|
|||||||||
![]() |
|
Thread Tools | Search this Thread |
![]() |
#1 |
Regular Crew
Join Date: Mar 2003
Location: Dublin, Ireland
Posts: 90
|
SYnchronizing of 2 PDX10s in a multi cam shoot
Hi All
Am about to get a 2nd PDX10 for multicam events. Apart from the old fashioned manual methods for syhcnhonising the footage from both cams, Is there a way using Timecode and U-bits on these cameras to create an automatic aligmnet in post? Many Thanks Regards P
__________________
Sony PDX10 by 2, DRS-11, Dual G5, FCX, DVDSP and Logic Plat. www.VarsityMusicVideo.com |
![]() |
![]() |
![]() |
#2 |
New Boot
Join Date: Mar 2004
Location: Chicago Subs
Posts: 19
|
why not just the old fashion way with a clap board and a smack.
line up the smack marks, all done! |
![]() |
![]() |
![]() |
#3 |
Wrangler
Join Date: Dec 2002
Location: Mays Landing, NJ
Posts: 11,791
|
I read an article where the author suggested shooting off a camera flash at the beginning of a concert for synchronization, or even using audience camera flashes for the same purpose.
|
![]() |
![]() |
![]() |
#4 |
Major Player
Join Date: Apr 2004
Location: Austin, Texas
Posts: 704
|
I second the idea of the flash.
The reason for that is that light travels much faster than sound. If the cameras are different distances from the sticks, then the sound could be off. I did a series of event videos for a cheerleading competition, and we had one camera at the top of the auditorium, and one of the floor. We tried syncing them up with audio, and couldn't figure out what was wrong at first (this was a long time ago). Finally we realized that the audio was reaching the back camera quite a bit later than the floor camera. The next time we shot we used a flash instead of a clap. We never had a problem syncing them when we used visual cues instead of audio cues. And to answer your original question, I don't think there is any way to use the timecode on the PDX10 to sync them up. -Luis |
![]() |
![]() |
![]() |
#5 |
New Boot
Join Date: Mar 2004
Location: Chicago Subs
Posts: 19
|
a flash is an interesting idea, but then where is the AUDIO cue?
a flash would only be a VISUAL and wouldn't work if you were recording to a seperate audio device doesn't a clapper have both an AUDIO and VISUAL cue? in the case of recording in a large room with enough distance between cameras to cause a delay, wouldn't the VISUAL cue of the clap be enough to sync to? (if you could see the clapper that is, or you would need a 2nd operator to zoom in to see it clear enough) I guess in a case like a cheerleading competition or something like that I would have had the cameras sync to the VISUAL cue and then use the only the audio from the closet camera. It would have the least amount of delay and echo in it. Most likely it would have the best audio with the most sound of the cheering and not a mix of cheering, crowd, and echo, unless you wanted that which then you could mix some ambient sound from the other cameras. I think this works best for most multicamera recordings, but I have seen things like a wedding done with 2 or more cameras, yet instead of using the best audio track they switch between cameras! ahhh one minute you can hear them close up, next minute you are in the back of the church listening to people talk about how long the marrage will last! PS did I use the right CUE?? hope so.. I'm tired |
![]() |
![]() |
![]() |
#6 |
Major Player
Join Date: Apr 2004
Location: Austin, Texas
Posts: 704
|
"I guess in a case like a cheerleading competition or something like that I would have had the cameras sync to the VISUAL cue and then use the only the audio from the closet camera."
That's exactly what we did. I was going on the assumption that Patrick was going to use the audio off one of his PDX10's, and not recording seperate audio. If you are recording your audio on a different device, then yes you would need an audio cue. But, if you are taking the audio off of the cameras, then a visual cue on its own will work just as well. Although, Michael does bring up a good point, which is to be on the safest side - have both. -Luis |
![]() |
![]() |
![]() |
#7 |
Wrangler
Join Date: Dec 2002
Location: Mays Landing, NJ
Posts: 11,791
|
As a rough approximation, sound travels about 1 foot per millisecond. So for each frame of 60i NTSC video that works out to ~33'. This might help you sync external audio in the camera flash scenario.
However the original question was how to sync two cameras, and the flash itself would let you do this without regard to audio. Presumably you would know how to sync the audio from at least one of those cameras anyway. |
![]() |
![]() |
![]() |
#8 |
Regular Crew
Join Date: Mar 2003
Location: Dublin, Ireland
Posts: 90
|
One option for synching audio (and hence video) between cameras is to send an audio test tone to both, but have one out of phase with the other (Some more advances mixing desks have a phase invert switch).
Back in the NLE the 2 feeds are in complete synch when the test tone becomes completely inaudible (as being out phase, hencethey cancel each other out) I still thought that the timecode on both cameras could be set equal to eachother in absolute time and then the tapes would be in synch automatically. (I was always told that this was one of the benefits of timecode!) Regards P
__________________
Sony PDX10 by 2, DRS-11, Dual G5, FCX, DVDSP and Logic Plat. www.VarsityMusicVideo.com |
![]() |
![]() |
![]() |
#9 |
New Boot
Join Date: Mar 2004
Location: Chicago Subs
Posts: 19
|
Here is a thought. I ran into this problem one day when filming a church play. Sometimes I will use my remote so that I don't stand up in front of someone. I soon found out that the other Sony camera next to me, not a PDx10 was affected by the remote also, so I had to manual do everything.
Why not use the remote to start both cameras at the same time? It will depend on there distance from the remote though, I found that the remote only travels 20 feet or so, but you can purchase a cheap IR repeater that can blast a IR signal hundreds of feet. Bad thing though you may turn on and off other Sony cameras! wouldn't that be a kicker, honey, why does the camera keep zooming in and out? I don't know, you must be doing something wrong, let me see that. next thing would be an irrate father jerking the camera around like Homer choking Bart. Can a signal be hard wired to the LANC controler to send a record single or to use the TimeCode on the camera? I'm no pro, but I noticed that one of the cables most HDTV camera have when doing multicamera shoots is a sync cable that is tied in with the engineers station and somehow link to the digital clapper. wonder if you could do the same with the LANC |
![]() |
![]() |
![]() |
#10 |
Wrangler
Join Date: Dec 2002
Location: Mays Landing, NJ
Posts: 11,791
|
That's an interesting thought about the remote, but I wonder if the time it takes to spin up the motor and thread the tape would be frame accurate between two cameras? It seems unlikely, but I suppose it's possible.
|
![]() |
![]() |
![]() |
#11 |
New Boot
Join Date: Mar 2004
Location: Chicago Subs
Posts: 19
|
how about a starting pistol?
that would give a good AUDIO and VISUAL |
![]() |
![]() |
![]() |
#12 |
Regular Crew
Join Date: Mar 2003
Location: Dublin, Ireland
Posts: 90
|
I thought that timecode meant that you could make both cameras imprint the current time and date on the (DVCAM) tape.
As long as the two or more cameras had the exact same time (down to frame accuracy) you could then align in post using timecode. Regards P
__________________
Sony PDX10 by 2, DRS-11, Dual G5, FCX, DVDSP and Logic Plat. www.VarsityMusicVideo.com |
![]() |
![]() |
![]() |
#13 |
New Boot
Join Date: May 2004
Posts: 10
|
Does anyone know how to set the timecode or user bit menus on the PDX 10? I assume this is the menu operation one needs to know if you're going to shoot a multi-camera activity and desire both cameras to have the same exact timecode.
How do you set two PDX 10 cameras to register exactly or can this be done? I had an opportunity to work on a multi-camera shoot using 3 Sony DSC 500 cameras and one of the technicians had an electronic device that downloaded the timecode to each camera. When the tapes reached the editor each camera was synchronized perfectly. |
![]() |
![]() |
![]() |
#14 |
Contributor
Join Date: Sep 2003
Location: Santiago, Chile
Posts: 932
|
PDX10 timecode
The PDX10 does not have any kind of dedicated timecode inputs, although perhaps there is a way to do it through LANC. Also, custom software could be written to control multiple cameras through Firewire... so you could even stop and start a camera array in sync. Would probably make an interesting project for a shareware developer. Reading this Graeme?
__________________
Ignacio Rodríguez in the third world. @micronauta on Twitter. Main hardware: brain, eyes, hands. |
![]() |
![]() |
![]() |
#15 |
Trustee
Join Date: Jul 2003
Location: Burlington
Posts: 1,976
|
There's no perfect solution with the PDX-10, but there are some things you can do to make it easier.
The simplest would be to set both cameras for free run code and start the code running on both cameras simultaneously. This won't be perfect, but it will be within a few frames and will allow you to keep track of the footage. Then use the other visual and audible cue methods that have been mentioned to achieve perfect sync. Don't forget to set different hours for each camera and put enough space in hours to allow for the entire time period you'd be taping, including setups and breaks. For example set camera one to hour one, set camera two to hour ten. The hour number will let you know which camera is which and the minutes, seconds and frames will let you know you're roughly in the right place. You could also get a device that extracts the timecode from the LANC port and send that signal to an audio channel on the second camera. You'd have to have an NLE that can interpret and use the TC on an audio track and you'd be losing one of your audio channels on the second camera. You can also use a timecode generator to send code to an audio channel of both cameras. Again you'd be losing audio channels for recording and you'd need an NLE that can handle it. Alternately, you can send the visual signal from a timecode generator to a larger display screen (or use a timecode slate) and shoot the slate or display screen for a few seconds with each camera simultaneously. Then keep rolling no matter what. Remember that it is possible for the ultra-short duration of a camera-flash to fall within the vertical blanking interval of the video camera and be invisible. It happens more than you'd think, so usually I flash it twice within a couple of seconds. |
![]() |
![]() |
![]() |
| ||||||
|
|