Android 13 update borked Poco X4 GT’s bluetooth with Smart Band 7 Pro :(

What the hell Xiaomi? Poco X4 GT and Xiaomi Smart Band 7 Pro worked like a charm, but ever since I received stable Android 13 update on the X4 GT, Xiaomi’s OWN Smart Band 7 Pro keeps somehow disconnecting and then failing to automatically reconnect back.

For some stupid reason I keep getting pairing dialog where I need to confirm pairing on both and to make matter worse, the damn thing doesn’t even work. I can confirm pairing and it won’t pair. I have to then manually turn off bluetooth on the phone, pairing fails and when I turn bluetooth back on, they suddenly get connected back together like nothing happened.

I’ve contacted Xiaomi support and unfortunately they weren’t very helpful. I did a factory reset for them and now NFC doesn’t work anymore with my banking app even though it did before and it didn’t fix anything. In the end they just directed me to a local service center. Ugh.

I really like the Poco X4 GT. Relatively cheap but really nice phone and Smart Band 7 Pro, while it has bunch of compromises, it’s also really nice and I enjoyed using both, but now the experience is so messed up it’s really not fun to use it when it keeps randomly bugging me with stupid pairing thing. Anyone else having same problem?

CDKeys “My Orders” page serving plain text gibberish?

Anyone else having this weird problem with CDKeys page? Entire webpage works fine, except “My Orders” menu inside user account profile. When I click My Orders button, it just serves me webpage in plain text instead of actually rendering it. Been talking to their support for days and apparently no one there knows how to go to that menu and fucking check it. I’ve tried in all the browsers that exist, on Windows, Linux, on Android, I’ve tried using 2 entirely different ISP’s, with adblocking and without, same fucking shit. It always serves me that page in plain text for some fucked up reason. And to make matters worse, I can’t redeem any game I buy because of this shit.

Can anyone who also uses CDKeys check this for me? Go to your profile and click on “My Orders”. Specifically this page: https://www.cdkeys.com/sales/order/history/

If anyone else has it rendered in plain text instead of seeing a normal page, please let me know. Thanks

Improving Creality Ender 3 V2 PSU cooling

For some reason designers of Creality Ender 3 V2 3D printer decided to make power supply unit (PSU) for it by slapping together two metal housings that don’t have overlapping holes, instead, the holes almost perfectly align in such a way that they block almost maximum possible amount of airflow into the PSU.

Spot the problem? So, I’ve decided to change that. But since my metal work skills are basic, I’d just cut away the middle part of the black shroud. But that would still be blocking large part of the airflow. So I asked my dad for help and with some of his metal working magic, this is the improved PSU fan intake now…

Spot the difference? Look, you can actually see the fan now! Now, the fan is still audible, but at least it’s not beating same air around trying to get it through those narrow gaps in metal that were there originally.

Next step is replacing that 40mm fan on motherboard that’s just 10mm thick with another 40mm fan with 20mm thickness. Should spin much slower while moving a lot of air. But that’s for another time…

HDMI is garbage for computer monitors

Yesterday, I’ve ranted about damaged DisplayPort cable and how it screwed up system badly with hard crashes and random reboots. Replaced it with HDMI and all seemed fine…

Until I noticed bad tearing in games. Hold up, I have FreeSync display, my image shouldn’t be tearing like that. Went to check settings if it somehow disabled itself and G-Sync settings panel on my RTX 3080 was just gone. Not even there anymore. What?! Checked Windows display settings and no Variable Refresh Rate settings either. Ok, now I’m confused. Done quick search online and read that HDMI 2.0 doesn’t support FreeSync or HDR. Now, RTX 3080 supports HDMI 2.1, but my monitor doesn’t as it only has HDMI 2.0 input. Well, fuck. So, I had to rush into a tech store for new DisplayPort cable on god damn Saturday just to get back freaking FreeSync. Ridiculous.

HDMI is such garbage connectivity standard for computers. Last monitor I had couldn’t do 1080p at 144Hz via HDMI. Only 60Hz. I had to use DisplayPort to get 144Hz refresh. Fast forward to now on larger 1440p display and same shit. Again some dumb limitations with HDMI. At least I could get 144Hz, but without FreeSync? Come on!

Well, now it’s DisplayPort and FreeSync is back again. Pretty annoyed by all this. If you’re on computer connecting higher end computer monitor, don’t bother with any other cable, go straight for DisplayPort, if you want to avoid dumb problems, because HDMI is apparently totally inferior.

Damaged/bad DisplayPort cable causing total system failures

Back to bizarro problem solving and here it is, another one I was dealing for a long time, but it never really crossed my mind it might be a god damn video cable causing this…

I’ve had really weird random system reboots for a while, we’re talking 2 years now because this AMD Ryzen 5800X system I’ve built just before the COVID pandemic hit with full force. But it happened so sporadically that I couldn’t really put a finger on it at any point. It could go entire month or two without an issue and then suddenly 4 reboots in a single day.

I’ve had my system just out of the blue in the middle of the night go “click” and reboot with fast fan ramp up that woke me up followed by 1 long beep and 3 short beeps. And then it would startup normally. What the F ?! Checking ASUS beep codes and 1 long and 3 short beeps means “No VGA detected”. Reseated my RTX 3080, reconnected VGA power again and it was fine. For some time. Recently I had just straight out power failure reboots. Like, I was using system and next moment it just blanked out and rebooted, but this time without the usual beep error code (but not always). Event Viewer gave me no insight on what happened other than “Unexpected power failure” error in the logs. Thought it was my Ryzen 5800X overclocking attempts. Just to have same problem on fully stock configuration. Ugh.

Then I had an “aha” moment as I was putting my system back together and I notice how wonky the DisplayPort connector is. It’s a plastic shell with a release button and I notice how connector can move slightly inside the connector shell and how it’s bent in the way how cable is hanging down from the system. Thankfully I had HDMI cable next for my PlayStation 2 “HD” module so I used that. It’s fine for 1440p at 144Hz so I totally unplugged the DisplayPort and I’m now on HDMI only for several days. No issues so far and it has been several days.

Now thinking about it, I’ve had this particular DisplayPort cable for ages (since GTX 980 actually and through GTX 1080Ti and now RTX 3080) and the way I have my PC case placed on a dedicated box stand next to my table, I regularly lift it up while turned on to put it on a table and tinker with it while everything is still connected. Apparently years of doing this messed up the cable which lead this cable to cause total system failures like reboots with graphic card detection failures as well as just absolutely random system crashes and reason they were happening so randomly is because I was trying to fix the issue, I was bending the cable some more and it sometimes fixed it for a moment and next time it made it worse. Seeing how flimsy that DisplayPort connector was after quick inspection, I have a strong suspicion this was really it, even if I don’t have weeks of system working to really confirm it. A god damn cable for 15 or 20€ causing such catastrophic problems. Wow. I’m generally used to problems like image blinking or green tint across entire image, not simple display cable crashing entire system like it lost power. And I think I’ll just stick with HDMI even though I generally prefer DisplayPort. Because all DisplayPort cables seem to have this wonky cheap plastic housing around the connector where HDMI has it really firmly molded into the housing, preventing any bending. You’d really have to yank it really hard to damage it.

Next time you’re having such weird issues, just try a different display cable as one of first things you try. I know I’ll do that from now on…

Update NVIDIA DLSS to latest version by yourself

A lot of games these days come with NVIDIA DLSS support which is nice. However, the DLSS support at the moment is sort of “baked” into games. The DLSS logic is not provided by driver itself, it’s provided by a DLSS DLL file inside game’s folder. So, you can update it and improve performance or resolve certain artifacting. However, there’s a catch. You can only update DLSS if it’s version 2.x. If game is still using original DLSS 1.x, then you can’t update it this way. You can check this by going into game’s folder and look for file named “nvngx_dlss.dll”, right click it and select Properties. Go to “Details” tab and look at file version…

On the left example from Battlefield V, you can see it’s DLSS 1.x which cannot be updated. However, the right example is from Doom Eternal which I already updated to the current latest version. As you can see from version number, it’s DLSS 2.x which can be updated by swapping the file with latest version.

You can always grab latest version from TechPowerUp and they even offer notification service via e-mail when new version is uploaded so you can always run games with latest DLSS version.

DLSS DOWNLOAD:

https://www.techpowerup.com/download/nvidia-dlss-dll/

Just download the ZIP archive and extract “nvngx_dlss.dll” file into game’s folder, replacing the file already located there. That’s all you have to do.

Don’t expect magic to happen, but I heard reports of newer version resolving certain visual issues created by DLSS itself and it’s possible that newer version might also provide higher performance. Or at least no performance degradation at better visual quality.

It’s still unknown if DLSS 3.x sometime in the future will behave the same or be updatable this way, but for DLSS 2.x based games, this is the method you can use to update DLSS.

Use “thermal throttling” to your advantage on PC

It’s no secret both, Ryzen 7 5800X and GeForce RTX 3080 are hot boyz. They both love to run really hot no matter how big coolers you bolt on them. And if you cool them better by ramping up fans they’ll just boost even higher. And noise is something I don’t tolerate well. I like my system quiet.

So I did something rather unorthodox this time around that actually doesn’t hurt performance much at all, but makes system sooooo much quieter. And cooler.

Temperature throttling the CPU…

Ryzen 5000 has a setting for thermal throttling in BIOS…

For a long time I treated this as something I need to avoid and it’s there just to protect CPU from burning out. But then it made me think. I’ve been at those temperatures by allowing it to pump as much power as it wanted and raising its clock limits to 5GHz. And all it did was drop clocks tiny bit and it’s not like it just drops from 4.4GHz on all cores down to 2GHz or something. The curve is rather progressive for it to stay at that temperature. So I dropped the thermal throttle limit down to just 65°C. My fans on AiO are set to run at 100% at 70°C. And behold, system was now so much quieter and observing the CPU temperature curve, it hits 65°C and remains there at all times, so fans never actually went to 100% as a result (I have Silent Wings 3 on my Liquid Freezer II 240mm AiO which are rather quiet until 100% speed).

What changes now is the cores clocks. And to my surprise, those haven’t changed much. In gaming, it runs at 65°C and still hits same clocks of 4.85 GHz. I kid you not. There was no change for games as far as clocks go. Played Necromunda: Hired Gun which is very new game and some Battelfield 1. Older, but still very nice looking game which is still relatively demanding. I also couldn’t tell a difference in gameplay smoothness. Probably something like GTA V would cause clocks to drop some more, but these CPU’s are really fast and they don’t need to max out for great performance. There was some change in Cinebench R23 where my usual unrestricted clock was 4.45 GHz on all cores and now it was around 4.2 GHz. Which is still quite respectable and given I generally don’t do much all core loads it’s still perfectly fine and I only lost few 100 points which is nothing.

Temperature throttling the GPU…

I’m using Palit GeForce RTX 3080 GamingPro which has a rather slim cooler that doesn’t extend past PCI rear bracket. It was only version at the time that I could get for normal price and it’s fine. I was playing Automachef yesterday and it was running full tilt even though game isn’t that demanding, making it quite noisy and pumping out tons of heat just because it was running at locked 144fps (my monitor refresh FreeSynced). And I applied same logic as for CPU. Lets just temperature limit the card…

Fired up MSI Afterburner and only thing I did was unlink power and temperature and gave TEMP LIMIT priority, dropped it to 65°C which is the lowest it can go. If temperature allows, it’ll still go to max clocks. And it sometimes does when not under heavy stress on entire GPU. And just like CPU, it goes up to 65°C and flatlines there. No matter what I do it stays at 65°C. Super quiet and it’s now only 1000 points slower than average RTX 3080 in 3DMark Port Royal test. In Necromunda and Battlefield 1, I couldn’t tell the difference once again. But my god it was quieter and under heavy loads clocks were fluctuating between 1500 and 1900 MHz. Quite lower than before when they were stuck at 1900 MHz and above, but drop in clocks doesn’t seem to affect framerate all that much even though in frequency numbers it’s huge.

Conclusion

Basically I prioritized my entire system on temperature and my choice was 65°C because it’s a good balance in my use case. And results are pretty amazing as performance wasn’t penalized as much as I was expecting, but temperature metrics changed dramatically. You can set any temperature limit you want so you can tailor it to your cooling capacity or noise tolerance that you can still deal with. Pretty useful settings.

Fixing MKV H.265 thumbnails in Windows 10

Had issues displaying thumbnails for MKV files that use H.265 encoding. Possibly H.264 MKV’s have same issue, but I was having this issue with MKV’s encoded with H.265. I only had generic media player icon. After digging around I managed to find a tool called “Icaros” made by a guy named Xanashi.

Icaros download:

https://shark007.net/forum/Forum-Icaros-Development

It’s pretty straight forward tool that solved my issue of thumbnails on MKV files. MKV files now display a thumbnail of its video content instead of just generic icon. Hurray 😀

Sidenote: Use 7-zip to unpack ZIP archive of the portable version. Windows is not able to extract it.

Call of Duty: Infinite Warfare movement with arrow keys is fucked up

Now I’m starting to notice a trend with something… It’s games I unknowingly kept in my backlog for years. Max Payne 3, Alpha Protocol, Call of Duty: Infinite Warfare… coz when gaming gods fuck me in the ass, they really go full force.

After I lost interest in Alpha Protocol because of stupid broken hacking “mini game” well knowing it’ll be this jarring trash experience every time I’ll have to hack any system, I dropped it and picked up Call of Duty: Infinite Warfare that was still waiting in my backlog. I mean, how can one fuck up CoD right? Well, they managed to do just that…

I’m old school gamer, back when we played games with arrow keys. And it stuck with me. I don’t like bunch of keys crowded underneath my fingers and arrow keys with surrounding keys with just enough space around provide comfort gaming for my big hands. Been doing just fine for almost 3 decades with some issues here and there. But I really haven’t had them for quite a while and certainly not in AAA games. Usually some indie games or from some B studios that had poor QA or whatever.

Well, started playing Infinite Warfare and I noticed how weird movement is. It’s like keys have a 500 millisecond pause in some cases, especially when trying to move diagonally with forward and left/right pressed at the same time. It’s really awkward and clumsy, like the legs aren’t connected to your body correctly or something. It’s really bad. So I searched online and apparently this is a known thing. Well, it was when game was released and apparently, nothing has changed in 2021.

Tried using WASD and behold, disconnected legs are not a thing anymore. Set back arrow keys and legs are behaving like they are all twisted and tangled up again. What the fuck? How can one fuck up a game in a way that actions you can bind to any key don’t behave the same depending on keys you use? WHAAAAAAAAAT? And to make matter worse, not fucking fix it at all?

EDIT:

Well, in the end I made a fix for a fuckup Infinity Ward couldn’t be bothered to fix.

DOWNLOAD

CODIW Arrow Keys Fix

Gigabyte G27Q monitor bricking itself after F05 update. But not really.

As per my usual Gigabyte tradition, things just have to go wrong with their stuff… Just the other day I was updating its firmware on my Gigabyte G27Q monitor to version F5 which brings some refresh rate fixes. So, I installed OSD Sidekick, downloaded firmware and applied it.

All went well until display went blank during update and monitor began doing what looked like a bricked loop state. Its power indicator was blinking in a weird way with two fast blinks and a longer pause, making weird high pitch noise every time it blinked. Since documentation doesn’t exist on Gigabytes site apart from entirely useless “manual”, I searched online and found someone else had exact same issue on Reddit. Apparently they fixed it by having it unplugged for a day. So I went with that fix. Left it unplugged from power as well as PC (I was using DisplayPort) for the night. In the morning, all excited to finally get to the desktop I plug it in aaaaaand the same thing. Blinking LED, dead monitor. Dafaq?!

Then as I was flipping it around thinking what to try next I realized I only unplugged power, its USB and DisplayPort cables. The HDMI going to PlayStation was still plugged in and I remembered the news about HDMI dongle fooling GeForce cards into allowing mining and I thought, what if presence of HDMI cable alone (even though PlayStation was turned off) is making it fail to operate. So I entirely unplugged it from all power and data sources. Waited a minute, plugged only power cable back and behold, monitor woke up normally. I stuck back in the DisplayPort cable and bam, Windows desktop appeared with OSD Sidekick app still there saying “Finished” in a popup.

Apparently, if you have HDMI plugged in, G27Q monitor will get stuck in a “boot” loop after firmware update. And since the F05 in particular has refresh rate fixes for PlayStation 5, that happening is a very likely scenario since it connects via HDMI.

So, apparently all you have to do to “unbrick” it is to disconnect ALL cables then only plug it into power and turn it on. Once it completes the first “boot”, it will work fine even with all the cables plugged in. What an absolutely bizarre issue that will surely freak out so many users who ventured down this firmware updating route. I just hope they’ll find this blog post and resolve the issue quickly. I can’t tell you have freaked out I was thinking shiny new 300+ € monitor just got bricked and I’ll have to deal with lengthy RMA process. Thankfully I didn’t have to after figuring out the issue myself. I also hope Gigabyte will write down a warning for F05 update that if this happens you need to disconnect it entirely, do a power only “boot” and then connect things back.