Genlock, Timecode, Wordclock -- What Are They? Do I Need Them? Are They Interchangeable? 

"Edit: when you were allowed one of those little yellow tweaker sticks to adjust machines... a proud moment lol." ah yes thos little tweak sticks...always getting lost! I started a few years later than you and ended up working for a subsidiary of major broadcaster. I also got to learn colour grading.

I'm not quite as young as I look ;) I've owned and used time base correctors, frame syncs, and color correctors myself. It wasn't that many years ago that I got rid of the last of them.

Most sound guys now timecode boxes like Tentacle Sync. If you roll for 2 hours no cuts like a concert or live event or documentary intv then Genlock keeps cameras rock solid sync no drift.

Yep. Been there.

I probably won't do a separate video. The difference between LTC and VITC is how it is stored with analog video. LTC is an audio track whereas VITC is stored in the vertical blanking interval as part of the video. Digital video usually stores timecode as metadata or as auxiliary data within a data stream. So they act differently in that way. The relationship between Genlock and time code doesn't change when you go from LTC to VITC or digital. Timecode is still just a number indicating what frame of video you're on and nothing more -- it still isn't used to tell a camera when to start a frame or line of video.

Thank you!!!

Decoded the audio signal, It matched the timecode ( 01:20:14:00) on the monitor behind you.

Nice work :)

It could vary by camera model, but the ones I’ve used will just ignore the genlock signal if it is at an incompatible frame rate.

If you have timecode syncing them up is pretty easy. I'll be doing a video on that soon -- hopefully next week, schedule permitting. Without timecode you may have to rely on being able to sync audio. Video editing tools these days have audio sync capabilities. They're not the most reliable though. If nothing else, you can slate (clapper board) when shooting and sync up the audio spike with the first video frame where the slate is closed.

What did I get wrong? Perhaps I just misspoke.

@@djp_video You were not pushing it too hard, but still: NTSC, PAL and SECAM has nothing to do with framerate or resolution. Also, if you are not running composite RF, there is no NTSC and PAL. So S-Video, component, Digibeta and ALL signals in the digital domain has no NTSC, PAL or SECAM.

Okay, technically yes, but it's certainly implied and people watching would know what I meant. If I was to refer to System M, System J, NTSC M, PAL M, etc. nobody would have any idea what I was talking about. Those terms aren't used in common vernacular.

@@djp_video A lot of people say NTSC today and mean 480i. And unless we use the correct terms, new people will not learn. 480i, 576i, 720p30, 720p25, 1080i25, 1080i30, 1080p25 etc. You get the picture. :-D Why am I saying this? I have been teaching a lot, and unless the teachers use the correct terms, we can not expect the students to learn the right things. A good example is PAL with 29.97 fps and 525 lines.

Oh, by the way - you channel is kinda technically, isn't it :-D

Sony PXW-Z150 primary camera, Blackmagic Design Micro Studio Camera for overhead shots (don't love it, but it works). Lighting is three LED panels -- a 60W panel just above me to my right, another 60W panel just above the camera, and a 30W panel just behind and overhead for a rim light.

@@djp_video Man! I'm amazed that your light looks so soft with panels. Are you modifying/diffusing them? I would love a BTS of your setup. I was also surprised by your camera choice which is very different from most RU-vidrs. I'm happy you can use different equipment and still get great results.

The panels come with a plastic diffuser panel that sits right in front of the LED elements, but that's it. I'm primarily a live event video producer. For the types of things that I do, camcorder-style cameras make a lot more sense than the typical camera options used by most RU-vidrs. I have to be able to shoot in a very wide variety of situations -- outdoor, indoor in well-lit rooms, indoor in poorly-lit rooms (almost always where we have no control over lighting); at varying distances from intimate interview closeups to conferences with the camera way at the back of the room, to sporting events with handheld roving cameras, and no other type of camera gets even close to handling all of those situations anywhere near as well as camcorders do. They were designed and built for that. ILCs and studio cameras might handle one or two of those situations alright, but they have some really significant limitations in others. The Sony PXW-Z150 camera that is my goto choice for most events, has a 1-inch class, 20MP sensor, so it's larger than your more traditional camcorders and broadcast cameras, giving it the ability to capture more light than other models, yielding less noise and more detail. But the sensor isn't so large that dealing with shallow depth of field with moving subjects is an issue either. The 20MP sensor lets the camera crop in on the sensor even in 4K without any loss of detail. The Z150 gives a 24X zoom range at f4.0 (12X optical plus 2X "Clear Image Zoom" which is amazing), so I can shoot both up close and far away and know that I'll get a good shot. The lens is also higher quality than the lenses that most people use with their ILCs, so the image detail is consistently a lot higher than I've seen (and gotten) from other cameras. The Z150 also costs far less than even a single, decent lens for a B4 broadcast camera. And when you do the types of events I do, having more cameras is almost always better than fewer "higher quality" models. Six cameras for an event is my general minimum. Having to purchase 6 or more of most other types of cameras with lenses and other required accessories puts you into big budget territory. For the handful of bigger budget events that come my way, where a true broadcast camera is required, I can rent them. I'm frankly shocked that anyone even attempts ILCs or studio cameras for mobile, on-location video work. I've tried them, and they are just so incredibly frustrating to use, and yield lower quality results. They just have so many more issues to deal with. Camcorders were designed specifically with these workflows in mind, so they tend to work really well.

@@djp_video If I may ask... What's the make and model of your light panels?

They aren't made any more. These are similar, though not exactly the same: amzn.to/3sIpCZd

If you want your devices to stay in sync so you can line up recordings in editing, you will need to use a wordclock signal to do so. You can use the Word Clock output on your M5402 and distribute that to all devices with a word clock input. You can use a word clock distribution amplifier to split that signal to any devices that may need it. Additionally, your DA-6400 has the ability to pass through the word clock signal so its output can be uses to sync another device. Though the manual on the DA-6400 doesn't specify, I suspect that you could set it to follow the Dante leader clock and then use its Word Clock output to sync additional devices. That wouldn't be my preferred way to do it, but it should keep things synchronized whenever it is powered on.

The Ref Out connectors are generating Tri-Level Sync, which is a type of Genlock. These are analog signals. But that output isn't going to line up perfectly with the signal at the video input -- it has its own timing, mostly independent of the input (aside from frame rate). The Ref In connectors all support Genlock -- both older blackburst-style as we as tri-level sync. On some devices (like the Ursa cameras) the Ref In can also take timecode. But that isn't "normal." SDI includes sync information as part of the data stream, but it isn't so trivial to extract to do anything with -- you can't take an SDI output and connect it to a Reference or Genlock input, for example. Consumer cameras and the ATEM Mini switchers don't support genlocking. That's something that's usually reserved for professional, broadcast style cameras, with a few other exceptions here and there in the Prosumer range. I'm curious your use case in which you'd need sync from the ATEM Mini?

The monitor is a Sony PVM-A170 and it has a feature to display timecode built in.

Genlockable sources usually have a timing adjustment. Honestly, I've never done long-distance syncing in HD. In the SD world we'd use a waveform monitor.

The EE (Electrical Engineering) answer is that you generally won't run into time-domain signal issues with cable length until you approach the quarter-wavelength of the signal. In other words: Transmission-line effects become significant when the length of the line is comparable to or greater than the wavelength of the signal. A more specific general rule of thumb is one-quarter of the wavelength. Given the HD 1080p Genlock signal demonstrated at 10:28, we can calculate the maximum cable length for which we will begin to have to worry about Transmission-line effects: Full-cycle wavelength of 33.7159 KHz signal = 29185.0432 feet = 8897.8790 meters. 1/4 wavelength of 33.7159 KHz signal = 7296.2608 feet = 2224.46976 meters. So for transmission delays on a long cable, we would need to have a maximum cable length just under 1.38 miles! That's quite a long cable! If we go over that distance we might begin to look at other transmission options such as fiber optics or radio frequencies. More likely the issue that you'll run into first with longer cables is dealing with noise picked up by the cable itself from nearby sources of electro-magnetic fields.

Genlock isn't SDI. I'm frankly surprised that anything recognized a signal from an SDI output. As covered in the video, there are two types of Genlock signal -- blackburst and tri-level sync. Blackburst is sometimes called bi-level sync. Blackburst is basically a black & white standard definition video signal. Tri-level sync is a simple series of pulses. SDI is very different and can't produce either of those. You'll need a proper genlock signal generator for any kind of compatibility with multiple types of cameras or other devices.

When you have similar audio across different cameras, yes, Premiere and other editing apps can synchronize based on the audio waveform. But this feature doesn't always work. I've had it fail as many times as it has worked. If the cameras aren't placed close to one another, or if one camera picks up more background noise than another, or if audio isn't recorded on a camera for whatever reason, or if the audio in any other way differs significantly, you're on your own.

Since most of the Studio cameras don’t have a Ref In connection, they can’t be synchronized directly to a Genlock source. I believe you can set them to synchronize to the SDI input, which would effectively take care of many (though not all) of the reasons to use Genlock, like making sure that cameras all stay in sync with one another. That won’t help with minimizing latency, though. Using the Ref output of your HyperDeck doesn’t really buy you much of anything. The ATEM has its own internal clock and frame synchronizer and it runs just fine without a Ref input. Having the HyperDeck supply a reference signal only prevents the ATEM from skipping or doubling video frames from the HyperDeck and nothing else, and that would happen so rarely that I’m not sure it’s worth adding that additional complexity and possible failure point by making that connection. The Focusrite isn’t going to recognize the Ref signal from the HyperDeck - it is strictly used for Genlock. Word clock is an entirely different signal.

Try setting the GEN10 to output sync for 1080i59.94. It's a different type of signal that I believe the Blackmagic cameras know how to use. The 720p sync signal has the wrong horizontal sync timing for 1080p/4K.

It's an Owon SDS7202 (amzn.to/3z7jjS5), but nearly any oscilloscope will work -- these signals aren't particularly high in frequency. I haven't done a switcher latency test. With genlocked sources, and for signals not being processed through a scaler, the ATEMs processing delay is less than 10 lines. But, if your video source isn't genlocked, add an average of a half of a frame. If your video source doesn't match the switcher's video format, add another frame. If the source is being run through SuperSource or a DVE, add another frame. If the signal passes through a converter (SDI to HDMI, etc.), those will add a millisecond or so. Figuring out the exact processing delay using an oscilloscope is going to be really difficult, as the video output signal (SDI or HDMI) isn't going to have obvious synchronization markers visible on an oscilloscope. One could connect the output of a switcher to a device which will generate a sync signal from the source (like the HyperDecks do), but those come with their own processing delay.

There isn't a hard limit. And it will depend on whether the signal is regenerated or simply passed through. It's better to use a distribution amplifier (any old-school video DA will work) and have separate connections for each device, but you can certainly chain one REF to another as long as you know those devices will be powered on.

When you say sync, are you referring to avoiding drift? Or lip sync where the same moment in time is seen on video as is being heard in the audio? They are different.

@@djp_video Thanks for your reply - I really love your channel. The first problem audio/video drift over time. It's a church, so we record for an hour. With all modern equipment, the drift is minor - like only a few frames. The second issue I'm having is more nuanced, but a significant problem. I am trying to use AES audio out of a Yamaha RiO stage box into a Blackmagic 12G Teranex Audio to SDI embedder. The result is that I get frequent audio dropouts on my recordings. My gut feeling is that it's a AES master/slave issue because the video system and audio system are not locked in time properly.

No. Time Code and Genlock are different signals. The only cameras I've seen that support time code on the same connector as Genlock are the Ursas from Blackmagic Design.

Blackmagic Design has their Sync Generator product. AJA has the GEN10. There are still quite a number around.

The timecode and genlock signals would have to come from the same source. Otherwise the timing won't stay in sync. Higher-end sync generators will create both signals for you.

Most likely it would. Video editing apps typically only look at timecode at the beginning of the file. So if you have missing frames the video doesn't get re-synced.

The tri-level-sync signal was intentionally designed to be simple. In theory, yes, they could include timecode in there, but they've elected not to. Genlock is distributed using old-school analog video distribution amplifiers. Equipment which has an output can also be daisy-chained. Timecode can be distributed using any audio distribution amplifier. Both types of devices are really easy to find.

@@djp_video I don't know if you have heard of this yet but it's really interesting for keeping time synced across servers and could be implemented in normal computers and was curious if you could see a way to apply this tech too the video world it is also open source. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-JK3eTGkX6qY.html

@@CYBERSTORM101X There are quite a few devices of a similar type for the video world. Many of the companies making genlock and timecode hardware have options that will use GPS as their time source; it is a common way to go.

The semicolon denotes Drop Frame format.

Some models like the current HyperDeck Studio models from Blackmagic can re-sync the input to a genlock source (but they obviously don't adjust the timing of the camera itself). But in most cases that doesn't buy you much -- most modern video switchers have frame synchronizers built-in which do the exact same thing.

If a camera has a jam sync feature you don’t technically have to maintain a connection to a master sync source, even though it would still be a good idea. But for cameras that don’t, you could get away with starting the beginning of each clip with timecode audio so the editor can sync the start of each file. The preferred way to go is to leave the master timecode source connected all of the time. Because timecode is just audio, you could split the signal from your timecode source to multiple cameras at the same time, the same way you’d split any other audio. I’ll be demonstrating one way of doing that in a future video.

@@djp_video looking forward to that video, thanks. I just picked up a Denecke SB-4 super cheap at a local pawn shop, will it’s BNC jack output TC audio? Can I just adapt to 3.5mm and jam a lumix camera?

Yes, you can connect to your camera. If your camera has a timecode decoder you can jam sync it. Otherwise it will just be an audio track.

All are 75 ohm.

@@djp_video Thanks! I guess 50 ohms RF was for analog antenna stuff, but they're everywhere in the stores, maybe for older analog signals or something like that.

@@the_newvoice Yes, 50 ohm is primarily used for radio applications. The one big exception is over-the-air TV antennas, which are either 75 ohm (coax) or 300 ohm (flat).

@@djp_video Thank you!

The Reference input on the HyperDecks is only used for playback. When recording it syncs to the incoming video signal.

@@djp_video thanks!

You do need to re-sync when you turn them back on. If you use your cell phone as the time source it's pretty easy to do.

Okay thank you, is it recommended on a full day shoot to even turn them off if you don’t need to? Or would new time codes not really make a difference in editing. Thank you again for the reply

Let's start from the base of information from this video. I'll build on that. (But I'm simplifying things at the same time too.) There are two separate synchronizations happening with Genlock ... Horizontal and Vertical. Vertical sync happens once per frame, and horizontal happens once for each scan line (including those that aren't part of the picture, which is kind of another discussion). Because the vertical sync is tied only to the frame rate and is independent of the resolution, many devices can use sync signals designed for lower resolutions when producing higher resolution video. So in theory a UHD camera could use blackburst, which is standard definition, for vertical sync. It would have to ignore the horizontal sync since those are tied to the resolution, but vertical sync is usually enough to line things up because the frame rates (24, 25, 30, 50, or 60 fps) is independent of resolution and therefore can be can be the same across SD/HD/UHD. So a 60 fps (frames or fields per second) SD sync signal can work for UHD running at 60 fps (frames per second). The other consideration is that blackburst was designed for interlaced video. And even though in the world of interlace video we usually refer to frame rates of 25 or 30 frames per second, each frame is made up of two separate fields, so the fields are being created at 50 or 60 frames per second. That also means that even though blackburst is syncing 25 or 30 frames per second, it is also syncing fields 50 or 60 times per second, and if you were to look at a blackburst signal you'd see vertical sync pulses 50/60 times per second. And HD and UHD cameras creating progressive video can use that field sync signal to synchronize progressive frames rather than interlaced fields, thus giving you compatibility between ancient blackburst signals and modern 4K cameras.

Thank you@@djp_video This is all very fascinating. I've noticed you have YT subscription available but have you been thinking about Patreon? I can imagine you have a lot of knowledge to share, I would love to learn more.

The semi-colon means that it is drop frame timecode (a regular colon is non-drop frame). Basically, since the frame rates, such as 29.97 and 59.94, are not whole numbers, every second is missing parts of frames (.03 or .06 in those cases), which will add up over time. So, with timecode, you can either ignore the missing frames and just keep counting or the timecode will drop frames on the counter so it accurately reflects the time, and so there are the 2 different types of timecode to notate that.

What Ryan said... ^^^

@@RyanBoni thanks Ryan, yes Doug did say he was running the system at 29.97fps so if understand it, the semi colon means that it is not running full frames per second (eg not 30). Thx!

@@lansleyONE ​ @Nick Lansley Sorry, I answered the question while the video was still playing before I heard Doug give a full explanation of drop frame/non-drop frame later in the video!! But, to answer your question, Nick, no the semi-colon or colon does not necessarily tell you one way or another if the system is running at 29.97 or 30. All that the appearance of the colon or semi-colon does is tell you if the timecode counter is counting all of the frames or not. Really, just think of it as 2 different counting methods. One precisely and accurately reflects the actual time elapsed and the other just keeps counting and will slowly drift from the actual time elapsed. The difference between the 2 is generally irrelevant, unless you are mixing DF and NDF together while editing things that were shot at the same time with different cameras, or if you need the timecode to accurately reflect the time of day or time elapsed for some specific reason. The general rule is to pick one or the other and just always use that one.

No. Timecode is only accurate to a frame (at 30 fps, 1/30 of a second, or 33ms). Dante's latency is usually well under a millisecond.

@@djp_video thanks so much for the quick response, this will definitely make life much easier on big multi cam OB setups. So in theory then if you used for example the Focusrite's RedNet PCI card on a PC and mapped one of the tracks to record the LTC you could use a computer as a secondary cheaper multi-track audio recorder with timecode accuracy? What software would you then recommend using to convert the timecode track and map it to the wav Poly file such that any editing software could then import that and timecode sync with pictures?

I'll be demonstrating how to do this in a video coming up shortly. The software that I'll be using is Tentacle Timecode Tool from Tentacle Sync. But many editing applications can decode LTC natively as well.

I'm not aware of any devices that will take a genlock signal and create word clock from it. But there are multiple options out there for devices that will create both though. Cameras which are genlocked should maintain audio sync with other genlocked sources. In terms of integrating with Dante, you'll want a mixer that has a word clock input, and then set that as the timing leader for the Dante network.

@@djp_video I use an ESE DV-321 which will derive a 48khz wordclock (and an AES11 signal) from genlock. It's a (relatively) inexpensive way to keep audio in sync with video over long periods of time.

The "proper" way to do it would be to use an audio distribution amplifier to split the timecode output from one camera for the timecode inputs on the other cameras. But, in reality, you can use a simple cable splitter and it will work just fine. Audio is a lot easier to split than video. The SDI output will not work. It is a video signal, where timecode is an audio signal. Please keep in mind that time code does not keep cameras in sync. It only makes sure that when you sync up footage in your editor that the *beginning* of each clips is lined up properly. Cameras can (and do) still drift. If you are shooting clips more than about a half hour at a time you're likely to see that drift and will need to adjust the playback speed to compensate. The only way to ensure that cameras stay synchronized is to use cameras that support genlock and provide them with a genlock reference signal. Go back and watch this video again. It has all of the detail you would need to achieve what you want.

@@djp_video Thank you so much!

@@djp_video I just wanted to thank you for your advice. We ended up jam syncing the cams one by one. We didn’t lose a single frame all day. Made editing a breeze!

When I got this the Broadcast version didn't exist yet. And I haven't gotten one because I've heard from many people that the low light performance isn't very good.

@@djp_video Oh I see. Hopefully blackmagic will renew their broadcast camera soon. The entry-level studio camera with the MFT mount and the pro-level with the B4 mount. Seems legit.

@@djp_video Hey Doug, the sensor does suck a lot of light, however when adding gain you don't get the usual graininess you normally see until +12 or even more. We have a few of the URSA broadcast, and do similar work to yourself, our crew love them, and would never go back to our older Sony cameras

@Mark Edge, what kinds of events are you shooting? We do a lot of indoor events without additional lighting. In those situations the Sony cameras with 1-inch sensors are the best thing I've found by a wide margin.

It isn't always audio, but it is usually basically just an audio signal. Dedicated timecode connections, including BNC, are essentially just audio. I don't have a Zoom F8 to test, but it probably is just an audio signal. But to connect it to a DSLR/mirrorless camera's microphone input you'd need to drop the level unless your specific camera can be set to accept a line-level signal. The F8 would be outputting something around a line level, but the camera is expecting microphone level. But that's easy to do.

@@djp_video thank you so much! So i need to add a resistor too. Thanks alot

Blackmagic Design SmartScope Duo.

@@djp_video Thaaaank Yoooou!!!

Both cameras support timecode or genlock.