GH2 Anyone successfully synced 80% slowmo up to sound?

Matthew P

Well-known member
Has anybody successfully synced the GH2's 80% slowmo mode with sound from an external recorder after conforming to 30p? No drift at all?

If no one has, could someone test it please? :D
 
I've synced audio in a music video that recorded in 80% but I think that's different than what you're looking for.
 
Well, all I want to know really is whether the 80% is infact true 29.970 fps, and that it wouldn't drift with sound synced up from something like a H4n. :)
 
Well no it's slowed down footage, so the recorder would have to be recording a little faster... the GH2 records 80% slow motion by conforming the 30fps to 24fps... if it was just recording 29.97 it should be fine but it's not, the footage is played back at a slower speed.

If I got my s&#$ straight...
 
Well no it's slowed down footage, so the recorder would have to be recording a little faster... the GH2 records 80% slow motion by conforming the 30fps to 24fps... if it was just recording 29.97 it should be fine but it's not, the footage is played back at a slower speed.

If I got my s&#$ straight...

Correct. But want I'm wondering is whether the sound will keep in sync with the video AFTER it's been conformed back up to its original (so it's believed) 30p. Hopefully it will, and there will be no drift at all.

Grisnjam, that'd be great, looking forward to hearing your findings. :) If you could test by clapping at the beginning and end of a fairly long clip (5 mins or so?), that should give us a good idea of how accurate it is, if it can sync up nicely.
 
I think then you're speeding up the footage and at that point your just using a 30fps clip at normal speed... is it worth hassle? Whats the end goal?
 
Has anybody successfully synced the GH2's 80% slowmo mode with sound from an external recorder after conforming to 30p? No drift at all?

Unless both pieces of gear are using an identical clock frequency ( not very likely ) there will always been some drift with very long takes.

For example, I shoot medical lectures that are often 60-90 minutes long, and record audio with an Olympus LS-10 recorder placed on the presenter's podium. The audio is always about 8 frames long every hour of video, so I bring the audio in to the GoldWave audio editor and use the "TimeWarp" function to shave off those extra 8 frames, then bring in the new "timed" audio which is now a perfect match to the video.
 
I think then you're speeding up the footage and at that point your just using a 30fps clip at normal speed... is it worth hassle? Whats the end goal?


The goal is 30fps. ;) 30fps beats 24fps hands down for internet use. It still looks relatively 'un-soapy", and the motion is lovely and even because computers run at 60hz.
 
You can adjust the audio lossless if you just microtune the sample rate. Just clap in and clap out, add up the samples, divide, multiply, set, render audio lossless (most good NLE DAW's do this)
For live recording, make the same calculation, then use a recorder that fine tunes the sample rate. For this, I strongly recommend any of the RME interfaces, like the FF400. After you have fine tuned the drift, just calibrate the RME.
Theoretically, you can use this to "undrift" any camcorder.
Just to be clear, this is NOT time stretching, which sounds dreadful.
 
Just to be clear, this is NOT time stretching, which sounds dreadful.

Time stretching or compression is audibly undetectable if done properly with the right software. I found that most audio editors do not have sufficient resolution to do the job properly, when you want to add or subtract 0.274 seconds out of an hour of audio to match the video. ( assuming that the clock frequency difference is linear for both )

Surprisingly the best audio editor I've found to date was also one of the cheapest. ( GoldWave costs about $19 to buy, and their "TimeWarp" function has resolution down to 0.0001 seconds. )
 
30fps beats 24fps hands down for internet use. It still looks relatively 'un-soapy", and the motion is lovely and even because computers run at 60hz.

Monitors can be set to all kinds of refresh rates depending on the resolution they're set to among other things. Don't count on people having them set to 60hz. My current AOC 22" HD monitor only goes as high as 60hz at 1920x1080, but LG makes one that will do 75hz at that same resolution. Benq has a monitor that will go as high as 120hz, but I don't know what resolutions it can work at. (Besides, I think HDMI won't output enough data to support more than a 75 hz refresh rate at 1920x1080, or at least I read that somewhere.) I haven't noticed much difference between 24P and 30p stuff I've uploaded except the 24p is a bit more filmish, but just barely. I certainly can't see any "hands down" difference bwteen the two. Not enough to worry about IMHO.
 
Has anybody successfully synced the GH2's 80% slowmo mode with sound from an external recorder after conforming to 30p? No drift at all?
:D

Yes, I've done it for a 3 minute clip and it was frame accurate - no drift. Can't say how it would hold up for really long takes, but any drift that occurs should be no different from whatever drift you'd get at 24P or 50i/60i.
 
Has anybody successfully synced the GH2's 80% slowmo mode with sound from an external recorder after conforming to 30p? No drift at all?

If no one has, could someone test it please? :D


It works:


"Here I am testing for audio drift on the GH2 when using the 80% slow-mo in 24p then speeding it up by 125% in post to get 30p. Audio is recorded with a H4n. The orignal clip was 30 mintues long and had no drift. I rendered it (at low res to save time) then reimported and cut it down to show the start and stop claps."
 
Now that's a real world test John. I'd like to see a lot more of that kind of thing around here. Simple and effective.
 
Back
Top