Netflix Dolby Atmos problems with new Apple TV 4k

It shouldn't matter which episode/season of Drive to Survive is used, except Season 2 up are DV, and I don't know if all the seasons/episodes have the same video bandwidth. What does matter is that the show is 4K 50Hz, and in my experience that causes the Atmos issues to occur far more frequently than all other content that is only 4K 24Hz.

I did get the dropouts on ATV+, so it isn't a Netflix only problem. I am sure different streaming providers allocated different bitrates to their video streams, and even between different programs on the same service. Also, the source material itself plays a factor in how much bandwidth is required to encode it. There are many variables in play.

I do believe the problem is in the Atmos MAT 2.0 encoding, but have no idea if it is encoding or timing errors, or if the MAT 2.0 stream randomly cuts out, but my guess is that is the latter is less likely. How the AVR/Atmos decoder handles the corrupted MAT 2.0 will have an affect on how the problem presents itself. The latest version of Netflix is using the native Apple player, so I see no reason it would be a Netflix problem, since Netflix stream fine on other streamers, including the Gen 1 ATV4K.

I look forward to your forced 1080p 50Hz results. I expect you will not hear any issues. I believe this is pretty solid evidence that video B/W is a significant contributing factor to the problem. If you are getting pretty consistent dropouts, I suggest you also try switching to a Wi-Fi connection. For whatever reason, Wi-Fi seems to significantly reduce the frequency of the dropouts, and I recommend Wi-Fi to everyone who has the Atmos issue.

I disagree 200%!

Do you have a better explanation for why Drive to Survive has far more dropouts than other programs? Can you explain why Drive to Survive forced to 1080p 50Hz has no dropouts? Have you tested the Atmos issue on both Wi-Fi and Ethernet? Apparently not.

Your example of Gen 1 units now having the Atmos problem doesn't support your statement that my test findings do not support my conclusions. I have stated numerous times that tvOS 15.x is an order of magnitude worse than tvOS 14.x on the Gen 2. This proves that S/W is a factor. I have even gone on to say that I believe the Atmos problem should be fixable or at least mitigated in S/W, for this very reason. Though I haven't seen enough credible reports to believe there is a problem on Gen 1 units with tvOS 15.4, I left mine on 15.3, just in case. It is very possible that tvOS 15.4 was a sufficient regression to also affect Gen 1 units. Nothing I have said suggests this couldn't happen.

If you are referring to my belief that the root cause is a timing or other error in the creation of the MAT 2.0, I didn't say that my observations supported that belief. A belief doesn't require proof. However, I have provided anecdotal evidence for why I believe this in prior posts, but the gist is that encoding is more resource hungry than decoding, and no other audio formats have any issue. It is also more likely the thing that is different about Atmos than non-Atmos audio, and that is the creation of the MAT 2.0 Atmos metadata stream that accompanies the LPCM 5.1 stream out of the ATV4K.

My conclusion that higher bandwidth video is a trigger for the Atmos problem is reasonable and supported by the evidence. I sent a letter to Ara Derderian, cohost of the HT Guys podcast. He replied that he had never noticed the Atmos issue on his Gen 2. However, when he played an episode of Drive to Survive as I suggested, he experienced the problem within the first 10 minutes. Ara reads my email and discusses the Atmos issue in this week's episode, #1051, if anyone is interested. I know of at least one other person who also tested Drive to Survive on both Wi-Fi and Ethernet, and also found far fewer dropouts on Wi-Fi. I haven't previously discussed why that would be the case, but my guess is that the Ethernet connection requires more processor overhead than the Wi-Fi, leaving fewer resources to the MAT 2.0 stream.

I was just trying to share what I believe are important observations regarding the Atmos problem, and some potential workarounds that I have not seen posted elsewhere. I trust that other readers will separate the wheat from the chaff. As an engineer I do a lot of troubleshooting, and part of that process is making educated guesses, some of which are wrong. I am used to rational arguments or test data to prove me wrong, not knee-jerk criticisms as is all too common in public forums, and I just had to vent. I put some thought and effort into getting this far with the Atmos problem, mostly for the fun and challenge of it, but also with the naive hope that some helpful information might trickle up (or down) to the engineers at Apple who can fix this. Though I would like to see this issue fixed, I have a few other streamers at my disposal that do not have an issue.

It looks like there was a software update today for the AppleTV 4K: I was on 15.4.1 and now I'm on 15.5. I figured I'd take a chance and see if Apple fixed the problem. The show we watched tonight was "Hanna" on Prime Video but it was 5.1 and I thought it sounded better and more solid than I remember. Afterwards, I played some of "Encanto" from Disney+ in Atmos and that worked fine. I also watched a few minutes of a 4k Dolby Vision+Atmos movie in my AppleTV library and that worked well -- no dropouts. Finally, I tried a bit of Netflix Atmos content and that was also fine. I didn't get a full hour of Atmos content in, but I'm cautiously optimistic since this is a major version jump. Fingers crossed.

Same issues with stuttering dolby atmos sound here…. Been trying to figure when it happens and why but seems totally random. Again drive to survive good example.

last night installed update of tvos to 15.5… was kind of hoping issues would be solved…. But to my disappointment they still appear.

It only occurs with playing Atmos content from Netflix and not from Disney+ for instance.

i really hope Apple will find a way to solve this very soon, as it is going on for a while now..

Oh yeah.. every show or movie with Atmos audio gets stutters in the audio.. strangely video image is not affected at all!

Also i can confirm that having an internet connection over wifi or cable does not make any difference.

Interesting. In my limited testing, about 2 episodes of DtS, a WiFi connection resulted in significantly fewer dropouts. I know of others who had similar results. I was surprised that WiFi made a difference. Try restarting your ATV4K and testing on WiFi again.

i updated to 15.6 today, and I watched the entire Season 1 ep 1 of drive to survive without a single drop out! Also tried 6 Underground and a few others.. no drop outs! Really hoping this is fixed now.

Well…. Just now i watched for 3 hours straight to netflix atmos content over a wifi connection…. To my surprise hardly any dropouts.. just 2-3 small ones…. Was watching drive to survive last episode and few eposodes of the lincoln lawyer…

before watching these over wifi i watched the final episode of drive to survive over cabled internet… completely full with dropouts until the point it was not funny anymore…

so wifi or cable internet makes a difference

Your new results mirror mine. Thanks for retesting and reporting back.

I would wager you will have no Atmos dropouts if you set the ATV4K output to 1080p (independent of internet connection). You may not even notice the difference in resolution, depending on how well your TV upscales to 4K, viewing distance, etc.

That does sound very promising. You were on a wired Ethernet connection for that testing? I look forward to testing 15.6 when it hits public beta.

I have believed all along that the Atmos problem was fixable in S/W, as tvOS 15.x was an order of magnitude worse than 14.x.

Why don't you perform this simple test for yourself and report back with your results?

I suggested a possible explanation for this in my last reply to you. Pity you didn't read it.

That explanation is consistent with my finding that higher bandwidth video triggers the Atmos dropouts, since both could result in insufficient processing power to support creation of the MAT 2.0 Atmos stream without timing or other errors.

I have also speculated that the Gen 2 hardware has a subtle difference from the Gen 1 that the S/W didn't take into account. Perhaps tvOS 15.6 finally addresses this. The first 15.6 report from roque98 seems very promising.

[Edited by Moderator]

It’s crazy, I haven’t had a single drop out since updating. For context, I’m running my ATV into a Denon 4500H. I have my ATV set to 4K SDR with match frame rate and dynamic range on, using a 48gb hdmi cable(not that this really matters).

similar config here, AVR is a Yamaha A1080. Will check later and report back. The appearance is complete random and unpredictable, that's why in my mind a few tests hard wired/WiFi, 8k/4k cables is not valid.

My go-to test has always been S01E01 of DtS, as it always has dropouts within the first 10 minutes, often within the first minute or so. I always attributed this to the fast action/cuts, but later found that the show is 4K 50Hz, which confirmed that bandwidth was the key factor. I only tested with S01E03 last night because I was getting tired of re-watching S01E01. I am sure that it made it 20 minutes in to the episode because there were not many fast action/cuts up to that point.

My repeated experience with DtS S01E01 suggested to me that making it to the end of the episode before having a single dropout on WiFi was significant. With several others having a similar experience, WiFi versus wired seems to make a noticeable difference. But I agree that a higher number of tests are needed to increase statistical confidence. I am just not terribly motivated to do a lot of testing as I have a Gen 1 that has never had a dropout, even when I tested it on S01E01 of DtS.

My setup is the same except I have a Denon 4300H and a cheap 3ft Cable Matters HDMI 2.0 cable, which is more than up to the task. DtS is the only meaningful test as it is 4K 50Hz and consistently produces dropouts for me. Nearly everything else is 24Hz, so significantly lower bandwidth, and can go many hours without a dropout.