I’m seeing this too. With the deinterlacer and audio set to experimental and video set to standard, it works fine. Once I turn on experimental 4K video, it seems to start stuttering off and on as mtnbkr60 mentioned.
When watching the same program recorded vs live, it seems to play back fine. It seems like when viewing live is where the issue occurs. I’ll do more testing this evening.
I experienced the judder issue without any experimental settings enabled - it was limited to TVE channels for me (most noticeable on sports due to the motion). In my case it would either clear up on its own or by exiting and retuning the channels. I did submit diagnostic information and send a message to support, though I can’t test any further since I ended up returning my Apple TV 4K (was disappointed with the deinterlacing - particularly on SD channels - compared to Chromecast, even with experimental enabled)
Upon further investigation I have found the audio breakup with the new audio driver only occurs on ota channels from my hdhr. Tve channels are fine.
So maybe an mpeg2 decode issue or the fact that ota has 6 audio channels in AC3 mode and the tve channels only have 2 in AAC?
I agree with this. I notice the audio problem on OTA. This is a recent issue. It works better on Default audio driver.
A few observations regarding Apple TV video problems:
It seems the problems I was having (flickering on interlaced channels, intermittent judder on TVE) are not specific to the new Apple TV model, as I’ve reproduced them on the Apple TV HD on the same Samsung 4K TV.
The flickering on interlaced channels seems to be caused by my Samsung TV’s Auto Motion Plus setting. I had it turned on but down to 0 to properly play 24p, but it seems even that is enough to cause issues. After turning it off, interlaced video looks as good as my Chromecast with hardware deinterlacing.
The judder issue seems to happen intermittently, on TVE channels only, without using any experimental video drivers or deinterlacer (the latter won’t even work on an Apple TV HD). It does seem like it may be better on the experimental video driver with the last update, though as this only happens intermittently it may be a bit soon to say.
Just thought some here might be interested - it still seems like the Chromecast wins in video quality on my TV due to being able to enable Auto Motion Plus (which definitely improves 30fps TVE) without causing issues for interlaced content, though this at least solves some of the issues. Still have to figure out whether I’ll get another 4K Apple TV or a Shield for my 4K TV (I’d like a device with more power than the Chromecast)…
I had been running the experimental deinterlacer for a while but yesterday I was seeing slight hitching. This time I didn’t see frequent frame drops in the stats like before, but I switched to the linear deinterlacer and the problem went away.
Of course it’s also possible I’m losing my mind and it’s all in my head. But I don’t think so.
I've been seeing the hiccup glitches recently too. Trying to figure out what's going on
To help us help you, is there an explicit place in the “show stats” on Apple TV that indicates whether the content is interlaced or not? There are a couple of stats that may indicate this but I’m not 100% sure. Would be nice to be able to tell easily. Sorry if I’m missing something…
If you go to your recordings on the server website (I.e. DVR-Manage at your IP:8089), it will say under Options-View Details whether a recording is interlaced or progressive.
Right, I was just hoping to be able to tell from “show stats” as well to help troubleshoot quickly.
I've been seeing the slight glitching on camera shots that were pretty still. It's really noticeable since there isn't much motion. I can go back and replay and then I don't see it again.
I'm uploading a new TestFlight beta now with more improvements to the experimental video driver.
New beta good so far. My only issue is TVE feeds that are sent 29.97 instead of 59.94. They take on that 3/2 pull down look. CNN is a good example.
I’m still seeing the issue on live tv on my Apple TV. It only occurs when I change the video driver to experimental.
Changing Deinterlacing mode doesn’t seem to effect it, just the experimental video driver. When I turn it back to default, the problem disappears.
Seems to be gone for me. Running all experimental modes now and using match content.
So for me it is MUCH better. I didn't realize how much I was relying on my TV's post processing to clean up the picture. I turned all of that stuff off and switched to the experimental video driver. Now TVE doesn't look like crap anymore, it's actually watchable.
I haven't looked for the microstutter issue yet, it takes longer for that to show but in my quick testing I didn't see it. (that could have been my TV's post processing getting in the way)
I noticed that the combination of settings is important. Experimental deinterlace along with experimental driver results in the ATV putting out 29.97 on 29.97 sources. But if you turn off experimental driver, then the stats for nerds says that 29.97 is going out as 59.94. So its a matter of do you want the ATV or your TV to do the conversion?
So, I just got the new ATV4k, add Channels and HDHomeRun 4K OTA today. What should the ideal settings be in the Channel Settings app? Appreciate the help!
My advice:
For channels:
Run the testflight build
Set the deinterlacer and video driver to experimental
For ATV:
Set to 4k SDR
Match frame rate
Match dynamic range
For HDHR:
Check your signal strength on the HDHR, they tend to not like hot signals. I'm 12miles from the towers and the signal reads 100%. I have 9db of attenuation that I added and it now reads 97% on most channels.
If you are far away from your towers then you probably don't have to fool with attenuation.
Thank you! Appreciate the guidance. Now going to setup my Harmony Elite Hub remote according to the tips from channels.