Critical Elements of Live Streaming

So if we look here, we’ve got video acquisition, live encoding and recording, recording for post production and then your file upload. So this is a studio and then from the studio I’ve got different cameras that are going to send a switched output and I can do a couple different things with that. We can actually send it live, so we can do a live stream. We can record it to a computer, we can record it to a hard drive and then we can edit it, put graphics in it, clean it up, do some other stuff, adjust our audio and then we can upload it. But in either case we have to start with the cameras. We have to start with what images am I going to capture.

So these are my keys to high quality acquisition. Number one, pretty straightforward, choose a high quality camcorder, use consistent, even lighting at the correct color temperature. If you have a lot of light with lights you bought at Home Depot and the guy on the camera looks really funny, it’s probably the lights. We laugh, but I’ve seen almost everything. You want to utilize interesting shot composition and proper set design. Some of you have probably seen some shots of a talking head with a pastor and the whole background behind him is black and the pastor looks like he’s a floating element in space. Okay, so we don’t want to go that direction.

And then the final thing that’s often overlooked is providing quality audio with your video. What a lot of churches will do is they’ll do their program audio that they’ve mixed and EQ’d for the room and they’ll send that same feed out to broadcast and they wonder why it sounds really bad when they watch it back on the internet. It’s because you need to put some headphones on and mix for what is actually not in the environment but what’s actually coming out through the signal chain. And if you don’t do that, you won’t have a good sound. Does that make sense to everybody, what we’re talking about? Good.

So here’s a really, really good rule. When it comes to producing video for streaming broadcast or archival, choosing the right cameras for your vision, environment and workflow is priority number one. And the reason we say vision, environment and work flow is because that’s different for everybody. Not everybody has the same environment. Some of you might have great lighting and you can put the cameras right there, some of you might have bad lighting and the cameras have to go far away, some of you have a lot of volunteers and you can have lots of manned cameras, some of you have a hard time getting volunteers and robotics might be a good idea.

The work flow, am I going to take it for post production, am I going to send it directly, all those things play into what’s important for you at your facility. Not everybody’s going to have the same process with that, but I’ll say this again, if you don’t start with a good image, guess what? All the rest of that doesn’t matter. So if the camera’s not capturing a good quality image, you ain’t gonna fix it. So the challenge is to choose the camera that is right for your application, at the right price, without short changing your ability to do better work in the future.

So even though you don’t have the budget to get the camera you really want, make sure you get something that’s at least going to be good enough that you can upgrade your system and add this maybe from a main camera you might put that over to a slash camera, there’s other options, but just going with the cheapest camera is not going to give you a good solution.

So, there are a lot of choices. So we have, upper left corner, that’s a three chip robotic camera. Down below it, that’s a single chip robotic camera. These are both Panasonics. Down here, this is a Blackmagic studio camera, this is another Sony studio camera, this is a little small Sony Handycam, this is a Canon HD Robotic camera, this is a Hitachi, well actually I think that’s another little Sony Handycam, and this is of course the Panasonic three chip, 2/3″ camera with a digi super lens on it. Every one of these cameras is an HD output camera. The image that comes out of all these cameras is wildly different depending upon the environment that you’re going to put them in.

I have a lot of people that say well gosh, we want to use this Blackmagic camera in the studio, because I’ve sene it at NABF I’ve seen it at the show, it looks great. Well, the issue with that camera is lensing. They’ve got lots of stuff for it if you want to put the camera twelve feet away, but when you want to put it fifty or sixty feet away now you gotta use all kinds of adapters and stuff and now your image gets real soft. So understanding the specifics of each camera and how it applies to your environment is absolutely critical to not wasting your money. So, are there any questions on that? Nobody’s got a question yet? Golly.

Yes, so that camera right there that’s on the upper left, that’s probably the camera that’s on me, that camera that’s in the upper right is that camera right over there with a different lens in the front. That digi super 22 is probably a thirty thousand dollar lens so unless you’re doing sports at your church, you probably don’t need that lens. Now if you’re doing sports, hey, go for that one right there.

So, here’s the kicker, and I’m going to go back to that slide real quick. These are all different cameras. Here’s the kicker. Make sure that the camera type, the output type, the recorded format and the resolution are consistent across all your cameras because if you mix different types of cameras, it’s gonna be exceedingly difficult to match them live or in post. And I’ve got plenty of churches that have gone with well see, we can afford one of these today and then now we’re going to buy a robotic over here, and we’re going to buy one of those other Handycams because somebody donated it and they wonder why every time they switch the camera the pastor’s suit changes from gray to green to brown. That’s because you have different cameras, and that’s a real pain in the neck to try to fix if not impossible.

So the camera chip size and the quantity are critical to displaying crisp images in low light settings. Each camera inside of it has a chip or chips and there are a quantity of chips, I actually took some of these highly technical slides out that explain how the camera actually works and takes images from light and creates it to an electrical environment off the back side of the chip.

But every camera will have a set of chips or a single chip of varying size. The bigger the chip, the more light that can expose on to it, the more detail you can get in the image. If you have multiple chips then you have more than one chip trying to decode the information. The reason that’s important is because the combination of chips size and number of chips is going to really indicate how well that camera’s going to perform specifically in low light.

So if you take a camera that has one chip and it’s a small chip and you put it five feet away from me and we can open the aperture of the camera all the way up so all the light’s coming into the camera, it’d probably look pretty good. You take that same camera and you put it back there forty feet away and you zoom in on me, you’ll see all kind of video noise and my shirt, image quality will look really bad, my shirt might look brown or blue because it can’t decode the color properly because there’s not enough chips and the chips are not big enough. Does that make sense to everybody?

The size of the chip, that’s something you can not overcome. If the camera is not the right camera for where you physically have to place that camera, you can’t fix that. And sometimes you say we want to go with this or this, it’s just not going to work for you. There’s no sense spending your money if you put the image out and you look at it and oh, that looks terrible. So just don’t spend the money.

So if you’ve got really good lighting, now what happens is that camera can go farther away because you’ve got good light on the subject. It’s the combination of the chips and the light and the distance that’s going to equate to how good is that image going to look. And again, the first thing we talk about is when you zoom in the aperture on the camera closes, right, to get that zoomed in spot and that lets less light into the sensor. Now if I put the camera further back and I have to zoom in, now I got a real small aperture for light to come in to start with. If I don’t have great lighting, the image quality’s going to be poor. It’s a combination of all those factors. So there’s no fundamental rule of thumb where I can put a one chip 4K sensor camera back 45 feet and I’m good, it depends on what your lighting is like too.

So there’s different types of ways video is captured and processed in the cameras and through the switchers. An interlaced signal which is the most common output out of most broadcast cameras and is mainly used with the live output because it’s bandwidth tolerant, it’s 1.85 gigabits, 1.485 gig per second. Whereas a 3G output 1080p is 3 gig per second processing. Now that all comes into the cable lengths and everything else so the common output of most cameras today still, any HD is an interlaced signal.

What that means is you get one field, the next field, one field, the next field and they come together very quickly to create a full field. Your eye never sees that, television’s been like that since the 1940’s. That fundamental technology of TV has not changed much since that day. What happens is when you take somebody with stripes on their shirt and you have that kind of an image capture and you don’t have the correct processing, what happens is as that camera’s shooting and that guy’s moving the fields, as they’re painting them and laying them back together, get out of whack and you see the shirt start doing that. Right? So what you want to do, there’s a couple ways you can get around that when you design the system. You can actually take this and convert it to a progressive scan, that fixes that, but that’s the general concept of what’s happening there.

Alright so we talked about set design, clothing, right? So here’s a couple things too and I’m just going to pitch Panasonic’s dynamic range stretch feature that’s on their cameras. And again, I don’t work for Panasonic, I don’t got a dog in the hunt, but the dynamic range stretch is an amazing feature that’s on most of those cameras over there and basically what it does is it looks at the image on the camera and at pixel group by pixel group and adjusts those on the output based on each group.

What most other cameras do is they look at the whole image and they make adjustments based on the whole image. So what the heck does that mean? Why is that important? Right? Let’s say you have a pastor who’s got really dark skin and he’s standing in front of a very lighted background, right? A lot of light behind him and a dark skinned pastor, maybe he’s got a light colored suit or he’s in front of a light background. What’s going to happen is with a traditional camera you either have to iris down because of the surrounding light and if you do that what happens is all the detail in his face is gone.

Now let’s say you got a light skinned pastor with a nice dark suit with some nice patterns on it and you’ve got him, same issue. You’ve either got him irised properly and you lose all the detail in the suit, or visa versa, so that dynamic range stretch is absolutely amazing when you have that type of situation, which always happens in churches.

The image quality that comes off of that camera, dynamic range stretch, helps with defeating some of those issues with patterns as well, because the camera’s picking up specific blocks and understanding what image it’s looking at. I haven’t seen any other cameras that do as good a job as the Panasonic. So it’s a pretty cool feature.

These guys over here, that light’s in my face, I can’t barely make you guys out. The dynamic range stretch works kind of like that. It can say block that out, focus on that. You guys are the same way. I can barely see you with those lights, but if I block that light out, now I can see you clearly. It kind of does that on a pixel block by pixel block process. Does that make sense? Really cool feature.

So here’s the other kicker, if you don’t get good capture with your video quality and your audio quality, you are stuck. I had a lady friend of mine who works at our church, she’s a professional video production person, she worked at CNN and she did a big project and they captured the audio and it was all over driven. All of it. So she called me and said hey, is there anything you guys know about [inaudible 00:23:25] that you can fix this thing because I shot the whole project and it’s all over driven. I said well, no. You gotta go back and reshoot it, sorry.

So, understanding on the front side, getting a correct, good quality image, getting your lighting and your cameras balanced, getting your audio sound properly leveled, you’ve gotta do that on the front side or you’re going to be stuck with garbage and you can’t fix it. Does that make sense to everybody? You can’t take a distorted audio and fix it, you can’t take a really bad grainy video image and make it look pretty. Even though on TV they show you these cool shows where the high tech guys can make a license plate all the sudden appear magically, that’s baloney, right? Doesn’t work that way. My kids don’t believe me, but I’m telling them.

So this is sort of a signal path. I like to do this since I do engineering sometimes, you can go to live recording or you can go to recording for post. So basically what this means is I’ve got my camera images coming out and rather than send it live, I’m going to capture it and do stuff to it and then I’m going to send it out. Does that makes sense to everybody? Okay.

So when you’re capturing it live, kind of the same thing. You could have a great image coming out of the camera, if you record it in a really compressed format and then you try to edit it and then post it and wonder why it looks bad, it’s because you didn’t have a good enough capture, the quality wasn’t good enough when you captured it. Same problem if you’ve got great images coming out of the camera and your record a bad image, you aint fixing that. So making sure you record in a high enough quality is really critical if you want to do any kind of post production.

Now it’s kind of a balancing act because here’s the deal, if you record uncompressed video off those cameras, that’s 1.485 gigabits per second, per second. Imagine how big that file would be after you recorded for an hour. It’s massive, right? So you gotta compress it, you gotta compress it some, but you gotta find the right balance between compression and not being able to use it afterwards.

So fortunately most of the products that we have back there or that we sell, you’ve got a couple industry standard settings that you can set that give you decent compression with still decent quality. And for most of the churches, that’s usually Apple’s ProRes LT. ProRes LT is a pretty high quality compression and it’s a reasonable file size. So when we record a 45 minute talking head thing, it’ll probably be about 60 gig. Now to those of you who don’t, think that’s still a big number, that’s actually pretty small for high definition video.

So that brings us to the next part of this post production thing which is I’ve got these files, what the heck am I going to do with them, right? Do I upload them to the web and let YouTube hang on to them forever, do I want to archive these, do I want to put them on a server? The back side of this is I got all these great digital files, now I have to have some place to put them.

Back in the day, I like to say that back in the day because I’m an old dude. Back in the day we would record our stuff to beta tape or some of the new digital DVC pro HD tapes that came out and we would put them on the shelf and you’d have all the sermons, you’d have all the teachings and the next thing you know after about five or six years you’ve got a giant room full of tapes and then somebody said wow, we can do that in DVDs. Then you’ve got a giant rack full of DVDs, right? Back in the day.

Now it’s all digital, now you’ve got a giant server full of stuff, right? It’s the same basic concept. You’re going to take that file, if you want to archive it you want to put it somewhere where you can pull it back later. There’s a couple parts to doing that that are really important. One is when you get 150 programs on your hard drive, how do you find the one you’re looking for? So if you don’t have the correct nomenclature there’s actually an ISO standard for writing file nomenclature that we typically would recommend. Otherwise, you never find what you’re looking for.

Pastor was talking about, oh what was he talking about. He was talking about Joshua and the battle of Jericho, when was that? Was it two years ago, three years ago, and you might spend six days trying to find that particular sermon that he wants his particular thing from. So if you don’t set up your architecture properly when you do that, guess what? You ain’t never gonna find it. So these are part of the things that you have to consider when you talk about archiving your video. When you archive stuff to the web, similarly. You gotta remember what video was that. So naming the things is really important.