Meta Purposely Reduces The Quality Of Your Videos. Here’s Why.
Date: October 28, 2024
Instagram Head Adam Mosseri explained during an AMA session why Instagram shows high-quality and low-quality videos.
Instagram is a place of mixed feelings, even for the best content creators. While every creator tries to upload the best quality, not only in content but also in resolution, colors, and pixels, the final output often remains low in quality. The same situation applies to viewers who experience a low-quality pixelated video right after swiping on a high-quality video while scrolling through Instagram feeds.
The actual reason behind fluctuating video quality on your Instagram feed is the backend algorithm that controls the quality of every video uploaded on the platform. Adam Mosseri, Head of Instagram, revealed the technical reason behind this behavior during an Instagram AMA session from his personal profile.
Mosseri responded to a query left on his AMA session by a user, “In general, we want to show the highest-quality video we can ... But if something isn’t watched for a long time — because the vast majority of views are in the beginning — we will move to a lower quality video. And then if it’s watched again a lot, then we’ll re-render the higher quality video.”
The extent of video quality depends on the performance of each video in the initial few minutes. Since Instagram hosts only short-to-medium form content, the virality factor can be determined in the early period of the upload itself. Based on the response received by the video, adequate video quality is decided, which can change later if the video resurges in views and engagement.
“It works at an aggregate level, not an individual viewer level. We bias to higher quality (more CPU-intensive encoding and more expensive storage for bigger files) for creators who drive more views. It’s not a binary threshold but rather a sliding scale,” said Mosseri.
Creators who drive more views get to showcase their content in the highest quality possible. This judgment factor is critical for optimizing resources to keep up with the increasing number of videos uploaded on the platform every day. Last year, Meta estimated 4 billion video streams per day on Facebook. Instagram would surpass the count any day.
Instagram provides the most basic encoding to the freshest uploads. Once they achieve sufficiently high watch time, the video gets a more robust encoding pass and higher resolutions through the slowest and most computationally costly processing of the video. For the remaining videos, the pixels may be reduced along with resolution and color banding to prioritize bandwidth.
By Arpit Dubey
Arpit is a dreamer, wanderer, and tech nerd who loves to jot down tech musings and updates. With a knack for crafting compelling narratives, Arpit has a sharp specialization in everything: from Predictive Analytics to Game Development, along with artificial intelligence (AI), Cloud Computing, IoT, and let’s not forget SaaS, healthcare, and more. Arpit crafts content that’s as strategic as it is compelling. With a Logician's mind, he is always chasing sunrises and tech advancements while secretly preparing for the robot uprising.
// Recommended
Pinterest Follows Amazon in Layoffs Trend, Shares Fall by 9%
AI-driven restructuring fuels Pinterest layoffs, mirroring Amazon’s strategy, as investors react sharply and question short-term growth and advertising momentum.
Clawdbot Rebrands to "Moltbot" After Anthropic Trademark Pressure: The Viral AI Agent That’s Selling Mac Minis
Clawdbot is now Moltbot. The open-source AI agent was renamed after Anthropic cited trademark concerns regarding its similarity to their Claude models.
Amazon Bungles 'Project Dawn' Layoff Launch With Premature Internal Email Leak
"Project Dawn" leaks trigger widespread panic as an accidental email leaves thousands of Amazon employees bracing for a corporate cull.
OpenAI Launches Prism, an AI-Native Workspace to Shake Up Scientific Research
Prism transforms the scientific workflow by automating LaTeX, citing literature, and turning raw research into publication-ready papers with GPT-5.2 precision.
Have newsworthy information in tech we can share with our community?
