|
Sagebrush posted:i am always really entertained by the "i am really good at my thing so i defy physics/biology" attitudes that are so pervasive among hobbyists buddy I read avsforum so I am intimately familiar with this, a bunch of dudes in their 50s and 60s arguing about high resolution music when they can probably barely hear over 12kHz
|
#
?
Aug 6, 2018 18:49
|
|
|
|
| # ? May 30, 2024 11:05 |
|
|
Yeah lol at the scale on that graph also packet size isn't that relevant to BTLE communication as long as the device is using btle 4.1 (? whichever revision added the extended pdu standard). previously it would have been very important to get your payload below 20 bytes which was the max size for a gatt pdu, and not as important below that unless you were doing """"connectionless"""" by putting data in advertising packets (do not do this). if you have the modern standards, you can get extended pdus. this is good because btle has a tdma scheme multiple connections to a single central, and you get one pdu exchange every time your connection interval comes around. the connection intervals are what determine the latency - your phy rate is 1mbps (or 2 on recent hardware) and with packets that even with the extended pdus are only 500ish bytes, that's not gonna contribute. the connection interval you negotiate with the host is what matters. this is a thing where the peripheral presents preferences but the host gets to pick, and the intervals go down to like 10 or 12ms. so what you see with the ds4 is basically the absolute minimum for btle latency and like people have been saying it's gonna be imperceptible especially on a console that is pushing frames to a real tv. the connection interval, and thus the latency, is going to be larger than that on other devices for a couple reasons - battery usage. the battery usage of a btle peripheral isn't super high, it's probably 5-10ma while transmitting the (again, quite short) packets and ua to na depending on how good the transmitter is otherwise; however, you do have to send something every connection interval and so this can be the difference between hours or days of battery life and months or years of battery life. so increasing the connection interval, or the number of connection intervals that can be missed before a connection dies, and talking less is a battery saving strategy. and these can vary quiiite a bit - connection intervals can go up to like 380ms and the timeouts before a device is considered disconnected much higher, to tens of minutes iirc - the drat central doesn't let you. the central is what eventually presents you connection intervals you have to use. this part of the central is usually controlled by the OS drivers, and applications written on the OS that want to talk to peripherals don't get any say in the matter. for laptop or phone centrals there's power saving here too that can matter a lot more, since they don't get a connection supervision timeout, they have to be monitoring advertising channels, and they have to connect to multiple devices - you're bad at developing btle peripherals. just rolled with defaults and your connection interval is at a lazy 100ms and you never investigated extended pdus even though your drat hardware is capable you just copy pasted the example code you lazy fucks i can fully understand why you were fired
|
|
#
?
Aug 6, 2018 18:52
|
|
|
Shifty Pony posted:here's the PS4 data format when acceleration and trackpad data isn't needed and it looks like they have it set up to jam all of the essential data into just 11 bytes. that’s pretty cool. the old Xbox 360 controller system used a lot of power and i’m pretty sure it killed a USB hub for two when i used the PC adapter. qirex posted:a 7ms difference in latency is not perceptible to a human that has a reaction time of 200-400 milliseconds i am in fact the same pleasurekevin that has argued that game streaming is fine and the lag is trivial. even when lag exceeds a few hundred milliseconds, we can adapt to it. but there are definitely bluetooth devices with and without perceptible lag.
|
|
#
?
Aug 6, 2018 18:56
|
|
|
jit bull transpile posted:What am I supposed to be looking at here the ps4 controller has lower latency via Bluetooth than it does when wired via USB. pretty good indication that they put a hell of a lot of work into optimizing the Bluetooth link.
|
|
#
?
Aug 6, 2018 19:02
|
|
|
The Management posted:the correct answer for this: yes, it would be loving amazing if apple made something like this, you know, like they did with the cinema display but no, let's just sell a bajillion dongles and hope loving belkin or lg can makes something that doesn't poo poo the bed in the presence of a wifi signal
|
|
#
?
Aug 6, 2018 19:03
|
|
|
Shifty Pony posted:the ps4 controller has lower latency via Bluetooth than it does when wired via USB. pretty good indication that they put a hell of a lot of work into optimizing the Bluetooth link. kind of but it’s more an indication that they didn’t put a hell of a lot of effort into optimizing the usb link. you could get a lot less latency than that with a usb2 isoch endpoint
|
|
#
?
Aug 6, 2018 19:04
|
|
|
jit bull transpile posted:What am I supposed to be looking at here people caring about 2d fighters in 2018
|
|
#
?
Aug 6, 2018 19:05
|
|
|
xbox one controller is the epitome of game controllers, no reason to ever use or buy anything else the 360 was almost perfect except for the d-pad and no usb port and they fixed both those things. nothing better
|
|
#
?
Aug 6, 2018 19:06
|
|
|
Cocoa Crispies posted:people caring about 2d fighters in 2018 does nidhogg count? because I will care about nidhogg until the end of time
|
|
#
?
Aug 6, 2018 20:37
|
|
|
Lambert posted:The PS4 controller absolutely does have less latency than other controllers, like the Xbox controller. Thanks for the graph showing Sony controllers are garbage
|
|
#
?
Aug 6, 2018 20:39
|
|
|
fart simpson posted:the actual correct answer is a computer with ports, op i bought one of these. it is good
|
|
#
?
Aug 6, 2018 20:40
|
|
|
Scott Forstall posted:does nidhogg count? because I will care about nidhogg until the end of time nidhogg is so good shame about nidhogg2 looking like some gross 90s cartoon
|
|
#
?
Aug 6, 2018 20:44
|
|
|
infernal machines posted:yes, it would be loving amazing if apple made something like this, you know, like they did with the cinema display you’re looking for the LG Ultrafine, it seems to have Apple’s seal of approval and is sold exclusively at Apple. Apple doesn’t make Apple-branded external displays anymore.
|
|
#
?
Aug 6, 2018 20:49
|
|
|
PleasureKevin posted:you’re looking for the LG Ultrafine, it seems to have Apple’s seal of approval and is sold exclusively at Apple. Apple doesn’t make Apple-branded external displays anymore. the lg ultrafine is a loving piece of poo poo i am aware that apple doesn't make 1st party external displays, that was sorta the point
|
|
#
?
Aug 6, 2018 21:15
|
|
|
infernal machines posted:the lg ultrafine is a loving piece of poo poo seems good to me
|
|
#
?
Aug 6, 2018 21:19
|
|
|
just fuckin' peachy absolutely the top quality hardware we've come to expect from lucky goldstar
|
|
#
?
Aug 6, 2018 21:22
|
|
|
that was over a year ago and it’s fixed. apparently thunderbolt carries so much data it can gently caress with electronics and they didn’t have adequate shielding.
|
|
#
?
Aug 6, 2018 21:30
|
|
|
i was gonna buy an lg 5k or two (and a macbook pro to drive them) but then apple announced they were working on a monitor for 2019 so welp
|
|
#
?
Aug 6, 2018 21:33
|
|
|
poty posted:i was gonna buy an lg 5k or two (and a macbook pro to drive them) but then apple announced they were working on a monitor for 2019 so welp i missed that announcement
|
|
#
?
Aug 6, 2018 21:41
|
|
|
PleasureKevin posted:apparently thunderbolt carries so much data it can gently caress with electronics and they didn’t have adequate shielding. yes, if your product isn't properly designed, it might not work. amazing.
|
|
#
?
Aug 6, 2018 21:42
|
|
|
duck duck go is failing me, i can’t find any announcement
|
|
#
?
Aug 6, 2018 21:42
|
|
|
PleasureKevin posted:duck duck go is failing me, i can’t find any announcement i hope this doesnt take much pleasure away https://www.macworld.co.uk/news/mac/new-apple-display-3474851/
|
|
#
?
Aug 6, 2018 21:49
|
|
|
poty posted:i hope this doesnt take much pleasure away on the contrary my pleasure has only increased
|
|
#
?
Aug 6, 2018 21:57
|
|
|
Regular brain: I'll use a wired input device that never runs out of battery, is cheap, and works great galaxy brain: I'll use a wireless input device that performs worse, runs out of batteries (thus contributing to ecological destruction and running the risk of literally exploding), is more expensive, and requires me to open up my computer to wireless attack vectors
|
|
#
?
Aug 6, 2018 22:03
|
|
|
quote:"As part of doing a new Mac Pro — it is, by definition, a modular system — we will be doing a pro display as well." ahahahahaha they've learned loving nothing from the trashpro
|
|
#
?
Aug 6, 2018 22:16
|
|
|
fishmech posted:Thanks for the graph showing Sony controllers are garbage learn2read
|
|
#
?
Aug 6, 2018 22:16
|
|
|
The fact it's taking this long to produce should make it pretty obvious the new Mac Pro is going to be another overengineered POS.
|
|
#
?
Aug 6, 2018 22:18
|
|
|
Lambert posted:The fact it's taking this long to produce should make it pretty obvious the new Mac Pro is going to be another overengineered POS. who are you?
|
|
#
?
Aug 6, 2018 22:20
|
|
|
It's me, Jony Ive.
|
|
#
?
Aug 6, 2018 22:22
|
|
|
Lambert posted:It's me, Jony Ive. 
|
|
#
?
Aug 6, 2018 22:26
|
|
|
you see, you've got your compute core here, then you connect your thunderbolt jbod/raid stack, and you connect your thunderbolt gpu stack, and you connect your thunderbolt PCIe enclosure for DSPs
|
|
#
?
Aug 6, 2018 22:33
|
|
|
everyone in yospos is constantly on a plane, I hope they get the thing they like
|
|
#
?
Aug 6, 2018 22:38
|
|
|
Lambert posted:learn2read the graph explicitly shows that sony controllers are garbage dude did you read it?
|
|
#
?
Aug 6, 2018 22:44
|
|
|
infernal machines posted:ahahahahaha we knew this ever since that article talking about how apple hired a bunch of "professionals" to study their workflow and the behind the scenes look had them running a recording studio using iPad pros as input devices for an iMac pro while a mbp with external GPU was doing whatever the gently caress on the side: quote:First, we visit the room where they record new instruments for Logic and Garage Band and then on to an edit bay used by the Pro Workflow Team to put Final Cut Pro through its paces. so yeah shits gonna suck because tim sees an opportunity to make users empty their bank accounts buying each modular bit separately.
|
|
#
?
Aug 6, 2018 22:54
|
|
|
infernal machines posted:just fuckin' peachy quote:the LG UltraFine 5K Display can be effectively unusable if placed near a Wi-Fi router with some kind of radio interference causing the monitor to regularly black out and hang the connected MacBook Pro. lmbo
|
|
#
?
Aug 6, 2018 22:59
|
|
|
turns out piping xxxtreme rfi down the pci-e bus is bad for stability
|
|
#
?
Aug 6, 2018 23:17
|
|
|
Shifty Pony posted:we knew this ever since that article talking about how apple hired a bunch of "professionals" to study their workflow and the behind the scenes look had them running a recording studio using iPad pros as input devices for an iMac pro while a mbp with external GPU was doing whatever the gently caress on the side: its because apple hasn't made anything good since december 1992, and gave up the pretense of making anything good in january 2011
|
|
#
?
Aug 6, 2018 23:17
|
|
|
I sincerely hope the next Mac Pro is essentially a Mac Mini with optional Thunderbolt eGPU that stacks on top of it
|
|
#
?
Aug 7, 2018 00:43
|
|
|
oh snap they leaked a pic of the Mac Pro 
|
|
#
?
Aug 7, 2018 01:16
|
|
|
|
| # ? May 30, 2024 11:05 |
|
|
Best Bi Geek Squid posted:oh snap they leaked a pic of the Mac Pro lol did they ever ship this?
|
|
#
?
Aug 7, 2018 01:19
|
|
