Frafs Benchmark Viewer
Lindsay Bigelow, an established member (raffriff) of FrapsForum.com, developed an uncomplicated tool for simply dragging and dropping Fraps frame time csv files into a program, named Frafs Bench Viewer, that quickly makes a chart. We would like to support his efforts by also using his tool and encouraging our readers to also make these simple charts.
Frafs Bench Viewer can be downloaded at SourceForge.com or from Softpedia.com as freeware. All you do is drag and drop a Fraps frametimes.csv file into the program and it automatically creates a customizable graph!
The Frafs Benchmark Viewer is a very useful an easy-to-use tool to display frame times and we encourage our readers to submit their own graphs and experiences on ABT forum. Because of the limitations of the program, we actually spend a lot of time comparing individual runs and charts made from them to pick a “representative” run. If we just averaged all of the results, there would tend to be unexplained spikes. In our case, comparing many charts can lead us to conclude if a spike or dip in the frame time is “regular” to the run or not. Jitter becomes noticeable to gamers at different levels. Some are quite sensitive to it while others are less so. A 2ms to 4ms spike is very hard to notice, even with two displays side-by-side and maybe then only in slow motion. However, a 16.7ms spike effectively cuts the frame rate in half on a 60hz monitor (with the appearance that the frame rate has suddenly dropped by 30fps for an instant) and beyond that, any spike or swing beyond 20ms is usually quite noticeable as unevenness – a stutter or jitter that disrupts the smooth delivery of the frames.
There are 1,000 milliseconds in a second, so if we divide 1000 by the specified frame rates below, we find the following times as expressed in a chart:
We also like Frafs Benchmark Viewer as a simple and easy tool to create nearly instant graphs out of Fraps frame time runs as well as the ability to show the frame time chart rankings. This function will become very useful in comparing smooth frame delivery of two video cards.
The numbers we will be comparing are:
- Average time: Average time across the entire run
- 1% time: Time it takes to draw 99% of frames
- 0.1%: Time it takes to draw 99.9% of frames
There is one issue with the way that Frafs Benchmark Viewer displays their charts. The upper range of the chart is determined by the highest spike and there may be some differences between the chart’s scale (also in the length of the runs between two video cards). In that case, the charts will be adjusted slightly as to give a more porportionate representation between them.
We plan to continue to use Fraps and Frafs Benchmark Viewer as we explore “smoothness” by measuring frame times, relating them to frame rates and perhaps posting Youtube videos of our own GTX 680 versus Radeon HD 7970 side-by-side runs at 60 fields per second, 120 fps, and 240 fps. We plan to test all of the games in our 30-game benchmark suite.
If you would like to weigh in on our future testing methods or to specifically discuss micro stutter and frame time testing methods, feel free to comment in the comment section, ask questions, make requests, or have a detailed discussion at ABT forum.
In the meantime, please continue on to Vantage as we compare fps and frame time runs of Jane Nash on SSD versus HDD.