Capture Game Video

Fraps and PlayClaw are the most powerful game video recorders and I want to know, which one is the best. In previous game video recorders test drive I’ve tested several recorders and generally Fraps was on the first place and PlayClaw on second. This test was made on old hardware (Intel Core2Duo, GeForce 8800GT) and with old versions.

Now Fraps is updated to 3.0.3 and PlayClaw to 1.7 build 699 with new features and bug fixes. And I can test them on new Intel Core i5 and ATI Radeon 5750.

Unfortunately, Fraps cannot start on my Windows7 x64 and I don’t know why. PlayClaw is working fine on Windows 7 x64. As I know such problem can occur on some systems. I’ve tried to find solution on the web but I had no luck. So, in this comparision I will not show you DirectX 10 and DirectX 11 games.

Hardware:

• Intel Core i5 750 (2.66 Ghz)
• 4Gb DDR3
• ATI Radeon 5750 1Gb
• Screen resolution 1440×900
• Windows is installed on HDD1, Game and video recording is on HDD2

Software:

• Windows XP Pro SP3
• Call of Duty Modern Warfare 2 (Direct X9)
• Fraps 3.0.3 trial
• PlayClaw 1.7 build 699 trial

So, I’ll make simple tests with new game Call of Duty Modern Warfare 2 on map Derail. Following screenshot shows you the road where I did recordings while walking.

Every video was recorded with equal duration – 30 seconds, because it’s trial version restriction in both tools. Without recording my hardware produced 88-91 fps.

As you probably know, Fraps limits game fps to video fps in Movies settings. So if you want to play smooth while recording video you have to set video fps to higher value (60 and more). As result, you get huge AVI file. PlayClaw doesn’t limit game fps to video fps so not all tests on PlayClaw can be done with Fraps.

Low compression of PlayClaw is very similar to standard Fraps’ compression, so we will see how optimized codecs are in these recorders. Info which I need from tests: ingame fps, real video fps, final file size. I’ve recorded video in Fraps with 90 fps, then in PlayClaw with 90 fps and then in PlayClaw with 30 fps (as I said PlayClaw doesn’t limit ingame fps and in most cases 30 fps in recorder video is enough).

First test results (Ingame fps, Real video fps, File size Mb) and graph:

Fraps, 90fps: 41 – 41 – 1000
PlayClaw, 90fps 39 – 34 – 900
PlayClaw, 30fps 65 – 30 – 800

As you can see PlayClaw has a bit slower low compression in comparision with Fraps and produces lower frame rate in final file. Ingame frame rates are almost equal. But everything changes if we run PlayClaw with 30fps video. No skipped frames in AVI file and great ingame fps boost!

At this point all tests for Fraps are ended. But not for PlayClaw. This tool has additional compression mode and video capture options. Let’s make a test with 90 video fps and high compression, using different number of cores used for encoding. My Intel Core i5-750 has 4 cores without HyperTheading and I can set 1,2,3 or 4 cores to compress.

Second test results and graph:

PlayClaw, 1 core: 40 – 24 – 150
PlayClaw, 2 cores: 40 – 37 – 250
PlayClaw, 3 cores: 37 – 53 – 360

Very interesting results. Almost the same ingame fps with very different video fps for every test. Well, in low compression mode in most cases the codec performance is limited by HDD linear speed. But take a look at file sizes in high compression – now HDD is not a problem. And with increasing number of used CPU cores for encoding we’re increasing overall codec performance.

Now I will test the high compression with 30 video fps and different number of CPU cores.

Third test results and graph:

PlayClaw, 1 core: 60 – 23 – 150
PlayClaw, 2 cores: 57 – 30 – 207
PlayClaw, 3 cores: 57 – 30 – 209

Awesome results! Playable ingame fps (60), almost no skipped frames with 1 core and completely no skipped frames with 2-3 cores. It looks like my CPU and PlayClaw’s codec have some reserve in video capture task. And don’t forget about final file size – 1 Gbyte versus 150-200 Mbytes !

Conclusions

I have simple conclusions – Fraps is good but if your hardware is not old, PlayClaw takes the first place. 3-4 core CPUs rules the world today 

P.S. I’ve made the same tests for PlayClaw in Windows7 x64 and got The Result! For 30fps video PlayClaw doesn’t lower ingame fps at all! I think it’s because of improved windows code for HDD and threads control.

PlayClaw – very powerful Fraps alternative

One Response to “Fraps vs PlayClaw competition I”