Hi Jim, no need to remove it (I can't do so easily due to cabling etc).
Anyway, as I suspected hardware is not the problem (I don't see how it could be when there are no ATMOS drop outs with MCP-BE or MPC-HC)
Here is what I did:
Preparation:Deregistered madVR / LAV
Uninstalled MPC-BE, MPC-HC, JRiver (complete uninstall, including library), LAV filters
Clean up registry, program files, program data, appdata/roaming etc. of any reference to the above.
Moved all test titles to cache SDD on the Unraid server to rule out any HDD glitch.
Bandwidth reported by UnRAID during playback between 8 MB/s and 13 MB/s, average around 10 MB/s. Can't be taken literally of course as UnRAID stats are not real-time or exact, but that's all normal and as expected.
Disabled all scheduled tasks in Windows
InstallationInstall JRiver MC 31 build 38 with default settings (RO JRVR).
Disabled background library import.
Hardware:ASRock X570 Creator, AMD 5950x 3.4Ghz, 32GB RAM, Zotac 3090 24GB (BUS interface confirmed PCIE 4.0 x16)
Software:Windows 11 x64 Pro 22H2 build 22621.1992 up to date.
nVidia Latest studio drivers v536.40
3D settings: Vertical sync: on, power management mode: max performance, prefered refresh rate application controlled
GSync: disabled
Desktop color settings set to reference
Video settings set to use the player
Desktop resolution set to UHD (native) 4K 23p 10bits RGB Full
Test title:John Wick 3 4K UHD Blu-ray, 1:1 rip from original disk to BD Folders, moved to SSD for testing
During playback (all tests / all video chains) JRiver reports 3840x2160 at 23.976hz VSYNC measured 23.976hz Overall 0.5ms avg / 0.5ms peak, so I think we have some rendering headroom
All good up to that point.
Playback chain for the first two tests:HTPC 3090 HDMI 2.1 Out > Denon X8500HA HDMI. 2.1 In > HD Fury VROOM HDMI 2.1 Out 2 > EZKOO HDMI 2.1 KVM > LG 4K Monitor HDMI 2.0 (supports 23.976hz refresh rate, single / primary monitor)
At the beginning of each test, I enable the JRVR OSD display (CTRL-J), reset the stats after a few secs to discard normal/expected initial frame drops (CTRL-R), and make sure I have 0/0 report at start of playback after the first ten seconds of playback. If it still drops frames after ten seconds of playblack, I don't reset. Obviously I don't touch the mouse or keyboard or pause or do anything during the test. Just play and watch the stats.
1) First test run with no change in JRiver settings at all (so LPCM audio, no bitstreaming, no bd menus) to get a baseline on the LG 4K Monitor (Audio PASS, Video PASS)Playback for 25 minutes (LPCM, no BD menus)
Results: 0 frame drop / repeat, no audio drop out (multi in) There are at least three known scenes where audio dropouts happen. Usually by that time I would have got at least one big audio dropout in JRiver and many frame drops. None with these baseline settings.
Stopped playback
Enabled HDMI bitstreaming to get ATMOS, changed default device set to Denon WASAPI
2) Second test run to establish baseline with bistreaming and ATMOS on the LG 4K (Audio PASS, video FAIL)Playback for 25 minutes (Bitstream ATMOS, no BD menus)
Results: 2 frame drops, 1 frame repeated, no audio drop out At the beginning, a few drops happened, I reset the stats a few times, then one happened and I left in on. This was in the first 30 secs of playback. No more frame drops after that in 15 minutes, then another one happened. Then a frame repeat about 20 minutes in.
So while there are no Atmos audio drop outs, the video performance is not perfect with 2 drops and 1 repeated frame. I don't really care about A/V performance on this monitor, I only use it for testing to save laser/bulb time, and the EZKOO KVM adds complexity that might cause these frame drops. So at this point I moved to the JVC. For these tests (audio drop outs with Atmos), it's still a PASS on the LG 4K.
Stopped playback, switched LG Monitor off
Changed the video chain to the PJ, which is what I use for normal playback (JVC NZ8, using a custom EDID in the VROOM to limit bandwidth to 4K120 4:2:0 and prevent FRL, to rule out one possible complexity, so only TDMS bandwidth HDMI 2.0 4K60 4:4:4 18gbs max) New Playback chain: HTPC 3090 HDMI 2.1 > Denon X8500HA HDMI. 2.1 > HD Fury VROOM HDMI 2.1 Out 1 > JVC NZ8 (single / primary monitor) This is my normal playback chain, the one I care about for video and audio performance. It's optimised and works flawlessly re Atmos bitstream with all other software (both MPC-BE and MPC-HC).
3) Third test run to establish baseline with bitstreaming and ATMOS on the JVC (Audio PASS, Video PASS) Playback for 25 minutes (bitstream ATMOS, no BD Menus)Results: 0 frame drops, 0 frame repeated, no audio drop out This confirms that the frame drop/repeat with Atmos/bitstream on the LG was specific to the LG 4K and that video chain.
Now that we have the Bitstream / Atmos validated with the JVC, we know that whatever causes Atmos drop outs in my set-up with JRiver is due to some JRiver setting, unless it was solved by the cleanup/uninstall/reinstallation.
I'm going to go back gradually to where I was settings-wise, checking playback after each change, to try to establish what causes the drop out.
At the very least these first tests rule out a hardware issue.Potential causes:BD menus
HDR / Scaling / Rendering settings
Conflict with other software that have been uninstalled.
I'll only introduce other software in the mix if I manage to get back to my full JRVR settings and still have no ATMOS dropout issues.
Stay tuned!
At least we know that after a full clean / reinstall and with default rendering / processing settings there doesn't seem to be any Atmos dropouts or frame/drop repeats here and no hardware incompatibility.
That's good news!