MARCH 2022

In this paper, our Business Development Director, Jonathan Regalado-Hawkey discusses the importance of using DSC compression.   

Estimated Reading Time: 8 minutes

It's fair to say that the last few years have been a challenge and the AV distribution residential market has experienced some big changes. However, despite economic difficulties and the manner in how end-users consume content, it's still managed to continue to grow. 

Despite AV over IP technology taking the industry by storm, for wired connectivity, HDBaseT™ has remained the most popular technology of choice for many professional Residential AV installers due to its end-to-end reliable performance and simplicity.

HDBaseT™ supports extension of full digital audio, HDMI® uncompressed video, 100Mb Ethernet, power, RS-232 and IR control over a single CAT5e cable, up to 100m. However, there has been a steady shift from extending HD resolutions to 4K (UHD) video without any noticeable effect on new HDBaseT™ installations. That is until HDR content is considered.  

"​DSC has become the de-facto compression standard for ultra-low latency compression"

What is DSC? 
Display Stream Compression (DSC) is a well recognised compression mechanism developed by VESA, an international trade association driving video standard initiative. Unlike other well-known and older compression standards in the industry such as H264, H265 and JPEG2000 to name a few, DSC technology utilises very light and efficient compression ratio (2:1 or 3:1) to provide a true visually lossless performance.

DSC has become the de-facto compression standard for ultra-low latency compression and has been adopted by major display interfaces including DisplayPort, HDMI®, MIPI and MHL.

Some other benefits of DSC include that it removes the need for lossy techniques such as Colour Space Conversion (4:4:4 to 4:2:2/4:2:0) and reducing colour depth (bit dithering).  

What is CSC?
Colour Space Conversion (CSC)—or more precisely, chroma sampling—is a video technique that reduces bandwidth required by a video signal by maintaining the luminance data but lowers the colour information. This technique was developed in the 1950's, along with colour encoding to support the shift from black-and-white television broadcast to the introduction on colour. 

The proven theory behind this technique is that the human eye is much more sensitive to changes in the shapes and contracts (luminance) than colour. However, when the technique is utilised in a digital transmission, it effectively reduces the pixel chroma bits (up 50% in 4:4:4—>4:2:0) which is information that can't be recovered at the RX side.  

Comparing DSC VS CSC
Display Stream Compression (DSC) standard was released in 2015 and quickly became an industry leading CODEC, recognised for delivering "visually lossless" mezzanine compression. Whereas CSC is an effective but simpler video technique that lowers bandwidth by reducing the amount of chroma information, which is unrecoverable. Below are some comparisons on video quality of the same source when comparing the two technologies.  

Colour Depth and High Dynamic Range
Colour depth is the number of bits per colour in a pixel. The standard video resolution distributed over HDMI® and DisplayPort has been default set at 8-bits per colour (24-bit per pixel); which translates as 8 bits of luminance (L), 8 bits of chroma red (Cr) and 8 bits of chroma blue (Cb).   There has been an increase in higher colour depths (10/12/14/16-bits per colour) otherwise known as "DeepColour" over the last few years, supported by HDMI® since their specification version 1.3 was released in 2006. 

The importance of higher colour depths is significant as it provides millions of more colours along with better gradients and significantly improved image quality, e.g. reducing those dreaded "banding" effects in a sunset picture.  

High Dynamic Range
Over the last few years, the latest production and recording techniques have offered further visual enhancements such as wider contrast, colour and brightness. These are generally classed as High Dynamic Range (HDR).   Currently, the most widely adopted HDR standard on the market is HDR10/HDR12. This requires a colour depth of at least 10-bits and 12-bits per colour respectively. There are other leading HDR standards that have rapidly been adopted across the industry such as HLG, HDR10+ or Dolby Vision, to name a few.   It's important to note that HDR content has been readily available in the past few years with new content coming from Sky, Netflix, Prime Video and others.  

Colour Bit Dithering
Colour Bit Dithering is a technique employed to reduce the colour depth of an image to decrease the link bandwidth required. The technique "throws away" the least significant colour bits (up to 33% bit saving when going from 12-bit to 8-bit) which is information that can't be recovered at the RX.   Bit Dithering greatly affects image quality when reduced below 10-bits per colour, much more than colour space conversion (CSC), because it affects the luminance of the picture. More importantly, HDR relies on 10-bit per colour and above so any bit dithering will affect the HDR content itself.   From an installation point of view, it's important to ensure that any video extension equipment being used is not utilising bit dithering techniques as this will ensure that the 4K HDR video content isn't displayed on the screen as intended.  

Testing visualization was conducted with the comparison tool “Beyond Compare”. The “image compare” view shows images side-by-side with their differences highlighted.

HDBaseT™ 2.0 + DSC
HDMI® 2.0 video resolutions (e.g. 4K@60/4:4:4/8-bit, 4K@60/4:2:2/10-bit) require ~ 18Gbps data rate, which means current HDBaset™ 1.0/2.0 solutions are only able to support up to ~9Gbps and do not have sufficient bandwidth.    

There are HDBaseT™ extension products in the market claiming "proprietary visually-lossless compression mechanisms" which rely on CSC and bit dithering to lower the video data required over HDBaseT™ link- e.g. 4K60/4:2:2/10bit (18Gbps)--> 4K60/4:2:0/8bit (8Gbps).

However, as we've seen, these products remove crucial luminance and colour information and provide poor quality compression that is damaging the image in a 4K HDR extension.   The only HDBaseT™ compliant mechanism to achieve HDMI® 2.0 resolutions (18Gbps) over HDBaseT 2.0 link (8Gbps) is by using DSC technology- described as "HDBaseT™ 4K60 Type A".    

By using equipment capable of supporting HDBaseT™ 4K60 Type A (DSC), it ensures backwards compatibility when connected to a standard with HDBaseT™ 2.0/1.0 (non-DSC).  

The Future of 8K  
And what about 8K? It's fair to say that there has been some marketing noise around the future of 8K with the release of HDMI® 2.1. However, it's unlikely that 8K resolution will become prevalent in consumer homes any time soon, or at least not until 8K content becomes readily available. It will come, but we are not there yet.   There are also some limitations to consider such as the much higher bandwidth required for supporting 8K resolutions and that even the latest advances in connectivity and technologies such as HDBaseT3 are still not able to cope with these new increases in bandwidth. Mezzanine compression techniques such as DSC, are likely to continue playing a key part when bridging this 'bandwidth gap' without affecting the picture quality.   With the advancing times, comes advancing technology however, for the time being, 4K is here to stay and 8K is the future. 

Jonathan Regalado-Hawkey


Jonathan has over 14 years experience in the semiconductor and ProAV industry. He has participated in all aspects of technical system architecture, customer support and sales. Over the years, he has developed long-term business relationships and is well regarded amongst industry luminaries.

With thanks to Silora R&D (www.silrd.com)