Jump to content

Live Streaming Experiences


Recommended Posts

Since I continually set up and strike in remote locations, it's important to keep track of which camera is which. Since my two PTZ cameras are otherwise identical, I pasted numbers on the side, also with the WiFi camera adapters for the Slingstudio. With most switch gear, port 1 is always port 1. The Slingstudio is different. The cameras are numbered in the order they are connected (paired) with the hub. For consistency, I keep the static camera (Sony, etc) as Camera 1, and the PTZ cameras left (1) and right (2).


I do something similar with microphones. I link stereo pairs together, for example Center LR, Side LR and wide LR. Typically the center mics are an ORTF (angled out) directional mics, the side mics are spaced (1 meter) omnis and wide mics are about 8 feet to either side, wide cards or omnis. Large choirs are generally spread wide but shallow. I use single mics spread across the stage, fairly close to the singers (~6') and high (8'), and number them in order, left to right. Having a "plan" helps keep things straight when mixing for FOH or live-streaming. Spot mics I usually grouped by section (e.g. winds, tenors, harp, etc).


My philosophy for classical music conforms to the old Deutsche Grammophone standard of one mic (or stereo pair) in the ideal place to establish the listener's point of view, and everything else just supplements that focus. Each mic is recorded in a separate channel, so the miracles can be created in post.


* FOH = Front of House (amplified reinforcement).

Link to comment
Share on other sites

  • 3 weeks later...
  • Replies 66
  • Created
  • Last Reply

Top Posters In This Topic

I'm currently using three systems for managing video from multiple cameras. Each has its advantages and disadvantages, and there are no universal solutions.

  • Slingstudio
    Up to 4 cameras, using HDMI/WiFi camera adapters which work up to 300' line of sight (no bulky SDI or HDMI cables across the floor, auditorium or gymnasium)
    Only one direct HDMI input
    Up to 6 Smart phone camera connections, using an app, "Slingstudio Capture"
    Maximum 1080p30 resolution
    Automatic switching at specified intervals (seconds by camera), so you can concentrate on other tasks (audio, stream quality, etc).
    High video latency, between 1 and 2 seconds, makes remote camera (PTZ) difficult.
    Unbalanced 1/8" (3.5 mm) stereo audio input (in addition to embedded audio from each camera)
    Flexible use of recorded clips and graphics
    Records program and ISO videos to an SD card or USB-C SSD
    Small tally light in each camera adapter showing the active camera
    Monitoring and control from a computer or iPad (tablet) via WiFi
    Records external and PGM audio, but does not add it to the ISO streams.
    Built-in WiFi and ethernet connections for streaming. Mini-HDMI (type C) port for external encoding
    Compact transport and setup. WiFi monitor and control
  • Black Magic ATEM Video Switch
    Up to 8 cameras, 4 HDMI and 4 SDI
    Maximum 1080p60 resolution, HDMI or 3G SDI
    Balanced L/R XLR audio inputs plus embedded audio with a virtual mixing panel.
    Almost zero latency (< 5 msec)
    Limited real time effects, transitions and keying (e.g., green screen).
    Some models offer front panel operation. All have complete function virtual panels for computers via ethernet (can be networked)
    Compatible with several tally light systems, via ethernet or RS422.
    Can use and control (via computer interface) up to 4 BM Hyperdeck recorder/players for video clips, via IP.
    No easy way to record PGM and ISO video streams
    Requires an external monitor for multiview and program video streams
    A lot of cables and connections! Hardware monitor and control (laptop and front panel)
    Requires use of an external encoder, either hardware or coftware (e.g., Wirecast or OBS)
    Not directly compatible with PTZ Optics 3G SDI output. Use HDMI or and SDI translator ro 1080p60.
  • Atomos Shogun 7 recorder/monitor
    Up to 4 cameras via SDI, plus word clock
    Maximum 4k60, 4,2,2 10-bit resolution (6G SDI)
    HDMI and SDI output
    Balanced (XLR adapter) stereo audio inputs and outputs.
    Almost zero latency
    Switching mode with touch screen operation in multiview
    Very high quality recording of PGM and ISO videos, ProRes or DNx
    Records embedded audio for each stream, with an option to add the external audio to each stream
    Global audio compensation for video latency (in msec).
    Integrated monitor and control (7" HD screen, usable in sunlight with minimal shading).
    Hard cut transitions in real time. Embeds optional cross fades in a Final Cut Pro XML file
    Requires an external encoder, either hardware or software (e.g., Wirecast or OBS).
    Fewer cables and connections than ATEM, more than for a Slingstudio
    Very difficult to insert clips and graphics.

I find the Slingstudio very useful for streaming concerts and events which require little or no post processing for re-broadcasting. The high latency precludes live control of PTZ cameras, deferring to presets to achieve a variety of angles.


The Shogun 7 is my preferred choice for high quality productions, mainly for uploading after grading and editing. Low latency means I can use PTZ cameras in real time, but I prefer to select and set up presets offline. However I can set up an angle in half the time as with a Slingstudio.


My last choice (not used for months) is the ATEM hardware switcher. It is very hardware intensive, with lots of connections and cabling. Ultimately the ATEM is the most flexible, and has broadcast quality. In addition to setup, it requires a lot of personal attention in use. Inability to record ISOs (without splitters) means you have to get it right the first time. At this point, the key advantage is the ability to use tally lights to apprise the talent of the ready/live state of cameras, so they know where to look.


What's next? I'm seriously looking into getting a teleprompter. I record and stream many programs which involve narration without notes. Not everyone is disposed to speak ex temp, and a teleprompter would add more polish to the production. Systems which use an iPad are simple and relatively inexpensive. The talent can pace the display with a foot pedal. I expect that its use will require practice by both the talent and the producer.

Link to comment
Share on other sites

Recent technology affects my use of an SDI ATEM video switch, affordable fiber-optic HDMI cables of 25' to 50', and inexpensive HDMI to 3G SDI adapters (also Black Magic) with dual SDI outputs.


While I can use an Atomos Shogun 7 to record and switch up to 4 cameras, it is difficult to add graphics and pre-recorded video clips, or use real-time transitions, lower thirds and bugs (small graphics advertising the production). Nor is it possible to directly control tally lights, which identify the ready and live cameras to the talent. These effects are readily implemented in a hardware switch (e.g., BMD ATEM), but without the ability to record the ISO videos. Black Magic pairs the ATEM inputs, but the second port is for control and communication with the camera, not for sharing the input (are you listening, BMD?). Every feature of the ATEM can be remotely controlled via ethernet, along with various memory and recording devices (e.g., BMD Hyperdeck carts).


Fiber HDMI cable, suitable for 3G signals, costs as little as $60 in 50' lengths. It's easy to handle in the field, being very flexible and only 5 mm in diameter. BMD just came out with HDMI to SDI Micro adapters, powered by USB, for about $50, which have dual 3G SDI outputs. I will use these outputs to feed the ATEM and Shogun simultaneously. Since Shogun channels are perfectly synchronized, post production is greatly simplified.


The downside is a dozen or so additional cables and connections and more hardware. However the additional power can be used to add professional flair to a live production with a relatively compact setup, with real-time control over up to 8 inputs and 4 Hyperdecks.

Link to comment
Share on other sites

  • 3 weeks later...

Using a Slingstudio (SS) last weekend proved to be just short of a disaster. I selected the SS in order to use the WiFi camera adapters instead of wiring three cameras, up to 50' away, in a tent with a dirt floor and foot traffic. Secondly automatic switching every 10 seconds is more entertaining to viewers and frees me to pay close attention to streaming status and camera manipulation. I used one static camera, two PTZ cameras, and 12 microphones for recording, shared with a FOH (Front of House) mixer. So far, so good.


The trouble began when my iPad would not pair with the Slingstudio for control. The only control surface is in an tablet or computer, so without that connection you can't setup or run the device. I ultimately had to cycle the SS power and re-pair the camera modules. I used a Teradek streaming encoder, which has a powerful WiFi transmitter. It's. possible it was too close the the SS, which uses it's own WiFi to connect to a control surface, causing interference.


An enduring nuisance is the 2 second latency in the SS video, which makes it impossible to control PTZ cameras efficiently. You have to bump the joystick and wait, ad infinitum in order to establish a preset. That done, you press one of the preset buttons to use a new angle, when that camera is not live.


It's nice to have a separate recording for each camera, in addition to the Program video sent to the encoder. That way you can patch the switching in post so that it fits the action better (e.g., focus on a speaker). The SS makes editing complicated because the individual recordings (ISO's) are out of sync by as much as 2 seconds.


The good news is I had a lot of help with the audio, before and after the concert (Jazz), which I appreciate greatly. Out of respect for COVID, musicians were widely spaced and in cubicles covered with plastic sheeting. I set up each musician with a separate microphone, at fairly close range. That makes for a better recording, and gives a lot of flexibility for FOH sound. The more live microphones at once, the more susceptible FOH is to feedback. Close miking makes feedback much less likely than with fewer microphones further from the talent.


Someday soon I should talk about setting levels. Cheers.

Link to comment
Share on other sites

I was hard on the Slingstudio in my last post. I had two hours to set up what is normally a 3 hour job if you do everything right, and not everything went right on Sunday. Backing up a bit, The Slingstudio works best if you use ordinary video cameras, not PTZ cameras, with built-in microphones which capture a sound track synchronized with the video. You can mix those audio tracks into the stream, but it's better to have separate microphones on stands, a mixer or recorder, with a stereo feed to the 3.5 mm TRS jack on the hub. ISO's are not recorded in sync with each other, but they're easy to line up if they have audio tracks. The program (switched) track is recorded too, but I usually like to touch it up in post to improve the timing and relevance.


Here is a breakdown of what happened...

  • At least six computer or tablet applications were needed at the same time: (1) Setup and control the Midas rack mixer, (2) record 16 channels of audio with Cubase, (3) Setup and control the Slingstudio, (4) Set up and monitor the Teradek encoder, (5) set up and monitor CORE cloud bonding and streaming service, (6) Initialize and monitor the destination channel (YouTube).
  • Cubase has to run alone, or else it is likely to crash. The good news is that tracks are saved every second or so, and can be recovered. The bad news is a crash loses some data which can't be recovered. I tried to run the Midas application, which is a virtual mixing panel, at the same time, which caused the crashes.
  • The Midas can only record a stereo mix to an SD card. That's not great, but better than nothing. That works even if the computer is disconnected.
  • A tablet or computer must be dedicated as the Slingstudio console.
  • A computer or table must be dedicated to setting up, enabling, and monitoring the YouTube destination. The channel must be enabled BEFORE starting the live-stream.
  • If the stream is interrupted more than about 5 minutes, YouTube ends the live-stream and archives the results. An unknown amount of time must transpire before you can resume streaming.
  • For some reason I could not reconnect to the Slingstudio after the initial setup. The SS acts like a WiFi access point with its own SSID, to which you connect your WiFi. Everything looked okay, but wouldn't connect until I cycled the power on the SS, which affects everything downstream as well as the camera connections.
  • Face-recognition doesn't work when you're wearing a COVID mask. You end up typing a lot of passwords, and your glasses fog up.

The happy news is I got an outstanding recording, once everything was working. I bought 16 Shure SM58 microphones so that every musician had his/her own, close enough to eliminate traffic and wind noise, and avoid feedback from the PA system. With Social Distancing, in a huge tent, musicians can't hear each other well, so a PA system helps both them and the small audience. Most of the audience is on YouTube, including quite a few in the parking lot..


In the future, I (or someone else) will set up a cell phone connected to YouTube, as a backup to get a stream going if other problems arise. Everything seems to happen in the last 15 minutes before air time, so that will be one less worry

Link to comment
Share on other sites

Premiere Pro has a function which is supposed to align audio tracks by similarities in content. I have not been able to make it work. It takes a very long. time for analysis, then fails. On the other hand, auto-alignment using embedded time code is fast and accurate.


It's also easy enough to do manually, using sound tracks (locked to the video). You only have to re-sync when there's a break in either the video or external audio. I shoot long clips and only break when absolutely necessary. There are ways to jam time codes in cameras and recorders to start at the same value. Cinematic cameras usually have sync ports which can be tied to a master clock by cable or wireless. For time code, you have to be handy with money.


The problem with PTZ Optics cameras is the lack of internal sound or time code support. The variance is usually 2 seconds or less, and can be aligned with video only if the same subject is covered by each camera at some point in the clip.


Different cameras tend to have different video latency, which means you may have to align video and sound in those clips before syncing clips with each other. Is this sounding like fun?


If all else fails, try to do it right the first time so you don't need to edit aftwards ;)

Edited by Ed_Ingold
Link to comment
Share on other sites

Sound for large groups with social distancing is proving to be a special challenge. Close miking and a lot of it seems to be the way to go. I just added 16 Shure SM58 mics to my collection, with another 8 on the way.



Link to comment
Share on other sites

I thought it would be useful to illustrate the recording situation which requires so many microphones, not to mention the challenges of lighting and camera placement. The main goal for a concert of this sort is (1) to show faces for the parents' sake, and (b) make them sound as good as possible. "Cinematic" quality doesn't make the list, but faces can be seen reasonably well in closeups (which I cannot post).


As you see, the concert is in a tent with a dirt floor. I'm considering taking my cables to a car wash ;) I'm using two PTZ cameras on 9' (13.2' max) light stands plus a fixed camera on a tripod for an overview (this shot). I'm using a Slingstudio with WiFi camera adapters, which eliminates the need for video cables. However the PTZ cameras need CAT6 ethernet cables for power and control. To reduce cable runs for audio, I used two 8-channel sub snakes, and 25' or 50' XLR cables to the mics. As much as possible, cables were routed away from foot traffic, or covered with runners. For sound, I used a Midas M32R rack mixer with a 32/16 digital snake/splitter. Inputs 1-16 were routed directly to the 16 outputs, subsequently to a FOH board.


This is a choir, but the members are in individual plastic enclosures per COVID requirements. There weren't enough microphones for everyone, so we tried to place them high, sharing between cubicles. Members can't hear each other, so we placed four stage speakers around the set, feed with a mix from these microphones. The digital splitter was used so the same mics could be used for recording and reinforcement independently. I was able to extract a useable mix for recording, but the experience prompted the purchase of enough mics to go around, as shown in the previous post.



Link to comment
Share on other sites

Between various devices and the internet, you need windows on a lot of things when live-streaming.


  • Destination Websites - When you stream using RTMP, you need to initiate live streaming in the destination sites, like YouTube and Vimeo. This requires internet access, preferably with a computer (type may be required) and more than one window or tab at once. The YouTube site must be made ready for streaming before streaming commences.
  • Encoder - You need to set up the encoder (e.g., Teradek Vidiu Go) and monitor its performance
  • CORE - I use Teradek CORE cloud service to bond two or more internet connections, and deliver the material to one or more destinations. Streaming is also started and stopped from the CORE dashboard, using a computer or iOS device. The same device can be used for both the Teradek and CORE.
  • Video Switching and Recording - When possible, I use an Atomos Shogun 7 for switching and recording up to 4 video cameras. It's a stand-alone device, SDI based, which serves as its own monitor. A Slingstudio, on the other hand, can only be controlled from an iPad or computer. There are no buttons to push, so its a full time job for the computer. In between, a hardware switcher can be operated in basic mode from its front panel, or more comprehensively with a dashboard utility in a computer.
  • Audio Mixing - When possible (8 channels or less), I use a Zoom F8n or Sound Devices MixPre 10ii. An iPad make mixing easier, but either can work well on its own. I can string two or more together, but only one can be monitored at a time. Consequently I use a Midas M32R rack mixer with up to 32 preamp channels, which requires a computer to set up and monitor. Normally, I feed the first 16 channels directly to corresponding outputs, and use that as a splitter for feeds to the FOH sound mixer.
  • Audio Recording - The F8n and 10ii are stand-alone recorders, using SD cards. However if I want to record from the Midas, I need a dedicated computer and DAW software like Cubase or Pro Tools, and record to an external SSD drive.

What this all means is that I need 2 to 6 computers or smart devices to run the show. When the lights start to flicker, you know I'm on the job ;)


I find I cannot run the Midas control panel and Cubase at the same time in the same computer. That means separate laptops. I'm considering an alternative. A full sized 40 channel mixer is too big for my type of operation. However JoeCo makes a 24 channel mic/line mixer recorder, with a virtual control panel. A Dante or MADI card can be added to use a digital stage box (or more) via CAT6 cable. This is a very expensive piece of kit (up to $4-6K), so I'll suffer with what I've got until Columbia Records gives me a call.

Link to comment
Share on other sites

  • 2 weeks later...

Not every Slingstudio job is quick and easy. Here, I'm using 9 mics, a standalone 8-channel recorder plus a 40 channel Midas rack mixer as a splitter for independent front-of-house sound. There are three cameras. One is static (Sony A7Siii) plus two PTZO 30x cameras and joystick controller. I'm using Cubase to record from the Midas, an iPad to control the SS, and another iPad to manage the streaming destinations. In lieu of ethernet, I'm using a Teradek VidkU Go streaming encoder with two cellular modems (HDMI Out from SS).


Last weekend expanded this setup to 26 microphones, requiring two 16-channel FOH mixers. This was for a choral program with plastic partitions and 6' separation for each member. A mic was used for each member, plus solos and announcements. Without reinforcement, the audience can't hear them and they can't hear each other. We added 4 floor speakers for the chorus, which made all the difference. I takes another person (shown) to handle FOH mixing in all but the simplest installations.


A word of caution. Keep the SS well away from other wireless transmitters. It doesn't want to pair with the control surface if there's interference. Also, you can't reliably do anything other than record with one computer, including a virtual mixing panel for the Midas.


It's handy to have a camera you can move around without then encumbrance of cables. In this setup, I used a Sony A7Siii with a PZ 28-135/4 lens on a tripod, with a large (190 WH) battery and a Teradek Bolt LT 750 wireless TX/RX pair. It has virtually zero (> 1 msec) delay, so it can be used together with wired cameras for live-streaming. Notice the dirt floor. I kid others that I need to take my cables to a car wash afterwards.



Link to comment
Share on other sites

  • 1 month later...

This is a job I did last Tuesday, a solo piano recital in a small studio. I used 2 mics under the piano lid and 2 main mics about 8' away, plus a handheld wireless mic for the soloist to speak. There are three cameras - 2 PTZ and a Sony A7Siii for a fixed wide shot. I shot this using a Slingstudio with 3 wireless SS Camera Links. One PTZ was shooting over the pianists shoulder, showing the full keyboard, and the other showing a 3/4 head and shoulders from the front. I didn't move the PTZ angles often. They make it easy to fine-tune closeups without walking around. About a dozen people attended, which complicates the setup a bit. Clockwise from the upper left, the console consisted of a POE ethernet switch, an 8 channel Zoom F8n recorder, a Zoom FRC-8 mixer attached to the recorder, and a PTZ Optics IP joystick camera controller. The PTZ cameras were mounted on aluminum light stands, which are taller than tripods and take much less floor space.


The main mics were in ORTF configuration, about 7.5' high in line with the piano lid to reduce cancellation. The piano mics were also an ORTF configuration, low and aimed toward the high and low strings respectively, mixed abut 3 dB below the main mics. A mixer with sliders makes it easy to adjust levels, expecially for speaking mics, instead of tiny knobs on the recorder.


IMG_0467.thumb.JPG.c217ea673f989b48e1ffe64e1cb50a6f.JPG IMG_0468.thumb.JPG.6cc98e2055bdaac5b5c6568aa8c8c879.JPG IMG_0471.thumb.JPG.dfdea1b4b2e9509ac2da2d03be83db27.JPG IMG_0474.thumb.JPG.43b46c2d2cb8c0a7e132784bcceb18eb.JPG

Link to comment
Share on other sites

  • 2 weeks later...

I find that I rely on iPads more and more in my endeavors. The key attribute is the ability to connect to the internet through cellular service, independent of WiFi and Bluetooth. None of my laptops (PC or Mac) have built-in cellular. You can use cellular modems, but they usually connect through WiFi, losing that independence. Why is that important?

  • WiFi is needed to control some of the devices I use, including a Sliingstudio and a Teradek VidiU streaming encoder. These devices act as access points, so you connect to them by WiFi using the devices' SSID. PTZ Optics cameras are controlled and powered through ethernet cable. Prior to a dedicated PTZ joystick controller (also PoE ethernet), I used software control in a laptop. By inserting a small access point (TP Link) into the PoE router, I could use an iPhone or iPad to control the cameras by WiFi connected to the TP Link SSID. Using WiFi in this manner means you can't connect to the internet by WiFi, only wired or cellular.
  • You need internet access to log into the destination website (e.g., YouTube) to set up an RTMP channel and monitor the results. I use a cloud service (Teradek CORE) to bond ethernet, cellular and WiFi connections, and to distribute the stream to one or more simultaneous destinations. This, too, requires an internet connection, but not necessarily a fast one (> 5 Mb/s) needed for 1080p video.
  • Unless you are streaming from a phone or webcam, you need to use RTMP protocol. When establishing an RTMP connection, you create or re-use a stream key, which is a 24 character password for that connection. The stream key must be copied into the encoder you are using (e.g., Slingstudio or Teradek), which is much easier to do by copy and paste within the same smart device.
  • There is a special caveat - The destination must be enabled BEFORE you begin streaming to that site. If your video fails to connect, you may need to stop and restart the stream. If the destination is designated "recurring," you can start and stop the stream as often as needed. If you stop longer than a few minutes (e.g., 5 min), the destination will end the stream, archive the results, and return to the ready mode. You may need to interact to respond to popups in the process (i.e., YouTube). If there's time, you can edit the name of the stream, making it easier for viewers to identify.
  • It is not always possible to establish a working internet connection through the venue's network, assuming there is one. I recently found a WiFi system I used before was unavailable due to a firewall gone berserk. As a result, my laptop was rendered dead weight. A cellular iPad connection saved the day. I used the MacBookPro power supplies (I have two) to keep the iPads running. An ordinary charger can't keep up.
  • Once things are running, it's not necessary to maintain a continuous connection to devices or the internet. However a lot of things can go wrong, and it's advisable to stay on top of things. I used two iPads for this recent job, one to control the encoder and the other to monitor two internet destinations (YouTube and Vimeo). Neither of these tasks requires much bandwidth or processing. The Teradek encoder was running on cellular data only, but with a special data account, dedicated to streaming.

Link to comment
Share on other sites

I have a job coming up shortly in which I need to stream a pre-recorded video prior to the concert. Basically, it started with a PDF of the printed program, which I parsed into 22 individual JPEG pages, using Adobe Acrobat, pasted them into a timeline in Premiere Pro, and rendered them into a video. At 15 seconds per page, the whole video is only 6 minutes, so I need to loop it for half an hour prior to show time.


Adding video (and/or graphics) is very easy with a SlingStudio (and most software solutions like OBS, Wirecast and VMix). The material can be saved in internal memory, on an SD card, or an external drive. Once you add them to the graphics page you can drag and drop them into the broadcast lineup. Looping is an option in the setup, to play once and stop, or loop repeatedly. The downside is the SlingStudio has only one direct HDMI camera input. Other cameras must be equipped with SS WiFi Camera Links. There is also a 2 second lag with respect to real time in the video.


I concluded that I need a Plan A and Plan B, the latter being the SlingStudio, because it's easier to work with hardware switches, which have near zero delay. In this case, I will use an Atomos Shogun 7 recorder/switch, which has 4 video inputs and records ISOs, audio and PGM files synchronously. The problem is how to play a video through the Shogun.


I have a small SD card video recorder/player, a Black Magic HyperDeck HD Mini, which interfaces nicely with a BM ATEM switcher and can be controlled through software. My thought was to switch between the Shogun and the HyperDeck with the ATEM. Call this Plan C for the moment. Until now I haven't figured out how to use a video not recorded on the HyperDeck. The manual says nothing useful. After researching on the web, I found you can create a QuickTime video in ProRes HQ and put it the root directory of an SD card. You must also change the HyperDeck's format to ProRes HQ, or it will not find the video. That done, you can hit the play button and the video will play and stop. If you hit the play button twice in succession, it will loop continuously.


When I connected the HyperDeck to the Shogun, the video played on the screen, but there was a warning that the Shogun was not compatible with Level B SGI (in the 4 input mode). It turns out that this HyperDeck is the only device ever to leave the BM factory with Level B only output. (Level B divides the frame into two for compatibility with interleaved video.) I needed a convertor. My first thought was to use a Decimator translator (a device no video engineer should be without). However a BM Ultrastudio micro HDMI to SDI converter does the job without setup. The HyperDeck has an HDMI output, which I convert to SDI Level A for the Shogun. Either the Decimator or BM Micro can be used with PTZO camera (Level A only), which are not compatible with an ATEM at 3G bandwidth. The BM converter is USB-C powered, unlike the Decimator which has an AC power supply.


If all goes well, the pre-quel video will go into one channel of the Shogun, while I use three cameras for the other channels. I'll have the Slingstudio ready to go, just in case, two PTZ Optics 30x robotic cameras and an unmanned Sony A7Siii in the balcony for overview. I will use a Teradek Bolt transmitter on the Sony in lieu of dropping cables over the balcony.

Link to comment
Share on other sites

I can't emphasize enough the importance of recording the event's audio and video locally. At very least record the program (switched) video as it is being streamed. Ideally you should record the individual videos (ISO's) as well, including the mixed audio which accompanies the program.


A successful live stream requires that every stage in the transmission and destination work perfectly. If any link in the chain is weak, the results will suffer. If you have a copy of the program, you can upload it immediately for viewing, or restream it for viewing in real time.


There are several ways in which this can be accomplished.

  • Insert a recording device in the SDI or HDMI stream before it is sent to the encoder.
  • Record in the encoder itself (e.g., SlingStudio)
  • Use a separate recorder for each camera
  • Record each camera internally (not all cameras allow simultaneous record and transmit via HDMI)

I generally use an Atomos Shogun 7, which has up to 4 video inputs, switch between them for the broadcast, and record the PGM and ISO files. All of the videos and audio are perfectly synchronized using this method. If you use separate recorders, or record in the camera, you must synchronize the files manually (in lieu of a master time code link to each device). A program recording between the switch output and the encoder can be used as-is, and sometimes that is enough.

Link to comment
Share on other sites

  • 1 month later...

The new concert season isn't as grim as for the last 18 months, but recording with multiple cameras in a larger venue presents new challenges. PTZO cameras require at least one cable, Ethernet, for power and control. For compatibility with most switching systems, you also need an SDI cable for video, or at least an HDMI cable. NDI video, which uses the same Ethernet cable, is a viable though expensive alternative, but requires a computer for switching and streaming to social media. Computers use so much CPU time for video processing that bandwidth for streaming suffers. Laying out and dressing cables for safety is a major task. Live-streaming also requires all video and audio control be consolidated in a central location. My goal is to place cameras in locations which can't be wired directly, including balconies.The search is on for wireless solutions.


Wireless video is pretty straight-forward. A quick and dirty solution (subject to limited availability) is to use a SlingStudio hub and SS CamLink WiFi transmitters at each camera. The CamLink devices use HDMI connections. They have a built-in battery, but the run time is so limited it's best to use an USB battery pack. CamLink transmitters work only with a SlingStudio hub, and the hub only has one HDMI direct connection. Every other video input must use a CamLink or be a smart phone running the SS "Capture" app.


A more flexible solution is to use TX/RX sets which work with both HDMI and SDI connections - HDMI for nearly any camera, and SDI for professional level switching and recording. I have a couple of Teradek BOLT kits which are reliable in urban environments at 400' or more, with nearly zero lag time. Alternatives include various Hollyland kits, which are less expensive but have up to 0.25 sec lag and less robust WiFi than Teradek. The Atomos Shogun 7 or Black Magic ATEMs work with this setup.


For power, you would like to have the option of battery operation, preferably with a high capacity V-Mount or Gold-Mount 14.4 VDC video battery. These batteries have D-Tap outputs for which there are adapters for nearly every video camera, as well as for Bolt and CamLink devices described above. I have not yet found a battery solution for PTZO cameras which I can trust (2.1 mm pin and barrel or PoE, 12 vdc).


There is a wireless control (Ethernet) solution for PTZO cameras, using their "Wireless Cable" system, which uses robust, encrypted WiFi TX/RX pairs. These devices are compatible with D-Tap power, and connect to an IP Joystick at the other end. One receiver can work with up to 4 transmitters, which is ideal for sports events with multiple, widely separated PTZ cameras.


There's no problem if AC power is available at the remote location. Full battery operation still wants a solution.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Create New...