Exploring “Frame time” measurement – Part 1 – Is Fraps a good tool?

Posted by apoppin On December 30, 2012 5 Comments

Test Configuration

Test Configuration – Hardware

Intel Core i7 3770K (overclocked to 4.5GHz); Turbo is on.
EVGA Z77 FTW motherboard (Intel Z77 chipset, latest beta BIOS, PCIe 3.0 specification; CrossFire/SLI 16x+16x using Plex chip.)
8GB Kingston DDR3 PC1866 Kingston RAM (4×2 GB, dual-channel at 1866MHz; supplied by Kingston)
Power Color Radeon HD 7970 (3GB, overclocked to GHz Edition speeds 1050/6000MHz)
Nvidia GTX 680 (2GB, 1006/6008MHz, reference clocks), supplied by Nvidia
Onboard Realtek Audio
Two identical 500 GB Seagate Barracuda 7200.10 hard drives configured and set up identically from drive image; one partition for Nvidia GeForce drivers and one for ATI Catalyst drivers
Kingston 240GB HyperX SSD used for AMD partition, supplied by Kingston.
Kingston 240GB HyperX 3K SSD used for Nvidia partition, supplied by Kingston
OWC 240GB Mercury EXTREME Pro 4G used for both AMD and Nvidia, on loan from OWC.
Thermaltake ToughPower 775 W power supply unit supplied by Thermaltake
Thermaltake Overseer RX-I full tower case, supplied by Thermaltake
Sapphire Vapor-X CPU cooler on loan from Sapphire
Philips DVD SATA writer
ASUS VG278 27″ Light Boost enabled 3D Vision 2 ready 120Hz display

Test Configuration – Software

Fraps 3.5.9 full version
ATi Catalyst 12-11 Beta11 drivers; highest quality mip-mapping set in the driver; use application settings; surface performance optimizations are off. Latest CAPs used.
Catalyst Control Center used to set power draw to maximum for the HD 7970; clocks, voltage and fan profiles are stock
EVGA PrecisionX used to set power draw to maximum for Nvidia cards; clocks voltage and fan profile are stock
NVIDIA GeForce WHQL 310.70 High Quality
Windows 7 64-bit; very latest updates
Latest DirectX
All games are patched to their latest versions.
vsync is forced off in the control panels.
Varying AA enabled as noted in games; all in-game settings are specified with 16xAF always applied; 16xAF forced in control panel for Crysis.
All results show average frame rates and frame times
Highest quality sound (stereo) used in all games.
Windows 7 64, all DX9 titles were run under DX9 render paths, DX10 titles were run under DX10 render paths and DX11 titles under DX11 render paths.

The Benchmarks

DX11

STALKER, Call of Pripyat
Metro 2033
Far Cry 2

Let’s head to the charts and see the frame time comparison between Fraps results and the Call of Pripyat and Metro 2033 engine results.

Pages 1 2 3 4

TOPICS FPS frafs bench viewer frame rates frame time Fraps gtx 680 HD 7970 measurement smoothness

ocre

wow. These results are extremely similar. More so than i ever thought. Fraps result do differ by a tiny amount but this is because the game engine and fraps grab their data completely different. Fraps monitors DX/openCL calls after the fact and it being an external program throws in its own latencies, however small they may be. But through all this the trend graphs show near identical trends for fraps vs the internal game engines. Its amazingly accurate and the results are like mirror images.

Some people may not know the point in these testings so to be clear i might add it. Most games do not have any way to gather frame times from the engine at all. Fraps can do this rather easily anytime while playing or benchmarking. The question was whether fraps was accurate enough to rely on at all. Whether the spikes meant what they were being represented as. Having these test with the game engines producing near identical trend graphs gives high credence to fraps being an acceptable tool to capture frame times. We also have the reviewers experience which seemed to align with the frame time data very nicely. He seen the small hitching or studdering that the trend charts represent.
Pancake

It’s remarkably easy, actually. Fraps and in game timing work by time stamping the ‘swap buffer’ command which indicates end of frame. As the timestamp is an internal CPU counter the only difference would be either side of procedure call so both would differ by at most a few nanoseconds. Fraps would be at most a few kilobytes of code mostly being a stub for forwarding calls to the ‘real’ driver.
Pingback: Exploring “Frame time” measurement – Part 2 – Is the SSD “smoother” than the HDD in Gaming? | AlienBabelTech
Pingback: Exploring “Frame time” measurement – Part 3 – the GTX 680 versus the HD 7970 GHz ed. | AlienBabelTech
Pingback: Resident Evil 6 Benchmark – Frame Time measurement – GTX 680 vs. Titan vs. HD 7970 GHz ed. | AlienBabelTech

Exploring “Frame time” measurement – Part 1 – Is Fraps a good tool?

ABT News

The Witcher 3 will have 36 different endings

E3 2013 Wrap-up

140 Characters Isn’t PR, It’s a Job Opportunity for Real Writers

ABT News for 6/12/13 – Mid-E3 Edition

Does the Kingston “Beast” 2133MHz DDR3 live up to its name?

Web News

Genius Introduces the F-1000 Arcade Stick

Thermaltake Group Presents World Premiers with Innovative Technologies at COMPUTEX Taipei 2013

Kingston Technology Ships Intel XMP-validated HyperX Memory to Support new ‘Haswell’ Platform

GALAXY LAUNCHES THE GEFORCE GTX 770 GC – FAST JUST GOT FASTER

NVIDIA GeForce GTX 700M-Powered Notebooks Shatter Records for World’s Fastest and Thinnest Gaming Notebooks

Categories