INTERACT FORUM
Devices => Video Cards, Monitors, Televisions, and Projectors => Topic started by: Manni on June 19, 2024, 08:27:40 am
-
It was established a while ago that this issue was specific to 3xxx GPUs, possibly 4xxx GPUs, and it might be a hardware issue. The reason why I'm creating this new thread is because I have some new information so I'd like to see if anything can be done in LAV or MC to fix it, despite the fact that it might be specific to the hardware.
I'm also trying to make this thread specific so that other similar but unrelated issues are not reported here. If you have a different GPU, no HDMI 2.1 AVR or if you experience this with other sound formats, or when not bitstreaming, or with non-4K HDR content, please DO NOT POST HERE. Find a different thread, or create another one.
If I'm the only one with this issue, then it will be easy for Hendrik to do nothing, and I'll be the first to understand :)
If you do post in this thread because you experience long freezes or micro dropouts when bitstreaming Atmos through an HDMI 2.1 AVR with a 3xxx GPU, please provide the following info, as I do, so that we can see if there are some common elements causing the issue:
Motherboard: ASRock X570 Creator
CPU: AMD 5950X
Memory: 32GB
GPU: Zotac 3090 (no OC/UC) with 555.99 Studio drivers
HDMI 2.1 AVR: Denon X8500HA
Anything else in the HDMI A/V chain: HD Fury VRROOM, but it's after the AVR (so not in the audio chain) and the issue also happens with the VRROOM out of the chain.
OS: Windows 11 23H2 with all updates installed
MC: MC32 build 55 (it happens here with both direct sound and WASAPI, and some buffer settings can improve the issue, but not fix it)
Renderer: JRVR (but it also happens with madVR).
The new information I have is that I have found two workarounds for this issue:
1) Compromise immersive audio quality and not use bitstream. That's an obvious one, and not new. If you set MC to not use HDMI bitstreaming, LPCM has zero audio drops with Atmos. Obviously, you lose the metadata for immersive sound (Atmos or DTS:X), so this is only a workaround if you don't have any height speakers in your setup or don't mind losing the native immersive audio and are happy to simply upscale the base DTS:HD or Dolby TrueHD layer using Neural:X or Dolby Surround.
2) Select 8bits in the nVidia control panel, instead of 10/12bits. That's less obvious, and the new bit of info. If you set the driver to 8bit, MC has to dither so you get a bit more noise near black with 10-bit content. You compromise the picture quality slightly, but you can bitstream Atmos and have zero dropped frames. At least with the native refresh rate. If I use 119p for 23p, I got one micro audio drop in a 153 minute films, which is still a lot better than in 10/12bits. [EDIT got Atmos freeze/drop at 4K119 with 23p content today, so native frame rate seems like the only workaround. I’ll keep testing that way.]
3) [EDIT 14-August-2024]: I'm adding this third compromise (see this post https://yabb.jriver.com/interact/index.php/topic,139110.msg967014.html#msg967014 on page two of this thread for more details), which is to compromise gaming (no 4K120) and use a HDMI 2.0 input of the AVR to force TDMS. This allows to keep 10/12bits at 4K23, so no compromise on audio or video, just a compromise on gaming (no 4K120). Selecting a HDMI 2.0 EDID while still using a HDMI 2.1 input might allow to reach the same results (as it would prevent the GPU from using FRL5/6), but I haven't tested this for now.
So if you experience this issue with long pic freezes/audio dropout or micro audio dropouts when bitstreaming Atmos, please could you 1) post your detailed set up as I did above and 2) post if these two workarounds work for you? Note that if these workarounds don't help, you migght have another issue in your system (network stability, HDMI cable, etc) so you might want to isolate that first, for example playing the test film from an SSD or replacing your cables with HDMi 2.1 Premium certified cables to ensure that your issue isn't network or cable related. Atmos needs a bit more bandwidth, so it might be enough to trigger a dropout in some cases.
Please note that you often have to wait at least 30 minute, frequently up to 60 minutes, and quite often up to 90 minutes or more before experiencing any of these dropouts, so please test when watching a film at least 120min long. You have to watch it because although the long freezes/dropouts will show a few hundreds dropped frames in the OSD stats, the micro audio drops won't show any dropped frames. You can't just watch 20 minutes once and report "no Atmos dropout issue".
If you don't have any audio dropout issue when bitstreaming Atmos through an HDMI 2.1 AVR using a 3xxx GPU (or a 4xxx GPU), please could you kindly 1) check before posting by playing a couple of full two+ hours films in full that you actually don't have the issue and 2) provide as much information as possible about your hardware and your software setup?
@Hendrik (and other devs), do you have any idea why the dropouts go away in 8bits? Also, I've noticed that increasing the buffering to 150ms instead of 100ms in the audio device settings helps a bit. Unfortunately, setting buffering to any higher value causes a crash / prevents playback from starting. Is there any chance you could fix this buffering issue, so that we could try higher buffering values?
Thank you!
-
I have this issue but I won't have time to test until maybe this weekend. I have a similar hardware configuration with an ASUS mainboard and an AMD cpu. I'll get the full specs of everything later. The difference for me is while I don't have an HDMI 2.1 AVR, what I DO have is the VRROOM in the chain which is HDMI 2.1. That is splitting audio out to my AVR through its HDMI out port. No audio is being sent to my display. In any case, I'll post more information about my setup and whether the suggestions change anything when I get time later this week. But you are not alone with the problem.
-
The buffer size is limited by the driver, and the absolute size depends on channel and sample rate used - bitstreaming uses the maximum however, as bitstreaming TrueHD uses 8 channel at 192kHz, so the maximum buffer length is limited accordingly.
-
The buffer size is limited by the driver, and the absolute size depends on channel and sample rate used - bitstreaming uses the maximum however, as bitstreaming TrueHD uses 8 channel at 192kHz, so the maximum buffer length is limited accordingly.
Thanks for the explanation, makes sense. You might want to add a check so that the user can't select a forbiden/excessive size, as it looks like a bug when MC then crashes during playback.
Any idea why the dropouts go away in 8bits? Is it just a bandwidth issue, as 8-bit does reduce the bandwidth for picture, which might leave more room for audio? Anything that could be done about this to make it work in 10bits? 4K23 10bits is still TDMS, so it's not FRL related. 4K120 8its is FRL5 and it stil works much better (single micro drop in almost 2.5 hours) than in 10bits, which is still FRL5. I was hoping this might give you some idea. Can you reproduce the behavior I'm reporting with your 4xxx?
-
I have this issue but I won't have time to test until maybe this weekend. I have a similar hardware configuration with an ASUS mainboard and an AMD cpu. I'll get the full specs of everything later. The difference for me is while I don't have an HDMI 2.1 AVR, what I DO have is the VRROOM in the chain which is HDMI 2.1. That is splitting audio out to my AVR through its HDMI out port. No audio is being sent to my display. In any case, I'll post more information about my setup and whether the suggestions change anything when I get time later this week. But you are not alone with the problem.
Thanks, looking forward to your feedback :)
-
Correction, I did more tests in 4K120 and I’m getting freezes/dropouts too.
At this stage, only 4K 8bits at native frame rate (TDMS) seems to resolve the issue (along with not using bitstreaming of course).
I’ve edited the first post to reflect this.
-
That's not an option for me. I do ATMOS and DTS:X so not going to happen as a "solution". 8 bit is also not great. I will do the testing this weekend to see if it works but it's by no means a solution in my book.
-
The only bitstreaming I do is when playing Netflix, which is playing through a browser, and that also has occasional drop outs. It seems like a fundamental problem, rather then any software issue.
-
That's not an option for me. I do ATMOS and DTS:X so not going to happen as a "solution". 8 bit is also not great. I will do the testing this weekend to see if it works but it's by no means a solution in my book.
I never said either was a solution, they are just ways to limit the issue by compromising either the sound quality or the picture quality.
8bit isn’t great, but there is no visible difference here, I haven’t noticed any additional banding, but I haven’t tested for this specifically. With madVR, using 8bit with HDR content only added a tiny bit of noise near black that was only visible with your nose on the screen. I suspect it will be similar with JRVR if 8-bit is confirmed as a workaround.
We need 10bit on the content side for HDR, not so much on the display side with good dithering. Not ideal, but good enough as a workaround if there are no visible downsides.
I find 8-bit less limiting than losing true immersive audio, so that’s what I’m testing at the moment.
I’ve created this thread to find a better solution, so I’m also hoping for a breakthrough. :)
-
Or you could save a lot of time and energy (but not money...) and buy a RTX 4xxx.
It solved all Atmos issues for me including this driver one : https://forum.doom9.org/showthread.php?p=1997341#post1997341
-
Or you could save a lot of time and energy (but not money...) and buy a RTX 4xxx.
It solved all Atmos issues for me including this driver one : https://forum.doom9.org/showthread.php?p=1997341#post1997341
Thanks, but the last time I tried a 4060 I still got Atmos dropouts, so I sent it back. That was before Hendrik improved things in LAV though, so I could give it another try. I would try my 1080ti first though. Which exact 4xxx (brand, model) did you get ? A 4070ti super (the first model at least as good as my 3090) would cost me more than $1,000, so not a cheap solution.
Note that I don’t have any of the issues you reported in that post with my 3090, so there is clearly something else that’s wrong in your setup. Please could you also post your detailed info, as I did in the first post, so we can have details of a non-working config with a 3090 and working with a 4xxx? Thanks!
-
I installed my 1080ti without changing anything (shut down, swapped GPU, restarted with the same drivers etc). I only set the correct settings in JRVR as they had reset to default.
No dropouts with 4K23 12bits. So this shows that it's definitely specific to the 3090.
I might just keep the 1080ti in the HTPC and use the 3090 in my E-GPU with the laptop when I need more GPU power for games or editing.
If someone comes up with an exact 4xxx model that's guaranted to work at 4K119 without dropouts, I might give it a try.
-
Just the HDMI port or does the DP port has same problem? Maybe a DisplayPort to HDMI adapter can reduce or solve the problem?
-
Just the HDMI port or does the DP port has same problem? Maybe a DisplayPort to HDMI adapter can reduce or solve the problem?
Nope, tried that a while ago and it didn't help.
-
I've ordered a Zotac Trinity 4080 Super, as that's about the only 4080 that can fit in my HTPC case. I should receive it tomorrow. Let's hope it will be as good as the 1080ti for MC duties, otherwise it's going back. I'll keep the 1080ti in the HTPC and will use the 3090 as an the E-GPU for the laptop for video editing and possibly gaming. Won't be as good as the 3090 in the HTPC, but it should be better than the 1080ti.
-
I'll be curious to see how that works out for you. Best Buy has an ASUS TUF 4080 Super in stock near me. I'm not really looking to drop a grand right now but if it solves these stupid issues then I'd at least consider it.
-
I’ve been using an MSI GeForce RTX 4080 Super 16G Gaming X Trio for 5 months now.
It literally solved all my issues. No Atmos cuts, no non-working weird combinations, no driver issues, everything just works as intended.
Obviously, I’m not using it only for video; I play games on TV with that PC, and now with DLSS 3 and frame generation, which is a game-changer.
The 3090 FE is now in my desktop PC.
-
Well if it solves all Manni's problems (they're the same as mine) then I'll have to contemplate a trip to best buy. I'll actually be right next to it tomorrow. I just don't really want to spend that kind of money. And that's complete nonsense on nVidia's part that an entire line of cards has a potential hardware problem.
-
I’ve been using an MSI GeForce RTX 4080 Super 16G Gaming X Trio for 5 months now.
It literally solved all my issues. No Atmos cuts, no non-working weird combinations, no driver issues, everything just works as intended.
Obviously, I’m not using it only for video; I play games on TV with that PC, and now with DLSS 3 and frame generation, which is a game-changer.
The 3090 FE is now in my desktop PC.
Thanks for confirming your GPU model. Please could you let us know your CPU, OS and driver version? Thanks.
-
i7 13700K, Win 11 and drivers always updated.
-
i7 13700K, Win 11 and drivers always updated.
Thanks. will update as soon as I receive and get a chance to test the 4080 super.
-
So, received and installed the Zotac Trinity 4080 Super today.
I did a straight swap again (power off, took the 1080ti out, put the 4080 super in, power on) because I didn't want to change any of the other parameters. This was with the latest 555.99. I only respected the same audio and video settings in MC. Obviously I also set the driver to 10bits.
I selected my longest Atmos movie as a first test, Titanic (3 hours and 14 minutes).
The whole film played without a single dropped frame or Atmos drop, micro or longer freeze.
So it looks like this is definitely a 3xxx specific hardware issue, given that it's solved by both the 1080ti and the 4080 Super.
I don't feel bad about upgrading as the 1080ti turned out to not pass HDR metadata and developed some random sparkling pixels (that neither the 3090 or 4080 have).
The 4080 Super isn't cheap at £959, but at least it's £350 cheaper than the 4080 was a few months ago, and it's the most efficient 4xxx. Almost completely silent, even under load (the 3090 was ok for video playback but sounded like a tornado under heavy load).
The 4080 Super is about 50% faster than my 3090, and it uses a lot less power as it's limited to 320W vs the 350W of the far less efficient 3090.
If this is confirmed in the next few days, I'll definitely keep the 4080 Super and the 3090 will end up in the EGPU for my laptop.
Anyway, not a cheap fix, but I highly recommend it (or another 4xxx) if you can afford to throw money at the problem. Looks like it solved it here.
I'll post some measurements for power use soon, and I'll confirm that all has kept working fine, but so far no issue at all.
By the way, I did my latest test with two Vrrooms in the chain, my usual one between the AVR and the display, and another one between the HTPC and the AVR, to allow me to extract the audio to my Sony HDMI Headphones. Clearly the Vrroom (or I suspect any HD Fury device) isn't related to these issues.
-
Ah man I was just over at best buy this morning, too. LOL I guess I'll have to make a trip cause we really need to be able to use the latest drivers and NOT have these kinds of stupid issues. This sucks. sigh but at least we know the fix.
-
Ah man I was just over at best buy this morning, too. LOL I guess I'll have to make a trip cause we really need to be able to use the latest drivers and NOT have these kinds of stupid issues. This sucks. sigh but at least we know the fix.
Yeah, I think from now on I’ll just upgrade my GPU when Hendrik does, and buy the same gen and model he does, it should save a lot of grief :)
-
ROFLMAO! Clearly! LOL
-
I told you so :)
-
This card is freaking giant! And it requires 3 power cables instead of the 2 I had on my 3080. This project is quickly becoming a PITA. As soon as I get another power cable to it I should be able to boot it up. That is one big ass card though. And I'm thinking a power supply upgrade at this point is in order. EDIT: As in right now. I just ordered a Corsair 850e. I don't like playing with that kind of power. Won't that be a fun tomorrow project. Now to see if I can get this thing running for tonight.
-
This card is freaking giant! And it requires 3 power cables instead of the 2 I had on my 3080. This project is quickly becoming a PITA. As soon as I get another power cable to it I should be able to boot it up. That is one big ass card though. And I'm thinking a power supply upgrade at this point is in order. EDIT: As in right now. I just ordered a Corsair 850e. I don't like playing with that kind of power. Won't that be a fun tomorrow project. Now to see if I can get this thing running for tonight.
Yeah, I had to get a third cable out for my 1000W modular PSU. Was in the box, so easy enough. You will laugh but the Zotac is the smallest 4080 super available. I wouldn’t have been able to fit any other in my HTPC case. I hope it will work for you as well as it seems to be working for me. Second film watched and still no dropout. Fingers crossed, but it looks like the real deal.
-
I told you so :)
Yes, looks like it’s the (expensive) magic bullet :)
-
Yeah, I had to get a third cable out for my 1000W modular PSU. Was in the box, so easy enough. You will laugh but the Zotac is the smallest 4080 super available. I wouldn’t have been able to fit any other in my HTPC case. I hope it will work for you as well as it seems to be working for me. Second film watched and still no dropout. Fingers crossed, but it looks like the real deal.
I can't get it to post. I have 3 power cables but the video card does not seem to like it at all. I may have to wait to finish this tomorrow when the new power supply comes and I can use a single 12v cable. I really hate this nonsense. I didn't even really want to buy a new card right now given that the 5xxx series will be out in a few months. But whatever. If it solves all our problems, it's fine. I'll eventually get it working.
-
Yeah, I think from now on I’ll just upgrade my GPU when Hendrik does, and buy the same gen and model he does, it should save a lot of grief :)
My HTPC actually runs a 3070 because I had to upgrade to something with HDMI 2.1 when I got the 4K screen a few years back, and that was it then. Not feeling like replacing that with a 40-series card over an occasional audio dropout thats exclusive to Netflix for me, since I dont otherwise bitstream. :P
My desktop on the other hand usually gets upgraded when a new gen comes out. :P
-
I can't get it to post. I have 3 power cables but the video card does not seem to like it at all. I may have to wait to finish this tomorrow when the new power supply comes and I can use a single 12v cable. I really hate this nonsense. I didn't even really want to buy a new card right now given that the 5xxx series will be out in a few months. But whatever. If it solves all our problems, it's fine. I'll eventually get it working.
The 3 PCI-E 8-pin to the 12-pin converter worked fine here, but each one has its own rail and the PSU is 1,000W. You'll need a 750W min PSU to run this.
Well, the 5090 at $$$$$ will be out in 2025, but by the time a 5080 comes out at an affordable time it might be next summer. I really didn't want to wait that long, or invest in a 5090. I had no plans to upgrade either, but if this 4080 Super does the trick, it will same me a lot of time, and time is $...
I'm sure you'll get it to work with the new PSU. Let us know.
-
I have an eVGA 850 Gold Plus so the PSU isn't the issue. I have 3 separate VGA cables connected to the PSU and the little adapter. But motherboard posts 00 code (ASUS 570E Gaming WIFI II with a 5900x cpu). All was working fine with the 3080 before I took that out. I'm really not in the mood to pull this ginormous GPU out and reinstall the 3080 to see what's going on at this point. Mounting that 4080 was a major PITA in my case. I think I'll just prep it for the new Corsair PSU coming tomorrow by removing the eVGA and all the cabling. God I hate this crap. I probably should have just resigned myself to building a whole new HTPC at this point but I don't have the money and I'm so not in the mood for system rebuilding right now.
-
My HTPC actually runs a 3070 because I had to upgrade to something with HDMI 2.1 when I got the 4K screen a few years back, and that was it then. Not feeling like replacing that with a 40-series card over an occasional audio dropout thats exclusive to Netflix for me, since I dont otherwise bitstream. :P
My desktop on the other hand usually gets upgraded when a new gen comes out. :P
Well then we have to get you to bitstream so that you suffer the same pain we do and are motivated to make it stop :)
I wouldn't have replaced the 3090 either if it wasn't for the dropouts, but these are all the time, it's really not usable like this and the two workarounds (8bits or no bitstream) are really not great.
Anyway, I'm happy with the new GPU: HDR metadata is fine, no Atmos dropouts [EDIT: unfortunately they are back] and much more efficient.
Here are some measurements I've taken over the years, and how the Zotac Trinity 4080 Super compares:
-
I have an eVGA 850 Gold Plus so the PSU isn't the issue. I have 3 separate VGA cables connected to the PSU and the little adapter. But motherboard posts 00 code (ASUS 570E Gaming WIFI II with a 5900x cpu). All was working fine with the 3080 before I took that out. I'm really not in the mood to pull this ginormous GPU out and reinstall the 3080 to see what's going on at this point. Mounting that 4080 was a major PITA in my case. I think I'll just prep it for the new Corsair PSU coming tomorrow by removing the eVGA and all the cabling. God I hate this crap. I probably should have just resigned myself to building a whole new HTPC at this point but I don't have the money and I'm so not in the mood for system rebuilding right now.
Sorry to hear this. The PSU shouldn't be the issue then, not sure the new one will make a difference. Do each of the PCI-E cable has its own rail on the PSU? Did you get the Zotac Trinity or another model? It's about the same size as my 3090, and it's not that much bigger than the 1080ti, so maybe to 3070 was tiny.
Good luck and keep us posted!
-
Sorry to hear this. The PSU shouldn't be the issue then, not sure the new one will make a difference. Do each of the PCI-E cable has its own rail on the PSU? Did you get the Zotac Trinity or another model? It's about the same size as my 3090, and it's not that much bigger than the 1080ti, so maybe to 3070 was tiny.
Good luck and keep us posted!
Yes, my PSU has 4 rails for VGA and I have 3 of them connected. I got an ASUS 4080 Super TUF card to replace my Asus 3080 TUF. It's probably 3" or 4" longer than my 3080. It fits my case fine (fractal design cases are awesome). I have absolutely no idea why it won't post. I also don't remember what code 00 is for my motherboard but it's likely power related. Quite honestly, I'm not impressed with the adapter anyway. Having a single 12v power cable makes way more sense. I just want to get the darn thing up and running again so I can test it out. And of course tomorrow I won't have any time at all to watch anything. Figures.
-
Yes, my PSU has 4 rails for VGA and I have 3 of them connected. I got an ASUS 4080 Super TUF card to replace my Asus 3080 TUF. It's probably 3" or 4" longer than my 3080. It fits my case fine (fractal design cases are awesome). I have absolutely no idea why it won't post. I also don't remember what code 00 is for my motherboard but it's likely power related. Quite honestly, I'm not impressed with the adapter anyway. Having a single 12v power cable makes way more sense. I just want to get the darn thing up and running again so I can test it out. And of course tomorrow I won't have any time at all to watch anything. Figures.
Oh yeah the ASUS 4080 TUF is a monster, no way I could fit this in my case (Silverstone Grandia GD-07B, same as the original Envy). Especially the height/length. The Zotac was my only option. With the power cable disconnected, make sure the GPU is seated correctly. It's so heavy that it might be causing an issue. Mine came with an adaptor but as the case is horizontal I didn't need it. You might even want to try another slot if you can.
I really don't think it has anything to do with the PSU or the adaptor if you have three 8-pin cables coming from three separate rails on an 850W PSU. I also have an X570 and a 5950X and I never draw more than 500W from the wall, so you should be fine with the 850W.
-
Oh yeah the ASUS 4080 TUF is a monster, no way I could fit this in my case (Silverstone Grandia GD-07B, same as the original Envy). Especially the height/length. The Zotac was my only option. With the power cable disconnected, make sure the GPU is seated correctly. It's so heavy that it might be causing an issue. Mine came with an adaptor but as the case is horizontal I didn't need it. You might even want to try another slot if you can.
I really don't think it has anything to do with the PSU or the adaptor if you have three 8-pin cables coming from three separate rails on an 850W PSU. I also have an X570 and a 5950X and I never draw more than 500W from the wall, so you should be fine with the 850W.
Yes it is a monster card. I have it laying down right now to alleviate weight issues (it has a cute little stick that supposedly will hold the rear end of the thing in place. That will be problematic in my case if I need it. I can't use another slot though. This thing takes up 3 slots. It really is huge.
I actually think it's the adapter. I did a little research on the ones that nViidia makes and they absolutely SUCK quality wise. It looks like people have had issues with the adapter shifting, not seating properly, etc etc. The Corsair I got is another 850. I have a 5950x, as well with 2 nvme SSD's and a water block for the CPU. I'm not drawing a lot of power, either. I mean the 3080 never had an issue even when cranking in games. I don't love having 3 VGA power cables running through the case anyway. Swapping those out for 1 12V power cable will make cable management nice, and I'll be ready for any future upgrades I may want to do. They don't change PSU designs all that often. I may toss the 3080 back in just to make sure I didn't break anything else while trying to mount that thing.
-
Looks like something on the machine has died. I hope not the CPU but that is possible. This is why I hate messing with perfectly working machines and why I tend to replace everything all at once when I do. So yea, now I have no HTPC. Fantastic.
-
Looks like something on the machine has died. I hope not the CPU but that is possible. This is why I hate messing with perfectly working machines and why I tend to replace everything all at once when I do. So yea, now I have no HTPC. Fantastic.
Sorry to hear that. I hope you'll sort it out.
-
Beware of 3080. I had a Seasonic 1000W PSU working perfectly for 3 years. A 3080ti killed it, as it has insane momentary current draw. It is a known problem of the 3080. Seasonic RMA the PSU and sent a newer model with the fix of the problem.
-
The 3080 has been running fine in that machine for a while. But the machine won't post at all so either I damaged the PSU or something else. I'll know Monday if it's the PSU. I switched my order from the Corsair to a new eVGA. I've had good luck with the eVGA over the years. But I want one with the new nVidia connector as I'm not jazzed about putting an adapter with 3 power cables attached to it in my machine. If when I get the new PSU it still doesn't boot, then I'll have to decide what I'm going to do about just building a whole new machine. With AMD Zen 5 releasing this year, I may just say screw it and not have an HTPC for a few months. Worse case I can always grab my long zeskit HDMI cable and run it to this laptop which has a 4080 in it when I want to use HTPC functionality for a few months. I'm just not wanting to spend the money right now for a full rebuild (cpu, mobo, and memory)
EDIT:
I threw the 3080 back in the machine I bought it in. Only way I could acquire a 3xxx card during the pandemic was to buy a whole machine. It's an ASUS machine with some intel i7-11700kf CPU but it at least boots. There's ZERO chance of fitting that 4080 in that case though. Might be a temporary solution if I can't get my HTPC working.
-
I moved this since it's now a purely hardware discussion. If you want the first part split and moved back, let me know where to split it.
-
I personally think people need to know about this issue with the 3xxx cards. If Manni and I weren't sure that it was NOT a JRiver issue, I can promise you that others might come to the conclusion that it's a JRiver problem. And that would be a shame given that nVidia quality control is at fault.
-
New power supply did not fix the issue, as expected. Something else has died. The manual states that q-code 00 isn't used. Um, what? That's ALL it displays. So it's not even trying to post. My guess is the mobo is dead. At this point I'll see what happens in July with the AM5 Zen 5 CPU's and motherboards. I'll just build a new machine and use the 3080 machine until I can make that happen. I tried re-seating the CPU and that made no difference so it really does seem like the mobo has died. Which is frustrating given that it was not a cheap mobo when I got it a few years ago.
-
New power supply did not fix the issue, as expected. Something else has died. The manual states that q-code 00 isn't used. Um, what? That's ALL it displays. So it's not even trying to post. My guess is the mobo is dead. At this point I'll see what happens in July with the AM5 Zen 5 CPU's and motherboards. I'll just build a new machine and use the 3080 machine until I can make that happen. I tried re-seating the CPU and that made no difference so it really does seem like the mobo has died. Which is frustrating given that it was not a cheap mobo when I got it a few years ago.
Sorry to hear that. Did you try using another slot for the GPU? As long as the power cord was disconnected when you first swapped the 3080 and the new 4080 (as powering off isn’t enough to cut off all power to the mono), the most likely thing that could have been damaged is the GPU slot due to excessive pressure when you took out the old / put in the new GPU. Worth trying another GPU slot on the mono with either GPU, probably the 3080 first in case your 4080 is DOA.
-
As I mentioned before, I can't use another slot. There's literally no room. When I say this thing takes 3 slots, I'm not joking. The only other slot is at the bottom of the case and has literally no room. So that's not an option. One thing I can do and I may consider it when I have a few hours is rip out the mobo from the ASUS machine and drop it in another case. I have an old intel machine kicking around (9900k) that has a water block that should work with the intel board from the ASUS. Then I should have enough room to mount the 4080 on that board. But for now I'll probably just install MC32 on the ASUS machine until I can rebuild my main machine. Then I can have the ASUS machine for a bedroom HTPC.
-
As I mentioned before, I can't use another slot. There's literally no room. When I say this thing takes 3 slots, I'm not joking. The only other slot is at the bottom of the case and has literally no room. So that's not an option. One thing I can do and I may consider it when I have a few hours is rip out the mobo from the ASUS machine and drop it in another case. I have an old intel machine kicking around (9900k) that has a water block that should work with the intel board from the ASUS. Then I should have enough room to mount the 4080 on that board. But for now I'll probably just install MC32 on the ASUS machine until I can rebuild my main machine. Then I can have the ASUS machine for a bedroom HTPC.
Well, I have bad news. It took four films but I got my first Atmos freeze :( so it looks like the 4080 might improve things but doesn’t fully resolve the issue. I need to investigate what’s happening before deciding if I send the 4080 back or keep it.
-
Well, I have bad news. It took four films but I got my first Atmos freeze :( so it looks like the 4080 might improve things but doesn’t fully resolve the issue. I need to investigate what’s happening before deciding if I send the 4080 back or keep it.
Well, that's still better than what we've dealt with prior on the 3xxx cards IMO.
-
Well, that's still better than what we've dealt with prior on the 3xxx cards IMO.
Yeah but not sure it’s worth a grand if it doesn’t fully solve the issue. I’m going to put the 3090 back and see if I can identify what’s causing these issues.
-
The other thing to try is a clean driver install. I know you dropped the 4080 in place of the 3090 to see if that would resolve it, but, a clean driver install may not hurt the efforts.
-
The other thing to try is a clean driver install. I know you dropped the 4080 in place of the 3090 to see if that would resolve it, but, a clean driver install may not hurt the efforts.
I did a clean install yesterday evening, after the first couple of tests.
-
I did a clean install yesterday evening, after the first couple of tests.
Ok. Well, that's not great news that's for sure. But still, it took quite a bit of effort to find another drop. I get that spending a grand on an imperfect solution isn't ideal but I'd really like to get away from having to deal with "oh this driver makes it worse so we have to use one that's 6+ months old"...
-
Ok. Well, that's not great news that's for sure. But still, it took quite a bit of effort to find another drop. I get that spending a grand on an imperfect solution isn't ideal but I'd really like to get away from having to deal with "oh this driver makes it worse so we have to use one that's 6+ months old"...
Well, the 3090 works fine with the latest driver, so that's not a reason good enough to drop a grand. If the 4080 had resolved the issue, it was an expensive fix. If it doesn't resolve the issue, I'd rather investigate the 3090 further while I can still return the 4080. If it's not the GPU but something else, I'll find out what it is. I'm looking at concurrent access to my unRAID server right now. After the first Atmos freeze earlier, I asked my daughter if she had just started using the server to watch a film and she said yes. It might be a dead end but it's worth investigating
-
Well, while you investigate that I'm figuring out how I can get the 4080 into service sooner than later. I may break down and buy a cheap mobo for my 5900x just to get the machine up and running again. Then I'd have the 3080 machine for a second HTPC/gaming machine which wouldn't suck.
-
I asked other members of the family not to use the servers and I was able to watch a full film tonight with the 3090, 555.99 in 4K119 10bits without any dropouts or frame drops. I'll keep testing over the next couple of days but at this stage it looks like the issue is NOT the GPU.
-
That would be odd cause I live alone and have 2.5gb ethernet everywhere. The HTPC sits next to the router which is directly connected to the NAS that sits below it. I can't imagine it's a network issue.
-
That would be odd cause I live alone and have 2.5gb ethernet everywhere. The HTPC sits next to the router which is directly connected to the NAS that sits below it. I can't imagine it's a network issue.
Well, in my case it's a bit more complex because my 4k content is on an Unraid server, which is also hosting a win10 VM that's the Mymovies server and the CCC (part of CMC) server, and the 4k server also hosts the Dune metadata used to display the Dune front end. So if someone tries to access the Dune metadata from a Dune player somewhere, it's entirely possible that it could disrupt the unraid server long enough to create a freeze/catchup. This wouldn't explain why I don't have the issue in 8bit, but that might have been a coincidence too.
I'm going to keep the family ban for a while, then I'll see if I can cause drops when they access the server. Then I'll try moving the Dune metadata to the other server (the one that holds my 2K content) and I'll see if it makes a difference.
I didn't think it could be server related because I did lots of tests initially with the films hosted on a local SSDs to rule out network storage issues, but that was a long time ago, before Hendrik improved things in LAV. So I think my drops/freezes might be back for a different reason/cause. I don't have 2.4G but I have wired 1G and I don't think it's a bandwidth issue.
Of course, if you're on your own, then that can't be the reason for you. I'll keep you posted. Still, it might be worth playing content from a local SSD to see if you still get drops or not.
-
Well, in my case it's a bit more complex because my 4k content is on an Unraid server, which is also hosting a win10 VM that's the Mymovies server and the CCC (part of CMC) server, and the 4k server also hosts the Dune metadata used to display the Dune front end. So if someone tries to access the Dune metadata from a Dune player somewhere, it's entirely possible that it could disrupt the unraid server long enough to create a freeze/catchup. This wouldn't explain why I don't have the issue in 8bit, but that might have been a coincidence too.
I'm going to keep the family ban for a while, then I'll see if I can cause drops when they access the server. Then I'll try moving the Dune metadata to the other server (the one that holds my 2K content) and I'll see if it makes a difference.
I didn't think it could be server related because I did lots of tests initially with the films hosted on a local SSDs tu rule out networ storage issues, but that was a long time ago, before Hendrik improved things in LAV. So I think my drops/freezes might be back for a different reason/cause. I don't have 2.4G but I have wired 1G and I don't think it's a bandwidth isue.
Of course, if you're on your own, then that can't be the reason for you. I'l keep you posted. Still, it might be worth playing content from a local SSD to see if you still get drops or not.
Yea I can certainly run some tests locally on the 3080 machine. It has some slower ssd's in it but they should be plenty fast enough for 4K MKV playback. I might have some time tomorrow night to watch a movie and see what happens.
-
I'm going to keep the family ban for a while, then I'll see if I can cause drops when they access the server. Then I'll try moving the Dune metadata to the other server (the one that holds my 2K content) and I'll see if it makes a difference.
Maybe one of your kids has figured out how to manage your router for you.
https://routerfaqs.com/how-to-give-priority-to-a-device-on-a-router/
-
Maybe one of your kids has figured out how to manage your router for you.
https://routerfaqs.com/how-to-give-priority-to-a-device-on-a-router/
As I said, not a bandwidth/priority issue, my kids only have a slow wifi connection to the router, I have the 1gigabyte access. The issue is (possibly) when a client tries to access info on the server, the server might just glitch while it tries to serve the info. They are not even streaming video at that point, just getting the metadata info for the Dune.
Anyway, I first need to confirm that this is the issue, then I will think of ways to remedy it, which might simply be to move the Dune metadata to the 2K server, so that they don't access the 4K server at all.
-
The speed of the connection is just one factor. And nominal speeds and actual speeds are different.
Network topology matters. NAS issues. Electrical interference. Connections to Internet servers. And so on.
This thread is filled with such problems: https://yabb.jriver.com/interact/index.php/topic,24031.0.html
-
The speed of the connection is just one factor. And nominal speeds and actual speeds are different.
Network topology matters. NAS issues. Electrical interference. Connections to Internet servers. And so on.
This thread is filled with such problems: https://yabb.jriver.com/interact/index.php/topic,24031.0.html
I'm not disputing any of this, this is very basic stuff. I just like to do things one step at a time:
1) Ascertain whether the issue is caused by other clients network acces or not
2) If network, identify the exact cause.
3) Take steps to mitigate/resolve this.
I have no interest in step 2 or 3 at this stage, network is just one possibility that I need to rule out or confirm first.
-
Well, troubleshooting this is easy. Just copy your movies to the local storage and disconnect from your network.
Yeah, thanks guys for all this unsollicited advice, but I don't need any help for this network testing, really. I know what I want to test, in which order, and how. I suggested the local SSD test to @SamuriHL, so it's not as if I wasn't aware that it's an obvious first step when/if needed.
The only thing that would have been helpful from you would have been detailed info about your set-up so that we could figure out why it works with your 4080, but I've asked for this every time you post in one of these threads and you never provide the full info, so I guess there must be a reason. At this stage at least here I don't think a 4080 makes any difference vs a 3090. Might be different for you or SamuriHL though. I need to do more testing.
-
I plan on testing that (local ssd) tonight on my 11700 Intel machine with the 3080. This same 3080 has given me dropouts in audio pretty consistently since I bought it. Manni, one thing that we do have in common is that my machine that I normally run the 3080 in is a 5900x on an ASUS 570e Gaming Wifi II board (that is now dead). I wonder if some combination of AMD and nVidia might be contributing to this issue?
-
I plan on testing that (local ssd) tonight on my 11700 Intel machine with the 3080. This same 3080 has given me dropouts in audio pretty consistently since I bought it. Manni, one thing that we do have in common is that my machine that I normally run the 3080 in is a 5900x on an ASUS 570e Gaming Wifi II board (that is now dead). I wonder if some combination of AMD and nVidia might be contributing to this issue?
That's something that crossed my mind, that's why I asked everyone contributing to this thread (those who confirm they have the same issue, as well as those who say that they don't) to post detailed info about that setup. That's one way to find out if there is a common failure point. It could also be the HDMI 2.1 AVR (or VRROOM if used before the AVR to route audio only to a non-HDMI 2.1 AVR, as you do).
-
That's something that crossed my mind, that's why I asked everyone contributing to this thread (those who confirm they have the same issue, as well as those who say that they don't) to post detailed info about that setup. That's one way to find out if there is a common failure point. It could also be the HDMI 2.1 AVR (or VRROOM if used before the AVR, as you do).
Or a combination of all the above which is what makes testing for it especially difficult. My AVR isn't 2.1 but everything is directly connected to the VRROOM which is 2.1. I know you have your VRROOM after the AVR in your chain. if when I test tonight I get no audio drops local, I'll have to test again off my NAS later this week given that I'm changing 2 variables at once. I don't love the idea of going back to an Intel platform. I'm really looking at the new Zen 5 CPUs next month with an 870 board as my upgrade path. That'll be a darn nice machine with that 4080. But not if I have audio drops....
-
Or a combination of all the above which is what makes testing for it especially difficult. My AVR isn't 2.1 but everything is directly connected to the VRROOM which is 2.1. I know you have your VRROOM after the AVR in your chain. if when I test tonight I get no audio drops local, I'll have to test again off my NAS later this week given that I'm changing 2 variables at once. I don't love the idea of going back to an Intel platform. I'm really looking at the new Zen 5 CPUs next month with an 870 board as my upgrade path. That'll be a darn nice machine with that 4080. But not if I have audio drops....
Yes, that's why I asked everyone to post ALL their details in a single post, as I did. No one has.
-
Yes, that's why I asked everyone to post ALL their details in a single post, as I did. No one has.
They just never mind the boss. ;)
-
Well it's important in this case. Manni's been trying to track down the issue with this for literally MONTHS now. The only way to narrow this down is to know all the variables involved.
-
Well it's important in this case. Manni's been trying to track down the issue with this for literally MONTHS now. The only way to narrow this down is to know all the variables involved.
What's made it harder is that a few software factors have changed over that period of time:
1) Hendrik did an update in LAV that really helped and the first "kind" of Atmos dropouts" (not long freezes, just Atmos sound drop outs) disappeared.
2) Then MS broke something in an update, which caused long freezes, and it looked like it was fixed at some point.
3) Suddenly the long freezes popped back up here, but this could be because I've given access to the network to the rest of the family, so that's what I'm going to rule out first.
That's why I keep asking details both on the hardware side and on the software side.
As soon as you've been able to test one config, please post all the details for it. Also test at least 4-5 films because it can take that long for the issue to pop up. That's why I kept testing after the first few encouraging results with the 4080, and why I tested FULL FILMS, each of them as long as possible. This issue can take a long time before it happens during playback, so testing 30-60min means nothing. I also think that if the Atmos track is anemic (as with older films), the issue is less likely to appear. It's more likely to appear in John Wick than in an old mono or stero movie that happens to have been remixed in Atmos.
-
Ah, man, I HAVE to watch John Wick...AGAIN? My heart is broken. LMAO :D Yea, I can't test 4 to 5 movies today but for me this issue was happening quite consistently in almost every movie which is why it was driving me nuts. If it was rarely happening I wouldn't care as much. The thing that's frustrating about this is that with my Ugoos, it NEVER happens. And that's going through the exact same chain as my HTPC...i.e. directly connected to the VRROM, audio to my AVR through the VRROM HDMI audio out port, video directly to the G2 with audio muted. So for me it doesn't seem like a network/nas issue given that the Ugoos is doing the exact same playback as my HTPC. In any case I'll see what happens tonight. The machine I'll be testing on is a prebuilt ASUS Intel 11700 machine. I dropped my corsair 32 gigs of DDR4 3200 in there from my HTPC since it'll never be used again in that machine. It's got a western digital 512 gig nvme SSD for boot drive and I dropped a 1tb SATA SSD as a second drive in the thing. I'll be playing off the 1tb SATA SSD tonight when I test. And obviously my ASUS 3080 TUF card.
-
Ah, man, I HAVE to watch John Wick...AGAIN? My heart is broken. LMAO :D Yea, I can't test 4 to 5 movies today but for me this issue was happening quite consistently in almost every movie which is why it was driving me nuts. If it was rarely happening I wouldn't care as much. The thing that's frustrating about this is that with my Ugoos, it NEVER happens. And that's going through the exact same chain as my HTPC...i.e. directly connected to the VRROM, audio to my AVR through the VRROM HDMI audio out port, video directly to the G2 with audio muted. So for me it doesn't seem like a network/nas issue given that the Ugoos is doing the exact same playback as my HTPC. In any case I'll see what happens tonight. The machine I'll be testing on is a prebuilt ASUS Intel 11700 machine. I dropped my corsair 32 gigs of DDR4 3200 in there from my HTPC since it'll never be used again in that machine. It's got a western digital 512 gig nvme SSD for boot drive and I dropped a 1tb SATA SSD as a second drive in the thing. I'll be playing off the 1tb SATA SSD tonight when I test. And obviously my ASUS 3080 TUF card.
Sounds good... Yeah I watch John Wick 3 ten times doing these tests, I had to pick other films after that, all I'm saying is that I've played very long OLD films thinking I was testing because they were Atmos tracks, and I don't think it's testing anything if the Atmos track doesn't need a significant bandwidth. At least with John Wick 3 we know that the Atmos track is lively :)
I never used to have ANY dropout with the Oppo 203 or the Dune Pro Vision 4K Solo either playing Atmos 4K bluray, or the ATV 4K 2021 / FireTV Stick 4K Max when streaming Atmos titles, but I did get ONE Atmos dropout when playing the Oppo 203 recently, so that's one of the reasons why I'm wondering if these new dropouts might be network related. At least it's something I want to rule out here before doing anything else.
I am also getting these drop outs all the time, at least one per title on the HTPC, which is why I'm trying to resolve this. I would still want to fix this if it was once in a while, because one drop out is enough to ruin a film if you're watching it for the first time. I'm only using the Oppo or the Dune to watch a film for the first time (if I care about it) because of these Atmos dropouts. All the tests I'm doing are with films I've already watched or with new films I don't care about (I've watched a LOT of crappy films recently on the HTPC to test this).
-
I don't tend to buy movies I don't care about. I've got a good 600+ UHD's in my collection now with more on the way this week. I have Ghost Busters Frozen Empire coming today which is probably another good candidate tbh. The Ugoos as we've discussed, I tend to use strictly for Dolby Vision titles although when I want to watch an HDR movie without audio issues I sometimes use it for that, as well. Would love to finally solve this HTPC issue once and for all.
-
I don't tend to buy movies I don't care about. I've got a good 600+ UHD's in my collection now with more on the way this week. I have Ghost Busters Frozen Empire coming today which is probably another good candidate tbh. The Ugoos as we've discussed, I tend to use strictly for Dolby Vision titles although when I want to watch an HDR movie without audio issues I sometimes use it for that, as well. Would love to finally solve this HTPC issue once and for all.
Yeah, believe it or not, I have to watch films (and series) for work, and a lot of them I don't care about. I don't buy them all though, I often rent them (for as long as that's possible, I'm sure at some point they'll stop renting physical discs) if they are not sent to me.
-
I've sent the 4080 back today for a refund. I can't justify paying £953 if it doesn't fully solve the problem. The 3090 is still a very decent card that does all I need provided I find a way to fix this Atmos droupout issue. Even if I don't, I can use the Oppo 203 and the Dune player instead for perfect video and audio when needed, so that's my plan B.
-
I've decided to keep my 4080 and build a new machine around it. I wasn't REALLY planning on upgrading the HTPC this year but hey why not. LOL The only reason is this gives me another machine I can put in my bedroom for HTPC and gaming duties which is nice to have. If I didn't have the second machine, I'd most definitely be returning the 4080 as well.
-
With the new AMD Ryzen 9000 inbound you can probably get killer deals on very good Ryzen 7000 CPUs so it may not be too bad of a time to be building a new machine. :P
-
Oh I'm definitely going for the 9000 Zen 5 at this point. When I build new machines I tend to go all out so they last a few years. This one not withstanding as it's only a couple years old unfortunately. Once I rebuild it I won't be touching it again for 4 or so years so I want the latest specs. Wifi 7 is definitely desired as I'm running out of 2.5gb ethernet ports and I've already got a wifi 7 router.
I'm debating between the AMD Ryzen 9 9900X and the AMD Ryzen 7 9700X. The 9950x is going to be insanely priced. The 9900x maybe a good place to be though.
-
I was able to test a second film tonight with the 3090 (and family network bar still in place). Zero frame drop, zero Atmos drop/freeze. Encouraging. I'll try to run at least one more test tomorrow, then I'll be away for a couple of days for work, but I'll resume testing on Thursday or Friday. If I manage to get five films without any drop, I'll start isolating the network cause. Right I suspect that the daughter's Dune is hogging the network briefly when she accesses the metadata, but it could be something else. I'll keep testing and will report back when/if I get anything conclusive.
-
I'm watching JW3 right now off the NAS (I didn't want to change more than one variable at a time and I'm already running this on a whole new machine). So far, half way through, no drops at all. Usually I have at least one by now.
-
Maybe you guys can increase the cache. When I use MPC-BE with MadVR, I use 4G cache.
-
Off the NAS on the ASUS machine...not one single drop. I don't remember the last time I was able to watch a movie on an HTPC without a single drop. The only change for me in my setup was going from the AMD 5900x on a 570e board to the Intel i7-1100kf on a ROG Strix G15CE. (The memory sticks are the same as I used in my AMD machine) I'm thoroughly amazed. Obviously far more testing will be required. But I've not been able to do a movie without a single frame drop in forever.
-
Off the NAS on the ASUS machine...not one single drop. I don't remember the last time I was able to watch a movie on an HTPC without a single drop. The only change for me in my setup was going from the AMD 5900x on a 570e board to the Intel i7-1100kf on a ROG Strix G15CE. (The memory sticks are the same as I used in my AMD machine) I'm thoroughly amazed. Obviously far more testing will be required. But I've not been able to do a movie without a single frame drop in forever.
Glad something positive is finally coming out of your ordeal, and it confirms that a 4xxx isn’t necessary to get a drop-free experience, though I’d watch at least 3-4 more films before drawing a conclusion. It’s happened to me quite often to have no drops for a while, then the drops would come back. But as I said, there has been various software factors at play, which explains why things have changed. Intel vs AMD might be the explanation your side, but it’s not here as I’m still using the exact same hardware.
-
I'll get back to it eventually. My 2 month old Alienware 4080 laptop just decided to no longer boot. So I'm dealing with a full windows reinstall at the moment. All my calibration software, davinci, etc all have to be reinstalled. I'm....."unhappy" right now...
-
Watching a movie in DTS-HD MA 5.1 and I got 2 drops. That's frustrating.
-
Not sure if this is helpful, but I finally was able to experience multiple excessive video/audio issues playing the UHD full disc backup of The Hobbit Extended - The Desolation of Smaug via UHD menu playback. I am bit streaming the audio and this is served off an SMB share.
At various times - and it's not consistent on any part of the movie, the frame drop jumps and the repeated frames jump. Causes stuttering for 3-5 seconds and then the movie resumes normally until the next event. CPU/GPU/Network are all normal. Other movies - like the first Hobbit Movie - no issues. This movie is a mess.
Playback of the full rip via SMB to the Oppo 203 works perfectly - no issues.
If I remux the video into a single MKV file, no issues
If I create a particle to 0800.mpls (the movie), no issues
It seems that when I playback via full UHD/BluRay Menu, issues.
Setting bit streaming to off, still issues
Don't know what is different with the full menu playback, but for now I'll just particles for titles that experience this.
-
All right, I wanted to update this thread because I did an in-depth debugging session with the 3090 and I *think* I've found a third compromise, which means no compromise on audio or video, but one on gaming... Here is what I did:
- I tried a clean install of Win10 x64 on a spare SSD (dual boot), with just MC installed. Got dropouts. This ruled out the software stack.
- I tried to take the VRROOM out of the chain, so I only had HTPC > X8500HA > TV. Got dropouts. This ruled out the hardware chain (besides the 3090 itself).
- I got a long audio+video freeze and realised that JRiver had just downloaded a new build while I was playing a film. I disabled auto-update and I've never had a long freeze/drop again (that was 10-15 films ago). If this is confirmed, I would call this a bug, because I don't see any reason for MC to do anything that could compromise playback. Could a dev confirm that MC only checks for a new build at launch, and not during playback? If it's possible during playback, could this be fixed?
Now, turning auto-updates off seems to have resolved the audio+video long freeze.
This means a few things:
1. The long audio+video freeze isn't specific to Atmos or bitstreaming, I experienced it (before disabling JRiver auto-update) on a DTS title using LPCM.
2. The 4xxx was most likely a fix for the micro audio drops, as what I remember was that I got long audio+video freezes, not micro audio drops. The possible network issue was likely a red herring. So if you have Atmos micro audio drops with a 3xxx GPU at 4K120, consider the 4xxx as a likely fix.
The 8-bit workaround mentioned at the beginning of the thread should have led me to test another workaround, which I used to use a while ago and forgot about, which is to have the HTPC on a HDMI 2.0 input of the AVR. This forces TDMS, and allows to have 4K23 in 10/12bits without any Atmos drops. You’re only limited to 8bits at 4K60.
This used to be an issue with some drivers newer than 532.03, as you would get a black screen when a HDMI 2.1 GPU was used that way, but it was fixed in 551.23.
Note that the output doesn't matter (both of the outputs of the X8500HA are HDMI 2.1 and I can use either without any issue).
I've now watched six full films with this set up (3090 >> HDMI 2.0 input >> TV) and have experienced no Atmos drops.
I'm going to do more checks to see if I can enable 8K Enhanced instead of 4K Enhanced in the AVR, and still get no audio drops.
I'm also going to see if I can bring the HTCP back to the HDMI 2.1 input and simply use a HDMI 2.0 EDID when I use it for video tasks (hence limiting it to a TDMS bandwidth) and then a full HDMI 2.1 EDID when I want to game at 4K120. But this will take a while and I wanted to share my third workaround right now, so others can test and see if it works for them too.
I might in fact keep the HTPC in a HDMI 2.0 input, as I can still game at 4K60 and the big advantage of doing this is I can use my Sony HDMI Headphones with the HTPC without having to put another VRROOM between the PC and the VAR to get the audio only to the Sony headphones.
One last thing, for those using a VRROOM, I kept having black screens with the latest fw, so I reverted to 0.34 (which happens to be the last fw with working JVC Macros) and the black screens seem to be gone.
So to summarise, here is my current workaround:
3090 HTPC with latest nVidia drivers (currently 650.81 studio) >> HDMI 2.0 input of the X8500HA >> VRROOM with fw 0.34 >> TV (Samsung S90C).
This gives me no audio compromise (Atmos + DTS:X), no micro Atmos audio drops, 4K23 at 10/12bits, no long video+audio freeze (auto-updates disabled in JRiver MC).
Main dowside is that I lose 4K120 for gaming, but I can live with that, at least until I get a 4xxx or 5xxx next year.
As always with HTPCs, YMMV
Good luck!
-
What happens if you set the VRROOM to matrix frl->tmds (mode 3)? Does that solve the problem?
-
What happens if you set the VRROOM to matrix frl->tmds (mode 3)? Does that solve the problem?
It depends where the VRROOM is in the chain, and on which output the TV is. I'm not going to do any of these tests for now, as I'm quite happy to have the 3090 going to a HDMI 2.0 input (it allows me to use my Sony HDMI headphone without having a second VRROOM in the chain).
However, given that setting the driver to 8bits (which forces the bandwidth to TDMS) resolves the micro-audio-drops issues, I would say that anything that forces a TDMS bandwidth on the GPU output should work:
- Setting the driver to 8bits
- Using a HDMI 2.0 switch
- Setting the AVR to 4K Enhanced instead of 8K Enhanced
- Using a 4K60 8bits EDID
Etc...
I have only validated using a 4K (HDMI 2.0) input in the AVR at this stage.
Having spent a week on this already, I need to get back to work before I don any further testing re possible other workarounds for the 3xxx ampere bug.
-
My HTPC is currently off at the moment. If I have time I'll maybe mess around with it this week. I have the 3080 in it as I don't have a way to throw the 4080 in yet. At some point, maybe by Christmas, I hope to get a new mobo and CPU to rebuild my proper HTPC and put the 4080 in it.
-
My HTPC is currently off at the moment. If I have time I'll maybe mess around with it this week. I have the 3080 in it as I don't have a way to throw the 4080 in yet. At some point, maybe by Christmas, I hope to get a new mobo and CPU to rebuild my proper HTPC and put the 4080 in it.
Thinking about it using the TDMS downscale on the VRROOM will most likely not work, as the whole point of it is to allows the source to send the full-fat 8K signal to the VRROOM so that you can pass it through to TX0 and so that it can be downscaled for a secondary display on TX1.
What you need is for the GPU to be limited to TDMS, so that it doesn't use FRL5/6 at all. For this, you need a HDMI 2.0 switch, a HDMI 2.0 EDID or the AVR to limit the bandwidth to HDMI 2.0 (if you're going through the AVR, I don't think you are).
-
I'm definitely not going through my AVR, no. I suppose I could force it with an EDID in the VRROOM. I'd rather just get my new system built but money and the fact that the mobos and cpu I want aren't out yet.
-
I worked around this problem by using both a high quality HDMI 2.1 cable and then setting GPU clocks so that they don't downclock all the way.
I noticed this as when I was gaming (clocks all running high), the dropouts didn't happen. Watching video though (clocks low, minimal GPU requirement), the problem was frequent.
By setting my memory clock minimum to 810 instead of 405, this runs the card on an extra 4W of power, and generates an additional 1C of heat, so negligible impact, but appears to have fixed the problem.
Depending on your GPU, the numbers and ranges may differ so some experimentation is likely needed if anyone’s reading this with the same problem, but the method’s pretty simple :
— Determine compatible clock speeds
Open an admin command prompt and type this command:
nvidia-smi -q -d SUPPORTED_CLOCKS
This’ll give you a long list of memory MHz speeds, and within each of those, a range of compatible GPU clock speeds. Ignore the batch of timings at the bottom of the list – those are the slowest downclocks which we don’t want as they cause the problem. For me, next group up is the batch for 810MHz.
Make a note of the memory clock batch number (810 in this case). This is your minimum memory clock.
Now make a note of the minimum gpu clock speed within that batch (it’ll be at the bottom. For me it’s 210).
So that’s your minimums. Now for the maximums.
Go to the top of the list for the fastest supported values. For me this is a memory clock of 10501MHz and a Graphics clock of 3240MHz.
Now you’ve got enough for these commands (which if you like you can also put into a BAT file and run as part of your startup routine).
The -lgc flag controls min,max values for graphics clock, and the -lmc for memory clock.
nvidia-smi -lgc 210,3240
nvidia-smi -lmc 810,10501
This will allow the GPU to ramp up clock speeds when it thinks it needs it, but it’ll stop it downclocking below the threshold required to supply steady signal through the HDMI.
Hope this is helpful!
-
I worked around this problem by using both a high quality HDMI 2.1 cable and then setting GPU clocks so that they don't downclock all the way.
I noticed this as when I was gaming (clocks all running high), the dropouts didn't happen. Watching video though (clocks low, minimal GPU requirement), the problem was frequent.
...
Thanks, that's very helpful and I might give it a try when I'm back from holiday in two weeks.
However, I can't see how the GPU could behave differently when using a HDMI 2.0 input and a HDMI 2.1 input, when both are playing 4K23 10bits RGB content within TDMS bandwidth, and I get drops with the HDMI 2.1 input and no drops with the HDMI 2.0 input.
Please could you kindly provide all the information listed in the first post, so that we can know more about your setup?
Also, surely this isn't relevant when using max performance power in nVidia, and I still get drops with that setting.