Tech conference videography: filming panels, demos, and networking
Tech conference videography - panels, product demos, and networking sessions - each requires a different camera approach. Here is what that looks like operationally, and what the brief needs to specify before filming starts.
By Vitalii Vakulchuk, DoP - We Stream
A tech conference is three different events happening in the same building on the same day. The panel sessions are formal, structured, and largely predictable in terms of where the action will be. The product demos are chaotic, spatially unpredictable, and often the most interesting content on the programme. The networking is informal, visually rich, and almost never filmed well because by the time the crew gets to it, the formally structured part of the day has consumed all available attention.
Covering all three requires treating them as distinct briefs within the same shoot - different camera approaches, different audio setups, different editorial priorities - rather than moving a single setup from room to room and hoping the footage is broadly useful. The conferences that produce genuinely good content from all three contexts are the ones where the brief specified each one separately.
We have covered the Berlin IT conference over three days - a client of four years, and one of the most consistent briefs we receive across all event videography work. We filmed Newsweek's conference with a two-person crew managing three cameras simultaneously across panel sessions while capturing highlights content in parallel. The operational requirements for both were specific and consistent. Here is what they were.
Panel sessions: the multi-speaker problem and why it is not solved by more cameras
A panel of five speakers presents a filming problem that is straightforward to describe and genuinely difficult to solve well. At any given moment, any one of the five might say the thing that the edit needs. The camera covering the panel from a wide position captures all of them; it captures none of them well enough for a usable single-speaker clip. A camera on each speaker would solve the editorial problem and make the shoot logistically impossible at any reasonable budget.
The practical approach involves a hierarchy of coverage: a wide position that captures the full panel and the audience response simultaneously, and a closer roving camera that follows the conversation - moving to whichever speaker is most active, staying with them long enough to capture a usable sequence rather than cutting constantly, then repositioning when the conversation shifts. The wide camera is the safety net. The closer camera is what makes the edit interesting.
The practical approach involves a hierarchy of coverage: a wide position that captures the full panel and the audience response simultaneously, and a closer roving camera that follows the conversation - moving to whichever speaker is most active, staying with them long enough to capture a usable sequence rather than cutting constantly, then repositioning when the conversation shifts. The wide camera is the safety net. The closer camera is what makes the edit interesting.
At the Newsweek conference, a two-person crew managed three cameras across all panel sessions while simultaneously producing highlights coverage. The three cameras were distributed - one wide, two with different focal lengths on the panel - and the highlights work happened in the transitions between sessions rather than by pulling a camera from panel coverage mid-session. The final delivery included a 1-minute-45-second highlight film and complete multi-camera panel recordings. Under a tight budget.
That was achievable because the coverage priorities were agreed before filming, which meant neither the panels nor the highlights were compromised by the crew making resource allocation decisions in real time. We've covered the brief decisions that determine what the footage can contain in detail - the mistakes that consistently waste conference video budgets and how to avoid them.
The audio question on panels is at least as important as the camera coverage and is treated as secondary far more often than it should be. A panel recorded on the room's PA system - the audio feed from the venue's installed speakers - is consistent but often processed in a way that sounds good in the room and compressed in a recording. A lavalier on each speaker is technically superior and logistically complicated at a five-speaker panel with unknown speakers who arrive moments before they sit down. A well-positioned directional mic on a stand between the panellists, combined with the PA feed as backup, is usually the practical solution. It requires knowing, before the event, that the venue's PA has a direct output and that the output is accessible during the session - not on the morning when the venue technician is setting up.
Keynotes: the one session that cannot be recovered
A keynote at a tech conference is the single session where failure is not manageable. It is usually the highest-profile speaker, the moment the rest of the programme builds towards, and - if something goes wrong with the coverage - the moment the client will notice first and remember longest.
The camera position for a keynote is not the same as the camera position for a panel. A keynote speaker owns the stage alone or nearly alone. The framing should be tighter than a panel wide shot - close enough to read the speaker's expression, not so tight that a step in either direction takes them out of frame. A second camera covering the audience - their reactions, the quality of their attention, the moments when something said produces a visible response - provides the edit with something that a single-camera keynote recording never has: evidence that what was said mattered to the room.
The camera position for a keynote is not the same as the camera position for a panel. A keynote speaker owns the stage alone or nearly alone. The framing should be tighter than a panel wide shot - close enough to read the speaker's expression, not so tight that a step in either direction takes them out of frame. A second camera covering the audience - their reactions, the quality of their attention, the moments when something said produces a visible response - provides the edit with something that a single-camera keynote recording never has: evidence that what was said mattered to the room.
At the Thames Freeport Launch at The Savoy, Rishi Sunak addressed a room that included several senior government officials and industry representatives. The full event video was delivered within five hours. That delivery was possible partly because the keynote coverage was planned specifically - not treated as a slightly more important version of the other sessions. The keynote determined the edit structure; everything else built around it. That framing decision was made in the brief, not in the edit suite.
The keynote is also the session where backup audio matters most. A PA feed that cuts out mid-sentence during a panel session is recoverable - the camera audio provides a workable fallback and the panel format allows for editorial flexibility. A PA feed that cuts out during the ten-minute keynote from the most significant speaker of the day is not. Two audio sources, confirmed before the event, is the minimum responsible coverage for any keynote session.
Product demos: filming something that was not designed to be filmed
Technology product demos at conferences were not designed with cameras in mind. The screen is usually too bright relative to the presenter. The viewing angle that makes sense for the audience in the room - looking at the screen from in front of it - produces footage that shows a screen and the back of someone's head. The interaction that demonstrates the product's value - the specific moment when the thing works in a way that justifies attending the conference - happens at the wrong distance, from the wrong angle, for approximately forty-five seconds, and is not repeated.
Filming a demo well requires being in position before it starts. Not when it starts - before, which means knowing that the demo is happening, approximately when, and what the spatial layout of the demonstration will look like. A crew that arrives at a demo in progress is already behind it.
Filming a demo well requires being in position before it starts. Not when it starts - before, which means knowing that the demo is happening, approximately when, and what the spatial layout of the demonstration will look like. A crew that arrives at a demo in progress is already behind it.
The DataBet Counter-Strike competition at SBC Summit Lisbon 2024 - where guests challenged a top CS player for prizes - was one of those activations where the interesting thing was happening in a specific space, in a specific sequence, and required the camera to be positioned with the right angle on both the player and the audience around them. The brief identified it as a priority. The crew was in position before it started. The footage was usable. At a trade show with multiple activations running simultaneously, identifying the priority and committing the camera to it is the decision that determines what the edit can contain.
For screen-based demos specifically, the practical options are: a clean HDMI capture feed from the presenter's machine, which gives the edit a clean screen recording to cut to; a camera positioned at an angle that catches the screen without the worst of the reflections and overexposure; or a close camera on the presenter's hands and face during the interaction, which communicates that something interesting is happening without requiring the screen to be legible. Each of these requires different equipment or a different camera position. None of them can be improvised from a standard event wide shot.
For screen-based demos specifically, the practical options are: a clean HDMI capture feed from the presenter's machine, which gives the edit a clean screen recording to cut to; a camera positioned at an angle that catches the screen without the worst of the reflections and overexposure; or a close camera on the presenter's hands and face during the interaction, which communicates that something interesting is happening without requiring the screen to be legible. Each of these requires different equipment or a different camera position. None of them can be improvised from a standard event wide shot.
Networking sessions: the hardest part of the day to film and the most valuable to get right
Networking footage is the most consistently underserved part of tech conference video production, for a reason that is operationally understandable and editorially costly: it happens at the moments when the crew is most likely to be setting up for the next formal session.
The morning coffee before registration closes. The break between the first and second panel sessions, when people who have just heard something interesting are discussing it while it is still fresh. The lunch period, when the conversations are less structured and more revealing. These moments are not on the running order. They do not have a scheduled camera position. They require a camera operator who is free to move through the room and read what is happening, rather than stationed at a tripod covering a stage.
The morning coffee before registration closes. The break between the first and second panel sessions, when people who have just heard something interesting are discussing it while it is still fresh. The lunch period, when the conversations are less structured and more revealing. These moments are not on the running order. They do not have a scheduled camera position. They require a camera operator who is free to move through the room and read what is happening, rather than stationed at a tripod covering a stage.
The Berlin IT conference highlight video included stage talks, panel discussions, round tables, and an evening networking session - all presented in a fast-paced, engaging format that the client noted consistently generates higher engagement and more reposts than other production companies they have used. The networking session footage was not an afterthought. It was planned coverage with its own brief: which moments to capture, what the camera needed to be looking for, and - critically - which crew member was available to cover it while the rest of the setup was configured for the next session.
The specific camera approach for networking is observational rather than directed. Directed networking footage - asking people to recreate a conversation, staging a handshake for the camera - looks like directed networking footage. Observational footage of a genuine conversation, caught at the right moment from the right distance, is the image that makes a conference feel worth attending to someone who was not there. It requires a camera operator who is comfortable working invisibly in a crowd, and it requires the brief to identify it as a priority rather than a bonus if time allows.
Three-day conferences and the coverage strategy that prevents day-three footage from being useless
Multi-day tech conferences have an energy distribution problem. Day one has the highest attendee energy - novelty, first impressions, the sessions everyone came for. Day three has the lowest - fatigue, thinner attendance, the sessions that were always going to be secondary. An edit that draws equally from all three days usually reflects that distribution honestly, which means the day-three footage drags the whole piece down.
The solution is not to avoid filming on day three. It is to know, before the event, what day three is for in the edit. At the Berlin IT conference, the three-day structure included a Female Founders opening session, a networking dinner, and then the main event floor - each day with a different character and a different editorial role in the final video. The Female Founders session opened the story. The dinner provided warmth and atmosphere. The main floor day gave the scale. That structure was decided before filming, which meant the coverage on each day was shaped by what that day needed to contribute rather than by a general instruction to film everything.
Same-day photo delivery across all three days - edited images delivered continuously while the event ran - required the photography to be structured the same way: which sessions on which days were the priority for social posting, and in what sequence. A photo of the main keynote speaker posted on day one performs differently from the same photo posted on day three, not because the image has changed but because the audience's relationship to the event has. The delivery schedule is an editorial decision, not a logistics one.
Audio at tech conferences: the one thing that most commonly destroys otherwise good footage
Tech conference audio is its own specialism and it is the element of conference coverage most likely to be acceptable rather than good - because acceptable audio from a room PA feed is the path of least resistance, and the problems with it usually only become clear in the edit.
Room acoustics at conference venues vary enormously. A room with hard walls and a high ceiling can produce a reverb tail that makes every sentence sound like it was recorded in a bathroom. A room with heavy soft furnishings absorbs so much ambient sound that the PA-fed audio feels disconnected from the visuals. The venue's own audio engineer optimises the sound for the people in the room, not for a recording that needs to be intelligible on headphones at one-quarter volume on a phone.
Room acoustics at conference venues vary enormously. A room with hard walls and a high ceiling can produce a reverb tail that makes every sentence sound like it was recorded in a bathroom. A room with heavy soft furnishings absorbs so much ambient sound that the PA-fed audio feels disconnected from the visuals. The venue's own audio engineer optimises the sound for the people in the room, not for a recording that needs to be intelligible on headphones at one-quarter volume on a phone.
Research published in Science Communication (Newman & Schwarz, 2018) found that degraded audio caused viewers to rate speakers as less credible and less intelligent - an effect that held even when the speaker's credentials were displayed on screen. The room audio engineer's job is to make the session work for the people physically present. The recording's job is different, and the standard it is held to is different.
For interview-led content at tech conferences - panel speakers captured for standalone clips, networking conversations filmed as testimonials - the audio is almost always better from a directional mic held close to the speaker than from any room feed. Close audio is clean audio. Clean audio makes the edit more flexible, because it can be brought up, reduced, or combined with music without the room noise becoming intrusive. For the DXC Technology and Luxoft partnership video - executive interviews shot with the London office as context - clean interview audio was the foundation that made the b-roll footage usable as a supporting layer rather than a distraction from conversations that could not quite be heard. The audio quality was an edit decision before it was a shoot decision.
The practical checklist for tech conference audio is short and worth running through with the venue before the event: direct PA output available and tested, room acoustics assessed during the pre-event access window, backup recording source confirmed on every interview setup. Each of those items is findable in advance. None of them is fixable in post-production if they are not addressed before filming.
The practical checklist for tech conference audio is short and worth running through with the venue before the event: direct PA output available and tested, room acoustics assessed during the pre-event access window, backup recording source confirmed on every interview setup. Each of those items is findable in advance. None of them is fixable in post-production if they are not addressed before filming.
A tech conference is three briefs in one building. The panel coverage, the demo coverage, and the networking coverage each require specific decisions - about camera position, audio setup, crew allocation, and editorial priority - that cannot be made well on the day. They need to be in the brief. If you're planning tech conference coverage and want to understand how we structure conference videography packages for events of this type, that is covered separately.
FAQ
How do you film a multi-speaker panel at a tech conference?
With a coverage hierarchy, not more cameras. A wide position captures the full panel and audience response simultaneously - that is the safety net. A closer roving camera follows the conversation, staying with the active speaker long enough to capture a usable sequence before repositioning. At the Newsweek conference, a two-person crew managed three cameras across all panel sessions while simultaneously producing highlights coverage - achievable because coverage priorities were agreed before filming, not decided in real time.
What audio setup works best for filming conference panels?
A well-positioned directional mic on a stand between panellists, combined with the room's PA feed as backup. A lavalier on each speaker is technically superior but logistically difficult when speakers arrive moments before sitting down. The PA feed alone sounds processed in a way that is acceptable in the room and compressed in a recording. The practical solution requires confirming before the event that the venue's PA has a direct output and that it is accessible during the session - not on the morning when the venue technician is setting up.
How do you film a technology product demo at a conference?
Be in position before it starts. The interaction that demonstrates the product's value happens at the wrong distance and wrong angle for approximately forty-five seconds and is not repeated. For screen-based demos the practical options are: a clean HDMI capture feed from the presenter's machine; a camera angled to minimise screen reflections and overexposure; or a close camera on the presenter's hands and face, which communicates that something interesting is happening without requiring the screen to be legible. None of these can be improvised from a standard event wide shot.
Why is networking footage so hard to film at tech conferences - and how do you get it right?
Because it happens when the crew is most likely to be setting up for the next formal session. Directed networking footage - asking people to recreate a conversation - looks like what it is. Observational footage of a genuine conversation, caught at the right moment, is the image that makes a conference feel worth attending to someone who was not there. Getting it requires a camera operator free to move through the room, and a brief that identifies networking coverage as a priority rather than a bonus if time allows.
What makes keynote coverage different from panel coverage at a tech conference?
Framing and consequence. A keynote speaker owns the stage alone - the camera should be close enough to read their expression, not so tight that a step takes them out of frame. A second camera on the audience captures the evidence that what was said mattered to the room, which a single-camera keynote recording never has. And backup audio is non-negotiable: a PA feed that cuts out mid-panel is recoverable; one that cuts out during the day's most significant speaker is not. Two confirmed audio sources is the minimum.
How do you structure video coverage across a three-day tech conference?
By deciding before filming what each day needs to contribute to the edit, rather than filming everything equally and hoping the energy distribution works out. Day one typically has the highest attendee energy; day three the lowest. An edit that draws equally from all three reflects that honestly - and day-three footage drags the whole piece down. At the Berlin IT conference, the three-day structure had a specific editorial role for each day: the Female Founders session opened the story, the dinner provided atmosphere, the main floor gave the scale.
How do you cover product demos and panel sessions simultaneously with a small crew?
By allocating each crew member a specific brief before the day starts, so the two tasks are not competing for the same resource at the same moment. At the Newsweek conference, a two-person crew covered three cameras across panel recordings and highlights content simultaneously - because the coverage priorities were agreed in advance. Without that specificity, the same two people covering everything would have covered everything adequately rather than covering each thing well. The distinction between adequate and usable is what the brief determines.
What goes wrong with audio at tech conferences?
Room acoustics are the most common problem and the least anticipated. Hard walls produce reverb that makes speech sound like it was recorded in a bathroom. A room's PA audio engineer optimises sound for people in the room, not for a recording heard on headphones at low volume on a phone. For interview content - panel speakers captured for standalone clips, networking testimonials - a directional mic held close to the speaker is almost always cleaner than any room feed. Clean audio is what makes the edit flexible; room audio recorded without assessment makes it rigid.
How much does tech conference video production cost in London?
Cost depends on crew size, number of filming days, the mix of panel recording and highlights coverage, and whether same-day photo delivery runs in parallel. A two-person crew managing multi-camera panel recordings and highlights for a single-day conference differs significantly from three-day coverage with same-day photography and multi-format social cuts. We Stream has covered both scales - for Newsweek and the Berlin IT conference across multiple years. For a full breakdown of day rates and formats, see our London video production costs 2026 guide.
What information does a video crew need before filming a tech conference?
Six things specifically: the running order with timings so coverage priorities can be allocated before the day; which session is the keynote-equivalent that cannot be compromised; whether product demos are on the programme, when, and what the spatial layout will be; the venue's PA output availability and room acoustics; the full deliverable list including social media formats; and which crew member is responsible for networking coverage so it is not defaulted to whoever is free between formal sessions. All of these are answerable in advance. None are manageable well under time pressure on the day.
© All rights reserved. We stream
team@westream.uk
team@westream.uk