How to choose the right bitrate for HLS video

A useful bitrate preserves the detail viewers need without spending delivery on data the source, screen or connection cannot turn into a better experience.

Published

A dark optical bench holding two horizontal glass conduits, one above the other, each threaded through a machined black collar. Inside the upper tube a strip of small illuminated film frames shows the same still scene again and again — snow-capped mountains, a pine treeline and a mirror-flat lake at dusk — passing through a narrow knurled collar. Inside the lower tube the frames are larger and every one is different: light streaks, swirling particles, an exploding burst, tumbling rock debris. They pass through a far wider collar banded with unnumbered graduation marks. Both strips are equally sharp; the busier clip's tube, frames and collar are all larger.
Same machine, two clips. A near-static shot repeats almost unchanged frame to frame and passes through the smaller path; a clip where every frame is new uses the wider path. What you are encoding, not just how big it is, affects how much capacity it asks for.Illustration · DynTube illustration

Bitrate is a quality and delivery decision

Video bitrate describes how much encoded video data is used over time, commonly expressed in kilobits per second. Raising it can preserve more visual detail, but it also increases the data delivered to each viewer. The right target is therefore specific to the source, rendition and audience.

In an HLS ladder, the player chooses among several renditions produced for the video. Each rung combines dimensions and bitrate. A lower rung provides a practical stream for constrained conditions; a higher rung preserves detail for larger displays and stronger connections.

Start with the source you actually have

Resolution alone does not determine a good encode. Frame rate, camera noise, texture, motion, animation, screen text and previous compression all change how much data is useful. Encoding a soft 720p recording at a very high bitrate does not restore detail that was never captured.

  • Talking-head lessons usually change slowly and can preserve quality efficiently.
  • Sports, gaming and handheld footage contain rapid motion and frequent scene changes.
  • Screen recordings need sharp text and interface edges even when motion is limited.
  • Animation and slides may have large flat areas but reveal ringing around text and lines.
  • Dark or noisy footage is harder to compress cleanly than a well-lit source.

Build a ladder instead of choosing one universal bitrate

Use several appropriately spaced renditions so the player has meaningful choices. Each rung should provide a visible quality step and a delivery step; tightly clustered rungs add processing and storage while giving the player little room to adapt.

DynTube derives the produced output from the source dimensions, project settings, format and processing policy. Treat any generic resolution table as a starting hypothesis, then inspect the actual output created for representative videos.

Evaluate motion, text and frame rate separately

For motion-heavy footage, inspect fast pans, particles, crowds and cuts. For instruction and product demos, pause on small labels and fine lines. Higher frame rates encode more frames each second, so compare motion clarity at the actual frame rate rather than borrowing a target from different content.

Connect bitrate to monthly bandwidth

At 5,000 kbps, one hour of continuous delivery is about 2.25 decimal GB per viewer before modelling how adaptive playback changes the average. Halving the delivered bitrate approximately halves that data; doubling viewers doubles it. Use watch time and a realistic average delivered bitrate in the bandwidth calculator.

Test the player, not only the encoded file

A visually strong encode can still be a poor delivery choice if the target audience cannot sustain it. Test initial playback, seeking and quality changes on the devices and networks your viewers use. Then review first-frame time, rebuffering and rendition behaviour from real sessions.

A repeatable bitrate review

  • Select representative low-motion, high-motion and text-heavy sources.
  • Upload them through the project configuration planned for production.
  • Inspect each produced rendition at normal viewing size and full screen.
  • Test on strong and constrained connections using the final embedded page.
  • Calculate normal and peak delivery using measured watch time.
  • Adjust project targets, process a fresh sample and compare again.

Use evidence from your own library

The best ladder is the one that keeps important detail while giving the player usable choices for your audience. Keep source quality, visual inspection, playback-quality reporting and monthly delivery in the same decision instead of optimising one number in isolation.

Include audio in the delivered bitrate

Video is the largest part of most streams, but audio is delivered too. Speech, music and multi-channel material have different quality needs. Include the audio bitrate when estimating total delivery, and listen on the devices your audience uses rather than evaluating only the picture.

Keep project settings aligned with each content family

A training library, a sports archive and a screen-recorded help centre may deserve different encoding profiles. Group videos with similar visual demands into projects, preserve a representative test title for each profile and document why its ladder was chosen. When source quality or audience behaviour changes, process a fresh sample and compare it with that baseline before changing the full workflow.

Put your video on your own terms

Host, protect, deliver and measure your video under your own brand, on the sites and apps you already run.

Start free trialView pricing