Marco Polo Season Two
- Thread starter Haywood
- Start date
HDR is not a slap you upside the head and hand you in next week kind of difference. It is more subtle, but it is still very impactful. The subtlety of light and the naturalness of the color is very immersive. Some people are calling this a DVD vs Blu-Ray kind of difference. It isn't. I do not watch it and then go back to a 1080p movie on Vudu or Blu-Ray and thing, "Wow, this looks like crap." What it represents is a very, very nice increment of improvement that takes us one step closer to that immersive experience we are all chasing.
For me, I have come to be very good at immediately noticing banding in shading around lights or solid surfaces. It is a very common issue with streaming video I have accepted and choose to blame on bandwidth constrictions. Supposedly, the hew HDR technologies will eliminate that issue, but I have yet to experience it.
Ditto. And it's why I just can't talk myself into upgrading my displays just yet.
I have yet to see any banding with HDR content. You can have an extremely bright light surrounded by dark and, not only is there no banding, the image retains shadow detail even very close to the light source. It provides that extra edge of realism.
I do not always get banding with 1080p sources. I suspect it is either a bandwidth problem that varies based on available bandwidth at viewing time or it is an encoding problem. I am in the midst of ripping my own Blu-Rays to put on my Plex server and have been experimenting with compression. If I am willing to wait 4 to 6 hours for my quad-core i7 to do the encoding and don't mind ending up with a 6-10 GB file, I can avoid the overwhelming majority of compression artifacts.
There are commercial (expensive to license) encoders which can get pretty small files (or relatively low bandwidth) outputs with little or no visible artifacts, and they are great for creating files, but for streaming, the bandwidth needs to vary to work with what's available, and as such it takes massive processing power and willingness to sacrifice quality in favor of break free and consistent playback.
The main reason to wait is price. The image quality improvement offered by the best of the new sets is well worth the coin, even with 1080p content. I don't know what set you currently have, but the image quality difference between my old Samsung LED Light Engine DLP and my new Sony XBR75940C is NOT subtle. The digital processing applied to upscale current content really delivers.
That makes sense and it also explains why I see artifacts in the same piece of content sometimes, but not others.
I am sadly stuck using MakeMKV and Handbrake, which produces a good result, but is very slow. Yesterday's project was Toy Story 3 and the end result was indistinguishable from the Blu-Ray. I would probably spot the differences if I viewed them side by side and stood close to the set looking for them, but the compressed file looked extremely good. The main downside was going from DTS HD Master to DTS. I'm playing with this because most playback devices cannot handle 1:1 Blu-Ray rips, I don't have a monster CPU in my media server for on-the-fly transcoding and compressed movies take roughly one fifth the space.
If you are talking about chapter markers, those can be retained.
I am using settings in Handbrake that the online community would consider extreme overkill, but it is not like I'm in a hurry or need super small file sizes. The main reason I'm dropping the lossless audio is that track alone can be the size of the entire movie without it. Other than that, I'm going for as close to the original as I can possibly achieve, even if it does take 7 hours to encode a live-action movie.
No... I-Frames in the compression process are those frames where 100% of the content is included. Then for the following several seconds, potentially even minutes, each subsequent frame pack is just what has changed from the previous frame. That's how they reduce the size and data rate. However, if those "P" and "B" Frames are being used to show a brand new scene, like going from a Beach scene to a Garage scene, it can appear the image is slowly forming into what it should be for many frames or until the next I-Frame arrives. If the encoder is told never to exceed a certain amount of data per second, then there may be too few I-Frames available to render something without serious artifacts.
So, if a human can mark exactly when the Image is completely different from the frame prior to it (like in a conversation scene where it switches quickly betweem two camera angles with different backgrounds), then the compression engine can make sure it looks better.
So, if a human can mark exactly when the Image is completely different from the frame prior to it (like in a conversation scene where it switches quickly betweem two camera angles with different backgrounds), then the compression engine can make sure it looks better.
It sure is... it will go through several seconds of frames trying to predict correctly to get the best results.
There are some great encoding tools out there which license hundreds of encoders known for special scene types, like moving water is very different from a spacious field of flowers which is very different from a fast action choppy scene. Those tools will actually encode several seconds of video with all 5 to 20 encoders then determine the one which worked best during playback and use the most appropriate encoder for each scene, which could be different for each scene. The playback decoder cannot tell the difference.
It is a crazy processing thing. For very serious projects, like encoding for Theater playback, for Cable/Satellite broadcast, or for BluRay, a human will make those decisions based on looking at the results. For near real time encoding of live content or random content being streamed from a master high resolution source file it needs to happen in near real time.
There are some great encoding tools out there which license hundreds of encoders known for special scene types, like moving water is very different from a spacious field of flowers which is very different from a fast action choppy scene. Those tools will actually encode several seconds of video with all 5 to 20 encoders then determine the one which worked best during playback and use the most appropriate encoder for each scene, which could be different for each scene. The playback decoder cannot tell the difference.
It is a crazy processing thing. For very serious projects, like encoding for Theater playback, for Cable/Satellite broadcast, or for BluRay, a human will make those decisions based on looking at the results. For near real time encoding of live content or random content being streamed from a master high resolution source file it needs to happen in near real time.
I just finished the first live-action movie with the settings I've been using. It took 7 hours, produced an 11 GB file and looked perfect. Granted, I only got 1:2 compression, but at least now the bit-rate is low enough to stream to most devices and there is basically zero degradation.
This is my next show.
I have to use the internal app on my Samsung SUHD to stream it with HDR because Roku 4 does not support HDR.
But I have to go into settings on my Samsung & Marantz to enable ARC etc so I can have DD+ (kinda of a pain for this foggy headed old fart).
Reason I stream with my Roku 4 is because I love the programming on Playstation's Vue.
I have to use the internal app on my Samsung SUHD to stream it with HDR because Roku 4 does not support HDR.
But I have to go into settings on my Samsung & Marantz to enable ARC etc so I can have DD+ (kinda of a pain for this foggy headed old fart).
Reason I stream with my Roku 4 is because I love the programming on Playstation's Vue.
Last edited: