Optimized .exe; builds of ioq3 engine
With GoogleTranslate, Bumpy, dmaHD, iKALiZER, Raw Mouse, Bug Fixes, ..
#950
Posted 26 July 2010 - 12:21 AM
The implementation of raw input is rather cookie cutter; you can do some minor adjustments and can deregister it (and re-register it during the course of the game) but to initially start it and have it give signals is generally a very predefined standard process can't easily be 'wrong' as soon as it works. Hence I'd imagine it's probable we're talking about xfire or driver or windows bugs.
Notice id software on quake3 live has raw input and have not set it to default to avoid quirks. I'm pretty certain they're confident the quirks don't come from their own implementation.
Also this technology is available only after XP and server 2003; It might behave better lately.
--
It might be interesting to see on what Windows version and what Mouse those issues occur.
This post has been edited by mitsubishi: 26 July 2010 - 12:48 AM
- Optimized exe; builds of ioq3 engine for urt With GoogleTranslate, Bumpy, dmaHD, iKALiZER, Raw Mouse, Bug Fixes, ..
- Networking, lag meter, and gaming consistency guide
#951
Posted 26 July 2010 - 08:09 AM
#952
Posted 26 July 2010 - 08:44 AM
Other than that, there's theoretically higher chance to see lower latency since the path is shorter (and busier). Some people say they see it but I can't claim I do. It may be there and help of course, but I wouldn't be able to say "there it was! this wouldn't go like that without raw.".
--
directinput (in_mouse 1 (+ in_rawmouse 0 here)) also avoids negative accel but it has quirks and latencies (latencies also according to Microsoft).
--
The reasons Microsoft claims make raw input better for gaming.
This post has been edited by mitsubishi: 26 July 2010 - 10:18 AM
- Optimized exe; builds of ioq3 engine for urt With GoogleTranslate, Bumpy, dmaHD, iKALiZER, Raw Mouse, Bug Fixes, ..
- Networking, lag meter, and gaming consistency guide
#953
Posted 26 July 2010 - 10:01 AM
mitsubishi, on 26 July 2010 - 08:44 AM, said:
Franckly I don't get it.
A long time a go I spent some of time studying game engine (mainly Q3) and the logic between sensitivity, DPI and screen resolution, so I'm quite puzzled when I read you:
"if in-game acceleration is low it usually means mouse sensitivity is high"
Ok, the user compensate the lack of acceleration but also loose a bit of accuracy. Understandable, why not ?
"that will produce negative acceleration since internal cursor hits the edge of the window often and that process introduces delays"
From what I remember the so bad named 'negative acceleration' is just the HARDWARE mouse sending too much datas to the HARDWARE buffer when moved fast so some moves are skipped and it seems like the cursor is slowing and/or jumping.
I'm even more puzzled when you talk about the 'edge of the window'... The data is processed by the engine and then decide if it should 'step' in ( m_pitch, m_yaw etc), I don't remember any kind of 2 dimensional system projection behaving like if a cursor was moving on the screen. And producing delays relative to an 'edge'... Do you have a link ?
"it will go away if driver sensitivity is set too low while DPI is still high"
Well... Isn't driver sensibility just the same than windows sensibility (being between mouse and Windows, but not providing any acceleration, just translating hardware DPI to software DPI)? Using low driver sensitivity with high mouse DPI is quite pointless don't you think ? Because faints movements would not be converted to pointer movements : you LOOSE accuracy on slight movements even though it seems more fluid with high speed movements.
Anyway "high DPI on high driver sensitivity" just seems to me like heresy : you get crazy uncontrollable movements AND a nearly total loss of accuracy, whatever the kind of movements you do (also I remember there was a fuss a year or two ago about a very high CPU usage due to high DPI mouse movements, I don't know if it's fixed now, my own simple tests show a very negligible impact on CPU usage, don't know if some mouse/systems are still affected).
Or I really missed something obvious.
This post has been edited by versus666: 26 July 2010 - 10:11 AM
#954
Posted 26 July 2010 - 10:12 AM
--
To see negative accel punch you in the face, get a 5700DPI mouse on very high driver sensitivity[e.g. 'Speed' in Logitech settings] and try to move in game when game's sensitivity is low.
--
It can't be hardware-related only since it goes away on raw or direct input.
--
OS sensitivity is not respected in raw input according to Microsoft though it gets quirky if I choose 'use OS native drivers' in Logitech. Apparently it's not respected in raw if that option isn't checked and it's respected if it is checked. In non-raw it appears to go only with driver's sensitivity in those mice. It's quite confusing (and it doesn't seem standardized).
This post has been edited by mitsubishi: 26 July 2010 - 10:41 AM
- Optimized exe; builds of ioq3 engine for urt With GoogleTranslate, Bumpy, dmaHD, iKALiZER, Raw Mouse, Bug Fixes, ..
- Networking, lag meter, and gaming consistency guide
#955
Posted 26 July 2010 - 12:06 PM
mitsubishi, on 26 July 2010 - 10:12 AM, said:
--
To see negative accel punch you in the face, get a 5700DPI mouse on very high driver sensitivity[e.g. 'Speed' in Logitech settings] and try to move in game when game's sensitivity is low.
--
It can't be hardware-related only since it goes away on raw or direct input.
--
OS sensitivity is not respected in raw input according to Microsoft though it gets quirky if I choose 'use OS native drivers' in Logitech. Apparently it's not respected in raw if that option isn't checked and it's respected if it is checked. In non-raw it appears to go only with driver's sensitivity in those mice. It's quite confusing (and it doesn't seem standardized).
So the negative accel you describe is just a buffer overflow ?
Logitech drivers are really a mess, I owed one, I saw the strange behavior with the driver, never installed it anymore, it seems they still haven't improved that in 10 year ^^
#956
Posted 26 July 2010 - 01:51 PM
- Optimized exe; builds of ioq3 engine for urt With GoogleTranslate, Bumpy, dmaHD, iKALiZER, Raw Mouse, Bug Fixes, ..
- Networking, lag meter, and gaming consistency guide
#957
Posted 26 July 2010 - 03:01 PM
Even if it's about mouse acceleration change between modes don't cl_mouseAccelOffset and cl_mouseAccelStyle can be used to fine tune ?
This post has been edited by versus666: 26 July 2010 - 03:08 PM