HPX250: Genlock/Sync Questions

B

Bassman2003

Veteran
Hello,

I have been shooting live shows over the past 10 years with a variety of cameras. I use three cameras recording independently and do all of the editing after the fact using Edius.

Right now I have two Varicam "H" models and am looking at a new third camera for purchase (HPX-250). Genlocking/syncing the cameras is on my mind so I thought I should ask some questions to those who know more than me.

1) For my uses, drift between the cameras is the only concern. I have never used timecode. Is Genlock the best way to avoid drift?

2) Can I genlock just using Varicams and an HPX250?

3) I have read about Genlock and have a grasp of what it does, but out in the field...does it work easily? So many things sound great on paper but have issues in practice.

4) Is there a wireless way to work with Genlock?

5) Is there any way to sync an AC-160 with the Two Varicams?

In the DV days I used three matching cameras and drift was not an issue. Since going HD I have bounced around between cameras manufacturers and have noticed I need a way to stop drift over long shows.

Thanks for your input!
 
The combination of genlock and timecode sync is in fact exactly what you want and would solve your problems. But you can't really do it perfectly with an AC160, as it doesn't have a genlock. Genlock synchronizes the frame creation process so they will be exactly in sync. You can timecode sync the Varicams to the AC160, but you can't keep them locked in sync; you can synchronize their free-run timecode and then break the link and go shoot, and it will work reasonably well. But if you genlock and timecode sync the HPX250 to the Varicams, it will always work perfectly well. Of course, that means running cables between the cameras constantly...
 
I use HPX250 as C-cams all the time. In the real world, it's not always possible to constantly feed the cams w/ BNC cables. Reality TV show prod. is a perfect example. To minimize frame drift, simply resync them whenever there's a battery change to any of the cam. Every 2 hrs I re sync them as the HPX250 needs change in battery. Whenever you shut off the cam, it goes into the internal oscillator and that is not in sync w/ the rest. Typically you can get away with +- 1 frame for about 2 hrs from the last cam shut off. Anything longer than that you'll see more than 2 frame of drifts. Not good when there's critical talking scene. Subjectively you can get away with up to 3 frames of drift before it becomes noticeable. 4-8 frames are definitely noticeable.

Ambient has a wireless TC slave receiver (ACD-301RF) to their smart slate for about $1,500. Useful for situation where you can't meet up with the cam ops to resync. Definitely worthwhile in a large, concert venue where pretty much it's not feasible to be moving around with a large crowd. Other than that, the good old BNC will do just fine.

I'd custom make or buy a short 5" BNC cables coming out of the master cam (2/3" likely) and have the slave TC cam op come to the master tc cam to update tc data often (during breaks or any battery change).

Pretty much any make or brand of camera that has SMPTE TC will work. Panasonic recommends to use a video as a genlock whenever the user sync TC. This means you'll need a BNC cable for TC and video to sync free run TC. I find that I can get away w/ only TC bnc cable and still get zero drift. I have not done any scientific test to see if genlock at all helps. In the frantic field environment, fewer cables are better. Within the past 5 years, I have not used genlock cable to sync TC. Worked for me as I've been able to multicam them flawlessly. Genlock doesn't cure the drift issue when any of the cam is turned off for battery change. It will drift about 1 frame every 1.5 hrs if you don't resync all the cams.

You need gen lock when you do live HD-SDI switching with stand along video switchers. But if ISO recording only, genlock for TC data is not needed. Some computer based switcher such as Broadcast Pix doesn't need genlock. I'd use that cable to constantly feed TC info as a substitute. This will give perfectly synced TC data on all the cams in a non-mobile environment.
 
Last edited:
There are problems w/ syncing w the audio waveform - delay in audio. If there are two rear cams at 300-500 feet away and 3 front cameras about 50 ft away from the originating sound source, the rear camera has a slight delay in audio from the source. This in turn makes the video 2-4 frames off. There are real world issues where any of the multicam is stopped in the middle, the software has to waste so much time analyzing and calculating the same characteristic of a song or dialogue. Sometime it crashes or hung. Another problem is when the cam ops are in different rooms or locations. Just not possible for audio sync as both of them doesn't have the same sound characteristic. Things get even more complicated when you add a dedicated 8-track audio recorder to the production. Here, SMPTE TC is unescapable unless the producer has time to waste or money to burn by sorting them out one by one. The tedious, labor intensive process sort and organize them out pretty much costs the same or greater than using SMPTE TC to begin with.

For most prods., OK to use audio. Other time - it's not and will cost much more in the end. FCP X'a multicam syncing via TC or sound has pretty much replaced Plural Eyes.

For me, I prefer solid, verifiable TC data and not letting some 3rd party app to do its magic behind closed door.
 
Last edited:
Thanks for the great replies. So, to get this straight, I can get a lot of the benefits of Genlock by only syncing timecode? If I would be using an AC-160, which port would I use to do the syncing? The Varicams have TC out but I do not see that on the AC160.

So how does syncing the timecode stop the cameras from drifting since they will not be hooked up to each other?

Also, if I use Genlock, the Varicam only has a Genlock input. Does this mean it can only be a slave to an external device?

BTW, Plural Eyes will not account for drifting of two or more cameras. It meraly shift clips around, not resampling them to match.

Thanks for your help.
 
Last edited:
>>So how does syncing the timecode stop the cameras from drifting since they will not be hooked up to each other?

When you're in a mobile field environment and is impossible to be tethered to BNC cable, frequent syncing (every 2-3 hrs) will ensure that all the cams' TC data will be pretty much identical. This is what editor would want - identical free run tc on all cams.

In other words, the frequent TC syncing with each other is a workaround to not being able to continuously feed TC signal on all the cameras. I make shooters come to me every couple hours to sync up their camera (if it's battery operated). Or there are times where I have to run BNC cable to a mid point, a slave tc camera somewhere close to the other ops and have shooters go there and resync their cam whenever there's break. Otherwise running BNC cables to the cam ops is just no fun.

If you leave the camera on all the time after tha last sync, it will likely to stay in TC sync even there's no continuous TC & gen lock feeding it in. I have yet to do a scientific test on this. But since the early days of the SDX900/DVCPRO 50 camcorders, I have yet to see any production not being able to sync up perfectly by doing the interval TC signal method.

There are details in 24P, 24PN, 24PA modes where user has to use non drop frame TC. It is also advisable to have TC locks in (about 3-4 sec) before recording to avoid TC error.

The manual recommends to use both GENLOCK & TC for progressive format shoots. If it's 60i, then it no genlock in is needed.

<< Also, if I use Genlock, the Varicam only has a Genlock input. Does this mean it can only be a slave to an external device?>>>

From what I've seen w/ virtually all 2/3" field cams (not counting the triax Ikegami or Sony type), there's only GENLOCK in. Genlock out is usually a dedicated video generator or any stable camera w/ a video signal signal.

So any 2/3" field cam with genlock in is simply taking an external SD or HD-SDI video signal to be used as reference to correct phasing (cable length and transmission differences), vertical timing, and color framing.
 
Last edited:
Thanks for your help. So the Varicam can not act as the source of any Genlock, only a recipient.

Further questions about timecode syncing:

Your experiences are reassuring. Is there something more that takes place during the sync because it seems you would be only creating a common starting point, not an ongoing correction. Even though you synchronize the timecode of separate cameras, wouldn't the internal recording speed of each camera still have an effect on if they drift apart over time?

Do you know which input is used for the AC-160 or HPX-250 to sync the timecode with another camera?
 
>>not an ongoing correction.<<
Yes. Once master slave are locked and all cams are not shut off, there' not much constant signal input is needed because the SMPTE TC is in FREE RUN mode. However, if you have the luxury to run BNC cables, then it doesn't at all hurt to keep a constant TC signal. Almost error proof if any of the camera gets shut off due to battery or faulty power supply.

I don't own the 160. But after briefly scanning thru the manual, I don't see that it can do FREE RUN TC, but only PRESET Free Run. This appears to be Panasonic's own TC protocol. Free RUN TC typically is associated with industry standard SMPTE TC and can fully be locked as master or slave and support various frame rates. Further reading reveals that the 160 can sync to other similar class of AG 160 series model only. This is kind of like the HVX200 series. I looked at the brochure and the TC voltage (via the VIDEO out) is a bit odd.

For the HPX250, there's menu setting where you have to specify the TC connection to be in or out. In my case, I always use IN because of my workflow. I find that the HPX250's internal TC generator is too slow to set up. W/ 2/3" cam, it's very fast to set your own custom value TC (usually real world clock) and be done w/. No more cumbersome submenus w/ the HPX250. I set it as slave for HPx-250.

To test SMPTE TC working properly, simply change the free run tc on the master cam to a different value. When you change it, the slave camera will immediately gets updated as soon as it's changed. There are concerns to frame rates w/ TC on the cameras. This applies only to 24P/24PN/24PA modes. User has to set to non drop frame. But if you're shooting anything other than 24p, drop frame or non drop frame. For me, non drop frame works out best as it accommodate all camera formats. I shot with a variety of sources ranging from the lousy h.264 dSLRs C5Dm2, Go Pro all the way to high end Reds.. My solution is to use any smart slate by Deneke or Ambient. You can still make the 160 to read time code by treating it like a dSLR and capture the tc values via a smart slate.

>> wouldn't the internal recording speed of each camera still have an effect on if they drift apart over time?<<

All cameras should have similar recording format to begin with and none of this drifting is of a concern. It's ideal to record all of them as 60i, 30P, 30Pn. Try not to mix 24P w/ 30p. You can do this if all time codes are set to non drop frame. But most NLE apps prefer same frame rate. For me, I set dSLR to 30Pn, Go Pro to 1080p-30P, AVC-I 100 30PN... But never any of the cam w/ 24P or 24PN. You can have different resolutions (480p,720p, 2k..), but matching frame rate is the key to keep a smooth editing and no nasty surprises.

If you notice carefully, virtually all high end 2/3" cams has TC in, TC out BNC connectors. This means you can loop through the slave TC to the 3rd or Nth camera. For example. The master cam is a Varicam HPX3700. I run to a HPX3100's TC in as a 2nd slave camera. The HPX3100's TC out then run to another camera. It can loop through quite a few. TC signal is regenerated each time it goes to a cam. Cable length is not a concern. But if that camera happens to be HPX250, you're stuck w/ out any looping capability as it can only accept in. Because of the small form fact, Panasonic didn't put TC in, TC out dedicated connectors to save valuable space. So use HPX250 as the last slave camera.

Other thing I'd pay attention to is not to record immediately as soon as the SMPTE TC is locking. This process takes only a few seconds and is not a big concern. But TC error occurs when user rushes to start recording while the tc is in the locking phase. There's no way of telling the tc is locked. Kind of risky. Mentioned earlier, change to a completely different FREE RUN TC value on the master cam and you'll know that the TC master and slave is working. All cameras should be set to FREE RUN TC. Overall it's very robust and virtually error-proof.

I assume that all recording media are P2 and not tape. If it's tape (SDX900, HDX900, Legacy Varicam), you'd want to record an extra 10 seconds of preroll for all FREE RUN TC. This is because when capturing, the deck will have to constantly search and update any break in TC. If no extra pre roll, it'll be very likely that you'll miss the first several seconds of the in point on tape based free run tc. And it takes 3-4x longer to capture a free run tc recorded tape. P2 is a big technological breakthrough in solving this annoying limitation.
 
Last edited:
lonewolf2koc@hotmail.com said:
. Whenever you shut off the cam, it goes into the internal oscillator and that is not in sync w/ the rest.

You need gen lock when you do live HD-SDI switching with stand along video switchers. But if ISO recording only, genlock for TC data is not needed. Some computer based switcher such as Broadcast Pix doesn't need genlock. I'd use that cable to constantly feed TC info as a substitute. This will give perfectly synced TC data on all the cams in a non-mobile environment.
Click to expand...

Actually, as soon as you unplug the TC and Genlock BNC, the cameras internal crystal oscillator runs on it's on.

Also, most of today's portable SDI switchers have TBCs on each input. Most live Red Carpet shows I've DP, TD and edited have been switched with Panasonic 400G's,etc. with no genlock. Even the old analog Panasonic 50 switcher has TBCs. That old workhorse could take cheap cameras and switch pretty good, as long as you used a 1 or 2 frame dissolve for cuts.
 
Ah, the internal osc is pretty good for several hrs. Never have the chance to do a real scientific test. 1-2 frames off is still very hard to detect unless you really look for it (like a camera flash in the scene). But gets worse as soon as the cam is shut off.

I have the old MX-70 with the SDI option board. It doesn't need any genlock as long as it's switched in the auto option. But when you press direct 1-8, it the video phase jumps just a bit and it shows on the screen. When auto take is on, then the frame sync takes over and no jump - absolutely clean. Others here complained that the picture quality is somewhat impaired compared to a Sony DFX-700 w/ required genlock. There's a quality hit.
 
Thanks once again for your help. Let me give an example of my recent experience:

I have been shooting with a Varicam, XF300 and an HMC-150 on the same shoot (720p60). The Varicam is recording to a Nanoflash and the XF300 & HMC-150 are recording internally. I did not connect the cameras in any way, just let them record. In my NLE, after visually syncing the two (dancer jumping and landing) the XF300 & HMC-150 were pretty close but both were one frame out of sync about every 10 min with the Varicam!

So with this example, if I had synced the timecode between thetthree would the drift been cured?

Or maybe I have a setting off in the Varicam? I will look around the menus today and see if non-drop frame is turned on etc...
 
I once did a test on a Deneke smart slate using drop frame and non-drop frame from several Canon 5Dm2 shooting at 1080/29.97. I noticed drop frame caused a lot head aches.

To cure this once and for all, simply take the TC out of the Varicam and run it to any smart slate's TC in. Make sure you set the correct frame rate on the smart slate. I usually do a manual instead of auto. There should not be any drift in frame on all 3 cameras. I have great success with this method on dSLRs, Go Pros, or any other non-SMPTE TC featured camcorders. No drifting problems after 10 min problem. This usually occurs when you "eye ball" the sync vs. doing it the SMPTE TC method.

Alternatively you can use the old school LTC SMPTE TC feed it to an extra audio channel on the XF300 & HMC-150. Here you'll have a nice time code track perfectly synced fr the Varicam. To read it, you would need http://www.videotoolshed.com/product/26/fcp-auxtc-reader to extract the TC and convert it to an AUX TC track. $450 for the Horita PTG and the $250 software are needed to get it work. ~$700 vs $1,400 for the Smart Slate fr Deneke or Ambient. You may not need the Horita PTG as the Varicam can serve as the generator. The trick is to use the Varicam's TC BNC (2-pin) and simply adapt it to a regular 3-pin XLR cable. TC signal is really simple & robust Bi-Phase mark type. I'm not sure if it's mic or line level on the Varicam. But there're adaptors fr. PSG to convert the correct signal type to record the audio time code: http://www.bhphotovideo.com/c/product/275822-REG/PSC_FPSC0010D_ALMP_Line_to_Mic.html. The Horita PSG's TC output is a flimsy RCA female at line level. You can get a cheap RCA-XLR adaptor and probably a D/A to feed it to multiple cam sources and be done with if there's no Varicam avail.

Smart slate method is arguably the easiest to setup and has greatest mobility over the LTC TC via XLR audio method as you'll be constantly tethered to the XLR audio signal recording TC instead of normal audio. However, the most TC accurate will be the XLR audio TC feed.

There's a slim chance that the Nanoflash is causing the delay. I had some past experiences doing a direct to a Decklink HD Studio Pro card on a Mac Pro. Sometime after 20-30 min it will drift even genlock IN from the same source is selected. It's rare, but did occur when using a recorder beyond the camera. I couldn't get the tc value to transfer over. It had to be manually entered after the fact. Luckily, the tape on the HDX900 saved the day as I was able to recapture it and all of the tc values are fully preserved. Only downsize was that it was DVCPRO HD EX instead of the much better Pro Res 4:2:2 HQ codec on the Decklink/Mac Pro setup.
 
Last edited:
Thanks for your reply. I checked the Varicam and it is set to drop frame, so no easy setting fix. I have to run out today but would like to post some more thoughts later.
 
Bassman2003 said:
Thanks for your help. So the Varicam can not act as the source of any Genlock, only a recipient.
Click to expand...
Of course it can act as a source; just use the video output (or SDI output) as your genlock source.

Your experiences are reassuring. Is there something more that takes place during the sync because it seems you would be only creating a common starting point, not an ongoing correction. Even though you synchronize the timecode of separate cameras, wouldn't the internal recording speed of each camera still have an effect on if they drift apart over time?
Click to expand...
Yes, of course, although it's usually pretty minimal. But that's the whole point of using a continuous timecode cable connection, which is possible on cameras that have TC IN/TC OUT jacks. You could leave the HPX250 connected to the VariCam and be assured that you would have perfectly synchronized footage for the entire duration of the production. But it would require them to be cabled together for that entire period.

Do you know which input is used for the AC-160 or HPX-250 to sync the timecode with another camera?
Click to expand...
The HPX250 has a proper TC IN/OUT jack. On the Ac160, it can use its "video out" jack to emulate an LTC timecode in/out, but only for purposes of initially synchronizing its free run timecode preset. You have to break the cable connection before you can actually use it as a camera again. It can emulate a TC generator (sending LTC out its video out port) and it can also go into a mode where it can receive timecode from any LTC device (to synchronize its timecode preset to the incoming timecode). Contrary to what was posted before, this isn't something that applies only AC160-to-AC160, the AC160 can sync to any LTC timecode device. But it can't do a continuous cable jam, it can only synchronize its preset (or send its timecode to another device for purposes of pre-synchronizing).
 
Thanks for your input Barry. Running cables for timcode between cameras is not a problem and sounds like the best way to go. I recently found a great deal on another Varicam "H" model, I will have two matching cameras and am trying to sort out the third. The XF300 & HMC-150 will be sold and am trying to decide between the AC-160 or the HPX 250. The HPX-250 looks like the best choice except a lot of the events I tape are long, some up to 4-5 hours per show or two shows in one day. P2 is pretty cost prohibitive in this environment. The AC-160 is a fine camera but I am unsure about drift as I am surprised how much the HMC-150 drifted away from the Varicam. I wish the HPX-250 had a better long-form recording mode and the choice would be easy.

BTW, it appears to me the Varicam does not have a video out, only two SDI outputs. Is there a way to get composite video out?
 
If there're limited P2 cards per camera, I'd consider the three card AJ-PCD30 USB3 P2 reader on location, it will get you by comfortably. You can offload 3 P2 cards simultaneously and it performs like a single high speed reader. Worse case is to switch to AVC-I 50 if offloads are not fast enough.
 
Last edited:
If I go with the HPX-250 I would probably buy two 64GB P2 cards and a Nanoflash. This is not optimal but I do not see a great value in spending $3,500 on P2 memory when I would still be using AVC-I 50 for some of the days. Plus, 100mbps just is too large of a datarate for long shows imho. I have been very pleased with the 50mbps setting on the Nanoflashes as the files are more manageable in size. Plus, I am not crazy about the non-full raster recording of AVC-I 50. I know it is probably no big deal but it is out of date and P2 is quite expensive to have something be out of date.

I am not a fan of offloading data on location. To me, this is more dangerous than using SD cards that so many seem to be afraid of.
 
Ultimately, P2 media investment will pay off in situation where it's not possible to offload in the field (tropical jungle, desolate desert, remote areas, etc). But if the bulk of the work is in urban locations, I do concur that the P2 media is way too costly.

I always copy to 2 sep. hd locations via Shot Put Pro app and be done with it. No second thoughts on whether all data are properly transferred or worry about hd won't spin up. There has never been any data lost issues for me for location offloading except of being a nuisance on the offloading process and finding a secure location w/ stable power supply.
 
I am more concerned with me making an error than a hardware issue! I think P2 is a great concept, but they forgot about us event guys. If Panasonic could include AVCHD for the times when you do need the record time P2 would hold a lot more value to me.
 
You must log in or register to reply here.
Back
Top