EDIT : whoops not real 120fps, upon further investigation
120fps @ 720p and I tried lower resolutions too. I've only got a slower Silicon Power Class 10, the file sizes never get past 28 MB in length. So with the lower resolutions I'm just getting longer record times but hitting the same file size limits
@mpgxsvcd what are your filesizes? I know you have a faster card than I have, justing wondering if it's a card or processor problem
It isn't real 120 FPS. I recorded extremely high speed motion(Cars Whizzing past). It is making a 120 FPS file. However, it is only sampling at 60 FPS. Then it plays back the 60 FPS footage at 120 FPS which just speeds everything up.
I am going to concentrate on getting either 720p @ 60 FPS or perhaps the 1080p @ 60 FPS to work without write errors.
It sets the frame rate for all MJPEG settings. Try 480p @ 60 FPS with MJPEG. It works very well with stock bit rates.
The GH2 sensor has a 120fps mode which is used during live-view for fast focusing - maybe it can be used for recording, but then again it might also be compromised in some way.
Im testing 720p @ 120fps at 1/250 of a second and I see all individual frames. I must be crazy, cause everyone else said it doesn't work. I'm trying to verify right now there is no frame doubling, but it sure doesn't look like it.
EDIT: I see now, its recording at 30fps but putting that in a 120fps mjpeg file.
I did tests with 1080p24p, 720p60p and then HD 1080p60p and I edited the footage and slowed it all down so that they all played only 24 fps, and then I noticed the 720p60p looked really good and smooth, but the HD at 24 fps playback was a bit rough. It didn't have that smooth look the 720p60p had.
Then I discovered the same "time dilation" issue Lpowell just proved. I recorded a clock with a seconds hand going by. I even counted with it. Every 2 clicks of the clock led to only 1 seconds passing by on the record meter on my Gh23. It was the same for HD, WVGA, QVGA etc.
Clearly the underlying MJPEG process is hooked to 30 fps. When we set that element in Ptools to 60 fps or 120 fps, it doesn't change the recording speed to 2-4 times higher, it only changes the file's "playback" speed, aka 2-4 times the normal 30 fps speed. This is so sad. For a moment I really thought I had 1080p 60p!?!
Yes that's right. The editings to MJPEG in the PTool do only change the things how the data should be saved into a file. But the real thing.. the information from the sensor (in our case 30fps 720p) do not get changed. It's just like you are trying to hack or to cheat on a MMORPG but you can't because all the settings are server-side. And the same for us.. All the settings are sensor-sided and can't be changed with PTool yet. You just change how it should be saved!!
Yea it is just doubling the playback speed even for 60 FPS. I think it is a no go on the variable frames per second. However, I did notice a clear difference in resolution. The video below has 1080p MJPEG and 720p AVC-HD mixed together.
Even on youtube I would prefer the 1080p @ 30 FPS MJPEG over the upscaled 720p AVC-HD for resolution alone. What do you all think?