Panasonic AF 100 + PIX 240 delay

[cinetj]

Hello everyone,

I'm going to start a shoot soon using 2 Panasonic AF 100 and 2 Pix 240 as the recording device for the image (the connection between the camera and the Pix is made by SDI). The workflow proposed is to jam sinc the 2 Pixes with my Nomad (a scratch track would also be provided for each camera - in case things go wrong with TC, Pluraleyes would be an option).

Yesterday, I ran this test with the Nomad, the AF 100 and the Pix. I jammed timecode from the Nomad to the Pix and using the electronic slate that comes with the Nomad pointed to the camera and marked a start beep and, after 10 minutes or so recording, an end beep. Watching the result, I observed that there is delay between the camera and the Pix of 7 frames. There is no TC drift since the delay is there from the start and it remains the same until the end of the recording.

We can warn post to assume a timecode offset of 7 frames and still get things going. But I would like also to know if someone here has ever experienced something like this using this setup.

[Mark Vesterskov]

I think this is to be expected using an external recorder - there is some processing time for the video being recorded to the drive. Does the PIX have an offset-option?

More reading here:

http://www.dvinfo.ne...ync-offset.html

EDIT:

And here:

http://forums.creativecow.net/thread/8/1152759

[rafaeldelauz]

You can adjust the delay on the pix to match the difference between the Sdi out and XLR input, it's in the audio menu. it goes from 0 to 200 mS, and its independent to each input on the PIX

[cinetj]

You can adjust the delay on the pix to match the difference between the Sdi out and XLR input, it's in the audio menu. it goes from 0 to 200 mS, and its independent to each input on the PIX

But even at the maximum delay rate of 200 ms there wouldn't be enough compensation for a 7 frames image delay (at a 29.97 frame rate, something like 230 ms). We would still be like 1 frame out (even though 1 frame difference is still tolerable for sinc).

Since I'm recording the audio at Nomad and the scratch audio track can be sent directly to the camera (therefore in sinc with the image), all I have to do is tell post about the 7 frames delay between the nomad and the pix timecode.

[cinetj]

But even at the maximum delay rate of 200 ms there wouldn't be enough compensation for a 7 frames image delay (at a 29.97 frame rate, something like 230 ms). We would still be like 1 frame out (even though 1 frame difference is still tolerable for sinc).

Since I'm recording the audio at Nomad and the scratch audio track can be sent directly to the camera (therefore in sinc with the image), all I have to do is tell post about the 7 frames delay between the nomad and the pix timecode.

And just to point it out the obvious: 7 frames is a whole lot! Maybe there is something else in this equation. We did this filming the nomad slate and I've heard in here that the lcd screen from Nomad has some latency also but I don't know of how much.

P.S: Mark, I've looked the links you sent and in of them someone pointed out a 92 ms delay from the AF 100 to the Pix - less than half of what i'm having.

[wolfvid]

call sounddevices - they know their shit wolf

[RadoStefanov]

ask camera department to figure it out. the name of the company "sounddevices" does not mean we have to deal with their video gear.

i was never asked to trouble shoot Aja recorder.

[VASI]

ask camera department to figure it out. the name of the company "sounddevices" does not mean we have to deal with their video gear.

i was never asked to trouble shoot Aja recorder.

[Matt Mayer]

Since audio is coming from the camera, it isn't an A/V offset issue, but a timecode offset issue.

Do not set any audio delay, you will knock the scratch audio out of sync with the picture.

The camera department needs to set a timecode offset. Encourage them to read Page 26 of the PIX 2.00 manual. Timecode offset of +/- 10 frames is available to make the file start times match.

[cinetj]

Since audio is coming from the camera, it isn't an A/V offset issue, but a timecode offset issue.

Do not set any audio delay, you will knock the scratch audio out of sync with the picture.

The camera department needs to set a timecode offset. Encourage them to read Page 26 of the PIX 2.00 manual. Timecode offset of +/- 10 frames is available to make the file start times match.

That seems to solve the problem, Matt. Thank you!

[badger]

Hi All

I have a gig on Monday where I'm recording straight into the pix to avoid using the Red Epic I/Ps. I don't suppose anyone knows what, if any offset would be required on the Pix Inputs to ensure sound was in sync with the picture on the camera. I'd be going into the Pix analogue from my 442.

Many thanks in advance.

Alex

[Matt Mayer]

Alex,

It is safe to say that some audio delay will be necessary for audio direct to PIX with SDI video input.

The offset is dependent on several things: Resolution, Frame Rate, Scaler and Codec, so it's hard to have a hard and fast rule for the offset required. The best thing is to record a slate clap in the exact config you are using and then take a quick look at the file in an NLE.

Not very practical in a lot of situations, so I usually figure on a 3-4 frame delay. Once I set the delay, I record a test and take a look and listen to see if lip sync looks reasonable. And remember the number of ms per frame will be different between 23.976 and 29.97. At 29.97, each frame = ~33 ms. At 23.976 each frame = ~42ms

[badger]

Hi Matt

Thanks for the reply, much apreciated. Turns out the director balked at the idea of recording audio seperately in any form and we are now using a F3. Makes things easier.

Thanks again.

Alex