api.video lets you quickly and securely deliver on-demand and live-stream videos directly from your website, software, or app. 

api.video

Glossary-Autoplay.svg

## What is the autoplay attribute in HTML5? 
Autoplay is an attribute commonly used in the HTML5 video tag. It tells the browser that the developers would like the video to play automatically when the webpage loads.

The actual implementation of autoplay is not as straightforward. We'll walk through the options in order of ease of implementation:

## General autoplay settings

* On mobile Chrome and Safari (and in most other browsers), videos will only autoplay when paired with the [muted](https://api.video/what-is/muted/) attribute. Of course, without audio, you might want to consider adding [captions](https://api.video/what-is/captions/) to your video.

* If a user has interacted with video on your website in the past, videos with autoplay set will play automatically with audio.  Each browser has different criteria to make this occur.  This is can hard to test, since once you interact with the page - you'll see the video autoplay.

## api.video autoplay settings

* api.video player: You can set the ['force autoplay'](https://docs.api.video/reference#patch_players-playerid) attribute to true, and all videos viewed with this player will autoplay (they will be muted).  

* Player SDK: Setting the autoplay attribute will result in the video playing as soon ias it is loaded.

```
   player.loadConfig({
      id: "<VIDEO_ID>",
      id: "<VIDEO_ID>",
      autoplay: true,
      hideControls: true,
   });
```

* api.video video: You can set the player to autoplay in the url by adding #autoplay to the end of the player url.

Glossary-Bit-depth.svg

## What is bit depth?

Bit depth describes the level of detail for the color information stored in an image. The higher the bit depth an image has, the more colors it can store. The more colors that can be stored, the smoother and more accurate an image looks. The lower the bit depth is for an image, the more likely you are to see sharp delineations between color changes in the images. Sharp changes that don't look like an image in real life are referred to as banding. 

An image depicting the difference between 24-bit and 8-bit color 

small_bit-depth.png

thumbnail_bit-depth.png

bit-depth.png

In the image presented above you can see how increased bit depth (the image on the left) yields a smoother image, while less bit depth results in what is known as banding - where you can see separate bands of color instead of a smooth blend between the colors. With techniques like [dithering](https://api.video/what-is/dithering/), you can conceal banding with varying degrees of success and avoid going up to a higher and more data expensive quality level. It depends on the image, the technique you use, and how high quality you need the final output to be. 

When discussing bit depth, you'll hear things like this codec or color space is 8-bit, or this is 10-bit, 16-bit or similar. If someone says this is 8-bit color, what they're saying is that there are 2^8 possible values for red, green and blue in the image. That means you can have up to 256 values per color. How many values are possible changes rapidly as you increase bit depth. For example 10-bit appears to be a slight increase from 8-bit until you realize that this is a way of saying there are 2^10 possible values (or 1024 values) per color (red, green, blue). It's a big quality increase, and therefore requires a lot more storage. 

As you can see, high quality color can become expensive quickly. So it's important to determine the level of quality you need. For example, you want the highest bit depth possible when you're capturing video footage because this gives you the most choices later when you process it. You can always drop information to create images with lower bit depth, but you can't add color information that you don't have. 

Glossary-Bitrate.svg

## What is video bitrate? 
The bitrate of a file is measured by the number of bits being transmitted over a period of time. For video it is typically measured in kilobytes per second (kbps) or megabytes per second (mbps).  

The bitrate of a video is most influenced by the size and quality of the video being transmitted. As an example a 4K video (3840 x 2160) has over 8 times as many pixels per frame than a HD video (1280x720) (8.2M pixels vs 921kpixels), and will therefore have a higher bitrate (assuming similar compression settings for both videos). (See table below for the bitrates delivered by api,video.)

Different compression [codecs](https://api.video/what-is/codec/) have a large effect on bitrate. As video compression algorithms improve, videos can be compressed more and more at the same quality, lowering the bitrate of the video.  As an example VP9 files are ~50% smaller than VP8, and AV1 files are [30% smaller](https://engineering.fb.com/2018/04/10/video-engineering/av1-beats-x264-and-libvpx-vp9-in-practical-use-case/#:~:text=Our%20testing%20shows%20AV1%20surpasses,however%2C%20due%20to%20increased%20complexity.) than VP9 For this reason, new codecs are under constant development.

If a video's bitrate exceeds the throughput of the network, the video will stall - it will playback faster than it can be downloaded.

## Adaptive Bitrate Streaming (ABS)

ABS is the ability of the video player to identify the display size, and the network capability of the device - and request a version of the video that will play for the end user.

## Bitrate at api.video

At api.video, all uploaded videos are converted into Adaptive Bitrate Streams. currently, our streams are encoded as follows:

|Resolution    |    FPS    |   Video Bitrate    |    Audio Bitrate    |
|---|---|---|---|
|240p|18|250 KBPS|96 KBPS|
|360p|25|800 KBPS|128 KBPS|
|480p|25|1400 KBPS|128 KBPS|
|720p|25|2600 KBPS|128 KBPS|
|1080p|30|4400 KBPS|128 KBPS|
|2160p (4k)|60|16000 KBPS|128 KBPS|  

&nbsp;  

This ensures that even on the slowest network connections, video can be played.  Typically video players will pick a video bitrate that matches the screen size or 80% of the available network speed.


Glossary-Bit-Rate-Control.svg

## What is Bit Rate Control?

Bit rate control is a crucial aspect of video encoding and compression, allowing content creators and video professionals to manage the quality, file size, and streaming performance of their video content. By carefully controlling the bit rate, users can strike a balance between visual quality and file size, ensuring their videos are optimized for various delivery scenarios and playback devices. 

## Understanding Bit Rate Control

The bit rate of a video file refers to the amount of data (in bits) that is processed or transmitted per unit of time, typically measured in kilobits per second (kbps) or megabits per second (Mbps). Higher bit rates generally result in better visual quality but larger file sizes, while lower bit rates can lead to reduced quality but smaller file sizes.

Bit rate control strategies allow video creators to manage this trade-off effectively. The two primary bit rate control methods are:
1. **Constant Bit Rate (CBR)**: In this approach, the bit rate remains consistent throughout the entire video, maintaining a fixed data rate. CBR is useful for ensuring a predictable and stable video stream, but it may result in quality fluctuations.
2. **Variable Bit Rate (VBR)**: VBR dynamically adjusts the bit rate based on the complexity of the video content, allocating more bits to scenes with more detail and movement, and fewer bits to simpler, less demanding scenes. This can lead to better overall quality and more efficient file sizes.

## Factors Influencing Bit Rate Selection
When choosing the appropriate bit rate for a video, several factors should be considered:
1. **Resolution and Frame Rate**: Higher resolutions and faster frame rates generally require higher bit rates to maintain visual quality.
2. **Content Complexity**: Videos with more movement, detail, and dynamic scenes typically require higher bit rates to avoid quality degradation.
3. **Delivery Platform**: The target platform or distribution channel, such as online streaming, broadcast television, or physical media, may have specific bit rate requirements or limitations.
4. **Bandwidth and Device Capabilities**: The available network bandwidth and the playback capabilities of the target devices can influence the optimal bit rate selection.

## Optimizing Bit Rate for Video Delivery
Striking the right balance between bit rate, quality, and file size is crucial for delivering high-quality video content efficiently. This may involve experimentation, testing, and adjusting the bit rate settings based on the specific requirements of the project, the target audience, and the delivery ecosystem.

Effective bit rate control can significantly enhance the user experience by ensuring smooth playback, minimal buffering, and the preservation of visual fidelity, making it a fundamental consideration in video production and distribution.  

&nbsp;  

At api.video, we understand the importance of bit rate optimization in delivering high-quality video content. Our platform provides robust tools and features to help you manage the bit rate of your videos, ensuring they are optimized for seamless playback across a wide range of devices and platforms. Whether you prefer the consistency of Constant Bit Rate or the adaptive nature of Variable Bit Rate, api.video empowers you to find the right balance between file size and visual quality, delivering an exceptional viewing experience for your audience. [Sign up now](https://dashboard.api.video/register)!

## What is buffering in video playback? 

**Buffering** is the state where video playback stops because there is no video available locally on the device. This state is typically reached when a video has partially downloaded, but playback has reached the end of the downloaded video. With no more video to play, there is nothing for the player to do, other than stop the video.  Often a spinner (or other symbol indicating 'please wait') is displayed. A common synonym for buffering is stalling.

**Buffer**: A video buffer is the amount of video that is downloaded and available on the device for playback. This is video that is ahead of the current playback mark.

In many video players, you can see the buffer as a differently colored line. Usually, you can see how many seconds of the video have been played, but also the amount of video seconds in the buffer. And of course, thow much of the video has not yet been downloaded.

In [HLS](https://api.video/what-is/http-live-streaming-hls/) video streaming, it is common for the player to have 2 segments of video in the buffer before video playback can begin.

The goal of adaptive bitrate streaming is to ensure there is enough video in the buffer by changing the video quality to adapt to the network speed. When the network slows down, the player begins requesting a lower quality video (smaller in size so will download faster). This prevents the video buffer from emptying - which leads to a stall or buffering state of the video. 

Glossary-Aspect-ratio.svg

## What is aspect ratio?
Aspect ratio refers to the ratio of the width and height dimensions of a video. You can calculate the ratio by dividing the width and height by the greatest common divisor of both.

For a video that is 1920 pixels wide and 1080 pixels in height, the greatest common divisor is 120, resulting in an aspect ratio of 16:9.

The two most common aspect ratios for video are 4:3 (e.g. 640x480) and 16:9 (1920x1080 or 1080p). With vertical video, popular for mobile device recording, the common aspect ratio numbers get reversed and 16:9 becomes 9:16.

The square aspect ratio, 1:1, is also becoming popular because of social sites like Instagram that allow square video as an option.

## What is the history of the aspect ratio?
When film was first starting out, everything was shot in 4:3. This was the standard for motion pictures, so it was easy to create everything like this. However eventually, people began experimenting with different aspect ratios, especially wider ones, which could help to create a more immersive experience for film goers. As it turns out, wider aspect ratios do a better job of mimicing how we see the world around us. To be able to stretch films wider and wider, film creators originally used anamorphic lenses. These lenses distorted captured images vertically, allowing them to be stretched out later and look 'right' on wide screens. These lenses are still used today. 

## What are the most popular aspect ratios?
The most commonly used aspect ratios today are: 
- 4:3 - It's what film started out on, but today it's more likely to be used for television and computer screens.
- 16:9 - Most cameras shoot with this aspect ratio, but this is more likely to be used for TV than any other medium. 
- 1.85:1 - This is slightly wider and shorter in height than 16:9. It's used for television, but it's used for TV shows that are going for a more cinematic look-- thus the slight additional width and smaller height. There is a slight letterbox (band of black at top and bottom of video) on video shot in this aspect ratio.
- 2.35:1 and 2.39:1 - These are used for films that are going to appear on the big screen. These aspect ratios always give a more immersive, filmlike quality to whatever's being shot. They're great for the cinema, but because they result in a bigger letterbox, people don't like this aspect ratio as often if they're watching something on TV.

## Why do aspect ratios matter? 
Aspect ratios evoke different kinds of feelings in viewers. Even an untrained viewer will feel different when they see a really wide, cinematic shot vs. a smaller, narrower television type shot. The wider aspect ratio can be used to create more dramatic landscapes while narrower shots can be used to give a more everyday feel. The psychological effects of aspect ratio on a story's theme, narrative and dramatic tension is a popular topic of debate. 

## Aspect ratio and api.video
At api.video, because content is transcoded to [HLS](https://api.video/what-is/http-live-streaming-hls/) so that it can easily be streamed across the web, the aspect ratio by default is 16:9. 

## What is an Artificial Intelligence?

Artificial Intelligence (AI) refers to the simulation of human intelligence in machines programmed to think and learn like humans. In the context of video technology, AI encompasses a wide range of techniques and applications that enable computers to perform tasks that typically require human intelligence, such as visual perception, speech recognition and language translation.

## AI Technologies in Video Processing
AI in video processing relies on several key technologies:
1. Computer Vision: For analyzing and understanding visual content
2. Speech Recognition: To transcribe and understand spoken words in videos
3. [Natural Language Processing](https://api.video/what-is/natural-language-processing/): For understanding and generating human language
4. [Machine Learning](https://api.video/what-is/machine-learning/): To improve performance through experience
5. [Neural Networks](https://api.video/what-is/neural-networks/): For complex pattern recognition and decision-making  

&nbsp;  

These technologies work together to enable sophisticated video analysis, editing, and generation capabilities.

## Applications of AI in Video Technology
Artificial Intelligence has revolutionized various aspects of video technology. In content analysis, AI can automatically categorize, tag, and describe video content, making it easier to search and organize large video libraries. For editing, AI-powered tools can suggest cuts, transitions, and even create rough cuts based on content and style preferences. In visual effects and enhancement, AI algorithms can upscale video resolution, reduce noise, and even generate complex visual effects.

AI also enables personalized content delivery by analyzing viewer preferences and behavior to recommend relevant videos and create tailored viewing experiences. Moreover, advanced AI models, particularly [Generative AI](https://api.video/what-is/generative-ai/) systems, can create entirely new video content from textual descriptions or combine elements from existing videos in novel ways.

## Challenges and Ethical Considerations

While AI offers tremendous potential in video technology, it also presents significant challenges.
- Bias and Fairness: AI systems can perpetuate or amplify biases present in their training data, potentially leading to unfair or discriminatory outcomes.
- Privacy Concerns: The ability of AI to analyze and understand video content raises questions about surveillance and personal privacy.
- Creative Control: As AI becomes more involved in content creation, there are concerns about maintaining human creativity and authorship.
- Transparency: Many AI systems, especially deep learning models, operate as "black boxes," making it difficult to understand their decision-making processes.

## The Future of AI in Video Technology
The future of AI in video technology is poised for significant advancements. We can expect to see more multimodal AI systems that integrate visual, auditory, and textual understanding more seamlessly, enabling more comprehensive video analysis and generation. Advancements in hardware and algorithms will likely allow for more sophisticated AI operations to be performed in real-time, even on mobile devices. We can also anticipate more tools that facilitate creative collaboration between human artists and AI systems, enhancing rather than replacing human creativity. Future AI systems may be able to understand and respond to emotional cues in video content, enabling more nuanced content analysis and creation.

As AI continues to evolve, it will undoubtedly transform the landscape of video technology, offering new possibilities for content creation, distribution, and consumption while also presenting new challenges to be addressed by technologists, policymakers, and society at large.

Glossary-Artificial-Intelligence.svg

Glossary-algorithm.svg

## What is an Algorithm?
An algorithm is a step-by-step procedure or set of rules designed to solve a specific problem or perform a particular task. In the context of video and AI, algorithms are the foundational instructions that guide how software processes, analyzes, and manipulates video content. These mathematical and logical constructs range from simple sorting procedures to complex decision-making processes that power advanced video editing and [generative AI](https://api.video/what-is/generative-ai/) systems.

## Types of Algorithms in Video Processing
Video processing employs a wide variety of algorithms, each serving specific functions. Some common types include:
1. Compression Algorithms: Reduce file sizes while preserving video quality.
2. Encoding Algorithms: Convert video into specific formats for storage or streaming.
3. Filtering Algorithms: Enhance image quality or apply visual effects.
4. Motion Estimation Algorithms: Detect and analyze movement between frames.
5. Object Recognition Algorithms: Identify and track objects or faces within video content.  

&nbsp;  

These algorithms often work in concert, forming the backbone of video processing pipelines in both production and post-production stages.

## How algorithms work in video AI
In the realm of video AI, algorithms are typically implemented within the framework of [deep learning](https://api.video/what-is/deep-learning/) models. These algorithms define how neural networks process input data, adjust their parameters during training, and generate outputs. For instance, a video recommendation system might use collaborative filtering algorithms to analyze user behavior and content features, providing personalized suggestions.

The efficiency and effectiveness of video AI systems largely depend on the underlying algorithms. Optimization algorithms, for example, play a crucial role in training neural networks, helping them converge on optimal solutions for tasks like video classification or generation.

## Importance in Video Production and Distribution
Algorithms are integral to every stage of modern video production and distribution. During filming, real-time algorithms assist with focus, exposure, and color balance. In post-production, editing software relies on sophisticated algorithms for tasks like color grading, noise reduction, and special effects rendering.

In video streaming, adaptive bitrate algorithms dynamically adjust video quality based on the viewer's internet connection, ensuring smooth playback. Content delivery networks use routing algorithms to efficiently distribute video data across servers, minimizing latency and buffering.

## Challenges and Considerations
While powerful, algorithms in video technology face several challenges:
- Computational Complexity: Many advanced video processing algorithms require significant computational resources, especially for real-time applications.
- Balancing Quality and Efficiency: There's often a trade-off between the quality of results and processing speed, particularly in compression and encoding algorithms.
- Ethical Considerations: As algorithms become more sophisticated, questions arise about privacy, bias, and the potential for misuse, especially in AI-driven video analysis and generation.

## The Future of Algorithms in Video Technology

As video technology continues to evolve, we can expect algorithms to become even more sophisticated and efficient. [Machine learning](https://api.video/what-is/machine-learning/) techniques are increasingly being used to develop self-improving algorithms that can adapt to different types of video content and viewing conditions.

The integration of algorithms with other AI technologies promises to unlock new possibilities. For example, the combination of video processing algorithms with natural language processing could lead to more intuitive video editing tools controlled by voice commands or textual descriptions.

Looking ahead, algorithms will likely play a pivotal role in emerging video technologies such as volumetric video, light field displays, and advanced virtual and augmented reality experiences. As these fields progress, the development of novel algorithms will be crucial in addressing new challenges and harnessing the full potential of these technologies. In essence, algorithms will continue to be the unseen force driving innovation in video technology, shaping how we create, distribute, and experience visual content.

## What is adaptive bitrate streaming? 
Adaptive bitrate streaming is a streaming technology based on the Hyper Text Transfer Protocol (HTTP). It detects a user's bandwidth and device capabilities (for example CPU capacity) and then adjusts the stream to match what they have available. In order to work, adaptive bitrate streaming requires an encoder that can take single source media and encode it at multiple bit rates. This allows player clients to switch between the stream qualities based on available resources, resulting in a smooth streaming experience. 

## How does adaptive bitrate streaming work?
Adaptive bitrate streaming starts with encoding video content at multiple bit rates. Then each stream is segmented into multi-second parts. Each part is between 2 and 10 seconds long. The client receiving the video begins by downloading what's called the [manifest](https://api.video/what-is/video-manifest/) file. This file describes the stream segments and their bit rates. When streaming starts, it's common for the client to use the lowest bitrate first, just to test things out. Then it moves to the appropriate bit rate for the network and device. Throughout streaming, the client adjusts the stream quality based on what's available resource wise. Normally, this allows streaming to continue smoothly, even if the quality might not be consistent. 

## Benefits of adaptive bitrate streaming
Adaptive bitrate streaming is based on HTTP, so unlike streaming technologies such as RTMP, it doesn't have issues with traversing firewalls and NAT (Network Address Translation) devices. It also doesn't require constant connections or session state on each client, making it easier to scale. This style of streaming is also compatible with edge servers and the software to handle it is simple. Edge servers use HTTP server software and can handle incoming adaptive bitrate streaming requests with ease.

## The history of adaptive bitrate streaming
Adaptive bitrate streaming was created in October 2002 by the DVD Forum at the WG1 Special Streaming group. The group was run by Toshiba and Phoenix, with collaboration from Microsoft, Apple Computer, DTS Inc., Warner Brothers, 20th Century Fox, Digital Deluxe, Disney, Macromedia and Akamai. Initially, they started with the concept of storing video in small files at two different quality levels that could be sent via HTTP to a media player. Over time this evolved into what we have today. 

## api.video and adaptive bitrate streaming
While there are many implementations of adaptive bitrate streaming, such as Adobe's HTTP Dynamic Streaming and Microsoft Smooth Streaming, none of these are as popular as Apple HTTP Live Streaming or HLS. HLS is supported everywhere and due to its popularity, this is what api.video chooses to transcode your videos to. All api.videos are in beautiful, high quality HLS. 

- For extra information, check this blog article about [bitrate](https://api.video/blog/video-trends/video-bitrate-importance/).
- Start building your own [live streaming experience](https://api.video/product/live-stream/) with api.video!
- Start using api.video now by [signing-up](https://dashboard.api.video/register) for a free sandbox account.

Video on demand

Live streaming

Analytics

Video infrastructure

Video player

AI features

See all features

Online learning & Corporate training

Social & Entertainment

Marketplace & E-commerce

Communication & UGC

Generative AI

Quickstart guides

Documentation

Ecosystem catalog

Clients & SDKs

Community

Blog

Demos

Customer success stories

Help center

Autoplay

What is the autoplay attribute in HTML5?

General autoplay settings

api.video autoplay settings