Originally Posted by jong1
Unfortunately, all your testing has proved is whatever "18Gb/s" cables you were testing did not meet their advertised speed which, unless they were Premium Certified, is actually quite common. Even if they were Premium Certified, it is not unknown for them to fail, as individual cables are not certified, only a design/model, based on sample testing
"4K 60Hz HDR12 4:2:2" (better called 4K/60 12-bit 4:2:2, since it does not need to be HDR) only requires 17.82Gb/s (https://community.cedia.net/blogs/da...tes-for-4k-hdr
) and no HDMI 2.0 source device will attempt to send more, whatever the theoretical bandwidth of the cable it is connected to.
It is possible that you may be able to make your PC card deliver over 18Gb/s (eg. send 4K/60 12-bit 4:4:4) but that would be outside of the HDMI 2.0 spec, not what any HDMI 2.0 conformant source would attempt and 4K/60 12-bit 4:2:2 will always require only 17.82Gb/s
That's interesting and makes sense. I would have hoped every cable had passed a performance quality inspection when it was manufactured though, but maybe not if the manufacturers are using the certification to prove the design and are then choosing to save money on performance inspection costs by simply not doing that performance testing because they think the certification means they'll never get quality problems? I know you'd never get away with that in the UK but maybe in Taiwan and China etc this lack of inspection is deemed ok and maybe even common practice to save money? That then makes sense why every cost saving effort I've known in Aerospace and Motorsport projects where the accountants have tried to save money by having parts manufactured in these cheaper countries has ultimately resulted in poor part quality resulting in huge production problems and typically several months of delays to projects before the decision was ultimately made to finally bring manufacture back to more local machine shops back in the UK.
When I work through the maths theory myself I keep reaching the answer that 17.92Gbps supports 4K 60Hz 12bit 4:4:4 so I'm not sure what's going on now because none of the cables, not even the 48Gbps cables give that performance..... thinking.....
I find it curious how every 18Gbps cable I've ever tried does not match up to the 27Gbps or 48Gbps cables though. I mean I've definitely purchased Premium Certified 18Gbps cables and those didn't work as good either.... so what's going on.....
There must be more than meets the eye going on inside a HDMI cable. I mean there's obviously sound information too but I'm sure this can only be a very small fraction of the data required compared to video information. Then there must be timing information to give the CPU on the TV end enough information to stitch everything perfectly back together again, but I'm not sure how much extra information this all takes up? Oh and it's digital too so the 17.92Gbps video information must be having to be sent at least twice...... WAIT, maybe that's it.....???
Maybe 18Gbps cables can only support a 9Gbps data "picture" but the information is sent down the cable twice... 9+9=18 ? That would actually make sense as a 4:2:0 picture requires exactly 50% of the data that a 4:4:4 picture requires!!! OK, so working through these numbers I get the following which seems to tie in nicely with the results being achieved with the cables I've tested:
4K 60Hz HDR12 4:2:0 picture = 17.92Gbps video data.
4K 60Hz HDR12 4:2:2 picture = 23.89Gbps video data.
4K 60Hz HDR12 4:4:4 picture = 35.83Gbps video data.
This would explain why all the 18Gbps cables I've tested are only achieving the 4:2:0 Chroma picture. And it would also explain why the 27Gbps and 48Gbps cables are achieving the 4:2:2 Chroma picture too. The only thing it doesn't explain is why the 48Gbps cables are not achieving a 4:4:4 Chroma picture, but I imagine this simply is a result of the actual female HDMI 2.0 port reaching it's maximum data capacity and so when the two devices are handshaking they're opting to choose the next best picture standard that does work which is the 4:2:2 Chroma one.
Otto, why do you keep plugging the RuiPro4K cables. I've just found them online and they don't appear to have sold more than 20 globally and have hardly any reviews? You never answered my question when I asked you what you do for a job so I'm now wondering whether you might work for RuiPro4K?? Are you a qualified engineer?
Ratman, what can I say, if I've offended you then I'm sorry and apologise, but you have to admit, it is a rare thing for a 65 year old to be posting in forums about technical subjects such as these so when you seem to be constantly degrading the words I'm saying and joining sides with a person who appears like they may work for RuiPro4K it begins to make me question everything about your profile, it's that simple, but if you are who you say you are then I do apologise.
All I've done is provide real world test information and try to provide answers as to why the test data I'm seeing is happening 100% of the time on my test setup and since posting this information I've been constantly told this information is not true from Otto, but it is true in my setup and I'm trying to look for answers why and hopefully help other members of the community who may also be having similar problems too.
Jong1, Hi, nice to meet you. You seem like you're willing to try and provide answers, so I'm interested to know if you have enough of a background to be able to comment on the numbers I've questioned above? I mean it really seems to make sense to me but I'll happily be taught otherwise if you know what is going on and can provide some simple logical reasoning to support it