass, and a high-efficiency desiccant to prevent condensation internally on the sensor itself.

Yes, I work with CCD and SCMOS cameras and many of our sensors are cooled via Peltier cooling in a vacuum sealed chamber[1]. I have seen condensation on the sensor window due to vacuum failure before. [1] https://andor.oxinst.com/learning/view/article/ultravac

About how much does a CCD like that cost?

He is not cooling the sensor though. He is cooling the CPU.

The guys a linustechtips watercooled one of their Red Helium 8k Cameras. And the actually did cool the sensor itself and not the processor.

Its a video series but the final assembly and test is shown here: https://www.youtube.com/watch?v=OEUCNh5g-2I

With that name, they should have liquid-helium-cooled them.

IIRC, they had to be careful not to _overcool_ the sensor because the sensors have an acceptable range in which they work.

Anyway, the entire series of videos is pretty fun to watch.

Yes, too low a temperature also messes with electrical properties. And makes everything very brittle, too.

A lot of the scientific sensors are Peltier cooled to -20C to reduce noise, particularly IR.

I have a friend who does amateur astrophotography and talks a lot about modifying DSLRs to cool the sensor and things like that. I never listen very carefully because I'm never going to devote the time and money the hobby requires, but I take it there's a dedicated community that does this kind of thing.

For anyone interested, the astrophotography guys on centralds.net have been doing similar things for a while now. If you're looking to accomplish something similar, they might be able to help you out.

Additionally, there are lots of people doing similar diy mods to DSLR cameras on youtube for astrophotography, usually with sub-zero cooling capabilities.

I am generally not of a conspiratorial mindset, but in the very particular case of these photography-centric cameras, I can’t help but wonder if these video features are purposefully subpar so as to not steal market share away from their more expensive cinema cameras.

The real reason is probably much more along the lines of “oh, that’ll be a nice bullet point to have on the marketing site, let’s throw it in and not really test it that much in the field”, but still, leaves a bad taste in the customer’s mouth.

Apparently at least some of the limitations are actually due to an EU import tax on video recorders but not on cameras. [1] Though this also says that that will be fixed in 2019, so I don't know what the current situation is.

[1] https://www.borrowlenses.com/blog/video-recording-limits-in-...

Manufacturers usually just disable or limit features to get around this in markets like the EU.

For example you can not sell a new device that has FM only radio without support for DAB. So many manufacturers of mp3 players etc. disable the FM radio feature firmware for those markets.

That’s well known, and not what I am thinking about. More like…

• Dumb frame rates.

• Lack of good flat color profiles.

• Lack of clean HDMI output.

• Dumb codecs.

There are so many things about a photo camera that can record video that makes a a workflow PITA for video only shoots. The small form factor alone makes these cameras difficult. There's no rooms for buttons to easily adjust settings. You get at best a single micro-HDMI port. Yes, it is getting better with all of the 3rd party devices that have been created to specifically address the fact that the form factor is so small it would be impossible to use without them. These cameras and their smaller form factor do lend to some very interesting things though, so please don't think I'm crapping on them. Their prices also make them nearly disposable for decent budget shoots meaning much more interesting shots can be captured.

TL;DR Big video cameras have their places that smaller photo-centric cameras can't replace.

All good points.

It just feels like all this video stuff…I don’t know, since the 5DmkII came out, I can’t figure out the marketing strategy. You’ve got big-sensor beautiful video, and then a whole bunch of weird limitations that make them a pain in the ass to use. (Going back to that camera: the literally exactly 30-frames-per-second frame rate.)

Video feels like a thing they put in there because it sounds good to influencers, but then it’s all half-baked and everybody ends up getting proper video cameras.

My impression, which could be wrong, is that the strong videographer uptake caught Canon somewhat by surprise. After that, depending upon your level of cynicism, they either decided to remain focused on still photographers and everyone else was incidental. Or they decided to deliberately nudge serious film people to their videocams. The EU tariff situation further complicated matters.

That’s about what I hear. The video on the 5DmkII was apparently something they popped in without much thinking. Like, just capture the Live View video and dump it to the card.

Professional cinema cameras have totally different interfaces and requirements optimized for what video needs.

Photographers and videographers have totally different priorities. Even if the sensor and lens is the same, everything else about it is different.

Zero conspiracy needed. Just meeting what the market wants.

When is Apple going to buy one of these struggling camera makers? They could pick up Olympus for nothing and ship a camera with a real CPU in it, and maybe even software that doesn't make you cry.

they -really- want to push iPhone (particularly Pro + Pro Max) as their camera. it even has Wi-Fi, mobile data and iMovie, so you can shoot video, edit it wherever you are, then push to Instagram/YouTube/whatever else.

i say when pigs fly. (sadly.)

They really want to push AirPods too (to the point where even the iPads no longer have a headphone jack, which is absurd), but they also have Beats, because sometimes you need/want more sound than wee little earbuds can give you.

Doing something analogous with cameras isn't an impossibility. It's not even going to cannibalize iPhone sales; the opposite if anything, although I'd imagine most people who are serious about shooting video on their phone have a 12 Pro in their pocket already.

But mostly, it would give an obvious upgrade path to get into cinema-quality and real lenses, for customers who cut their teeth on what the iPhone has to offer.

It's a natural fit for Apple, although I think I understand why they haven't: it's hard for that company to get out of bed for less than a million units sold, of anything.

From your mouth to Tim's ear!

I picked up a Sony this year, and I'm very pleased with the quality of the photos and video I'm able to achieve.

The software is a constant, grating annoyance. It feels like being transported 15 years into the past, and (this one is Apple's fault), I have to pick my video format carefully if I want to import onto my iPad, which generally, I do.

Everything, from the menu, to importing, remote control, all of it: it just doesn't work great, the ergonomics are terrible.

Great camera though. Just a terrible computer.

Obviously we can only speculate here but I'm guessing there isn't a big enough market they could actually target. The average person will never carry a dedicated camera again and the average company won't be satisfied with something apple would produce so the only market they could sell to is youtubers which are probably fairly satisfied with what they have now.

I have a sony camera and I find the UI and physical interface to be pretty good. What its lacking for me is some of the more advanced features like video stabilization (on their lower models) 4k 60 (also on their >$2000 models) and long exposure photos with shake reduction so you can take it from your hands.

The average person doesn’t need Final Cut either, but Apple participates in that market for some reason.

Youtube + Vimeo alone is a huge market.

Apple made R&D facility on Tsunashima, Japan and they hiring engineers from camera industry.

I too wish for more smartphone makers to get into the discrete camera business.

Unfortunately, it is not a growing market segment. [1] Sales are down 54% YoY in 2020 (and were well in decline before COVID). [2]

Samsung used to have their own line of discrete cameras, with quite a unique feature set for their time, but these were discontinued in 2015. [3]

15 million units a year is not a large enough market for any new entrant to bother investing the R&D. They would then have to license an existing lens mount or design their own, and photographers with thousands of dollars in lenses are very conservative about moving to a new system. Third party lens makers are also slow to redesign their products to support a new system.

[1] https://www.statista.com/chart/5782/digital-camera-shipments...

[2] https://petapixel.com/2020/11/06/camera-sales-in-2020-have-p...

[3] https://en.wikipedia.org/wiki/Samsung_NX_series

The original video:

https://www.youtube.com/watch?v=X1u-9YqrIJc

This is the same chap who built Archimede's death ray in his bedroom. That video was even more impressive. https://www.youtube.com/watch?v=6bqBsHSwPgw

I have the compact g7x mark iii, and its terrible at overheating when shooting 4k. After 4-7 minutes of shooting 4k 24p, it overheats. Its terrible at autofocus, and exposure too. And its not even cheap, $700 latest Canon compact camera. They need to do something about the overheating issues, because much smaller phones don't have these problems. And apple can now do 4k 60p, and colors are pretty good.

This was super interesting. I've torn apart some video cards and laptops (Mac/Intel) to redo the cooling assembly. Just simply applying better thermal paste makes things run a lot cooler. This example of a Camera was awesome and a great hack to unlock some serious value.

>Canon responded to the controversy by pushing out a firmware update that changes how the R5 monitors internal temperatures.

This is the most spectacular part of the story. No marketing department could have come up with a better plot.

I've been using my mirrorless camera as a webcam while working from home and it overheats after a while. I solved the issue by laying an ice pack on top of it before long meetings.

If I were Canon, I would make this guy a generous job offer to integrate a better stock heat spreading solution and optional external cooling into future camera models.

Why? He’s just hacked on an external cooling system that is in no way meeting the basic quality requirements of a production device. For that you’d want to hire a heat transfer engineer with experience of designing small-sized passive solutions, and with both a valid work visa for Japan and fluency in Japanese.

Kudos to the guy for debugging his particular problem, but that doesn’t make a product.

His practical business solution to it is quite sensible and likely overlooked: "Matt also believes that Canon could have designed the camera to dump the heat toward the bottom of the camera and then sold a separate base attachment for videographers that provides vastly extended recording times."

In the video he more specifically describes how it could be possible to use the mounting point's metal base to transmit heat to a mounted cooling attachment.

Its quite clear that canon engineeres or managers are aware how to build it like this.

I haven't even started watching this video and had the same thought of transporting the heat through the bottom.

Whatever made canon made it like this, was on purpose.

Either the concentrated on picture for the most time and realized to late and it would have required rebuilding core parts and not getting it out in time or they will bring out a camera/pro version to this with unlimited 8k recording and with a higher price.

The assumption that a hobby engineer can solve this issue while canon can't, is not realistic.

Based alone on the original firmware approach, i would argue that time or priority was the issue.

Yeah, no way they overlooked something like that. I had my Sony α6100 overheat within 20 minutes on mere 1080p video when the ambient temperature was at least 45°C. When these things run hot, it’s easy to feel where the hottest areas of the body are, and attaching a heatsink is obvious, and at that time, dumping heat to the tripod mount was the first thing that occurred to me when I contemplated how I might do it.

The flip side of this is that Sony (finally) made an APS-C mirrorless, in the α6600, which has a recording time that's only limited by power or storage.

So it was never impossible, and it isn't just a Nerf to segment customers into the cinema bracket. It's just an area where they're willing to compromise.

Without IBIS, the αN < 6500s aren't great for many video applications, they're photo cameras which can also take some video. I suspect that Sony just didn't realize the degree to which the YouTube generation would be using their entry-level mirrorless systems for extensive video. Consider that it took them a generation to make flip screens standard, and that they bumped the battery size and ability to record indefinitely on (only) the model which has IBIS: that's a nudge, saying "If you want to record a lot of video, the α6600 is the model you want".

It was the α6000, α6300 and α6500 that were time-limited, because they generated more heat in their video encoding. The newer α6100, α6400 and α6600 don’t run so hot, using a newer and better chip for the encoding. You’re wrong about the ability to record indefinitely: this is not specific to the α6600; all three of the newer generation get it. I’m recording 100 minute videos weekly on my α6100 with no sweat now (plugged in by USB, otherwise the battery will be down to 5–10% by this time—with the camera fixed in place for these recordings I also don’t care about image stabilisation, whether in-lens or in-body), it was only having trouble in the middle of summer when the ambient temperature was at least 45°C.

Alright, good to know.

I think that's more evidence that the overheating problem wasn't a deliberate attempt to segment customers into the cinema bracket, but rather something Sony didn't realize would be such a problem, because who is going to want to film 100 minutes on their photo camera? Practically everyone as it turns out.

Yeah, they could have and should have added better heat dissipation: but they can and should make (much) better software as well, and don't, because... it's Sony.

It doesn't have to be the engineers who overlooked it, I am sure there were a lot of great cooling solutions thought of once the overheating issues were discovered.

However, the management did overlook it. Now the machine overheats and requires hours of cooling down to shoot continuously for more than 20-25 minutes. Is that an acceptable amount considering any kind of film work?

Its not build as a film camera and its not a film camera. Its a picture camera.

The market segmentation is totally different for that.

I would still buy it. I have a Canon 80d and 90d and i rarely do any video recording at all. why? Because just a recording ready body doesn't give you good video. You need audio equipment etc. and 8k video is huge like wtf huge.

Canon has better heat spreading already - you just have to buy the more expensive camera models ;)

Really? Thats interesting, have you got a link?

Cinema cameras: https://www.usa.canon.com/internet/portal/us/home/products/l...

SLRs are optimized for handheld, photography use. The body just doesn't have the surface area for proper long-term cooling, and even if it did, it could be difficult to get it to work well while maintaining water resistance and retaining the compact grip.

The r5 is a mirrorless design, not a (d)slr. Sony sells plenty of video optimized cameras in both mirrorless and DSLR designs, in particular the a7s line are explicitly video cameras which can take stills. They can overheat eventually in very hot weather, but it takes some doing.

None of them do 8K raw though.

More pixels more better? I'll take 4k raw at 16-bit with larger pixels over 8k 12-bit and needing more light. To each their own.

Same, but the R5 will do 4K 30p cropped for as long as you want, or 1080 60p. As will my older EOS R. The part that’s unique and that involves heat issues is 8K recording. We’re in the megapixel war again, this time between Canon and Sony, which is the only reason anyone’s complaining about 8K video performance in the first place.

I believe it was mentioned in the video that scaling 8K to 4K gives better quality (maybe by averaging out the color mosaic filter pattern?).

Canon has the EOS C70 that is more or less what you called a mirrorless design. Its just recording 4k and has a smaller sensor. So its not that canon isn't able to come up with a camera like the Alpha 7S series. And the Alpha 7s is also not able to record in 8k.

It’s much larger and I believe it has a fan, though.

It's absolutely clear that Canon made the camera work like this deliberately, to not canibalize their line of cinema cameras with a body intended for stills that is half the price. So yeah, it's a nice hack to force the design to deliver more than the manufacturer allowed it, but don't expect them to be happy about it.

The fact that the two thermal pads only partially cover the heat-spreader on the main processor is pretty damning evidence against Canon's claims that they did not deliberately cripple the camera's video functionality. It seems they want to encourage film makers to buy their more expensive cinema line of cameras.

Of course, this has been their MO since DSLRs started being able to compete with dedicated camcorders in video quality. Nikon is an even worse offender.

It will still record 4K 30p just fine for as long as you want.

Nikon's poor support for video (even though being one of the first to market) boggles the mind. They don't even have professional video cameras that could be cannibalized, I don't think they even make camcorders.

That seems like a pretty strong evidence against this theory?

I don’t think it’s intentional that not all DSLRs don’t also make great video cameras. The requirements are very different for great stills cameras vs great video cameras. Size and heat dissipation are the most significant.

It might not be intentional, but it sure doesn't hurt them. Buy a DSLR for photos, even if it can do great video, just get a separate camera for video.

Probably the same reason laptop manufacturers design their cooling to be barely adequate. It would cost less than a dollar to make it dissipate the maximum TDP at full load (not just "normal" load).

To make it easy, let's ignore ultrathin laptops - even mobile workstations have had the same terrible heat dissipation for over a decade, with very few exceptions.

> It would cost less than a dollar to make it dissipate the maximum TDP at full load

How do you figure?

> It might not be intentional, but it sure doesn't hurt them.

I just explained why video cameras have different requirements than DSLRs. Besides which, 4K30 or 1080p60 is plenty of unlimited video capability for most consumers. The camera can even do 8K30 and 4K60 for longer than most people take videos.

I don't have any skin in this game, I'm just sharing my thoughts. Besides, in Europe they're limited to 30 minutes of recording (otherwise they're classed and taxed differently iirc).

The laptop thing - using a bigger heatsink, or more copper, or better thermal paste, or multiple heatpipes instead of one (that was actually HP's solution to their overheating ZBooks) would cost pennies and would actually keep the CPU/GPU under throttling limits.

I've repasted and modified enough laptop heatsinks to know it's a very easy thing to do. But the manufacturers don't care. I get it, a dollar saved on one machine is a lot of money overall. I'd rather have it cost $10 more but have a decent heatsink.

Seems more likely a sign of being assembled quickly on an assembly line and not having been designed with the thermal mass to properly dissipate the heat of the device recording in 8K raw. This is primarily a stills device, after all.

Canon and Nikon have had these "issues" for over a decade. In markets where they're not limited to 30 minutes of video recording by law, they would overheat in less than an hour.

Yes, they're working as intended.

I highly recommend DIY Perks YouTube channel. Check out other videos, like the artificial sun and the DIY projector.

Is the dark magic pocket camera is actually internal perri cool? And many astro as well.

I have a Sony RX1r (have had both mk1 and mk2) and it does the same thing recording high bitrate 1080.

How the hell are companies allowed to sell a video recording feature where it only records for a short period of time before just shutting down?!

If you sold a car like this there would be lawsuits.

DIY Perks is a channel what I love. I suggest to check other videos.

I would like to see a modular full frame sensor connected to iOS. Can I dream?

A friend of mine and I were working on building exactly that, but unfortunately he died last year and I don’t have the funding on my own to continue building it.

Care to elaborate more on this?

Sure, you can find my contact info in my profile.

Sony made steps in this direction with ILCE-QX1, and with their current compact full-frame ILC progress they could decide to reboot that product line. Would be interesting to have a compact photo/video powerhouse out of that lens-style camera with direct connection to iPad Pro.

Though they are notoriously bad on the third-party integration side.

> Matt decided to use a thin strip of copper to pull heat away from the processor

Errr...this ain't watercooling man.

recall coming up for this $4,000 camera?

Yeah I’m in awe that this is considered acceptable that the camera just can’t record video more than 20 minutes for all the money you pay.

20 minutes at 8K. At lower sizes, it’s not such a problem: https://en.wikipedia.org/wiki/Canon_EOS_R5#Overheating_issue...

but you probably can get a $100 camera that can do 4k videos (I got a 1080p video camera for $30 5 years ago)... I would expect a $4,000 camera to be able to do 8k reliably...

The resolution is not everything. Compression and bitrate matters as well. No $100 Camera is able to handle that bitrate. Raw 8k results in 2.6 Gbps. That is not something that a small and low prower CPU can process. And thats what causing the heat.

I checked for a 100€ 4k action cam and tried to find out the bitrate they use. I found none where it was specified but this one https://www.goxtreme-action-cams.com/de/goxtreme-manta-4k/ requires at least a U3 class SD card. That means 30MB/s so 240mbps. And thats what causing the heat. Canon could of course but a better cooler into the camera but that was either a economic decision to protect their professional camera gear, a a design decision to keep the case small or a mix of both.

> The resolution is not everything.

If you talk about 8k, 4k or 1080p, the only thing that matters is the resolution. Whatever resolution that is advertised should work.

You are quite wrong, ignoring both the megapixel wars and how video encoding works. There are enormous trade-offs involved here, and one chip may produce files twice the size and half the quality of another chip even with the same sensor and lens—to say nothing of the variation possible in sensors and lenses! If I record 1080p on my Surface Book with its rear-facing camera, for example, it runs quite a bit hotter than a typical phone with very similar optics, but produces substantially better quality for a given bit rate—in doing rough visual comparison earlier this year, I estimated that it was producing obviously better results at a quarter of the file size, comparing it to a mid-range phone of a generation two or three years newer than it.

Also the other aspect of resolution not being everything is the optics; my Sony α6100’s 1080p is miles ahead of most phones’ 1080p in quality, and I think there’s a high chance it’d be better than a $100 4K video camera’s 4K.

In short, not all 1080p or 4K or 8K or whatever is created equal.

One other thing to remember: all else being equal, 8K video will be consuming four times as much power and generating four times as much heat as 4K video, which will be consuming four times as much power and generating four times as much heat as 1080p video. (Of course, all else never is equal, but it’s a fair approximation, especially within the context of a single device; but when comparing between generations, as part of quadrupling the pixels pushed, they probably put more effort into efficiency to try to balance power and heat things out too.)

> I would expect a $4,000 camera to be able to do 8k reliably...

It already is reliable for it's stated goals. Unlimited 8k recording is neither a stated feature nor a goal of this camera. If you peg a smartphone CPU at 100% the system firmware will throttle it down to a safe limit. This is no different. The R5 is a tiny full-frame weather-sealed camera that's designed to be hand held. Which other camera that is comparable does things better?

It is a camera for taking pictures.

It is stated on their website.

The main market for this camera is probalby not 8k video.

Its also hard to determine if this is a real issue for canon or if its just an issue in a circle of people who are more vocal about it.

> Its also hard to determine if this is a real issue for canon or if its just an issue in a circle of people who are more vocal about it.

he sure went through a lot of trouble if it was not a real problem (maybe he works for Nikon)

He did it because it gave him, as a YouTuber, a great video camera for a very good price.

It is a real problem. The questions is for whom? For the majority of canon R5 buyers? I doubt that.

And even for a youtuber 4k 60fps is more than enough. How many people are watching youtube videos in 8k 60fps. How many people even have the internet speed for that let alone the monitor.

Not many but as Matt explains in his video, you can downsample the 8K footage to 4K to get higher detailed 4K. The same is often done with 4K to 1080p footage for the same reason.

Meh, all mirrorless cameras are like that, it's physics, they're passive, small and extremely powerful, even if video capabilities are getting better they're still primarily designed to shoot stills.

If you want to record over long period of times you get a proper movie camera, which will be much better and even cheaper: https://www.blackmagicdesign.com/products/blackmagicpocketci...

> Meh, all mirrorless cameras are like that,

They all suck so they all need to keep sucking?

Idk, try taking pictures with a microwave, it might suck too. Use the r5 for what it's intended for and it's a perfectly fine camera

I had some cheap phones and photo cameras that could take videos non-stop... But my old Droid I think used to overheat a bit while it was taking videos non-stop while plugged in

> I had some cheap phones and photo cameras that could take videos non-stop

If you have a "cheap phones and photo cameras" that can take 8K footage non-stop then I'm very interested....

If they could do it at high resolutions, I'd be surprised. My old Lumia 950 would overheat after about 20 minutes at 4K, but record indefinitely at 720p (most I ever tried was about two hours; I don't think I ever got around to testing it at 1080p, and IIRC both were 60fps).

You can also take non stop videos with the r5, just not in 8k ... Just like you can use your car non stop, just not red lining the first gear on the highway

Damn, I'm not a field that requires camera use at all, but that Blackmagic Pocket looks very very impressive. Makes me want to impulse buy just to mess with it.

Notably the (linked) Blackmagic camera maxes out at 6K. I don't think any "action cam" vendors have an 8K offering either.

Just shows that 8K remains a cutting edge feature and isn't surprising that a still-camera has marginal support for it.

Reply on news.ycombinator.com