Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

YouTube Format IDs

gist.github.com

Post reply on HN
Print view View original

81–87 of 87 posts

Re: YouTube Format IDs

#81

Post by ErroneousBosh »

account42 wrote: post #73
ErroneousBosh wrote: post #37

Earlier quoted context omitted.

There is absolutely no reason for any sane video editing software to be able to read VFR files. It should, however, reject them gracefully but positively.

Of course there is. A lot of older TV shows are effectively variable frame rates when properly detelecined/deinterlaced. And mixing two fixed frame rate sources with different framerates would naturally produce a VFR video and with current editing software you need to destroy some of the source information or interpolate some slop to make it work. There is no good reason for editing software to have these limitations…

> A lot of older TV shows are effectively variable frame rates when properly detelecined/deinterlaced.

How do you figure that?

Broadcast TV is very very very very firmly locked to 25 interlaced frames per second (or 29.97ish in the US where they have to do everything with funny numbers), interlaced to 50 (59.94 US) fields per second.

How could that in any way be VFR?

Re: YouTube Format IDs

#82

Post by ErroneousBosh »

shric wrote: post #52
ErroneousBosh wrote: post #25

Earlier quoted context omitted.

> monthly (mobile) data caps Are we back in the early 2000s or something?

There are only 5-10 countries in the world that I know of that don’t have mobile data caps. The vast majority of the HN demographic do not live in those countries. The remainder either have an explicit hard cap or some kind of fair use clause that throttles when you reach some undisclosed amount.

Right, but in most countries the data cap is so high you'd never hit it.

I'm not on even a particularly expensive or specialised deal on my phone here in the UK and the cap is something like 1TB.

Re: YouTube Format IDs

#83

Post by Balooga »

echoangle wrote: post #79
Balooga wrote: post #59

Earlier quoted context omitted.

Are you joking? Most people don't notice a difference between 480i and 1080p when blown up on a large TV in their living room. I know this from experience having worked at a satellite TV provider and hearing actual customer feedback.

Are most customers 70 year old? You should get your drivers license revoked if you don’t notice that difference.

Not at all. If you toggle between SD and HD in front of them, it's plain to see.

But if you tuned to a random channel, most people couldn't tell you if what they are looking at is SD or HD. "It's all digital, right?"

Again, if you then swapped to the HD version, they would immediately notice.

But we had so many stories of installers visiting customers homes on service calls to be confronted by massive TVs displaying 480i or 480p. And the customer as happy as can be with the picture quality.

And there were two version of this; (a) Two physical channels, one channel SD and one channel HD, and (b) a single HD channel and the DVR is can be set to down convert to SD.

Re: YouTube Format IDs

#84

Post by zahlman »

account42 wrote: post #72
zahlman wrote: post #56

Earlier quoted context omitted.

Especially retro console games, where sprites might flicker at 60Hz (either to create an illusion of translucency or because of hardware limits on simultaneous sprites displayed); viewing footage at 30fps can end up catching all the "off" frames and make screen elements disappear completely.

Yes, or in general everything that was originally 30 FPS interlaced should be de-interlaced to 60 FPS progressive video and not 30 FPS.

Oh, I somehow completely forgot about interlacing having been a thing.

Re: YouTube Format IDs

#85

Post by mengyuxuan »

throw0101d wrote: post #35
echoangle wrote: post #24

Earlier quoted context omitted.

If you don’t even see the difference between 1080p and 720p, there’s something else going on.

> If you don’t even see the difference between 1080p and 720p, there’s something else going on. If I open a browser window and click on a YT video suggested on the main page, it plays on small portion of the screen of my 14" MacBook: there is little practical visual difference between 720 and 1080 at that 'video box' size. If I open the same video on my 27" 5K iMac and full screen it there, then yes, there is a diffe…

[flagged]

Re: YouTube Format IDs

#86

Post by hulitu »

sebmellen wrote: post #58
dietr1ch wrote: post #30

Earlier quoted context omitted.

Last time I went to the movies I didn't have a good time watching "motion" at 24Hz on a large screen while used to watch motion at 240Hz on a small screen. I felt I was looking at a powerpoint presentation or an animation made at the edge of a (paper) notebook.

The cinematic obsession with 24 FPS is so stupid…

But those are 24 real frames. Not predicted ones like in your favourite Youtube video.

Re: YouTube Format IDs

#87

Post by hulitu »

echoangle wrote: post #23
throw0101d wrote: post #4

> HFR stands for "High Framerate", which means up to 60 FPS, whereas non-HFR is limited to 30 FPS Personally I find most ((almost?) all?) >30fps generally unnecessary and just chews through bandwidth, monthly (mobile) data caps, and processing/batteries. Are there specific cases where "HFR" videos are really improving the viewing experience of particular (types of) content? (Note: in the table it's about support for…

How much extra bandwidth does it even take to double the framerate? The frames should be highly correlated and as long as you only add p-frames, the data needed should be pretty small, no?

> How much extra bandwidth does it even take to double the framerate?

For ? No extra bandwidth. They just increase the number of predicted frames.

Post reply on HN