Starlink’s raw speed numbers are more than enough for any video call. The problem isn’t bandwidth — it’s jitter, micro-drops during satellite handoffs, and things happening on your local network that video calls can’t tolerate the way streaming video can. Most of these have real fixes. A few are inherent to how satellite internet works and can only be managed, not eliminated. This guide tells you which is which.
The frustrating part about Starlink and video calls is that the numbers look fine. Your speed test shows 100 Mbps. Your ping is 30 milliseconds. But Zoom still freezes mid-sentence and Teams audio turns robotic. Here’s why that happens — and what to do about it.
Speed — the number your speed test gives you — is not what matters most for video calls. What matters is jitter (how much that latency varies millisecond to millisecond) and packet loss (data that never arrives at all). A 100 Mbps connection with high jitter will produce worse calls than a 20 Mbps connection with stable, consistent delivery. Video calls are real-time — they can’t buffer ahead the way a Netflix stream can. When packets arrive late or out of order, your audio turns robotic, your video freezes, and the platform eventually disconnects you. Starlink’s jitter during satellite handoffs is the core issue — not missing bandwidth. Every 15 seconds, your dish transitions to a new satellite, and that handoff can cause a brief burst of jitter or lost packets that Zoom and Teams can’t absorb.
Yes — on paper and in practice, when conditions are stable. Zoom requires 3.8 Mbps upload and 3.0 Mbps download for a full HD 1080p group call. Microsoft Teams requires up to 4.0 Mbps upload for best performance. Starlink delivers 10–20 Mbps upload and 50–200 Mbps download under normal conditions, which comfortably exceeds both platforms’ requirements. Starlink’s median U.S. latency of 25.7 milliseconds at peak hours is well within what both platforms need. The issue is consistency, not capacity. During a satellite handoff or a brief obstruction event, your 15 Mbps upload can spike to 200ms latency or drop to near zero for 2–5 seconds — and that’s enough to make Zoom kick you off the call even though your average speed is fine.
Audio is more sensitive to packet loss and jitter than video, which is why it degrades first. When packets are delayed or arrive out of order, Zoom and Teams use error-concealment algorithms to fill in the gaps — which is what produces the robotic, underwater, or stuttering sound. Video can tolerate more loss because the human eye is more forgiving of a frozen frame or lower resolution than the human ear is of scrambled speech. Microsoft’s own Teams documentation states: when bandwidth is insufficient, Teams prioritizes audio quality over video quality — meaning Teams is actively making video worse to preserve audio. If audio is still suffering, the problem is severe enough that video quality reduction alone isn’t compensating. Jitter and packet loss during satellite handoffs are the most common cause.
It will make a substantial difference — and this is the single most impactful change most Starlink users can make for video calls. Benchmark testing found that switching from Wi-Fi to Ethernet reduces jitter by approximately 40% on Starlink connections and eliminates micro-stutters almost entirely. Here is why: Wi-Fi adds its own layer of variability — interference from neighboring networks, signal fluctuation as you move, competing devices all sharing the same wireless channel, and the Wi-Fi radio on your device having good and bad moments. All of that variability stacks on top of whatever variability Starlink’s satellite connection introduces. Ethernet removes that entire second layer. Gen 3 Starlink routers have a built-in Ethernet port. Gen 2 routers require a Starlink Ethernet Adapter (available at starlink.com), which costs around $25.
Starlink’s network is divided into geographic coverage cells. When many users in the same cell are active simultaneously — which peaks between roughly 6:00 and 10:00 PM local time — available capacity is shared and latency spikes. This is the same reason any shared network slows down at peak hours, amplified by satellite’s capacity constraints. Video calls during peak congestion experience higher jitter and more packet loss than during off-peak hours. If your calls are much worse in the evening than during the day, peak-hour congestion is almost certainly a contributing factor. Starlink’s Priority data plans move your traffic to the front of the queue during congestion — worth considering if you have regular evening calls that you can’t reschedule.
Background processes compete for upload bandwidth even when no one is actively using the internet. The most common culprits: cloud backup services (iCloud, Google Photos, OneDrive, Dropbox) that upload large photo or file batches; automatic software updates on computers or gaming consoles; smart home devices pinging their servers; and Microsoft Teams’ own OneDrive file sync, which downloads shared files in the background during calls. Upload bandwidth saturation is particularly damaging because it directly degrades the quality of your outgoing video and audio — the stream you’re sending to other participants. Identifying and pausing background processes before an important call takes two minutes and frequently resolves drops that have nothing to do with Starlink at all.
Yes — and this connection isn’t obvious to most people. An obstruction that causes brief 2–5 second signal interruptions will barely register on a speed test or during casual browsing because those uses can absorb a short gap. But Zoom and Teams cannot. A 2-second packet burst during a satellite handoff through an obstructed portion of sky is enough to trigger Zoom’s “Internet connection unstable” warning and cause audio dropout. If you’re experiencing drops that happen on a recognizable schedule — approximately every few minutes — and you haven’t run the Starlink app’s obstruction check recently, that is the first thing to do. Even a small obstruction covering 2% of the sky can cause 20 or more brief signal events per day, each one harmless to web browsing but disruptive to any live audio or video call.
Yes — both platforms have settings that reduce their sensitivity to connection instability. In Zoom: turning off HD video reduces upload consumption from 3.8 Mbps to under 1.5 Mbps, giving the connection more headroom to absorb momentary drops without triggering a disconnect. In Teams: disabling background blur and video effects reduces CPU and bandwidth load simultaneously; turning off video completely during audio-only portions of a call frees nearly all upload bandwidth for audio. The single most effective in-app setting for both platforms: disabling automatic quality adjustment when it is sending video at higher resolutions than your connection can consistently support, and instead locking to a lower but stable quality. Stable beats sharp when the connection isn’t perfectly consistent.
Starlink’s average performance exceeds the requirements for every video call scenario listed below. The problem that causes drops is not missing bandwidth — it’s brief instability during satellite handoffs layered on top of whatever else is competing for your connection.
| Platform / Call Type | Min. Upload Needed | Min. Download Needed | Latency Target | Starlink Typical |
|---|---|---|---|---|
| Zoom · 1-on-1 HD (720p) | 1.2 Mbps | 1.2 Mbps | Under 150 ms | ✅ Exceeds · 10–20 Mbps up |
| Zoom · Group HD (720p) | 2.6 Mbps | 1.8 Mbps | Under 150 ms | ✅ Exceeds |
| Zoom · Group 1080p HD | 3.8 Mbps | 3.0 Mbps | Under 150 ms | ✅ Exceeds |
| Zoom · Screen sharing | ~1.0 Mbps | ~1.0 Mbps | Under 150 ms | ✅ Exceeds |
| Teams · Audio only | 0.1 Mbps | 0.1 Mbps | Under 150 ms | ✅ Exceeds significantly |
| Teams · HD Video (up to 1080p) | 1.5 Mbps | 1.5 Mbps | Under 150 ms | ✅ Exceeds |
| Teams · Best performance (1080p) | 4.0 Mbps | 4.0 Mbps | Under 100 ms | ✅ Exceeds on average |
| Teams · Screen sharing + video | ~2.0 Mbps | ~2.0 Mbps | Under 150 ms | ✅ Exceeds |
| Starlink typical (for comparison) | 10–20 Mbps up | 50–200 Mbps down | 25–60 ms average | ✅ Well above all requirements |
Zoom’s official minimum for a basic call is 0.6 Mbps. Teams can technically connect on even less. But those thresholds assume a perfectly stable, single-user, zero-competing-traffic network under laboratory conditions. Real-world testing consistently shows that you need at least 10–25 Mbps download and 5 Mbps upload for reliable calls in a normal household with other devices on the same network — far above the official minimums. The point isn’t Starlink’s average speed; it’s whether Starlink’s connection remains stable for the 15–45 seconds of continuous, uninterrupted data flow that audio and video calls require at every moment.
Video call drops on Starlink rarely have a single cause. Most drop events happen when two or three factors overlap at the same moment — a satellite handoff, a background upload, and a marginal Wi-Fi signal all happening at once. Understanding each one separately lets you eliminate them individually.
Every 15 seconds, your Starlink dish hands off from one passing satellite to the next. During that handoff, latency can spike and packets can be lost for 1–5 seconds. Web browsing, streaming, and downloads tolerate this invisibly — they buffer ahead and retry automatically. Zoom and Teams cannot. They require continuous, low-latency data flow with no gaps. When a handoff causes a 2-second burst of packet loss, Zoom’s “Internet connection unstable” warning appears and audio cuts out. If the handoff is bad enough, Teams or Zoom drops the call entirely. This is not something you can prevent — it is a consequence of how LEO satellite networks work. What you can do is remove every other variable (Wi-Fi jitter, background uploads, obstructions) so that when handoffs do occur, the impact is as small as possible.
Most people connect their laptop or desktop to the Starlink router via Wi-Fi. Wi-Fi introduces its own jitter — signal fluctuation from interference by neighboring networks, competing devices, microwave ovens, cordless phones, and the simple fact that radio signals through walls are inconsistent. This jitter stacks directly on top of whatever jitter the satellite connection introduces. The result: a connection that has two independent sources of variability instead of one. Benchmark testing showed that switching from Wi-Fi to a direct Ethernet connection reduced jitter by 40% on Starlink, and that heavily loaded Ethernet outperformed idle Wi-Fi — meaning a busier wired connection was still more stable than an unloaded wireless one. Gen 3 Starlink routers have a built-in Ethernet port. Gen 2 routers require the Starlink Ethernet Adapter.
Upload bandwidth — what you send out — directly determines the quality of your video and audio as other participants receive it. Starlink’s upload speeds are lower than download (10–20 Mbps versus 50–200 Mbps download), and that smaller pipeline saturates faster. The most common background upload consumers: iCloud Photo Library or Google Photos backing up new photos; OneDrive or Dropbox syncing files; Windows Update or macOS automatic updates downloading and uploading data; Teams’ own OneDrive integration syncing meeting files in the background; and gaming consoles uploading game saves or downloading patches. If your video looks fine to you but others say your audio is cutting out or your video is pixelated, upload saturation is likely the cause. On Windows, open Task Manager and watch the Upload column in the Network section. On Mac, use Activity Monitor.
A physical obstruction — tree branch, chimney, roofline, new construction nearby — creates brief signal interruptions every time a satellite passes through that blocked portion of sky. These events can happen 20 or more times per day from a single small obstruction. During regular browsing or streaming, you’d never notice them. During a Zoom call, each one triggers an audio stutter or freeze. The identifying symptom: drops that happen on a schedule — approximately every few minutes at the same interval, or that correlate with specific times of day. Open the Starlink app, go to Visibility or Statistics, and look at the outage log. If entries say “Obstructed,” that is your confirmation. The fix is always physical: raise the dish, relocate it, or trim the specific obstruction if possible.
Starlink’s network divides coverage into geographic cells — when many subscribers in your cell are online simultaneously, available capacity is shared and latency increases. This manifests as higher jitter and more frequent micro-drops during peak usage hours, typically 6:00–10:00 PM local time. If your calls are noticeably worse in the evenings than during working hours, congestion is likely contributing. You can partially address this by upgrading to Starlink’s Priority plan, which moves your traffic to the front of the queue during congestion. The practical alternative: schedule demanding calls for off-peak hours when the cell is less loaded. Morning and early-afternoon calls on Starlink are almost universally more stable than evening calls in high-density subscriber areas.
These are practical changes you can make to your setup — ranked by impact. Start at the top. Each one addresses a different layer of the problem, and several of them can be done today without purchasing anything.
This is not a minor convenience tweak. Switching from Wi-Fi to a direct Ethernet connection is the single most impactful change most Starlink users can make for video call stability. Ethernet eliminates Wi-Fi interference, competing wireless devices, signal fluctuation through walls, and all the other variability Wi-Fi introduces — reducing jitter by approximately 40% in benchmark testing on Starlink specifically. Gen 3 Starlink routers have a built-in Ethernet port on the back of the router — plug in directly with any standard Cat 5e or Cat 6 cable. Gen 2 routers require the Starlink Ethernet Adapter, available at starlink.com for around $25. If you can’t run a permanent Ethernet cable to your desk, a Powerline network adapter (which sends network data through your home’s electrical wiring) is a practical alternative for rooms where running cable isn’t feasible.
Two minutes before any important call, do a quick background upload audit. On Windows: open Task Manager (Ctrl+Shift+Esc), click the Network tab, sort by Upload. On Mac: open Activity Monitor, click Network. Look for anything consuming more than 1–2 Mbps upload. Common offenders to pause: iCloud Photo Library, Google Photos, OneDrive, Dropbox, and any console or PC that’s downloading or uploading updates. Also pause any cloud backup software (Backblaze, Carbonite). On Windows, right-click the OneDrive icon in the system tray and select Pause Syncing. For Teams specifically: Teams syncs meeting files and shared documents via OneDrive during calls — this can consume significant upload bandwidth in enterprise environments. Disabling OneDrive auto-sync during scheduled calls is a reliable fix many IT departments now recommend.
Quality of Service — QoS — is a router feature that tells the network to serve certain types of traffic first when the connection is under load. When properly configured, QoS gives your video call packets priority over a simultaneous Netflix stream or background download, which means even when the connection is busy, call quality is protected. The Starlink router itself has limited QoS capabilities — but if you’re using a third-party router in bypass mode, most modern routers (ASUS, Netgear, TP-Link) include QoS settings that can prioritize real-time communication traffic. In bypass mode: open the Starlink app, go to Settings, and enable Bypass mode. Connect your third-party router directly to the Starlink router’s Ethernet port. Then configure QoS in the third-party router’s settings, prioritizing “video conferencing” or the IP address of your work computer. This is a more technical change but worth doing if you have regular, high-stakes calls.
Open the Starlink app and navigate to Statistics, then the Outages log. If you see entries labeled “Obstructed,” a physical object is clipping the dish’s sky view during satellite passes. The Visibility section shows which portion of the sky is affected — wait at least 24 hours after any change for the map to populate accurately. Target: under 2% obstruction. If you’re above that, the fix is physical: raise the dish, move it to a more open location, or trim the specific obstruction if it’s a tree branch. Seasonal changes matter too — a dish that had clear sky in winter may now be partially blocked by summer foliage. Re-run the check every spring. Even a small reduction in obstruction percentage significantly reduces the number of brief signal events that interrupt calls.
If running Ethernet to your work device is genuinely not feasible, the next best Wi-Fi option is forcing your laptop onto the 5 GHz band exclusively. By default, the Starlink router combines 2.4 GHz and 5 GHz into one network and automatically assigns devices. 2.4 GHz is slower and more prone to interference from neighboring networks and household electronics. 5 GHz is faster with less interference but has shorter range. Open the Starlink app, go to Settings, and enable “Split 2.4/5 GHz Networks” — this creates two separate network names. Connect your laptop or desktop only to the 5 GHz network. Keep less demanding devices (smart speakers, thermostats, older tablets) on the 2.4 GHz network. This won’t eliminate jitter from satellite handoffs, but it removes a common source of Wi-Fi-specific instability that compounds call drops.
Zoom has several settings that reduce how much bandwidth and connection stability it demands. These don’t fix the underlying network — but they lower the threshold at which a momentary satellite handoff will disrupt your call.
Zoom’s HD video setting (Settings → Video → Enable HD) increases upload requirements from around 1.5 Mbps to 3.8 Mbps for group calls. On a stable fiber connection, that’s trivial. On Starlink, where upload bandwidth during a satellite handoff might briefly drop to near zero, that extra demand increases the chance of a call drop. For internal team calls, casual check-ins, or any meeting where your appearance on camera isn’t the focal point, turn HD off. Zoom will automatically lower your outgoing video quality to a level your connection can sustain, which means stable 480p is better than 1080p that freezes every three minutes. Path: Zoom app → Settings → Video → uncheck “Enable HD”.
Screen sharing and live video simultaneously is one of the most bandwidth-intensive things Zoom does — you are uploading both your screen content and your webcam feed at the same time. On Starlink, this doubles upload demand at the moment the connection may be least stable. For screen-sharing presentations, stop your video and share your screen only. Your participants see your screen and hear your audio — which is almost always sufficient for a work presentation. Audio-only mode during a presentation also reduces CPU load, which frees processing headroom on older laptops and can itself reduce call stability issues unrelated to the network. Path: click Stop Video during a screen share session.
Zoom’s background blur and virtual background features use significant CPU to process your video in real time. On laptops where the processor is already managing Zoom’s encoding, network traffic, and other apps, this CPU overhead can cause frame drops that look identical to network problems. Zoom may also reduce bitrate when the CPU is overloaded, making the video quality issue look like a connection problem when it’s actually a processing problem. Disabling background effects — especially on older or lower-powered laptops — reduces CPU load enough to noticeably stabilize calls. Path: Zoom app → Settings → Backgrounds & Filters → select “None”.
Zoom applies audio processing — noise suppression, echo cancellation, automatic gain control — to your microphone by default. On a poor or noisy microphone, this helps. On a quality USB or XLR microphone, this processing can actually degrade audio quality and make voice sound more robotic during packet-loss events by over-compressing the already-damaged signal. If you’re using a dedicated USB or condenser microphone, enable “Original Sound for Musicians” in Zoom’s Audio settings. This disables the processing chain and lets your microphone’s output pass through directly. For people whose audio sounds robotic specifically to other participants, this setting often produces a more natural sound even when the connection itself is slightly unstable.
Teams is designed to prioritize audio over video when bandwidth is constrained — which is the right call, but there are settings that help it make those adjustments more gracefully and prevent full call drops when the satellite connection briefly fluctuates.
Teams defaults to adapting video quality automatically, but it pushes for higher resolution when it thinks bandwidth supports it. On Starlink, where average bandwidth looks plentiful but occasional drops occur, Teams may attempt 1080p just before a satellite handoff degradation — then abruptly downgrade, which causes the stuttering and freeze that looks like a call drop. Manually setting a lower cap prevents Teams from reaching too high. Path: Teams → Settings → Devices → under Camera, disable “HD video.” Teams’ own documentation states it prioritizes audio over video when bandwidth is insufficient — but setting a lower video ceiling means that bandwidth-saving action happens earlier and more gracefully before a drop occurs.
In organizational Teams environments, shared meeting documents, attachments, and recordings automatically sync through OneDrive during calls. If you open a document during a meeting, Teams may begin uploading a local copy to OneDrive while you’re still on the call, consuming upload bandwidth at exactly the wrong time. Before calls: right-click the OneDrive icon in the system tray (Windows) or menu bar (Mac) → Pause Syncing for 2 hours. This prevents OneDrive from competing with your outgoing audio and video for upload bandwidth. This is one of the most commonly overlooked Teams performance fixes in work-from-home environments — Teams’ own background file operations consume upload capacity that users don’t realize is being used.
Teams is not the only Microsoft 365 application that uses network resources. Outlook syncing new email, Word or Excel autosaving to OneDrive, and SharePoint syncing local folders all consume upload bandwidth in the background. None of these are heavy users individually — but on Starlink’s limited upload pipe (10–20 Mbps), they collectively add up. Before an important call: close Outlook, pause OneDrive, and close any Excel or Word documents that are actively syncing. On Windows, you can see exactly what’s uploading in Task Manager under the Performance → Ethernet or Wi-Fi tab — anything showing consistent upload activity while you’re supposedly idle is consuming headroom your call needs.
Many organizations require a VPN connection for Teams or work systems. VPNs route your Teams traffic through your company’s corporate network before it reaches Microsoft’s servers — which can add 50–100ms of additional latency and route your data through a path not optimized for video. On Starlink, which already adds satellite latency, a VPN can push total latency from a manageable 40ms to an unstable 130–200ms. The fix: ask your IT department whether Teams has been configured for “split tunneling” — a VPN mode that routes Teams traffic directly to Microsoft rather than through the corporate VPN. Microsoft explicitly recommends split tunneling for Teams in their network optimization documentation and many IT departments have already implemented it. If yours hasn’t, this one configuration change can eliminate most Teams call problems on any satellite connection.
For professional calls where reliability is non-negotiable, build the setup in layers. First: Ethernet cable from laptop directly to the Starlink router. Second: close every background app and pause all cloud syncing 10 minutes before the call. Third: run the obstruction check in the Starlink app to confirm under 2% sky blockage. Fourth: test your connection immediately before the call using Starlink’s built-in speed test plus a third-party tool like fast.com — look not just at speed but at consistency across three successive tests. Fifth: have a cellular hotspot on standby. A 4G or 5G hotspot connected to your phone costs nothing to keep charged and ready — switching to it takes under 60 seconds if your Starlink call drops at a critical moment. For regular client calls, this is the most practical backstop available.
When audio holds while video freezes, Teams and Zoom are doing exactly what they’re designed to do — protecting audio at the expense of video. This means the connection is borderline: enough stable bandwidth for audio, but not consistently enough for video. The fixes in order: switch to Ethernet if you’re on Wi-Fi; turn off HD video in the app settings; close background upload processes; lower your camera’s resolution (some webcam software lets you cap at 720p instead of 1080p). If video continues to freeze after these steps but audio remains clear, the connection is fundamentally inconsistent — the satellite handoffs are causing periodic drops that aren’t severe enough to cut audio but are enough to interrupt video encoding. Reducing to 720p video is the most direct way to give the connection more headroom.
Robotic or underwater-sounding audio is almost always caused by packet loss — data that left your computer but never arrived at the other end. Zoom and Teams use audio reconstruction algorithms that produce that characteristic robotic sound when they’re filling in missing packets. The most common cause on Starlink: upload bandwidth saturation during a satellite handoff. Check what’s consuming upload bandwidth with Task Manager (Windows) or Activity Monitor (Mac) and pause anything that isn’t the call. Then switch to Ethernet if you’re on Wi-Fi. If the robotic sound happens consistently regardless of what you try, open the Starlink app Statistics page during a call and look for the packet loss graph — even 1–2% consistent packet loss (not just during handoffs) indicates an obstruction or hardware problem that is causing ongoing signal degradation.
This is a clear congestion pattern. Your Starlink cell is getting busier as the day progresses and reaching peak load in the evening hours. For the morning calls: your current setup is probably fine — keep using it. For afternoon and evening calls: the most direct fix is a Starlink Priority data plan, which moves your traffic to the front of the queue during congestion. A more immediate workaround is scheduling your most demanding calls — large group meetings, client presentations, screenshares — before 4 PM when possible, and using afternoons for email, documentation, and lower-bandwidth tasks. If your employer provides a data allowance or expense reimbursement for home internet, the cost difference between standard and Priority plans may be coverable as a work expense — worth asking your employer before paying out of pocket.
This combination — Starlink’s satellite latency plus a VPN adding another routing hop — is one of the most common causes of severe Teams drops for remote workers. The fix must come from your IT department, not from your end. Contact your organization’s IT helpdesk and ask specifically: “Has Teams been configured for VPN split tunneling per Microsoft’s guidance?” Reference Microsoft’s published documentation on network connectivity principles for Teams, which explicitly recommends split tunneling for video call traffic. Most corporate IT teams have not yet implemented this for remote workers, and a direct request with a specific reference to Microsoft’s own guidance usually gets action. Once split tunneling is enabled for Teams, call quality on satellite internet typically improves dramatically — often resolving most or all drop issues without any other changes.
Yes — and having a backup is simply good practice for any critical call on satellite internet. A cellular hotspot is the most practical option: your smartphone’s personal hotspot converts 4G or 5G cellular data into a Wi-Fi connection your laptop can use. In rural areas where Starlink is common, cellular coverage varies — check coverage maps for your carrier at your address before assuming this will work. In areas where you have at least 4G LTE signal, cellular hotspot provides a reliable backup for audio-only or lower-quality video calls. The handoff process: when Starlink drops, connect your laptop to the hotspot network in under 60 seconds, reconnect to Zoom or Teams, and continue. Many Starlink users keep a prepaid cellular SIM specifically for this purpose — even a basic $20–$30/month prepaid plan with a modest data allowance is enough to cover call-only backup use for most remote workers.
This guide is for general informational purposes only. Platform bandwidth requirements, Starlink network performance, and software settings change over time — verify current specifications directly with SpaceX, Zoom, and Microsoft. Bandwidth figures for Zoom and Teams are sourced from their official published documentation and Microsoft Learn. Starlink latency data reflects publicly available SpaceX network reporting. Jitter reduction figures are from third-party benchmark testing. This content is entirely original.