In a recent tendering process for an AV system on a superyacht, one of the bidding parties mentioned to me that they were planning to include a 1Gb/s based HDMI over IP video distribution system in their proposal. It met the specification requirements ([email protected], HDR, etc), so why not? Given the fact that the client wants the highest quality video possible, I recommended they looked into a solution that was uncompressed, which they did. Interestingly, the manufacturer of the compressed system looked me up to ask why I had advised the integrator not to offer their system. It’s a valid question, so I thought I’d state my case here for all to see.
Let me be clear about one thing: I have nothing against compressed solutions. They make total sense in certain situations and the performance they achieve compared to even a few years ago is striking. The compressed solution provided by the manufacturer that looked me up is a very nice one indeed, but I still feel that in this situation, and most others when dealing with a new build, uncompressed is the way to go. Also, I make a point of trying to ensure my client’s requirements are met, and when a client is looking for the best possible video quality, that’s what I will try to ensure they get.
“But our solution is visually lossless! Here look…can YOU see any difference with the uncompressed feed?” I’ll admit, these technologies are very good, and in many cases the viewer will never realize they are watching a compressed feed. This is what the manufacturers of the compressed systems are banking on of course. Despite this, there is no denying that cramming a 4K HDMI signal in 1Gb/s of bandwidth requires a serious amount of compression. Lots of information is being thrown away in the name of getting the signal down to a much lower bandwidth. No mater what marketing terms you throw at it, this compression is far from lossless.
Video compression relies on very clever algorithms, but you can create content that absolutely throws these algorithms for a loop. Granted, that is a rather contrived situation, and those types of content are not typical fodder for an entertainment system, but all it takes is one scene in a movie that happens to trip up the codec, and you may have a very difficult conversation on your hands. It may be somebody’s stripy shirt, or a closeup of a fine grained architectural pattern in a panning shot, or a multitude of other possibilities that prove very tricky for video codecs to handle. But if the codec does trip up, the effect is glaringly obvious. Given that these codecs are typically designed for video (motion picture) compression, non-video content, such as spreadsheets or other PC applications can prove challenging as well. If your client spots these flaws, they may be less than enthused, and in a situation like that, answering “…yes but this system is state of the art and uses visually lossless compression” is not exactly going to bail you out.
Perhaps the simplest argument is this: In a shootout between a compressed and an uncompressed system, the performance may subjectively be equal in many cases, but the uncompressed solution will perform measurably and therefore objectively better in all cases. There is simply no arguing the fact that for pure video quality and performance, uncompressed always wins. That alone is enough reason for me to recommend the uncompressed solution for most situations, especially when you are starting from scratch.
How about cost? Interestingly enough, when comparing the high-end compressed solutions with the uncompressed ones, the total cost is quite comparable. That’s all I’ll say a bout that.
Now, in defense of compressed solutions, and to make this a more balanced article, they are definitely the way to go in situations where you are dealing with legacy (1Gb/s) infrastructure. There may also be certain features available in the compressed solutions that are unavailable on uncompressed ones. It is also fair to mention that the current generation of uncompressed solutions relies on 10Gb/s infrastructure, and the highest HDMI bitrates are significantly higher than that. In other words, for the highest bitrates, even the ‘uncompressed’ solutions turn to compression to get the signal across, but the amount of compression applied is significantly lower, and up to a certain bandwidth it’s even objectively lossless.
In conclusion, I firmly believe that in new build scenarios with high-end systems and demanding clients, uncompressed is the way to go, but it’s not always that clear cut. Weigh up the pros and cons, be open about them with your client, and do what works best in that situation.
Originally published on July 28, 2017, by Edwin Edelenbos on LinkedIn.