4K video is awesome. Probably wasted on a phone, but I can instantly see the difference between 1080p and 4K on anything larger.

I'm pretty sure that it's right on the diminishing returns margin, however. I don't have an 8K resolution device to try it out on, but that's because everything I use is in the 'retina' zone of pixel density, I never see a genuine pixel under normal circumstances. (Actually my TV isn't 4K but if I cared at all it would be).

For what I use my phone for, the only thing which could improve in a way I would notice is the camera. That won't be true for everyone, games in particular have an essentially unlimited appetite for specs, but this is just to say that the state of the art as of four years ago is still just fine by me.

The biggest difference with 4K IMO isn't the resolution itself, but the features that typically come with 4K content. HDR and 10-bit colour especially make for a noticeably better picture than the pixels ever will.

That and the higher bitrate! e.g. youtube

And, of course, the higher bitrate directly ties into our topic at hand. The increase resolution is just a side effect.

Based on back of the envelope math I did based on known ideal viewing distances, most TVs are too far away to be able to get all the benefit out of even 4K. Something between 4K and 1080P would be the “ideal” balance between bandwidth and resolution.

For monitors, the maximum resolution is a bit behind 4K (since monitors are closer than TVs), and I suspect that’s why Apple offers a 5K iMac.

I didn’t estimate this for mobile devices.

what if I told you I voluntarily set youtube to 480p or even 360p ?

I would say: Rock on with your potatocam aesthetic!

Downsampling a high-resolution video is always possible, the reverse is an... unsolved problem.

IIRC there were 4k upscaling demos but I'm past the point of pixel resolution, my interests are not in pristine visuals but a blend of surprises in content and composition.

Upscaling is 'real' in the sense that it exists, and for animation in particular the results can be quite satisfying.

But the technology to add missing detail will never exist; we must be content with a realistic facsimile.

> But the technology to add missing detail will never exist; we must be content with a realistic facsimile.

You are right in a sense. But there are some details that are in the pixels but that your eyes can't see. You can upscale to make them visible to the eyes.

See https://en.wikipedia.org/wiki/Compressed_sensing

I normally browse the comments on certain channels while watching, and with the video taking up only the upper third of the screen, I can’t tell a real difference between 480 and 1080.

If I flip to landscape though it can be noticeably, significantly worse.

I bet we haven't even come close to realizing the potential of having extremely fast/high bandwidth cellular connections. Streaming 8k video to a phone seems silly but there are some brilliant and clever people out there who will find a use for that kind of mobile bandwith and apply it in ways you or I never thought of.

Bill Gates infamously said sometime in the 80's "640K is more memory than anyone will ever need" , and for the scope of computer applications at the time he was correct.

But (stipulating the attribution is correct) was Bill right?? Okay 640K is a little too far, but have we gotten 25,000 times more value out of the ~25,000 times more memory we now have? One could argue that we could have made much more ingenious use of, let's say, 1/100 of the memory and CPU we now have. We are so far at the end of the value-add curve, that we're running super-computers so we can clear up our complexion or make ourselves look like mice or potatoes in real-time video -- it's absurd.

> One could argue that we could have made much more ingenious use of, let's say, 1/100 of the memory and CPU we now have.

The counter argument is that this is over-, premature, or unnecessary optimization. If we can put 2GB of memory in a computer for $20, why spend millions on R&D that will be rewritten to make it work on 128 MB?

There is also an environmental argument for your point, but everything will always be a tradeoff to companies trying to build and sell devices quickly.

You say absurd, I say awesome.

Why would we expect value to scale linearly with memory and transistor count? I would in fact expect it to be logarithmic: each doubling would lead to a roughly linear increase in cool stuff we can do.

I think that's what we see. Yes, our computers could be radically better experiences if it weren't for the bad habit of squandering improvements with layers of unnecessary abstraction and laziness. But the only thing standing in our way is those unnecessary and that laziness, so maybe we should stop doing that, now that real gains in speed are slowing down.

Gates never said that; the quote is nothing but an urban legend.

I think in one of his AMAs on reddit he confirmed that he never said it and no software developer during those times would say such a thing that 640K is more than enough.

Yep, when implementing a simple OS for a class, it is clear that the 640KB limitation was imposed by the hardware at the time, not the DOS. So it wouldn't even make sense for him to say it, unless it was more akin to "it's enough for us".

I don't know who said it, but it was in one book I had back in the day. That is a book from the era where 640k was something big companies might splurge on but only for systems that proved a need.

Yeah, VR comes to mind. But also speed gets reduced even with a small deterioration in signal. If you're getting 10-20mbps at full strength, you might get 1-3 Mbps with a weak signal. I would presume that 5G would be faster in weaker signal scenarios.

I just got 200Mbps down on 4G. Granted it’s late at night and I’m in an office building by the highway next to a grave yard and a school, but still.

Yeah, with 1 bar of signal strength though? I work remote as a nomad and I frequently have to depend on 4G for connectivity. I don't get anywhere near 200 mbps in most parts of the U.S. I'm lucky to get 20, and often limited to the low single digits.

The higher your max bandwidth, the quicker it collapses under signal deterioration.

In 2011-2012, when 4G started becoming mainstream, did you make similar claims? For example, did you say something in the lines of “why would I want to watch videos on my phone? 3G is more than enough for the browsing I am doing now?”

4G might be good enough for you right now, but your needs will increase, way more and way faster than you expect. Remember this the next time you try to do a video call / watch an HD video / etc. If we had stayed in 3G, you would not be able to do any of these, but when 4G was introduced you did not feel any need for them.

Idk, I still don't know why anyone would want to watch videos on their phone, and I have a 4G phone capable of showing videos... (watching videos on other devices is way better).

I mean, since the invention of semi trucks, you could technically can send me ~40 tons of postcards to me all in one shipment, which would be impossible in the days where postcards were sent via a guy with a satchel on horseback, but then why aren't you sending me 40 tons of postcards right now? Why would you even want to?

Don't get me wrong, semi trucks certainly have it's uses and has enabled many things, so would 5G provided it does something better than 4G, but I can see why some people would go "Why would you want more of that?" Not everyone has the same wants and needs.

These days, latency seems more problematic to me than bandwidth. Unfortunately though, short of optimizing networks, and reducing the distance between server and client, there's not much that can be done for latency...

> Not everyone has the same wants and needs.

And that's why 4G is useful for a lot of people even if it's not useful for you specifically.

5G will be similar. _You_ might not benefit, but a lot of other people will.

You've never watched a video on your phone in your entire life?

No, I did not say that - 3G was slow enough to be bothersome for simple tasks. Not everything is relative. There are absolutes, and there are diminishing returns on investment, and a differences between creating tools that solve problems and creating infrastructure in search of problems. But I suppose I should be happy for anything that drives growth, or even appears that it may, at this point.

> but your needs will increase, way more and way faster than you expect.

Forever? Is there no bandwidth so great that most users would be happy with it? Can new uses for extra bandwidth always be found, and do at least some of those uses always significantly improve the user experience?

> Remember this the next time you try to do a video call / watch an HD video / etc. If we had stayed in 3G, you would not be able to do any of these

I wouldn't? News to me! I just disabled 4G on my phone, which fell back to HSPA+ with a fast.com result of about 7mbps. I then successfully tested a Zoom test meeting with video (worked fine) and watched a YouTube video at 1080p (buffered for three seconds at the start, noticeably longer than usual, but then played fine).

So what's going on there?

> Forever? Is there no bandwidth so great that most users would be happy with it? Can new uses for extra bandwidth always be found, and do at least some of those uses always significantly improve the user experience?

The fact is that every year mobile data consumption is increasing significantly and it is expected to continue this trend. I do not have an opinion on whether more data equals better user experience.

> I just disabled 4G on my phone, which fell back to HSPA+ with a fast.com result of about 7mbps. I then successfully tested a Zoom test meeting with video (worked fine) and watched a YouTube video at 1080p (buffered for three seconds at the start, noticeably longer than usual, but then played fine).

Without knowing the context of your set up, I can only guess. I would say that not that many people use 3G in your area, so you used an uncongested cell.

With datacaps that could run through 5G in seconds, it makes little sense in opting for 5G instead of 4G. Which costs more and with less battery life.

Processing power on phones is useful with video editing and AR. Social media is moving towards videos instead of photos. Lot of it can be attributed to the capabilities of current phones.

We'll hit the wall on the technology eventually, just as we have for desktop computing. A 10-year old desktop is still perfectly serviceable. Imagine saying that in 2000.

I can’t disagree more with the idea that we shouldn’t have tech like 5G “thrust” upon us.

Sure, AT&T 5G isn’t great now, but it’s all about spectrum allocation.

That spectrum allocation will only become more and more 5G-leaning over time. 5G isn’t the same great leap as 4G was, but I’d still take it over 4G.

Increasing capacity via 5G will allow higher data caps too.

those things are completely disconnected