System for watching 3d movies on OLED displays that support BFI (black frame insertion at 120 hz)

[peter64]

Hey Guys,

So I'm looking for beta testers and volunteers to help with a system I'm working on I have it working and tested with 2 TV sets already
LG OLED C1 and Sony XBR 48a9s the C1 has zero ghosting/crosstalk whilst the Sony has about 5% at top and bottom of screen, but it's fine for letterbox content.
I need one interested beta tester with a LG CX OLED to help with testing, if you have an LCD TV with 120hz that supports pulsed backlight provide details and you may be eligible for a free testing unit still too.

TV:
I am looking for anyone who wants to help test the unit with LG OLED CX, or other Sony OLEDs from 2019-2021
Your TV must support strong BFI at 120hz, by that I mean it must insert a dark frame for a large portion of the time (preferably as close to or greater than 50% of the time).
If your TV BFI duty cycle is not close enough to 50% you will experience cross talk / ghosting, but it may still be ok for watching letterbox content (with black bars at top and bottom).
I have confirmed that unfortunately the LG C2/G2 and LG C3/G3 OLED's will not work as they have removed 120hz BFI support.
There may be a way to hack it back in with firmware hacks, but that is beyond the scope of this project. If anyone is interested and able to obtain the required data I'm happy to try and locate the firmware changes required.

3D Glasses:
You will also need to own a pair of 3D glasses as I will not be shipping them with the unit.
The unit should support Sony TDG-BRXXX glasses, Panasonic TY-EW3DXXX glasses, 3D Vision glasses.
I may be able to add support for any other 3D glasses that include 4 token signalling (eg open right, close right, open left, close left) if anyone wants to supply a pair of such glasses for testing and analysis.
Additionally I have a custom designed firmware for some versions of the Panasonic glasses that totally eliminates resync issues reducing resync delay from 5 frames to 0 frames.
I may consider making a custom firmware for the sony glasses too, but they are harder to open and program. The pansonic glasses can be programmed through the USB port.

Computer requirements:
Now the best part, this system uses software page flipping in open gl directly so it doesn't require 3d vision or quad buffer opengl.
I have tested it in Windows and Linux.
It works extremely well on debian based linux distros like Ubuntu and you can easily run it from a bootable live usb drive. (occassional glasses resync that appears like you are blinking once every 5 mins or so when working properly).
It is also working very well on windows 10 with very few dropped frames.
There are performance issues on windows 11 with lots of dropped frames.
A cheap laptop with AMD integrated graphics is sufficiently fast to perform the software page flipping at 120hz 1080p.
In theory one could use this system for 4k 120hz oled 3d output, however I don't have a video card capable of outputting 120hz 4k.
There is an issue I ran into with newer windows systems on AMD where it is not possible to disable ALLM (adaptive low latency mode) from the control panel.
There is a workaround for this but it requires using CRU tool to override the OLED's HDMI config to hide ALLM support from the computer. I can provide direction on how to do this if required.
This is not an issue with Linux as far as I can tell, and I don't think this is an issue for nvidia or intel either.
I hope to resolve the performance / dropped frame issue with windows 11 if there is demand.

Hardware:
You will need to mount a tiny device on the top left of your TV screen, it will monitor signals embedded into the video feed that trigger the glasses.
I have done my best to try to alieviate burn in concerns by using soft edges and allowing configuration of the brightness for the embedded signal.
I have 3 devices I will be willing to ship out to beta testers (for their feedback and help) if they cover the cost of shipping via air mail.
If you have any python programming ability and are interested in adding functionality to the software that is a plus

Released so far:
https://github.com/open3doled/open-3d-oled
1) Software for PC to play videos using software page flipping whilst embedding signals for tv mounted sensors.
2) Software for tv mounted sensor, based on atmega32u4 micro controller
3) Hardware design for the tv mounted sensor including Gerber, BOM, etc.
4) Parametric 3D openscad designs for mounts to attach the tv sensors to various shapes and thickness of TV.

Future plans:
1) Release some youtube tutorials and demonstration videos (late February).
2) Put up a webstore where I will sell kits including the hardware and glasses with modified firmware optimized for software pagefilipping. (mid March)
3) Release open source firmware to replace that on sony and pansonic glasses to support improved interaction with software page flipping (dropped frames, and OLED update jitter). (date as of yet not determined)

Background:
I had been using a 3D DLP projector with 3d vision to watch movies but 3d vision isn't supported by new cards and if I'm not mistaken intel and amd also have offerings which are not supported any longer.
I also was using a 3D VIP hardware adapter for a while to convert SBS into 120 hz page flipped, but it wasn't working well with my new OLED TV.
So I've come up with the solution presented above.

If anyone has any questions let me know.

[savannah]

What specific feedback or assistance are you looking for from beta testers, especially those with Python programming ability, and how can they contribute to improving the functionality of your system? build now gg

[peter64]

Hey so as I mentioned above,

1) I'm looking to understand what models of TV's it works on. So far I know it works on LG OLED C1 and Sony XBR 48a9s perfectly, but only with the strongest BFI setting. Since BFI functions differently on different OLED TV's both between generation and between manufacturers I'm looking to get the precise timing settings that give the best results on different TV's. The timing parameters are also related to the glasses you are using I have tested with 3d vision first gen, panasonic ir glasses 1-3 gen, and sony glasses (out of sony glasses it works best on the latest TDG-750 variant, Also kind of works with earlier ones too, but earlier ones don't fully support the 4 token IR protocol they claim to support) and have different profiles for each that work well with my LG C1 OLED.
2) I'm looking for bug testing the open source python software to try and improve performance on certain systems and also to add functionality that people may find useful (or allow others to add said functionality).

Right now the software has native SRT stereo subtitle support with depth settings, and afew other stereo specific features, FullSBS, HalfSBS, FullOverUnder, HalfOverUnder, Flip left and right eyes, etc....
It supports any video format that gstreamer gst-play can parse properly on the target system but gstreamer sometimes has issues with mkv matroska packed files with 6-10 Multilanguage DTS5.1 streams so you sometimes need to strip out all the audio streams you don't want using ffmpeg. But I have some simple commands to do that.

Also if we can get this working well I think this will be a killer app for the TONS of people with LG and Sony OLED TV's even if only 0.1% want to watch 3D movies. It's basically a system that runs decently on almost any moderately specked PC at least under linux currently, (I'm considering to try to get it running adquately for 1080P on a rasperrypi 5 but unfortunately the PI doesn't have hardware H264 decoding so I'm not too hopeful it will work well) with no requirement for 3d vision or quad buffer opengl. And the hardware is an order of magnitude cheaper than VIP 3D or other FPGA based SBS to FrameSequantial systems.

[peter64]

Also another big plus is if you can run the software under both Windows and Linux as it's probably best to nail down optimal settings in Linux then if possible provide any help one can to improve the windows performance after.
As for providing contributions, as I mentioned above I plan to put all the necessary code on a github repo in the next few months with the open source license details.

[AmelieTurner]

Because it's usually better to find the sweet spot on Linux and then, if possible, offer any assistance one can to improve Windows performance afterward, being able to run the software under both Windows and Linux is also a huge plus.
Regarding making contributions, I intend to post the required code on a GitHub repository along with the specifications of the open source license in the coming months, as I indicated earlier.hill climb racing

[toedreadful]

As of right now, the software supports native SRT stereo subtitles with depth settings in addition to a few other stereo-specific capabilities like Flip left and right eyes, FullSBS, HalfSBS, FullOverUnder, and HalfOverUnder.
It supports any video format that gstreamer gst-play can correctly read on the target system. However, gstreamer can occasionally have problems with mkv matroska packed files that contain six or ten Multilanguage DTS5.1 streams; in these cases, you will need to use ffmpeg to remove all of the audio streams that you do not want. But I can accomplish that with a few easy instructions.

[laurawoods]

Because it's usually better to find the sweet spot on Linux and then, if possible, offer any assistance one can to improve Windows performance afterward, being able to run the software under both Windows and Linux is also a huge plus.
Regarding making contributions, I intend to post the required code on a GitHub repository along with the specifications of the open source license in the coming months, as I indicated earlier. Buckshot Roulette

I don't understand your explanation.

[brianallen]

The best way to find participants or beta testers may be through technology forums, technical communities or websites that specialize in TV, 3D and technology. You can also use social networks to share information and find interested people.

[peter64]

Because it's usually better to find the sweet spot on Linux and then, if possible, offer any assistance one can to improve Windows performance afterward, being able to run the software under both Windows and Linux is also a huge plus.
Regarding making contributions, I intend to post the required code on a GitHub repository along with the specifications of the open source license in the coming months, as I indicated earlier. Buckshot Roulette

I don't understand your explanation.

Thanks guys, so I've solved most of the performance issues I think, at least to my satisfaction for now. Probably another month or so and I'll have the github online.

To the person above who said they don't understand, it will probably be easier to understand when I have images and a full detailed blog and/or video showing how it works.
The main point is to confirm which TV's it will work with and generate timing parameters for those TV's for various types of glasses to get optimal performance.
This is akin to what the nvidia driver would do automatically and why they are only supported certain certified displays (because not all displays will work properly and those that will, will require proper timing parameters).
I have a step by step process one can go through to find the proper timing parameters for a given display though so it shouldn't be too bad.

Still looking for anyone interested. I'll be cross posting elsewhere too soon.

[3DNovice]

I added your original post to a more active sub-forum, where I started a thread about a product doing this very thing called Athanos.

I think it it would be huge if you could add Checkerboard support, then theoretically, 4K DLP Projectors that are using a two flash/pixel shift DMD could be used
for stereoscopic video playback as well.

It would be fantastic if you could make a plug-in for 3DMigoto as well, plus the fish tank viewing.
FishtankVR for The Witcher 3 viewtopic.php?t=23645

Also FWIW, at least one of the post above is spam, refer to
viewtopic.php?t=26680 Countering Forum Spam

example, use the quote feature for toedreadful, you will see a malicious link

Anyhow, I do think the addition of Checkerboard would be huge, just like Samsung and Mitsubishi TVs did stereoscopic 3D via wobulation using the Checkerboard format.

[Robert256]

Hi Peter,

This looks very much like a project I was planning, using a light sensor and digispark (ATTiny85), but never got the timings right, probably because it wasn't fast enough. Would love to contribute to your project! I mostly have Win32 C/C++ experience with OpenGL.

Best regards, Robert

[peter64]

Hi Peter,

This looks very much like a project I was planning, using a light sensor and digispark (ATTiny85), but never got the timings right, probably because it wasn't fast enough. Would love to contribute to your project! I mostly have Win32 C/C++ experience with OpenGL.

Best regards, Robert

Hey Robert,

Fantastic man! It sounds like you have soldering skills too! If you are interested I can send you a working detector board in the mail.
Please let me know what types of shutter glasses you have. It does work with off the shelf shutter glasses from Panasonic, Sony and 3D Vision.
Also let me know what type of TV you are planning to test on so I can confirm that it will most probably work, or at least has a chance of working.
I'm also keen for someone to test it using a high speed LCD panel with strobing backlight.
Compatible monitors would be those listed on the following page https://www.rtings.com/monitor/tests/mo ... -insertion with a "backlight strobe frequency picture" similar to this one https://i.rtings.com/assets/products/gM ... -large.jpg
I tried it on a regular monitor without strobing and the scrolling refresh made it unworkable, I suspect strobing might be sufficient to get it working with normal monitors.
As I said above any OLED with high BFI settings will work at least to some extent.

In terms of getting the timings right, I'm working on a somewhat simple process that will kind of guide you through the process of setting up the timings while you view ghosting and cross talk to get the best image.
I should be done that hopefully in the next week or so.
About half of my time was spent reverse engineering the Panasonic 3d glasses and the Sony 3d glasses so that I could re-write my own custom firmware.
This was really important because the OLED timing seems to drift a little bit and you need to make the glasses more responsive to changes in timing, than the default firmwares.
The default firmwares take about 5-10 frames to resynchronize onces they detect any drift (or dropped frames in the case of software page flipping), which makes them have a jitter effect at the top and bottom of the screen.
The custom firmware I worked on updates timings instantly and will never show the wrong picture to the wrong eye (at least not completely, sometimes it might start to show the wrong picture to the wrong eye, but it will detect the dropped frame immediately and correct itself.
It can cause a brightness spike on one eye for 1 frame though as the right eye will receive two frames in a row while the left eye will miss a frame for example.
it's not too bad though if it only happens once every 1-2 minutes as it's basically like blinking, and it shouldn't happen on faster PC's at all at least from my testing under linux.

Anyways the sensor I am stuck using is unfortunately the OPT101 (https://www.ti.com/lit/ds/symlink/opt101.pdf) sensor. I looked into digital sensors, specifically the OPT4001 (https://www.ti.com/lit/ds/symlink/opt4001.pdf) which has digital thresholds to trigger interrupts which would be ideal to offload all the ADC processing required, but unfortunately it's fastest conversion time is 600 us. Using the OPT101 with an ADC fast sampling I can get conversion times of around 40us which is 10x better at reducing jitter and improving responsiveness to dropped frames.
Given 120hz corresponds to 8.333 ms per frame, you can't really afford to have a jitter of 600us so the opt4001 was no good so I am stuck with the OPT101 which unfortunately costs $7-8 per unit and you need two for the device to work right.
The cheapest I found them was on aliexpress for $5 if you buy in bulk, but avoid anything under $5 as they are all scams (I had 3 different sellers sell me bad chips).

Anyways let me know your TV and your glasses type and I'll let you know then I can send you a board to help me test with and work on the software.

[peter64]

I added your original post to a more active sub-forum, where I started a thread about a product doing this very thing called Athanos.

I think it it would be huge if you could add Checkerboard support, then theoretically, 4K DLP Projectors that are using a two flash/pixel shift DMD could be used
for stereoscopic video playback as well.

It would be fantastic if you could make a plug-in for 3DMigoto as well, plus the fish tank viewing.
FishtankVR for The Witcher 3 viewtopic.php?t=23645

Also FWIW, at least one of the post above is spam, refer to
viewtopic.php?t=26680 Countering Forum Spam

example, use the quote feature for toedreadful, you will see a malicious link

Anyhow, I do think the addition of Checkerboard would be huge, just like Samsung and Mitsubishi TVs did stereoscopic 3D via wobulation using the Checkerboard format.

Thanks for adding it to another thread.
It may be possible to add checkerboard support for displays that support it as that would entirely eleviate any issue surrounded with dropped frames from software page flipping. As you would never end up needing to quickly resynchronize the glasses.
Also the Athanos stuff sounds interesting I had never heard of it and don't have time to look right now, but once I get this code up on github so it's in the public domain I might check it out. I want to make sure this is out and in the public domain before I look at anything related.

[Robert256]

Hey Robert,

Fantastic man! It sounds like you have soldering skills too! If you are interested I can send you a working detector board in the mail.
Please let me know what types of shutter glasses you have. It does work with off the shelf shutter glasses from Panasonic, Sony and 3D Vision.
Also let me know what type of TV you are planning to test on so I can confirm that it will most probably work, or at least has a chance of working.
I'm also keen for someone to test it using a high speed LCD panel with strobing backlight.
Compatible monitors would be those listed on the following page https://www.rtings.com/monitor/tests/mo ... -insertion with a "backlight strobe frequency picture" similar to this one https://i.rtings.com/assets/products/gM ... -large.jpg
I tried it on a regular monitor without strobing and the scrolling refresh made it unworkable, I suspect strobing might be sufficient to get it working with normal monitors.
As I said above any OLED with high BFI settings will work at least to some extent.

In terms of getting the timings right, I'm working on a somewhat simple process that will kind of guide you through the process of setting up the timings while you view ghosting and cross talk to get the best image.
I should be done that hopefully in the next week or so.
About half of my time was spent reverse engineering the Panasonic 3d glasses and the Sony 3d glasses so that I could re-write my own custom firmware.
This was really important because the OLED timing seems to drift a little bit and you need to make the glasses more responsive to changes in timing, than the default firmwares.
The default firmwares take about 5-10 frames to resynchronize onces they detect any drift (or dropped frames in the case of software page flipping), which makes them have a jitter effect at the top and bottom of the screen.
The custom firmware I worked on updates timings instantly and will never show the wrong picture to the wrong eye (at least not completely, sometimes it might start to show the wrong picture to the wrong eye, but it will detect the dropped frame immediately and correct itself.
It can cause a brightness spike on one eye for 1 frame though as the right eye will receive two frames in a row while the left eye will miss a frame for example.
it's not too bad though if it only happens once every 1-2 minutes as it's basically like blinking, and it shouldn't happen on faster PC's at all at least from my testing under linux.

Anyways the sensor I am stuck using is unfortunately the OPT101 (https://www.ti.com/lit/ds/symlink/opt101.pdf) sensor. I looked into digital sensors, specifically the OPT4001 (https://www.ti.com/lit/ds/symlink/opt4001.pdf) which has digital thresholds to trigger interrupts which would be ideal to offload all the ADC processing required, but unfortunately it's fastest conversion time is 600 us. Using the OPT101 with an ADC fast sampling I can get conversion times of around 40us which is 10x better at reducing jitter and improving responsiveness to dropped frames.
Given 120hz corresponds to 8.333 ms per frame, you can't really afford to have a jitter of 600us so the opt4001 was no good so I am stuck with the OPT101 which unfortunately costs $7-8 per unit and you need two for the device to work right.
The cheapest I found them was on aliexpress for $5 if you buy in bulk, but avoid anything under $5 as they are all scams (I had 3 different sellers sell me bad chips).

Anyways let me know your TV and your glasses type and I'll let you know then I can send you a board to help me test with and work on the software.

Hi Peter,

I've only basic soldering experience, but enough to play around with some simple Arduino/digispark stuff. I've a 144Hz Asus monitor and MSI laptop with 4k 120 Hz LCD screen. I have several pairs of 3D Vision glasses (first and second generation) as well as a 3DV compatible model. Would love to try out your board for 3D Vision and can hopefully contribute to the software and firmware. It's not a problem to pay for the board and shipping, just send me a PM.
Btw. I was planning to use a simple analog photo cell. Isn't that possible?

Best regards, Robert

[peter64]

Hey Robert,

Fantastic man! It sounds like you have soldering skills too! If you are interested I can send you a working detector board in the mail.
Please let me know what types of shutter glasses you have. It does work with off the shelf shutter glasses from Panasonic, Sony and 3D Vision.
Also let me know what type of TV you are planning to test on so I can confirm that it will most probably work, or at least has a chance of working.
I'm also keen for someone to test it using a high speed LCD panel with strobing backlight.
Compatible monitors would be those listed on the following page https://www.rtings.com/monitor/tests/mo ... -insertion with a "backlight strobe frequency picture" similar to this one https://i.rtings.com/assets/products/gM ... -large.jpg
I tried it on a regular monitor without strobing and the scrolling refresh made it unworkable, I suspect strobing might be sufficient to get it working with normal monitors.
As I said above any OLED with high BFI settings will work at least to some extent.

In terms of getting the timings right, I'm working on a somewhat simple process that will kind of guide you through the process of setting up the timings while you view ghosting and cross talk to get the best image.
I should be done that hopefully in the next week or so.
About half of my time was spent reverse engineering the Panasonic 3d glasses and the Sony 3d glasses so that I could re-write my own custom firmware.
This was really important because the OLED timing seems to drift a little bit and you need to make the glasses more responsive to changes in timing, than the default firmwares.
The default firmwares take about 5-10 frames to resynchronize onces they detect any drift (or dropped frames in the case of software page flipping), which makes them have a jitter effect at the top and bottom of the screen.
The custom firmware I worked on updates timings instantly and will never show the wrong picture to the wrong eye (at least not completely, sometimes it might start to show the wrong picture to the wrong eye, but it will detect the dropped frame immediately and correct itself.
It can cause a brightness spike on one eye for 1 frame though as the right eye will receive two frames in a row while the left eye will miss a frame for example.
it's not too bad though if it only happens once every 1-2 minutes as it's basically like blinking, and it shouldn't happen on faster PC's at all at least from my testing under linux.

Anyways the sensor I am stuck using is unfortunately the OPT101 (https://www.ti.com/lit/ds/symlink/opt101.pdf) sensor. I looked into digital sensors, specifically the OPT4001 (https://www.ti.com/lit/ds/symlink/opt4001.pdf) which has digital thresholds to trigger interrupts which would be ideal to offload all the ADC processing required, but unfortunately it's fastest conversion time is 600 us. Using the OPT101 with an ADC fast sampling I can get conversion times of around 40us which is 10x better at reducing jitter and improving responsiveness to dropped frames.
Given 120hz corresponds to 8.333 ms per frame, you can't really afford to have a jitter of 600us so the opt4001 was no good so I am stuck with the OPT101 which unfortunately costs $7-8 per unit and you need two for the device to work right.
The cheapest I found them was on aliexpress for $5 if you buy in bulk, but avoid anything under $5 as they are all scams (I had 3 different sellers sell me bad chips).

Anyways let me know your TV and your glasses type and I'll let you know then I can send you a board to help me test with and work on the software.

Hi Peter,

I've only basic soldering experience, but enough to play around with some simple Arduino/digispark stuff. I've a 144Hz Asus monitor and MSI laptop with 4k 120 Hz LCD screen. I have several pairs of 3D Vision glasses (first and second generation) as well as a 3DV compatible model. Would love to try out your board for 3D Vision and can hopefully contribute to the software and firmware. It's not a problem to pay for the board and shipping, just send me a PM.
Btw. I was planning to use a simple analog photo cell. Isn't that possible?

Best regards, Robert

hey Robert,

An anaolog photo-cell might work fine I should probably have a try to see if it does or doesn't. I went for one with an amplifier built in but it's worth a shot because it would be cheaper. I'll PM you shortly

[peter64]

just send me a PM.

Hey Robert, I tried to send you a PM but it seems it's just sitting in the outbox. Not sure if that means you didn't read it yet or the forum is broken. Let me know if you didn't get it yet. Thanks!

[peter64]

Bad news I have confirmed that 120hz BFI was removed from LG C2/G2 and LG C3/G3 OLED TV's so this technique will not work on them. Also i have not confirmed but according to information online all 2022 Sony OLED's also have had 120hz BFI removed. It looks like this was probably a pre-emptive attempt to engineering market segmentation for this years release of high refresh rate OLED gaming monitors. They probably didn't want the 48 inch C1 OLED with effect 310hz motion clarity competing with there overpriced Alienware 360hz OLED gaming monitors, as well as other OLED gaming monitors from ASUS

Good news, it works flawlessly on the LG C1 OLED and should also work on the LG CX OLED (I need to test this still) which have 41% pixel duty cycle on BFI Motion Pro HIgh.
On the Sony XBR 48a9s I have measured a 61% pixel duty cycle on Clearness high. So for the Sony you will get some ghosting/crosstalk unless you are viewing letterbox content like I was in my original tests.

Finally all the source code is up along with all design files licensed under creative commons non commercial license at https://github.com/open3doled/open-3d-oled
I will put the store up in afew weeks along with some youtube tutorial videos.

[Robert256]

just send me a PM.

Hey Robert, I tried to send you a PM but it seems it's just sitting in the outbox. Not sure if that means you didn't read it yet or the forum is broken. Let me know if you didn't get it yet. Thanks!

Hi Peter, I was away from home for a few days, I got your message!

[3DNovice]

By chance, have you tried looking in the service menu of those TVs not supporting BFI?

I know that in days past, you could enable the feature sets of some higher end models on
entry level models by enabling them in manually the service mode.

For some TVs, a service remote from Amazon can be had for under $10.

[peter64]

Someone with the latest G3 OLED 77 inch from LG checked for me and it's not in the service menu either, at least not that they could find.
Someone else told me that they have actually even tried flashing the C1 firmware to a C2 LG OLED and it didn't return the 120hz BFI feature.
So they speculated it's actually locked out from the panel itself, they didn't provide any further details though about how they tried to mod a C2 into and C1, but I expect it may not have updated the panel specific firmware or perhaps FPGA bitcode if there is an FPGA in there.
It's possible they stripped it out so they wouldn't compete with the OLED high refresh rate gaming monitors they are releasing now.
Also apparently it was a rarely used feature as not many people had graphics cards powerful enough to push a consistent 120hz at 4k which is required for 120hz BFI (at 4k at least) because BFI and VRR (or FreeSync/GSync) can't be enabled at the same time.
If anyone wants to do a second check though and finds something different please do let me know as I would be very pleased if it were there.

[18000rpm]

I have an 85" Sony X95K which supports BFI. I use it at 4K120P and when I change the Motionflow clearness setting (which is what Sony calls BFI) from 0 to 3 I can see the display get progressively darker so BFI should be working. I'm happy to help test but only have access to Windows 11 though. I also have tons of 3D content in .mk3d (SBS/OU), Blu-ray ISO rips etc.

[peter64]

I have an 85" Sony X95K which supports BFI. I use it at 4K120P and when I change the Motionflow clearness setting (which is what Sony calls BFI) from 0 to 3 I can see the display get progressively darker so BFI should be working. I'm happy to help test but only have access to Windows 11 though. I also have tons of 3D content in .mk3d (SBS/OU), Blu-ray ISO rips etc.

Hey I'de love for you to help me test it. I can probably send you a unit next week I'll PM you shortly with some questions regarding what you have available in the way of 3d glasses and the dimensions of your TV for the mount to attach the sensor unit to the TV. I just looked into your TV and it seems it's the latest microled TV and it seems it does indeed support 120hz BFI as your tests seem to have indicated. I'm super excited as this means there may actually be a new TV that works with this adapter and perhaps even a slew of future TV's that use microled will be compatible. I'm going to PM you now with the questions.

EDIT: Some absolutely fantastic stats and and analysis for how your TV backlight works over at https://www.rtings.com/tv/tools/compare ... 1220/39007 under the sections regarding "Flicker Free" and "BFI".
It has a native 720hz flicker rate and then it kind of modulates that by the framerate so if your running 120hz content each frame is going to get 6 flickers it seems like.
You can see in this image what I mean for the X95L at 120hz BFI there are 5 humps that go in a bell curve for each frame and one frame that appears blank
https://i.rtings.com/assets/products/pY ... ormat=auto
They have a similar image for the X95K but it looks like they only captured the example at BFI 60hz not 120hz as well so it's harder to count the ridges.
https://i.rtings.com/assets/products/Ke ... ormat=auto
I can only assume that mini LED tv's probably still do a scrolling update just like LCD and OLED, but it's unclear if the PWM timing of the 720 pulses is locally coupled, as in the pulsing is timed to match when a specific horizontal portion of the screen updates or if it just generally flashes. I think regardless it should be possible to adjust the timing parameters to get it working, but it will be important that the opt101_block_signal_detection_delay parameter on the emitter be set to something much closer to 8333 us something like perhaps 8000 us. Otherwise the flashing of the backlight might prematurely trigger the emitter to think it already received the next frame and it was a duplicate frame.

[peter64]

Hey everyone, I've had afew people PM me interested in helping testing, but it seems they are having difficulty to figure out how to read replies to PM's. If you are one of these people, please take a look at your inbox as I have replied to all PM's requesting beta test units up to this present point in time. Thanks!