GH4 I have no confidence in free run timecode from the GH4

M

mcbob

Veteran
[note: edits added in brackets]

There was a previous glitch mentioned here:
http://www.dvxuser.com/V6/showthread.php?326139-Timecode-Glitch&highlight=gh4+timecode

I recreated that as well, but seems that's not all.

Based on another user's issue in the AF100 forum, I did a bit of fiddling to determine how best to sync a GH4 with another free-run timecode enabled camera (in this case, an AF100) for prospective multi cam shoots. Basic findings: the GH4's free run TC mode is not reliable.

I did the usual free run setup: both cams in the same rate/frame drop (in this case, 1080p23.976), free run, cue up TC reset and hit as close to simultaneously as possible. While there's often a resulting +/- error, it's usually always close enough and remains consistent throughout the timecode run. GH4 firmware v2.0, AF100 firmware v1.24. The YAAAAAAAAAAGGGHHHHH!!!!!!!!! was not utilized.

I let the AF100 roll away with its timecode and shot the display with the GH4 to see if the time codes maintained the same spread. 5 tests were run in UHD .MOV, 1080-200 .MP4, and 1080 AVCHD. I offloaded the GH4 clips, threw them into Premiere, and tallied the displayed timecode difference.

Good news is [WITHIN INDIVIDUAL CLIPS] there was no drift in timecode difference from the first frame to the last: there was a direct 1:1 correlation. Average wait time between recorded clips was 5 seconds. The longest clip recorded were one UHD and one 1080-200 of approximately 5 minutes each.

[However, there was a change in the timecode difference each time a new clip was recorded on the GH4]

The difference in displayed TC in the UHD .MOV clips were:
1 - GH +09
2 - GH +05
3 - GH +19
4 - GH +04
5 - GH +15

The difference in displayed TC in the 1080-200 .mp4 clips were:
1 - GH +13
2 - GH +15
3 - GH -01
4 - GH +04
5 - GH +11

AVCHD was impossible to test, because Premiere did not preserve the timecode metadata for the GH4 AVCHD. Likely I didn't import this properly.
[edit: I cannot find a way to import GH4 AVCHD into Premiere CC with timecode metadata. I do not have this problem with AF100, HMC150, or C100.]

In the GH4, there is a noticeable delay when hitting the record button until the timecode begins to tally. This delay occurs when using either the shutter button or rear video button to initiate recording. Since there was no observed drift in long clips (5 minutes), I'd surmise that the issue arises from the factors that cause this delay.

My conclusion is that any multi camera shoot with the GH4 will have to rely upon standard visual/audible sync methods, and that Free Run Timecode with the GH4 is only useful to very roughly synchronize clips from different cameras (the difference did stay within one second, at least). It is not as reliable as, say, free run on multiple AF100's/HMC150's/etc.


Curious what others may have found regarding free run or TOD TC on the GH4
 
Last edited:
less inspiring follow-up:

I checked again, and can't for the life of me get Premiere CC to import GH4 AVCHD with the timecode metadata. No problems with the AF100. Le sigh.
 
First should say "there was no CHANGE in time code difference." So if a clip began with G@+05, it ended with G@+05
 
Thanks, but I'm still a bit confused. If two clips started with a 5 frame difference in timecode due to the vagaries of manually setting them, and both clips ended with that same 5 frame difference (a situation I would entirely expect to happen), then what is it about the GH4's timecode you find unreliable? Seems to me it's working precisely as it should.
 
If I set free run time code with my AF100 and HMC150 or two AFs, they might have a few frames offset that remains consistent throughout the work. I can stop and start either and that offset doesn't only not drift within a clip, but doesn't drift between clips.

In this test, with free run on the AF and GH, the TC offset changes each and every time I trigger the GH4. When syncing clips on a timeline, I can't just multiclip them by TC with a single known offset... I'd effectively have to go in and manually sync each and every clip. It rather defeats the purpose of free run TC.

Ifothers have had different experiences, I'd be eager to hear them.
 
mcbob said:
In the GH4, there is a noticeable delay when hitting the record button until the timecode begins to tally.
Click to expand...

Thanks for sharing you results!

FYI starting with the GH2 I noticed a delay between hitting record & the camera actually capturing footage. Similar problem when cutting the camera where the last bit of footage was lost also. Very frustrating in documentary situations.
 
No problem... it's a remarkable little camera, but sadly not without annoyances like these.

On a secondary follow-up test this morning, I placed AF100 and GH4 on free run to confirm that the AF didn't drift between clips if I kept the GH4 running and stopped/started the AF100 repeatedly.

Long story short, it didn't. BUT...

I had to take the extra step of importing the GH4's .MOV file to verify this because the GH4 doesn't display the actual timecode on the monitor while recording. It actually holds a value (say, 00:00:13;21) for 2-4 frames before displaying the next value, so the previous method of just checking through the recording on the interrupted clips didn't work.

Curious why timecode on this little guy is such a mess...
 
mcbob said:
the TC offset changes each and every time I trigger the GH4.
Click to expand...

Ah, now I get it. That is weird.

Well, I can't solve that issue, but I can offer my standard advice.

When shooting multicam, don't stop recording until the event is over. Doing so makes post much more difficult.
 
That's good advice on shorter events (>~2hr), but for longer events, doc-type-stuff mostly, free run classically helps bridge those times when one of the cameras has to swap batteries or cards without losing sync.

It's not a deal-breaker by any means, I can get by just fine with visual and audio slating cues in most situations, but am mildly miffed that its timecode in general is broken on several levels.
 
You must log in or register to reply here.
Back
Top