Is Apple really not innovating anymore?...

[Kuroz]

I am skeptical of claims of near zero latency. People have just taken the word of apple enthusiasts/tech reviewers about this claim. Please google reddit for interesting screen captures disputing this claim of check below--

https://www.reddit.com/r/apple/comments/3kaapx/do_you_think_apple_pencil_will_be_a_viable/

A lot initial impression claim the Apple Pencil is lag free. From the hands on videos, I can conclude the Apple Pencil has ~60ms (+-20ms) lag.

(reference video with guy claiming lag free)

http://imgur.com/a/hA9qm (screen shots of lag in action)

Yup i noticed the lag in the disney video a few days ago.

 

[engineerben]

I am skeptical of claims of near zero latency. People have just taken the word of apple enthusiasts/tech reviewers about this claim. Please google reddit for interesting screen captures disputing this claim of check below--

https://www.reddit.com/r/apple/comments/3kaapx/do_you_think_apple_pencil_will_be_a_viable/

A lot initial impression claim the Apple Pencil is lag free. From the hands on videos, I can conclude the Apple Pencil has ~60ms (+-20ms) lag.

(reference video with guy claiming lag free)

http://imgur.com/a/hA9qm (screen shots of lag in action)

Well, Apple says "virtually no lag" and "only milliseconds between the image you have in your mind and the one you see on the display."

Is 60ms "virtually" no lag? I don't know. I remember in early VOIP systems that somewhere between 100 to 200ms conversations became difficult because of the latency, but once the latency was consistently below about 100ms it just didn't matter anymore.

So what latency is acceptable in a stylus system? I don't know that either, but I'd suspect it's somewhere in the low tens of milliseconds. If they're sensing position 240 times a second, that's a little more than 4ms per sample, and that sets a floor on the responsiveness. I'd expect them to average a few samples to make a smooth experience. But if the lag is really closer to 60ms, that implies the graphics subsystem is doing some seriously heavy lifting between sensing stylus position and expressing the mark on the screen.

 

[Serban]

since pencil has at most 60ms and the new stylus from micrososft surface pro 4 has 122ms...let's face the fact..that the frist generation pencil is the perfect one from the start, better than others 3 or 4th generation.

 

[rowspaxe]

To me, the real innovation is happening behind the scenes; for example, this: how does a manufacture of consumer goods manage to challenge the number one builder of CPUs on the planet on raw horsepower? .

By violating patents held by the University of Wisconsin? Check today's MR news

 

[Krevnik]

So what latency is acceptable in a stylus system? I don't know that either, but I'd suspect it's somewhere in the low tens of milliseconds. If they're sensing position 240 times a second, that's a little more than 4ms per sample, and that sets a floor on the responsiveness. I'd expect them to average a few samples to make a smooth experience. But if the lag is really closer to 60ms, that implies the graphics subsystem is doing some seriously heavy lifting between sensing stylus position and expressing the mark on the screen.

However, I'll point out that in iOS 9, touches are coalesced so that you can get better resolution, and predictions are available to reduce latency. (Text summary from WWDC on this topic)

The end result is that you have one thing that actually raises latency a bit, but another that is meant to let you bring the perceived latency down. But the high sample rate is key for not getting "kinked curves" that need to be smoothed out in software.

So while the sample rate is higher than 60Hz, iOS won't deliver touches to the app any faster than 60Hz (never has). Or about 33ms between updates to apps. And then you have latency on drawing/etc on top of that. However, there is also touch prediction available which is supposed to help you get latency down by telling where the next touches should be. But using these touches is totally up to the app developer, so it is possible that different apps will lag more than others with this setup.

That said, I'm a bit surprised that Apple's own app doesn't seem to be using the predicted touches at all. But 24-30fps video is going to introduce some issues actually measuring the real lag involved. I suspect that in this case, the 60ms is a bit high. 45ms lag actually sounds pretty close to what I'd expect from an app that's not using the predicted touches.

EDIT: After reading the whole summary, iOS 9's tweaks should help get apps to 1.5 frames of latency, which is still bigger than 33ms, but less than 66ms. So that's the "expected" range, if we use the WWDC talk as a guide... so higher than I was mathing out.

 

[rowspaxe]

since pencil has at most 60ms and the new stylus from micrososft surface pro 4 has 122ms...let's face the fact..that the frist generation pencil is the perfect one from the start, better than others 3 or 4th generation.

surface pen latency, anandtech, june, 2015

surface pro 3 : max 141ms, min 75ms , avg 97.8ms


www.anandtech.com/show/8077/microsoft-surface-pro-3-review/2

 

[engineerben]

By violating patents held by the University of Wisconsin? Check today's MR news

Yeah, that's the one.

Look, I'm not going to defend Apple in this forum - they have plenty of lawyers and can defend themselves. I'll just observe that the patents in question are held by one of the largest patent aggregation outfits out there - right up there with Rockstar, of which Apple is a member. Welcome to the world of elephants dancing.

 

[Kal-037]

This topic grew, awesome I think personally Apple is still doing what they always have, making things different and in many cases better. That doesn't mean other companies suck or aren't innovating as well, I think Microsoft and Samsung have some good and unique products out there too.


Kal.

 

[Krevnik]

Yeah, that's the one.

Look, I'm not going to defend Apple in this forum - they have plenty of lawyers and can defend themselves. I'll just observe that the patents in question are held by one of the largest patent aggregation outfits out there - right up there with Rockstar, of which Apple is a member. Welcome to the world of elephants dancing.

Well, it also gets weird. General practice around patents is ignorance. Treble damages if you willingly violate a patent tends to lead to weird discussions. Ones where a company would rather unintentionally violate a patent through re-invention than get hit for extra damages.