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

iPadOS

techcrunch.com

Post reply on HN
Print view View original

291–300 of 337 posts

Re: iPadOS

#291

Post by sobani »

_gjrn wrote: post #273
headsoup wrote: post #260

Earlier quoted context omitted.

Actually I think it highlights how well Apple manage their message and keep people focused on them: most people probably don't notice or actually care about the difference in latency, and 9ms is not the best in the market, yet it becomes a 'good news story' validating Apple

Latencies bellow 16ms require a refresh rate higher than 60Hz. For 9ms you need at least 90Hz. Don't know many who do that. Specially if you take in consideration the battery hit that the faster update rate requires.

Don't those latencies add up?

In the worst case 60Hz + this new pen will add up to (16 + 9) 15ms of latency, right?

If the screen refreshes at T=0,16,33,50 and you create input at T=8 then the input will be processed at T=17 (8+9), so you will need to wait for T=33 before it show up for a total latency of 25ms (33-8).

Which is a lot better than the potential total latency of 36ms with a 20ms pen.

Re: iPadOS

#292

Post by ksec »

pcr910303 wrote: post #227
fumar wrote: post #200

Earlier quoted context omitted.

I have the 10.5 in Pro and Apple Pencil 1. I wonder if the 20 ms vs 9 ms is detectable by the human eye.

It can be very detectable, I'd say. Try watching this: https://www.youtube.com/watch?v=vOvQCPLkPt4 It shows the difference between from >100ms to ~1ms. It's mind baffling.

I wonder how did they manage it. Even the Screen itself has a 1 - 4ms latency. That is excluding the Input latency and Software in between.

Re: iPadOS

#293

Post by coldtea »

ksec wrote: post #289
davemp wrote: post #250

> Apple Pencil latency is dropping from 20ms to 9ms, This is why I respect Apple. Most tech forums would have been flooded people whinging about how mere humans can’t perceive such low latencies and how improving them is a waste. Meanwhile Apple knows not to leave anything on the table for UX.

I am guessing this is pure Input latency? what I want to know is End to End latency, from Input > OS > Display. I wonder if we will ever get sub 5ms in End to End Latency.

Why would it be important to know? Previous situation was already industry best (comparing other tables, Wacom, etc), the new one is several ms better.

Re: iPadOS

#294

Post by coldtea »

headsoup wrote: post #260
davemp wrote: post #250

> Apple Pencil latency is dropping from 20ms to 9ms, This is why I respect Apple. Most tech forums would have been flooded people whinging about how mere humans can’t perceive such low latencies and how improving them is a waste. Meanwhile Apple knows not to leave anything on the table for UX.

Actually I think it highlights how well Apple manage their message and keep people focused on them: most people probably don't notice or actually care about the difference in latency, and 9ms is not the best in the market, yet it becomes a 'good news story' validating Apple

>and 9ms is not the best in the market

Really, who does it better? This is already < 60hz, so we're talking about 120 hz devices to begin with...

Re: iPadOS

#295

Post by coldtea »

Mertax wrote: post #37
AdmiralAsshat wrote: post #16

This feels like it could become a repeat of Android Honeycomb, but iOS is a much more mature operating system now than Android was at the time, so I guess we'll have to wait and see.

It is an interesting strategy; where others have tried to unify their OS (WinRT/UWP, Honeycomb->Ice cream sandwich). Honeycomb and WinRT were viewed as fragmentation, I think somehow it won’t be viewed the same with iPadOS

>Honeycomb and WinRT were viewed as fragmentation, I think somehow it won’t be viewed the same with iPadOS

iPadOS runs in at most 10-20 devices (and all are iPad models). HoneyComb and WinRT run on hundreds and hundreds of devices, with widely different capabilities.

So there's that fragmentation there too.

Re: iPadOS

#296

Post by hu3 »

NikkiA wrote: post #261
criddell wrote: post #55

Earlier quoted context omitted.

The one you want is PCalc. It isn't free, but it's not expensive either.

PCalc does its math wrong (using floating point), and probably isn't what you want. Calculator (infinity-symbol) and Calc HD Pro do it the right way. (Check '(0.1x1024-102)x10-4' on a physical calculator and on PCalc/others, it should return '0' if the math is done using bcd (almost all physical calculators) or ~5.68e-14 on broken floating point math)

This is amazing! I've seen 2 recommendations for PCalc in this thread alone.

If an erroneous calculator app using floating point can have success, this gives me hope to have success as well. I mean no disrespect towards the app, it just gave me a bit of extra motivation.

Re: iPadOS

#297

Post by iSnow »

tinus_hn wrote: post #286
_gjrn wrote: post #273

Earlier quoted context omitted.

Latencies bellow 16ms require a refresh rate higher than 60Hz. For 9ms you need at least 90Hz. Don't know many who do that. Specially if you take in consideration the battery hit that the faster update rate requires.

If you’re clever you only have the higher refresh rate while drawing or animating, so the battery hit is limited.

Apple has the technology in place for that: https://www.gsmarena.com/understanding_apples_promotion_disp....

Re: iPadOS

#298

Post by b1daly »

ben_w wrote: post #282
jordache wrote: post #60

Earlier quoted context omitted.

what's the solution? not implement multi-finger gestures due to complexity and just omit the feature altogether? I'd rather have those capabilities? My drawing apps on iOS all have tons of multi-touch gestures. I appreciate they are there.

Making things discoverable and easy to learn is an art in its own right, completely separate to the art of programming.

Well yeah, but don’t you think this is exceptionally hard with a touch screen?

Didn’t microsoft try to do this in Windows 8 with little icons in the corners? The problem is that little icons provide only marginally more information, if at all, than the existence of the corner of the screen.

It’s a very hard problem and Apple works on it. They have abandoned previous standards on discoverability because of the inherent limitations of a touch interface on a small screen.

Sacrificing discoverability for usability is the right choice.

I think one issue is that some UI conventions have not been settled on, so I often have to look up how to do something on iOS.

The most recent example of this was “shake to undo.” After looking this up, not only did I learn how to undo typing, I also learned why I would occasionally get these inexplicable warning windows asking if I wanted to undo!

Re: iPadOS

#299

Post by georgyo »

coldtea wrote: post #293
ksec wrote: post #289

Earlier quoted context omitted.

I am guessing this is pure Input latency? what I want to know is End to End latency, from Input > OS > Display. I wonder if we will ever get sub 5ms in End to End Latency.

Why would it be important to know? Previous situation was already industry best (comparing other tables, Wacom, etc), the new one is several ms better.

Amazing. The top level comment is about improving performance even when many people say it isn't important.

Someone says it would be nice to get even lower, and you say it isn't important.

Full circle in three comments.

Post reply on HN