BUG in 353.30+ (incl. 384.76): 2560×1440 instead of 4K after Stand-by
[size=XL][color=orange]Summary[/color][/size] With nVidia drivers 353.30+ under Windows 7, the OS-level resolution sometimes switches from native 4K to 2560×1440 after monitor’s Stand-by, and 2560×1440 becomes the MAXIMUM available resolution. DDU (and then installing solely graphics driver and PhysX) does NOT help. What [i]temporarily[/i] helps is turning the monitor off/on, or rebooting the PC, or reinstalling the nVidia driver. Older nVidia-driver versions such as 353.06 and 350.12 are NOT affected. But they are incompatible with GeForce Experience 3.x: recording videos and taking screenshots don’t work, so I’m forced not to use GFE 3.x. [size=XL][color=orange]Details[/color][/size] All versions of the nVidia driver since [b]353.30[/b] (including 353.30 itself, the latest 384.76, and all versions in between) have a [b]serious bug[/b]: once the monitor is returned from Stand-by mode, the OS-level screen resolution gets [b]2560×1440[/b] (instead of the native [b]3840×2160[/b] which the monitor was running at before going to Stand-by), and this is the [b]maximum[/b] available resolution until I [b]turn my monitor off and then on, or reboot my PC, or reinstall the driver[/b]. This happens not always, but sometimes: e.g. I was testing 361.43 for 3 days before the bug has been reproduced. [size="S"][i]Fwiw, Stand-by mode is a mode initiated by operating system after some time of idle; in this mode, the monitor’s backlight is turned off.[/i][/size] I’m currently forced to use the latest non-buggy nVidia driver 353.06. Previous versions — e.g. 350.12 and 352.86 — are also [b]not[/b] affected by the bug. Please fix the issue. 353.06 is [b]not[/b] buggy, the next 353.30 [b]is[/b] buggy: it should be easy to find what exact resolution-related change in 353.30 compared with 353.06 has introduced the bug. DDU (and then installing solely graphics driver and PhysX) does NOT help with this issue. My hardware: [list] [*] [i]4K monitor:[/i] [url=http://www.dell.com/en-us/shop/accessories/apd/210-agnk]Dell P2415Q[/url] connected via DisplayPort. [*] [i]Graphics card:[/i] [url=http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-650ti-boost]GeForce [color=orange][b]GTX 650 Ti BOOST[/b][/color][/url] (2 GB) (Gigabyte [url=http://www.gigabyte.com/Graphics-Card/GV-N65TBOC-2GD]GV-N65TBOC-2GD[/url]). [*] [i]Mainboard:[/i] [url=http://www.gigabyte.com/Motherboard/GA-Z68X-UD4-B3-rev-10]Gigabyte GA-Z68X-UD4-B3[/url]. [*] [i]CPU:[/i] [url=https://ark.intel.com/products/65525]Intel Core i7-3770T[/url] (4 cores). [*] [i]RAM:[/i] 16 GB (4×4 GB, Hynix). [*] [i]Power:[/i] [url=http://www.corsair.com/en-ww/professional-series-gold-ax850-80-plus-gold-certified-fully-modular-power-supply]Corsair AX850[/url] (850 W, Seasonic-based). [*] [i]OS:[/i] Windows 7 Home Premium SP1 (build 7601) (64 bit) at 200% OS-level zoom. [/list] Thanks.
Summary

With nVidia drivers 353.30+ under Windows 7, the OS-level resolution sometimes switches from native 4K to 2560×1440 after monitor’s Stand-by, and 2560×1440 becomes the MAXIMUM available resolution.

DDU (and then installing solely graphics driver and PhysX) does NOT help. What temporarily helps is turning the monitor off/on, or rebooting the PC, or reinstalling the nVidia driver.

Older nVidia-driver versions such as 353.06 and 350.12 are NOT affected. But they are incompatible with GeForce Experience 3.x: recording videos and taking screenshots don’t work, so I’m forced not to use GFE 3.x.

Details

All versions of the nVidia driver since 353.30 (including 353.30 itself, the latest 384.76, and all versions in between) have a serious bug:

once the monitor is returned from Stand-by mode, the OS-level screen resolution gets 2560×1440 (instead of the native 3840×2160 which the monitor was running at before going to Stand-by), and this is the maximum available resolution until I turn my monitor off and then on, or reboot my PC, or reinstall the driver.

This happens not always, but sometimes: e.g. I was testing 361.43 for 3 days before the bug has been reproduced.

Fwiw, Stand-by mode is a mode initiated by operating system after some time of idle; in this mode, the monitor’s backlight is turned off.

I’m currently forced to use the latest non-buggy nVidia driver 353.06. Previous versions — e.g. 350.12 and 352.86 — are also not affected by the bug.

Please fix the issue. 353.06 is not buggy, the next 353.30 is buggy: it should be easy to find what exact resolution-related change in 353.30 compared with 353.06 has introduced the bug.

DDU (and then installing solely graphics driver and PhysX) does NOT help with this issue.

My hardware:

Thanks.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#1
Posted 12/08/2015 04:58 PM   
The latest 361.43 (2015.12.21) driver is still affected by the bug.
The latest 361.43 (2015.12.21) driver is still affected by the bug.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#2
Posted 01/01/2016 08:45 PM   
The latest 376.33 (2016.12.14) is still (again?) affected. The bug didn’t appear for a while, but now has appeared again.
The latest 376.33 (2016.12.14) is still (again?) affected. The bug didn’t appear for a while, but now has appeared again.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#3
Posted 12/24/2016 03:12 AM   
The latest 378.66 (2017-02-14) is affected.
The latest 378.66 (2017-02-14) is affected.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#4
Posted 02/15/2017 10:22 PM   
The latest [url=https://forums.geforce.com/default/topic/1002635]381.65[/url] (2017-04-06) is still affected by the 2560×1440 bug.
The latest 381.65 (2017-04-06) is still affected by the 2560×1440 bug.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#5
Posted 04/08/2017 05:20 PM   
The latest 381.89 (2017-04-25) is still affected by the 2560×1440 bug.
The latest 381.89 (2017-04-25) is still affected by the 2560×1440 bug.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#6
Posted 04/27/2017 05:57 PM   
This comes down to an issue with the connection handshake between card and tv and something your TV is not doing right (or simply not fast enough, where the case of a failed handshake). You do not mention what cable is used to connect the screen (in this thread). A non vesa compliant TV or cable can (many cases does) result in the wrong resolution when coming out of sleep. There are a large number of cases of this happening because of screen being in deep sleep and not waking fast enough. Windows will drop back to its current virtualised desktop setting in this case (Note i said WINDOWS.) as the display/monitor and reoslution api is managed by windows itself. nvidia are very particular about keeping to standards, hence the lack of gen3 pci-e on sandybridge-e mainboards. An intermittent delay in the wake up behavior of your TV would be the only possible reason you sometimes do not experience the issue.
This comes down to an issue with the connection handshake between card and tv and something your TV is not doing right (or simply not fast enough, where the case of a failed handshake).

You do not mention what cable is used to connect the screen (in this thread).

A non vesa compliant TV or cable can (many cases does) result in the wrong resolution when coming out of sleep.
There are a large number of cases of this happening because of screen being in deep sleep and not waking fast enough. Windows will drop back to its current virtualised desktop setting in this case (Note i said WINDOWS.) as the display/monitor and reoslution api is managed by windows itself.

nvidia are very particular about keeping to standards, hence the lack of gen3 pci-e on sandybridge-e mainboards.

An intermittent delay in the wake up behavior of your TV would be the only possible reason you sometimes do not experience the issue.



In Memory of Chris "ChrisRay" Arthington, 1982-2010

Specs:
CPU:Intel Xeon x5690 @ 4.2Ghz, Mainboard:Asus Rampage III Extreme, Memory:48GB Corsair Vengeance LP 1600
Video:EVGA Geforce GTX 1080 Founders Edition, NVidia Geforce GTX 1060 Founders Edition
Monitor:BenQ G2400WD, BenQ BL2211, Sound:Creative XFI Titanium Fatal1ty Pro
SDD:Crucial MX300 275, Crucial MX300 525, Crucial MX200 250
HDD:500GB Spinpoint F3, 1TB WD Black, 2TB WD Red, 1TB WD Black
Case:NZXT Phantom 820, PSU:Seasonic X-850, OS:Windows 7 SP1
Cooler: ThermalRight Silver Arrow IB-E Extreme

WIP:
CPU:Intel Xeon x5660, Mainboard:Asus Rampage II Gene, Memory:16GB Corsair Vengeance 1600 LP
Video:EVGA Geforce GTX 680+ 4GB, Palit Geforce GTX 550ti
Monitor:Pending, Sound:Pending
SDD:Pending
HDD:Pending
Case:NZXT Guardian 921RB, PSU:Corsair 620HX, OS:Windows 7 SP1
Cooler: ThermalRight True Spirit 120M

#7
Posted 05/02/2017 08:13 AM   
384.76 (2017-06-29) is affected too.
384.76 (2017-06-29) is affected too.

Nonblurry integer-ratio upscaling:
Article | Demo | Petition | GeForce Thread | AMD Thread

#8
Posted 06/30/2017 06:04 PM   
Scroll To Top