Genlock for 2-cameras doing long takes - hardcabled or Sync box

[PCMsoundie]

It's been a while for a thread on Genlock.

Actual camera sync (not just timecode identification on frame1 in the XMP metadata of a file)

 

Sure

1. live video switching needs Genlock,

2. 3D camera rigs -- cameras need Genlocked for  in phase shutter and timecode sync.

3. shooting concerts, all-day corporate conferences, or long form documentaries with takes being 1-2 hours before media change means multicameras need to stay in sync and not drift.

 

Other than the old RED camera issues and ANYTHING older than 2011....

When you've done a shoot with 2 or more cameras with very long takes. 

Were there any very productive conversations in pre-production that ended up in production requiring and agreeing to using a Sync Gen?

Yes Genlock can interfere with the shutter and scanning processes of cameras and that's why I am asking since it involves Sound/Camera/Post.

Have production wanted to do a test with 3-cameras with Sync Gen and long takes during prepro when shooting longer than 40 min. takes?

Have any of you got the dreaded call from post about camera drift even though you know you have a great sound recorder with high accuracy?

How did that go? Were you blamed because 1 or 2 cameras out of 5 were not in sync?

 

With new smaller lighter TC boxes have Genlock such as 

Timecode Systems UltraSync ONE with a temperature controlled XO it is going to be more accurate than a camera itself. 

I started using daisy-chaining 2 cameras on non-stopdown 2 hour sit-down interviews so they would both stay in sync. I was also feeding them both TC. Then I started using a AJA GEN10. I have found the GEN10 was very specific for PsF vs P settings for certain cameras before they would see it and lock.

For most modern cameras that have GENLOCK IN have some of you other soundies used a AJA GEN10 (to hardcable Genlock) or Timecode Systems UltraSync ONE when using 2 or more cameras?

Have any of you gotten a AJA GEN10 and billed for it to provides a frequency clock reference to "synchronize the cameras" on a long take show? 

 

Did you also feed AES10 to your recorder via AES IN and stay in-sync with the AES sync generator?

Have any of you experienced Camera Dept. or DIT providing a GEN10 ?

 

On shows with HDSDI hardcabled to camera Is a Genlock cable and also a TC cable really that awful these days that you've gotten a lot of pushback from Camera Dept. when camera is on a tripod in a 1 or 2 location only day?

Has anyone experienced more accuracy with hardcabled Genlock instead of using a little box providing Sync and TC?

 

 

[edward chick]

I do a quite a bit of network news multi cam interviews with takes that go for an hour or two and have always used tc boxes and a smart slate. I have never used Genlock and have never had any calls from post regarding drifting. Cameras vary from Panasonic Vaticam LT to Eva Cams, to old school style HPX 3700 series and sony fs and A7s with canon c series tossed in. Basically whatever cam the network bureau camera guy is provided and if there is a freelance cameraman on the job, he brings what he has. This has been my experience for quite a few years. Concerts are an entirely different beast, and I could see why Gen Lock would be necessary due to hours long takes.

[PMC]

As a guy who grew up seting up multicam shoots with BetaSP cameras recording ISO and sometimes feeding a switcher I always tried to connected camera genlock to a master generator. We had to for the switcher, of course. NTSC - never the same color. 

 

Today, and for the past 8 years or so, I very rarely genlock the cameras. We put the cameras in PsF, and rely on the frame syncs in the switcher to handle the timing. This is fine for the live mix but doesn't sync the cameras to each other. But I have never received a complaint about drift over time.

 

As an example, I just AC'ed a two camera documentry shoot this week for HBO using one Sony Venice and one Canon C500. It was a three hour continuous rolling interview while swapping out cards as we rolled. I put a Tentacle on each camera. I monitored the Tentacles along with the SD664 timecode on my phone. No drift. I guess I could have performed a clap or tone at the end of the shoot to test but as I say, I haven't had complaints. 

 

As an aside, if cameras need genlock I feel that is a camera department issue. When I am mixing sound, 40% of the time, I only worry about the mix and timecode. Well, and helping out the youngsters setup their shinny new camera that they haven't read the manual for, lol. 

 

Others may have a different opinion about who's responsibility genlock is and that's okay. 

[codyman]

In the last 11 years, I've never been asked for gen lock a single time (then again, I don't usually do live events).  In the last couple years I've been using Denecke JB-1 boxes and I've literally had zero sync issues and not a single complaint from post about timecode drops or anything of that sort. 

Now that I've bragged about this, I'm bound to have problems soon though...

[Philip Perkins]

With modern pro level cameras, good sync boxes and takes under 10 min or so we've quit genlocking.  For longer, esp concert set type takes, or with projections etc....still like genlock.

[RunAndGun]

In almost 25 years, I honestly don’t believe I can think of a time we genlocked field cameras for multi-cam shoots.  Even in the Beta days.  For any type of live stuff, the cameras(when using ENG style cams) are just hooked up with SDI going through frame stores/frame syncs.  These days, outside of old-dogs(like me) and engineers, most people probably wouldn’t know what you were even talking about if you asked about it.  Today, SOP is Tentacles on everything and/or slate or even just a hand clap.  No complaints.

 

I don’t think I’m exaggerating too much when I say that I think the last time I remember of a complaint about sync, was back when we were shooting Beta and someone’s camera was in NDF instead of DF.

 

 

[WaterlooNorm]

We record a lot of live event/conference productions and have gotten away with not genlocking cameras. That said, our switchers will frame sync all input/outputs and we typically record iso outputs from the switcher although we have also used the camera internal 4k files in post as well and timecode has been sufficient so far. Frame sync is needed between cameras and LED walls during live stage or volume vfx production. 

[Wandering Ear]

On the long form I've done, concerts, comedy, events, etc. I've had the conversation about genlock, and if I can get to someone who knows what I'm talking about I'm usually told something like "I'll just slip edit it".  I don't do a lot of live work, and it seems like slip editing is good enough for content going through post.  Long form in not my primary work though, so could just be the jobs themselves.

[Ontariosound]

Genlocking cameras should be a camera department decision in my opinion.  A senior DOP warned me that improper use of Genlock can

screw with the video processing, therefore I don't even bring it up on multi camera shoots.

[PCMsoundie]

On 7/15/2023 at 12:02 PM, Ontariosound said:

Genlocking cameras should be a camera department decision in my opinion.  A senior DOP warned me that improper use of Genlock can

screw with the video processing, therefore I don't even bring it up on multi camera shoots.

Would you say on any concerts, comedy performances, musical theatre performance, and live events where multicam is there is a video engineer that would handle that with long BNC cables for Genlock or else the camera dept would provide a portable Genlock device for a dolly, slider, Steadicam? Do they ALSO provide a BNC to the stationary sound table with synced from the the Master Clock system on-set providing Genlock OR as an additional sync output 48kHz Wordclock or AES11? or would that be a portable Genlock device the camera dept would provide for sound dept. only as Genlock ONLY?? Does anyone have experience with how the camera dept. would be handling this ?  No I am not talking timecode.  Would 1st Camera ACs have any experience with it or would it strictly fall under video engineer/DIT? (Not a media manager but a Digital Imaging Technician). IATSE Union or Non-Union. I was just looking to see what experiences people have had with cameras and sound being in true synchronization for 2 hour+ takes. 

[Philip Perkins]

I can only speak for the jobs I do, but in nearly all cases in recent years there is no overall video TD as there was in the '90s and before, and sync seems to fall to the sound dept.  Thus my rig becomes the master clock, having the most accurate clock+TC gen on the shoot, and my TC jams the TC boxes that are put on the cameras, like most sync sound double system shoots these days.  Even on jobs where the "record span" is really long, like hours and hours between camera reloads with audio rolling through, Genlock isn't requested or done, and the TC boxes in use don't provide it.  It seems that the cameras, looking at the external TC from the sync boxes are stable enough without external Trilevel sync being fed to them.  The idea of running separate TC+Genlock lines in addition to a monitor video line (if video transmitters aren't being used, which they usually are anymore) is kind of a non-starter on my jobs anymore.  I generally find that ACs are not very well-versed in multicam sync matters.   Further complicating a concert shoot for sync is how the show sound system is being clocked, especially if video audio is required to take a Dante etc feed from their board, with that board becoming our ADC vs our own rig.  Lotsa questions to ask at the scout!

[PCMsoundie]

On 8/2/2023 at 6:17 PM, Philip Perkins said:

there is no overall video TD as there was in the '90s and before, and sync seems to fall to the sound dept. 

Quote

really long, like hours and hours between camera reloads with audio rolling through, Genlock isn't requested or done, and the TC boxes in use don't provide it.

Quote

It seems that the cameras, looking at the external TC from the sync boxes are stable enough without external Trilevel sync being fed to them. 

Phillip since Genlock sync (and since everything is HD and above we are talking about Tri-level sync) is what this thread is about and very long takes and all video devices does fall under camera department does that include sync and the tool to do it properly: Genlock with Tri-Level Sync?

As you point out there is no technical director/DIT/video engineer usually sync of sound to picture is our department. Single camera is simple. Multi-camera is where it gets grey area.  If Genlock sync is not requested is it up to the sound department to ensure ALL cameras are in sync WITH the sound recording? No not just on the first frame with metadata stamp in the file header. Where does IATSE camera union fall on this topic of having a video engineer/DIT (NOT a media manager) to handle technical camera things like Genlock sync?  Can sound department handle Genlock sync? Is sound department on a IATSE show NOT allowed to handle Genlock at all?  Does is end with out department suggesting a slate clap at heads and tails? If the cameras are out of sync at tail slate it is not up to sound department for why it is not in sync if no Genlock sync was used? I would guess if it is being recorded using LOG and no IMAG (image magnification) then no need for video engineers (shaders) to shade cameras live and a live video switcher/TD. At what point on productions is a TD/video engineer officially needed, hired, involved in pre-production and the show is using Genlock? It seems like a discussion before a location tech scout.

 

Can anyone here speak about Netflix/HBO/Amazon comedy specials shot on HD or 4k recorded over 2-3 hours or music concerts and how sync was handled for all cameras specifically? Are the timecode boxes enough for say an 8-camera shoot recording 3 hours at that level of production? What about a mix and match of camera manufacturers and models? If there was a rackmounted House Video Sync/Master Clock Generator like with with video outputs. word clock outputs, two AES3/11 used? Did you take wordclock or AES3 [AES11 =the connection is referred to as a Digital Audio Reference Signal (DARS)] for sync? How did the pre-production discussion regarding sync and timing go?

 

Quote

 a concert shoot for sync is how the show sound system is being clocked, especially if video audio is required to take a Dante etc feed from their board, with that board becoming our ADC vs our own rig.  Lotsa questions to ask at the scout!

I think the music concert shoot and Dante Audio shared resources is a different topic though related. Since All Dante-enabled devices use Precision Time Protocol (PTP) a Dante Primary Leader Clock must be chosen in advance and cannot change during the show that is part of a discussion of sharing mic inputs as resources on the Dante network for the broadcast mixer to use along with crowd ambience mics to mix for a livestream and record separate from the event house PA audio. That all gets into sometimes more than 64 ISO tracks, DANTE/MADI routing, backup multitrack recorders and also secondary networks and more layers. A separate thread should be for that and Dante master clock.

[Philip Perkins]

I "handled" genlock sync for many years: our earlier TC boxes like Ambient "greenies" and Denecke SBT all made trilevel sync as well as TC.  That's what we put on the cameras, and, yes, we soundies were expected to do that.  WHO supplies the sync boxes is always a matter for discussion--it seemed like the bigger (LA based for me) jobs got the sync boxes with the camera packages, the local (non LA) jobs expected me to to deal with it, so I did.  A truly huge number of jobs were done this way, including multicam long-span recordings of concerts, events etc., successfully.  The Dante issues are VERY related to all this now, in that it is increasingly unlikely that a video audio recordist will be able to get any analog feed from a large FOH sound system other than a 2-mix anymore.  The only way to derive a full track-per-mic split is probably via getting on their Dante (or MADI) network.  At that point the clock of the FOH system has to be questioned, since in many FOH setups the board looks at its own internal clock, only, and that board is now your ADC. 

[tourtelot]

I believe that in a Dante rig, no matter what is listed as the "Leader" clock, it is that device's Dante card that is providing the actual timer.  Dante card clocks have been rock solid across the board in my experience. Not ALWAYS the case ("sync from external") but there is barely a need to do that that I can imagine.  And even then, I believe the Dante clock re-clocks the device's sync anyway.  I'd be interested in others' experience as well.

 

D.

[Philip Perkins]

I'm not really clear on what you are saying here, can you elaborate?  If an FOH board is running on it's internal clock, and has a Dante card for communicating with its stage boxes and other systems, are you saying that the master clock of that rig is the clock in the console's Dante card?   That the board's ADC (in the stage boxes) is under the Dante card's clock?  And the clocks in Dante cards are "video audio accurate", ie what we're used to re: sync sound on movie jobs?  And that a movie sound rig that has taken a Dante feed from the FOH Dante system could then successfully jam sync TC boxes etc and the sync would hold as well as if the video audio recorder (SD, Zax etc type), now creating the timecode for the video shoot aspect of this gig, were the master clock?

[Matthew Steel]

Dante clocking is very similar to using word clock, at least in the sense that every device must ultimately derive its clock from the same source.  Dante has the nice feature that it can choose the "best" Dante device to be that source and all the Dante devices on the network will sync to it.  So, in the scenario with a FOH board and Dante card, one must clock the other.  Generally the recommendation I have heard is that the Dante network be the master for all other devices, including the FOH board, monitor board, etc.  As I understand it the primary advantage is the ability to choose the "best" of the devices as clock master, and the fact that if the elected "Leader" device goes away the Dante network will elect a replacement and everybody will still get a clock signal without intervention and perhaps even without any audio interruptions for the remaining devices.

 

However, the quality of the clock is a separate discussion from simply getting a clock.  And, it is important to note whether we are talking about jitter (sound quality)  or frequency accuracy (drift with respect to picture).  I think @tourtelot may have been talking in the context of jitter, but the main discussion in this thread seems to be more a question of drift.

 

Video equipment (and audio equipment targeted toward use for picture) typically has a clock that is accurate to within fractions of a part per million.  On the other hand, typical FOH or studio equipment does not require that kind of absolute accuracy and the clock specs I have seen have been 10, 20, even 50ppm.  Notably, I have never seen a frequency accuracy spec published for any Dante interface card for a mixer, or for any of Audinate's chipset modules.  I haven't looked at all, but anywhere I looked the best I could find was a statement to the effect that the clock was high quality.  Given that lack of information, I cannot expect that the ppm spec is any better than that of a good FOH board.  So, @Philip Perkins concern seems valid to me.

 

If the camera has a +/-0.1ppm clock, and the Dante audio is clocked by a +/-10ppm FOH mixer, then they may be jammed to the same timestamp at the beginning of a shoot but will drift as much as a frame per hour.  Fine for some scenarios, definitely an issue for others.

[tourtelot]

Mathew, you make a very good point.  I will reach out to tech at Audinate and see if I can get an accuracy spec outta them.  More soon.

 

And Philip, in re-reading my post, it might seem that I was saying "Sync from external" was less reliable that using a Dante card clock in one device as the master.  Not what I meant.  Only that "sync from external" is more rarely used, at least in my experience.

 

Maybe the answer IS to "Sync from external" for the Dante network using the most accurate clock in the house?

 

BTW, here is a link for an extensive article about using Dante to shoot a music event across great distances.  I have not digested it all but it might be interesting.

 

https://www.audinate.com/webinar-peltrix-takes-dante-over-distance-with-blue-note-entertainment-group

 

D.

[tourtelot]

And BTW, I was speaking about drift.  But things like house consoles and stage boxes on a Dante network get their clocks sync'd from their Dante cards.  Dante makes it a point of controlling digital clocks in their network unless told to do otherwise.

 

That is where "sync to external" comes in.  I just talked to the folks at Grace (I use their preamps) about the accuracy of the digital clocks in m108s.  Waiting for an answer.  I can, and sometimes do, sync my Dante network externally from one of my m108's clock thinking (and maybe falsely) that the Grace clock should be of very high quality.  I will know soon enough.

 

Obviously, for me, even providing audio for a stream, I am not too concerned and never have had any push-back about sync issues.  Most of the streaming video guys I work with don't know or don't care.  I have never had a discussion.  Everyone seems more interested in delay.  FW that IW.

 

I just heard back from Alex at Grace.  He says the clocks in the m108 preamps are spec'd at 10ppm with a 15ppm frequency stability.

 

How accurate or gen-lock clocks in the real world?

 

D.

[Philip Perkins]

I am old school enough I guess that if I was to get on a multicam concert gig where I was expected to be the source of TC for cameras (via sync boxes with accurate clocks), I would want to know how to align whatever I was using as the master TC generator to make that TC, as well as word clock for devices I'm using that need to do their own ADC (ie feeds not coming from FOH, like audience etc mics).  I guess the answer might be that this is an "in for a nickel, in for a dollar" sort of deal, where said TC generator/masterclock for audio etc would have to be a Dante device also, so that everyone and everything was on the same clock.  I don't like it--too much of a rig like this is out of my control: I'd rather start with the analog sources pre-FOH (like analog splits) and be completely independent of what they are doing, but that is an increasingly unpopular idea with FOH folks I've been finding.  And anything Dante is pretty damn expensive.

[tourtelot]

Lots of advantages to a Dante network that mitigate the cost, not the least of which is I can carry 1000' of fiber to a venue in one hand that will allow me to send and receive hundreds of tracks of audio.  Think of 1000' of, even, 24 channel copper and you can see why I consider anything but Dante almost never.

 

D.

[tourtelot]

15 hours ago, Matthew Steel said:

If the camera has a +/-0.1ppm clock

Is this a real number?  Can you cite a device(s) spec'd at 0.1PPM?  Sincere question.  I find what are called "high precision digital clocks" spec'd at 1PPM.  There is a device that is touted as specifically being a "Dante Leader Clock" and is rated at 1PPM for clock accuracy.  One part per million over the length of one year.

 

No affiliation: https://studio-tech.com/wp-content/uploads/2022/04/ds_m5401a_5.pdf

 

And what clock accuracy is REALLY necessary for any work that you or I might do.  I have never seen any program that doesn't cut away to some other shot over the long length of time that has been mentioned.  At least in my world, picture gets cut to sound so a PPM among friends shouldn't matter.  Please correct me if my thinking is off.

 

Oh and Philip's thoughts about joining two Dante networks up for a multi-camera concert shoot?  Me neither!  I don't want the hassle of a pissing contest about who controls the joint network.  It was bad enough in the day arguing about who took the transformer side of an analog microphone split.  I'll use my analog splitter and take the transformer side.  You power the mics.  Or the other way around.  I really don't care.

 

D.

[tourtelot]

Antelope claims 0.02PPM for its Antelope Audio 10MX Rubidium Atomic Master Clock.  Almost $7K.

 

D.

[Philip Perkins]

I think Brainstorm has a pretty sexy box for this too (https://brainstormtime.com/products/dante-dxd-firmware-update/), that has a boffo clock, makes TC, WC, Dante etc etc, and is also Mighty Expensive.  I get the argument re: lightweight snakes for stage boxes etc w/ Dante, but 1000' runs don't happen much for me any more.  And in truth, if some FOH people are talking a Dante feed for me they are also probably saying "why do we need him?" to each other, and I'll be bumped off the job forthwith with them recording the output of their board however, themselves.

[tourtelot]

1 hour ago, Philip Perkins said:

 I'll be bumped off the job forthwith with them recording the output of their board however, themselves.

Yeah and they make movies with iPhones these days.  What are ya gonna do?    We are dinosaurs these days; the kids have taken over.

 

D.

[OB1]

@tourtelot Any feedback from Grace yet? I use their equipment as well.