USB Audio & Video 03: USB Doesn’t Just Capture Video—It Outputs It Too

A Symmetry We Tend to Overlook

The first two articles have been entirely about capture—how images flow from camera into computer.

But audio and video are always two-way. What comes in, must also be able to go out.

Have you noticed: today, you can take a single USB-C cable, plug one end into your laptop and the other into an external monitor, and the picture just appears. Plug another cable into a USB-C projector—picture comes up. Some USB-C docks can drive two 4K monitors simultaneously from a single connection to your laptop.

This was unimaginable a decade ago—back then, connecting a monitor meant hunting for the right VGA, HDMI, or DVI port. Now, the same USB-C port can take a charger on the left, a monitor on the right, and a flash drive when you swap the cable.

How did USB suddenly “learn” to carry display output?

The answer is a little surprising: USB itself never actually learned to carry video. It did something much smarter: it lent its physical cable to someone else.

Output and Capture Are Two Different Worlds

Let’s establish a fundamental distinction—

Capture is “data stream” logic. A camera is an independent digital device. Internally, it senses → encodes → packetizes, then sends a standardized video data stream to the computer over USB protocol. The computer receives it and parses it according to UVC standards. This path runs entirely on USB’s data communication protocols.

Output is something else entirely. What your graphics card outputs is a signal called DisplayPort (DP) or HDMI—a high-speed, low-latency electrical signal specifically designed for video display. It’s electrically a completely different standard from USB data transmission.

So how does a USB-C port “transmit” a DP signal?

The answer: the USB-C connector physically reserves multiple data lanes that can be switched in purpose. When the computer detects a display on the other end, it switches a few of these lanes from carrying USB data to carrying DP signal. This mechanism is called DP Alt Mode (DisplayPort Alternate Mode).

This is the most important thing to grasp about this article.

DP Alt Mode: USB-C’s “Lane Borrowing” Trick

To understand DP Alt Mode, take a look at the USB-C connector’s physical structure.

A USB-C connector has 24 pins. They’re allocated roughly as follows:

• 4 high-speed differential pairs (8 pins total)—for high-speed data transmission, used by USB 3.x protocol
• 2 USB 2.0 data lines—for backward compatibility with USB 2.0 protocol
• 2 CC (Configuration Channel) pins—for handshake, capability negotiation, orientation detection
• 4 power/ground pins—for power delivery (up to 240W with USB PD 3.1)
• Remaining pins for auxiliary signals

Those 4 high-speed differential pairs are the key. Physically, they’re just 4 independent high-speed channels—they happen to default to running USB 3.x protocol. But since they’re high-speed differential pairs, why couldn’t they carry other high-speed signals?

They can. That’s the core idea of Alt Mode.

When the computer detects, via the CC pins, that the other end is a DP display, both sides negotiate via USB PD (Power Delivery, which also handles communication) to enter DP Alt Mode. Once in this mode:

• 2 or 4 of the 4 high-speed differential pairs are switched to carry DisplayPort signal—no longer running USB 3.x, but directly running DP’s electrical standard
• USB 2.0 data lines remain available, still handling keyboards, mice, USB drives, etc.
• Power pins continue to work, capable of reverse-charging the laptop (PD charging)

So in DP Alt Mode, a single USB-C cable simultaneously handles three things:

1. 4K video signal transmission (on the borrowed differential pairs running DP)
2. USB data transmission (on the remaining USB 2.0 channels)
3. Bidirectional power delivery (charger → laptop, or laptop → display)

This is why today’s USB-C docks can deliver “one cable for everything”—they’re essentially the combination of DP Alt Mode + USB Hub + USB PD stacked together.

Key insight: DP Alt Mode doesn’t “encapsulate” DP signal inside USB protocol. What it does is make the USB-C physical cable switch roles, ceding some lanes to DP.

The elegance of this design: the monitor side receives a standard DP signal, indistinguishable from a direct DP connection. No transcoding, no extra latency.

Why DP Instead of HDMI?

You might ask: HDMI is the most ubiquitous video interface in consumer markets. Why did USB-C choose to borrow lanes for DP instead?

Historically, HDMI Alt Mode did exist—released in 2016 by the HDMI Forum together with USB-IF. Same idea as DP Alt Mode, but with HDMI as the borrowed protocol.

But it was adopted by almost nobody. Every “USB-C to HDMI” cable and adapter you can buy today actually works on DP Alt Mode + a DP-to-HDMI protocol conversion—DP signal runs through the USB-C cable, then gets converted to HDMI at the adapter or monitor end.

Why did this happen? A few reasons:

First, DP is an open standard; HDMI is a licensed standard. DP is managed by VESA and royalty-free. HDMI is managed by HDMI Licensing, and every HDMI device pays licensing fees. Phone, tablet, and laptop manufacturers naturally chose the lower-cost option.

Second, DP’s technical structure is more amenable to “being borrowed”. DP uses packet-based transmission, which aligns with modern high-speed serial protocol design philosophy (including USB 3.x, PCIe). It’s engineeringly easier to multiplex on the same high-speed differential pairs. HDMI uses TMDS clock+data parallel transmission, which is harder to multiplex.

Third, DP scales better in bandwidth. DP 1.4 delivers 8.1 Gbps per lane, 32.4 Gbps across 4 lanes. DP 2.1 hits 20 Gbps per lane, 80 Gbps across 4 lanes—enough to drive 8K 60Hz displays. HDMI has consistently trailed DP by half a generation to a full generation in bandwidth.

So when we say USB-C outputs video today, we essentially mean USB-C borrowing lanes for DP. Apple, Dell, Lenovo, Huawei, Xiaomi—every mainstream USB-C video output device runs on DP Alt Mode under the hood.

DisplayLink: A Different Path, A Different Trade-Off

But DP Alt Mode isn’t the only solution for USB video output. There’s another interesting approach called DisplayLink.

DisplayLink’s philosophy is the exact opposite of DP Alt Mode—

DP Alt Mode is “borrow lanes”—it cedes the USB cable’s physical channels to DP.

DisplayLink is “shove it through”—it compresses video on the computer in real time, transmits it through standard USB data channels, then decompresses it back to video signal on the other end.

This path runs entirely on standard USB data transmission. It doesn’t need Alt Mode, doesn’t need special physical lane switching. Which means—

Any USB port can drive a display via DisplayLink. Even non-Alt Mode USB-A ports, even older USB 3.0 ports, even computers that have no DP output capability at all—DisplayLink works.

Its strength is extreme universality—an old laptop with no USB-C and no DP output can still drive an external monitor via a DisplayLink dock.

But its costs are equally clear:

First, it requires a driver. DisplayLink isn’t part of the USB standard. OSes don’t natively support it. The DisplayLink official driver must be installed on the computer to register a virtual graphics adapter.

Second, CPU usage. The compress-transmit-decompress chain consumes CPU resources. In multi-display 4K scenarios, DisplayLink CPU usage can hit 5~15%.

Third, latency and quality. DisplayLink’s compression introduces latency (typically 30~50ms), and fast-motion content (gaming, video) shows compression artifacts. It excels at “static office display” scenarios but is unsuitable for video editing, gaming, or professional graphics work where quality and latency matter.

So DisplayLink’s positioning in reality is clear—it’s a compatibility solution for devices that don’t support USB video output natively. On mainstream USB-C + DP Alt Mode devices, DisplayLink has essentially no use case.

But conversely—in enterprise IT environments, DisplayLink still has its place. A company with a fleet of laptops varying in age and configuration can standardize on DisplayLink docks for external display, avoiding the “this laptop supports USB-C DP output, that one doesn’t” compatibility headache.

Audio Output: UAC, Clean and Simple

Video output covered. Audio output is delightfully simple—it uses the same standard as audio capture: UAC.

USB audio devices are fundamentally bidirectional. USB microphones push audio into the computer (capture). USB headphones play audio out (output). USB audio interfaces do both. At the protocol level, these are all UAC devices, the computer talks to them through standard UAC interfaces, and input streams and output streams share the same protocol family.

This is why audio output has never been the “complicated” thing video output has been—

• No “lane borrowing” needed, because UAC runs comfortably on standard USB data channels (audio bandwidth is tiny, as covered last time)
• No “compression shoving” needed, because there’s no bandwidth pressure
• No “virtual driver” needed, because OSes support UAC natively

Plug in a USB headset, and a new audio output device appears in system settings instantly—click and use. This is the silky result of UAC’s twenty-plus years of standardization polish.

A Fascinating Asymmetry

Looking back at this article, an interesting pattern emerges—

USB video input and output take two completely different paths.

On the input side, UVC is genuinely “USB-protocol video”—video data is packetized according to USB standard data packet format, the OS parses it according to UVC standard. Here, the USB cable is both the physical channel and the protocol layer.

On the output side, the USB-C cable is just being borrowed for DP. USB protocol itself has no “video output” capability—it gives up physical channels through Alt Mode and lets DP run on them.

The fundamental reason for this asymmetry: post-compression video input data is reasonable (30~50 Mbps), fitting into USB protocol just fine. But video output raw data is too large, and the latency/stability requirements are too strict (4K 60Hz raw is about 15 Gbps)—forcing USB protocol to carry it would be uneconomical. So the output side simply chose to “lend the lanes”—letting DP use the underlying physical cable directly, bypassing USB protocol overhead.

This is a very engineering kind of thinking—do your own job well, but let the specialist do the specialist’s work. USB doesn’t insist on “solving everything itself.” It knows when to step aside and let a more capable protocol take the stage.

This restraint is, in fact, the deep reason USB has lasted thirty years, iterated to USB4, and continues to evolve.

Next time we zoom out to a more macro layer—

USB audio and video is, in the end, just a tiny slice of USB’s vast capability. From day one, this cable was given an “connect everything” ambition: flash drives, keyboards, mice, printers, external hard drives, network adapters, sound cards, gamepads, graphics cards (yes, USB4 supports external GPUs), 5G modules, lidar, medical instruments, industrial controllers… nearly every digital peripheral can connect through USB.

But this achievement has a critical hidden premise—one we’ve been relying on throughout the past three articles, but never named directly.

This premise is what determines why UVC and UAC ended up the way they did. And it’s also what makes “let the professional camera in the meeting room get into the laptop” a far harder problem than it seems.

See you next time.