A bitrate ladder is the structured set of encoding tiers that sits behind every adaptive stream. When a viewer presses play, the player doesn't pick a fixed file. It picks from a ranked list of versions, each encoded at a different resolution and bitrate, and switches between them as network conditions change. The ladder is what makes that possible. Build it carelessly and you waste bandwidth, degrade picture quality, or both.
What a bitrate ladder actually contains
Each rung of the ladder pairs a resolution with a target bitrate. A simple ladder might look like this: 360p at 400 kbps, 540p at 800 kbps, 720p at 1,500 kbps, 1080p at 4,000 kbps, and 1080p HDR at 6,000 kbps. The player's ABR (adaptive bitrate) logic steps up and down between rungs based on available throughput.
Two things make this more complicated than it sounds. First, the right bitrate for a given resolution depends heavily on the codec. H.264 needs roughly twice the bitrate of HEVC (H.265) to produce equivalent quality at 1080p. AV1 goes further, delivering comparable quality to HEVC at around 30 percent lower bitrate in many test conditions. Second, content complexity matters enormously. A locked-off talking head compresses cleanly at 1,200 kbps. Fast-moving sport or a particle-heavy visual effects sequence can look soft and blocky at three times that rate.
This is why borrowing a generic ladder from a platform's documentation is only a starting point. The right ladder for a studio delivering cinematic brand content is different from one serving a corporate training library. Both are real production decisions. Variable bitrate encoding adds another layer to this, allowing the encoder to allocate bits dynamically within a rung rather than spending them evenly across every frame.
How content type changes the calculation
Not all content compresses equally. The encoding community uses the term "complexity" to describe how much information changes from frame to frame. High complexity content includes anything with fast motion, grain, natural textures, or fine detail. Low complexity content includes animation, screencasts, and slow-paced interview footage.
A studio producing a product campaign with high-contrast motion sequences should test each rung against a representative clip, not just the opening five seconds. The worst-looking moment in the asset, not the average, determines whether a bitrate tier is acceptable. Set the threshold too low and a single complex shot will expose compression artefacts the rest of the video avoids.
Resolution and bitrate don't scale linearly. Jumping from 720p to 1080p increases pixel count by a factor of 2.25. The bitrate required to maintain equivalent quality doesn't simply double. It depends on the encoder, the content, and the codec profile. Testing with VMAF (Video Multimethod Assessment Fusion), a perceptual quality metric developed by Netflix, gives a more reliable signal than PSNR alone for comparing rungs.
Device constraints that shape your rung choices
A bitrate ladder built only for desktop browsers will underserve mobile viewers and overserve connected TVs. Three device classes require specific attention.
- Mobile on cellular: 4G LTE delivers 10–50 Mbps in ideal conditions, but real-world throughput on crowded networks drops considerably. A mobile rung at 720p and 1,500 kbps is a reasonable ceiling for most cellular contexts. Anything above that risks stalling on a busy network.
- Connected TV on Wi-Fi: This is where your highest rungs matter. A 65-inch OLED at close viewing distance reveals compression artefacts that a 15-inch laptop screen conceals. 4K at 12,000–16,000 kbps (HEVC) is a realistic target for premium connected TV delivery.
- Desktop browser: The middle rungs do the heaviest lifting here. Most desktop viewers are on broadband, but browser-based players often cap codec support. H.264 remains the safest cross-browser choice; HEVC and AV1 require feature detection and fallback logic.
The gap problem: rungs too far apart
One of the most common ladder mistakes is spacing rungs too widely. If the gap between your 720p rung and your 1080p rung is 2,500 kbps, the ABR algorithm faces a binary choice on a mid-tier connection. It either holds at 720p when the viewer's bandwidth could support something better, or it jumps to 1080p and risks buffering.
Adding intermediate rungs at finer intervals gives the ABR logic more room to move. A 900p rung sitting between 720p and 1080p isn't standard, but it's technically valid and often produces better perceived quality on mid-range connections than forcing a hard jump. The trade-off is storage and transcoding cost. Each additional rung adds to the packaging bill. Studios need to weigh this against the proportion of viewers sitting in that mid-tier bandwidth range.
This is closely related to how adaptive bitrate streaming handles transitions. A well-spaced ladder means transitions between rungs are subtle. A poorly spaced one produces visible shifts in sharpness that attentive viewers notice.
Testing and validating your ladder
Building the ladder is one step. Validating it under real viewing conditions is another. Three checks matter before a ladder goes into production.
Run VMAF scores across a representative sample of your content at each rung. Target a minimum VMAF of 93 for premium content and 85 for utility content (training videos, screencasts). Below 80 is where viewers start to consciously notice quality loss. These thresholds aren't universal rules, but they give a usable benchmark.
Simulate network throttling. Most major browsers support network condition emulation in developer tools, and dedicated tools like Apple's Network Link Conditioner allow desktop-level simulation. Test your top-to-bottom ladder under 3G, mid-tier 4G, and congested Wi-Fi conditions, watching specifically for how the player transitions between rungs.
Check the lowest rung on the smallest device you expect in your audience. A 360p floor rung looks acceptable on a 5-inch phone screen. It looks unusable on a 27-inch monitor. If your analytics show a meaningful share of viewers on large desktop displays, your lowest rung should probably sit at 480p or higher.
Codec decisions that change the whole ladder
Adopting a newer codec can let you collapse the top rungs without sacrificing quality. A studio currently delivering 1080p at 4,000 kbps in H.264 might achieve equivalent perceptual quality at 2,200 kbps in HEVC. That's a 45 percent bandwidth saving on the most-used rung, with direct consequences for CDN costs and viewer buffering rates on congested networks.
AV1 is increasingly viable for on-demand delivery. Chrome, Firefox, and Samsung TVs have supported it for several years. The encoding time cost is still significant: AV1 encodes 5–10 times slower than H.264 at equivalent quality settings, making it impractical for live or near-live workflows but reasonable for pre-packaged VOD content where encoding happens once.
HEVC remains the practical choice for connected TV delivery, particularly on Apple devices. Safari's support for H.264 and HEVC, without AV1 support until very recently on older device versions, has historically meant studios maintaining parallel H.264 and HEVC ladders for broad compatibility. That's a real infrastructure cost that needs to sit in the packaging budget.
Keeping the ladder current
Audience device mix shifts over time. A ladder optimised for your viewer base in 2023 may be mis-weighted today. If your analytics now show 40 percent of plays on connected TV where two years ago it was 15 percent, the top rungs deserve more investment. If mobile share has grown, the gap between your cellular-safe rung and your Wi-Fi rung matters more than it used to.
Reviewing your ladder alongside your video compression settings at least once a year keeps delivery aligned with how your audience actually watches.