Things I’ve learned, published for the public benefit
Hope This Helps header image

Persistent arm pain after getting the Pfizer Covid-19 vaccine – case study

October 26th, 2024 · Health

In 2021, I got an unusual adverse reaction to the Pfizer-BioNTech Covid-19 vaccine. After my first dose, my arm started to hurt. A little arm pain is a normal reaction, but in my case the arm remained sore for three months. I also felt unexplained sensations of burning, stinging and tightness. After my third dose (in the opposite arm), both of my arms started hurting. This time the soreness was more intense. It gradually subsided over the next year or so, and then reappeared when I got Covid. Thankfully, as I write this in 2024, the symptoms are 95% gone.

What follows is a detailed account of my case, compiled from the notes I took in 2022. As I could not find any reports of a similar adverse effect, I’m writing this post simply so that there is a public record of the reaction I experienced. This post is long overdue (blame my procrastinating nature), but even now, I think it could be useful to medical professionals, researchers and patients with similar problems. If you’re none of the above, this post will probably be extremely boring!

First vaccination

On 28 April 2021, I received the long-awaited Pfizer-BioNTech Covid-19 vaccine (also known as Comirnaty or BNT162b2). I got a vigorous immune reaction with a fever of 38.2°C (100.8°F), severe chills, fatigue, and pain in my left deltoid muscle. All of the symptoms quickly went away, except the muscle pain, which persisted at a mild-to-moderate level.

Second vaccination

My second Covid shot was scheduled five weeks later, on 2 June 2021. Because my left arm was still sore (to which the nurse reacted with disbelief), I asked for the second injection to be given in my right arm. Although it was the same vaccine, this time my immune reaction was barely noticeable.

After the vaccination, there was no new pain in the right arm, but my left arm kept hurting.

Description of the symptoms

  • What I refer to as “pain” was actually a range of sensations. Most of the time, it felt like soreness (as if there were a painful bruise there) or burning pain, sometimes resembling a nettle burn. At other times, it could manifest as a feeling of tightness (as if a knot had been tightened in the muscle) or sudden “jabs” (for example, when lifting an object). Sometimes lightly brushing a fingernail on skin would result in weird discomfort.
  • There were no outward signs, such as redness or swelling.
  • The sensations were contained to the deltoid area, centered precisely on the injection site. Usually, they seemed to come from “just under the skin” rather than deep inside the deltoid muscle.
  • The pain got much worse whenever my arm was exposed to heat. The effect was very prominent. I often felt the pain when resting on my arm or wearing heavy clothing. Conversely, applying ice brought instant relief – what’s more, the beneficial effect persisted for hours. For this reason, I used cold compresses regularly and started dressing more lightly, making sure my arms were uncovered whenever possible. This made an unreasonably large difference.
  • The pain also got worse whenever I used my deltoid a bit more. This flare-up would not occur immediately, but rather on the next day. It was as if the muscle had become very intolerant of exertion. For this reason, I had to avoid lifting my arms or pushing heavy objects as much as possible.
  • I accidentally noticed that hydroxyzine (an oral antihistamine) seemed to reduce the pain significantly. I am not completely certain, though, because I did not take it for very long.

Third vaccination (booster dose)

The left arm pain from the first Pfizer vaccine dose gradually eased. After three months, it was almost gone. Because my second shot (in the right arm) did not produce any adverse effects, I decided to get a booster shot in the right arm again.

On 10 December 2021, I received another dose of the Pfizer vaccine. The immune reaction was fairly intense, with fever and chills – definitely closer to what I had experienced with my first dose than with my second dose.

Remembering my painful adventures, I decided to baby my deltoid for 2-3 weeks after the shot. In that period, I did not notice any symptoms, save for some short-lived pain at the injection site. On 23 December 2021, I went swimming and followed it with fairly intense housework. The next day, on Christmas Eve, I was lying on a couch in a warm room when I started feeling burning pain in my left arm (not the one that received the latest shot), along with all the other symptoms I described above. A week later, the pain appeared (for the first time) in my right deltoid.

Not only was I now feeling the symptoms in both my arms, but the soreness was also more intense this time around – to the extent that it started to affect my everyday activities. For example, I would feel burning pain when lifting a mug, turning the steering wheel and vaccuming the floor. To spare the deltoid, I had to change the way I did those activities – lifting with my bicep, pulling the steering wheel down rather than pushing it up, pulling the vaccuum toward me rather than pushing it away. I got used to the fact that my deltoid was very delicate. Sadly, this also meant I had to stop going swimming. The worst part about it was not that the pain was unbearable – in fact, I would classify it as “moderate” – or that I had to forgo some activities – it was not knowing what was wrong with me or how long the symptoms were going to last (a month? a year? forever?).

I dealt with the issue by applying ice and generally keeping my arms as cool as possible (for example, taking off my hoodie during a winter run). After 2 months, I was even able to go swimming from time to time (at the cost of a small flare-up), as long as I took it very easy. My symptoms gradually subsided, and after about 9 months I was basically pain-free at my normal level of activity – though I suspect putting serious strain on my deltoids would have been a bad idea.

Covid infection

On 25 November 2022, I went down with Covid-19 – probably variant BA.5, based on genomic data from that period. The first night, I got a fever and probably the worst chills of my life. I stupidly didn’t drink enough water, and in the morning, I passed out due to dehydration. Other than the brief loss of consciousness, the course of the disease was fairly benign. After 8 days, my home Covid tests started coming out negative. After 14 days, I felt mostly fine, if a little low in energy.

What is more relevant to the topic, three days into the disease, I started feeling pain in both my deltoids at the same time. It was the same soreness accompanied by various sensations as described above. The right arm hurt much more than the left arm. Over the next few days, it ramped up in intensity, reaching perhaps 40% of the maximum severity that I had ever experienced (which was after my third vaccine dose). This time, the recovery was faster. In April 2022, I wrote in my log that I was “basically pain-free in everyday life”, meaning that I felt no pain as long as I didn’t do anything that strains my deltoids (like swimming front crawl or moving desks). A few months later, I was able to stop babying my deltoids altogether.

Common cold

In September 2024, I got what could be described as a “serious cold” (with muscle pain, some fever, cough, etc.). Covid tests came out negative. The illness resulted in a very slight flare-up of my deltoid pain (primarily the right deltoid), which started a few days after symptom onset and lasted about 1 month. I am not entirely sure if the trigger was the infection or the heat (I was feeling cold and spending a lot of time in warm rooms).

Current status

As of October 2024, my deltoid pain has gone down to almost zero and I’ve almost forgotten about it.

Since my third dose, I have not had any other doses of any Covid-19 vaccine. After each subsequent vaccination, my nerve symptoms were more severe, so I think it is likely that another dose would cause even more intense and longer-lasting arm pain. I don’t know if the adverse effect I experienced was caused by the Covid-19 spike protein or by the mRNA vaccine technology, so I don’t know if getting another type of Covid-19 vaccine (e.g. viral vector or protein subunit) would be safe for me. I also don’t know if getting an mRNA vaccine against another disease would be safe.

Some theories

As it is impossible to get medical care for a nonstandard issue like that, I tried to get some answers on the Internet. I have compiled some of my thoughts below.

If you Google something like “vaccine arm pain”, you get a lot of pages about SIRVA, which stands for “shoulder injury related to vaccine administration”. This is an injury that can happen when the nurse botches the job and makes the injection too high on your arm. The vaccine then sets off an immune reaction inside your shoulder joint, which can cause permanent damage to the joint. But SIRVA is not what happened to me. Here’s why:

  • Both of my arms hurt, which would imply that two separate nurses made the same, very rare mistake.
  • I saw where the injection had been made and it was near the center of the deltoid muscle.
  • One of the hallmarks of SIRVA is limited range of shoulder motion, which I did not experience.

The great variety of sensations I experienced – burning, stinging, tightness – are typical of nerve pain, as is the protracted recovery (nerves take months to regenerate after injury). It certainly does not sound like ordinary soft tissue pain. I am quite confident that something bad happened to the nerves in my arms.

The intensity of the pain correlated with the intensity of the immune reaction following vaccination. Recall that after my second dose, I did not feel cold-like symptoms – and the arm pain did not get any worse, either. This suggests the symptoms were caused by an immune reaction to the vaccine. The apparent responsiveness to hydroxyzine could have been a coincidence, but if not, it would suggest some type of inflammation was present, which would be consistent with immune system involvement.

What happened cannot be explained by the simple theory that the immune reaction to the vaccine produced lasting local inflammation. Note that only my first Covid-19 vaccine injection was given in my left arm; the other vaccinations were given in the right arm. But my third dose (right arm) was able to “reactivate” the pain in my left arm. Furthermore, one year later, a Covid infection (which has nothing to do with arms) was also able to bring back the pain in my left arm. (I don’t think I’ve had any major colds since 2022, so I’m not sure if another respiratory infection would have caused the same type of flare-up.)

Let’s try again. Is it possible that my initial vaccination sensitized my immune system to the nerves in my arms? It was meant to attack the spike protein, but since my nerves were around when the immune response was underway, they were mistakenly identified as enemies. From then on, every time I was exposed to the Covid vaccine or the Covid virus, they got attacked. That makes more sense, but it leaves some questions unanswered. Why didn’t my immune system attack both arms after my first vaccination? It only attacked my left arm, so it seems there must have been something different about my left-arm nerves. I don’t believe that left-arm nerves innately present different antigens than do right-arm nerves, so it must have been some extraneous element that made them look different. Similarly, you could ask, why was the pain mostly contained to a section of the deltoid – what was special about that section?

How about this: My immune system repeatedly targeted nerves in my arms, because those nerves (or something near them) had somehow been marked by the vaccination, as if someone had painted a bullseye on them. So the timeline was this:

  1. After the first dose, the left arm nerves were marked and attacked.
  2. The second dose did not matter because it didn’t set off a significant immune response.
  3. The third dose (given in the right arm) alerted my immune system and reactivated all the anti-Covid immune cells that I had, including those which targeted the “bullseye mark”. That’s why the left arm was attacked. In addition, the right arm was marked and attacked.
  4. When I got Covid, the same thing happened, except both of my arms were now marked, so they got targeted again.

How might this marking have happened? I have no idea, but one possibility (which I am not smart enough to disprove) is that mRNA vaccine nanoparticles got into some of the cells that surround and support nerves (Schwann cells, epineurial cells, etc.), causing them to produce and express a SARS-Cov-2 spike protein, which serves as the bullseye. The spike must have remained in some places where it can be detected, but cannot be fully removed. Another possibility is that some long-lived immune cells took up residence around my nerves. Every time the body detects the presence of something Covid-like, those cells wake up and start wreaking havoc.

Reports of similar symptoms

The CDC has an official database of adverse reactions to vaccines called the Vaccine Adverse Event Reporting System (VAERS). Its official character notwithstanding, the VAERS is not the most reliable resource for doing research on vaccine safety. Many of the reports in the VAERS are made by patients, so event descriptions are extremely inconsistent, ranging from the laconic (“3 months of persistent arm pain in the deltoid muscle”) to three-page write-ups describing everything that was wrong with the person’s health in the past three years, including long lists of food sensitivities and blood work results. Unless an event is classified as “serious”, the CDC doesn’t verify it in any way.

Furthermore, the VAERS is not a very convenient resource because, apart from being unintuitively designed, it doesn’t allow you to search by symptom duration. If you try to search for “arm pain”, you will get hundreds of thousands of irrelevant reports of people who had typical arm pain for a few days after receiving the vaccine.

I tried to improve the signal-to-noise ratio by including month, months, week, weeks, etc. in the search query, though I’m sure that excluded some relevant reports. With my imperfect methodology, I’ve found 14 VAERS reports (see this text file) which describe arm pain lasting more than one month and do not include other symptoms like fatigue or leg pain. My search included reports from other countries as well as the USA. Considering the huge number of people who have received Covid-19 vaccines, this is a very small number. For comparison, Guillain-Barré Syndrome, a much more serious neurological disease, occurs in about 3600 VAERS reports.

In case you’re curious, as of 2024, there have been almost 1 million VAERS reports of adverse events after receiving a Covid-19 vaccine in the United States. For comparison, about 600 million doses have been administered in the US.

→ No Comments

AOC Q27G3XMN monitor – a personal review

June 1st, 2024 · Technology

Photo of the AOC Q27G3XMN monitorI’ve got a new computer monitor – the AOC Q27G3XMN – and it has revolutionized my desktop.

For the past 15 years (!), I’ve been daily driving the Dell 2209WA, an early 22″ IPS screen with a fluorescent backlight. In 2016, I tried to upgrade to a 24″ Dell U2415 and, while I liked its size and energy efficiency, it had much worse uniformity (both white and black), so I ended up returning it. Since then, I’ve mostly been whining about the sorry state of the monitor market and waiting for something that would be a really convincing upgrade. The AOC Q27G3XMN is what ends the wait for me.

Here are all the boxes that it ticks for me:

  • contrast ratio that outclasses IPS displays
  • no backlight bleed
  • high refresh rate (180 Hz)
  • manageable size (27″)
  • not curved
  • good text clarity

Contrast ratio

The AOC Q27G3XMN has a VA panel with a high native contrast ratio (4300:1, where IPS can manage 1000:1). On top of that, it has a mini-LED (full array local dimming) backlight with 336 zones, which further enhances the contrast. Subjectively, it blows my old IPS display out of the water. Even when I’m looking at a white page with small black elements like boxes or icons, I can immediately see that those elements are actually black as opposed to dark grey on the IPS. When the image is darker, the difference only gets larger, to the point of making the IPS look like a grey washed-out mess.

I was happy to see that AOC went for the subtle approach when it comes to local dimming. The algorithm is clearly tuned to minimize blooming, at least on Low and Medium settings (I use Medium most of the time). I can just barely notice it in some situations, but it is easier to ignore than to notice. I am of course talking about videos and games – in desktop use, or in any other situation where you have uniformly colored boxes, the blooming will be quite obvious.

It’s important to remember that local dimming cannot switch off individual pixels – it can only dim larger sections of the screen, perhaps a few square centimeters in size. In addition, individual backlight zones leak light into adjacent zones. This means that this technology can only provide macrocontrast (contrast between sections of the screen), not microcontrast (contrast between individual pixels). If we’re talking about large areas, for example when 3/4 of the screen is dark and 1/4 is bright (or vice versa), a mini-LED will render a picture that rivals an OLED, possibly even exceeds it. On the other hand, it does not do so well with a fine pattern of bright and dark shades – for example, stars in the night sky, a glittering, half-lit rock in a cave, or white subtitles under a movie. If an area is mostly dark, the display will dim it, which helps create depth across the entire scene, but stars, subtitles and other smaller bright objects will be dimmed, too. For what it’s worth, I have tested a variety of SDR games and video content with local dimming on and off, and every time I have found that I prefer to get a deeper image with true black backgrounds, even if it means some brights will be muted.

I took some photos in Cyberpunk 2077 to show the effect of local dimming and how the AOC Q27G3XMN compares with my old Dell monitor. To see the differences most clearly, I suggest you view the image on a high-contrast screen (for example, an OLED phone). Even so, the images do not quite show what it actually looks like. In real life, the highlights are not blown out (this is due to the poor dynamic range of digital cameras) and the dark parts are more dark. But the image does a decent job of showing how the different displays/modes compare to each other. Enabling local dimming darkens the image primarily in the shadows, producing a somewhat more dynamic image. The dark wall gets more inky, but the graffiti and other shadow detail get dimmer (but still quite legible).

Interestingly, the contrast on the AOC Q27G3XMN can even compare favorably to my old plasma TV. Sure, the plasma TV is a self-emissive display with per-pixel illumination, which means it beats the AOC on microcontrast. But it has one weakness – it cannot completely turn off its pixels. It can only dial them down to an “almost black” level. In other words, in a very dark room, you can see that black isn’t black. Therefore, in scenes where larger sections of the image are black, the AOC has an advantage, because it can completely turn off the backlight in those areas. Add to that the much higher maximum brightness of the AOC and you may actually decide (like me) that the AOC produces a more dynamic and exciting image, at least in some movies.

Viewing angles

The AOC Q27G3XMN is very much a one-person display because it has terrible viewing angles. It’s not quite as bad as a TN display (if you’ve seen one, you know what I mean), but – at a 70 cm viewing distance – to experience the full contrast of this monitor, I have to be sitting dead in the center of the screen. If you’re looking over my shoulder, you will see a washed-out image, so this is definitely not a display for social gaming sessions. I even had to slightly tilt the screen up – otherwise, the Windows taskbar appeared washed out.

Uniformity

I’ve had a fair bit of experience with IPS displays, so whenever I get a new monitor, I check it for backlight bleed. Even with local dimming switched off, the AOC Q27G3XMN does not have bright patches in the corners that you might see when watching a movie with black horizontal bars. It even acquits itself well on the toughest possible test – a pure black background with brightness turned up to 100 (over 500 nits!). If there is any backlight bleed there, it is completely masked by VA glow. Of course, with local dimming on, backlight bleed becomes a non-issue.

As I mentioned, the display exhibits VA glow – this is like IPS glow, except it is much darker. Subjectively, it feels about 1/3 as bright as on my old IPS Dell. Here is a photo, which shows more or less how it feels with local dimming disabled. Because the camera exaggerates the glow on the IPS, I had to shoot the displays at unequal brightness levels so you could see what it looks like to the eye. The white dots are actually in the image and are not broken pixels.

Photo comparing IPS glow and VA glow

Because of the poor viewing angles, the AOC exhibits a type of vignette effect – if you’re sitting at a normal distance (70 cm), the image will be slightly dimmer and less contrasty in the corners and along the edges. This is most noticeable in desktop work – for example, black-on-white text will gradually turn dark grey as you get further from the center of the screen. It’s not ideal, but at least I cannot notice the effect when gaming or watching movies.

Brightness

I’m going to confess something to you: I used to hate bright monitors. When I got my old Dell, which has 300 nits, I decided that this sort of searing intensity had no use other than burning out your retina and must be a marketing trick for stupid consumers who believe that “bigger number = better number”. I ended up using the Dell at 10% brightness for years.

So when I saw the measurements for the AOC Q27G3XMN, putting it at 560 nits in SDR mode, I thought “Great, yet another monitor with pointlessly high brightness – what a waste”. But after spending some time with the AOC, I have seen the light, if you’ll pardon the pun.

My mistake was failing to distinguish between desktop use and media use. In desktop use, you’ve got a lot of 100% white backgrounds. If you make the display put out 560 nits, or even just 300 nits, it will indeed feel deeply uncomfortable. However, videos and games are a completely different animal: there have no large pure white areas. Even in a daytime scene, the brightest large areas you might see are a piece of light-blue sky and perhaps some kind of light-grey wall. This drastically reduces the amount of nits that reaches your retinas.

In practice, when gaming or movie-watching in the daytime on the AOC, I’ll use 70% brightness, which would be probably around 400 nits for pure white content. Not only is it not blinding, but colors appear more saturated and the whole scene feels more like looking through a window and less like looking at a backlit photograph. It’s just more impressive, the same way a garden looks more impressive when the sun comes out from behind a cloud. Is such a powerful backlight necessary? No. If I go down to 40% brightness (which is around 250 nits – what you’d get on a maxed-out 2024 QD-OLED), I can certainly see everything I need to see. It just feels less vivid and lifelike.

So if you are a brightness skeptic like I was, don’t knock high-brightness monitors until you’ve tried them for video and games because I’ve done a complete 180 on that. In fact, now I feel brightness is a serious reason to reconsider my longtime desire to get an OLED display. (And the fact that Sony is moving from OLEDs to mini-LEDs on their flagship TV models suggests that OLED panels will not be able to catch up in the next few years.)

Color accuracy

Out of the box, the AOC has a slight red tint, which is easily eliminated by switching to the User color mode and dialing the red channel down by 2 units.

High refresh

I have fond memories of playing Company of Heroes on my 17-inch Philips CRT monitor. For those unfamiliar with the game, the core gameplay loop involves scrolling from one part of the map to another to give orders to your units. With a 100 Hz vertical sync frequency, the Philips CRT rendered deliciously lifelike camera movement with virtually no motion blur.

The AOC Q27G3XMN does not quite reach such heights of motion clarity. It will do 180 Hz, which – on a sample-and-hold display – is not nearly enough to match a 100 Hz CRT, but!.. I’d say it gets you about halfway there. It’s like going to a restaurant and being served chicken soup. You know it’s not quite the same as your mom’s chicken soup, but it’s good enough that it reminds you of it, and for that you are grateful.

The 180 Hz refresh rate is not only beneficial in entertainment use – it also enhances the desktop experience. Moving the mouse, scrolling pages or moving windows around all feel much smoother than on a standard 60 Hz screen. You feel more like you’re manipulating a real object. There is a perceptible decrease in delay between pressing a key on your keyboard and characters appearing on the screen – though this is most apparent in text editors which prioritize low input lag (such as Sublime Text) – most editors are so sluggish that the 6 ms difference in input lag between 60 Hz and 180 Hz barely matters.

One weakness of a VA panel is ghosting. Occasionally, fine visual elements will change color or disappear in motion. For example, grey text on black background will often dim when you scroll it. If you have seen videos showing things like disappearing foliage when moving the camera in a videogame, bear in mind that this only happens with very specific combinations of background and foreground colors. So far, I haven’t encountered anything as dramatic as that in any game that I’ve played. When playing an FPS, I’ll notice some small ghosting perhaps once an hour.

Variable refresh rate

It’s undeniable that the AOC Q27G3XMN has VRR flicker. RTINGS currently ranks the AOC as having the worst VRR flicker out of the 38 monitors that they’ve tested for it. In the current version of their review, they even describe the flicker as “terrible” – though it’s perhaps worth noting that, in the first published version (before they had their new panel of VRR flicker tests), they didn’t seem to notice this “terrible” flicker and even recommended the monitor. On the other hand, Adam Simmons (PCMonitors.info) has said (in YouTube comments under his review) that the AOC definitely does not have the worst VRR flicker that he has seen and, unlike some other displays, it was not bad enough to make him want to disable VRR.

So far, I have played three games on this monitor – Cyberpunk 2077, Jagged Alliance 3 and Hitman 3 – and, although I notice VRR flicker in all of them, I haven’t found it necessary to disable VRR. I only seem to notice flickering in static scenes with a lot of uniform dark or midtone areas – for example, in menus, inventory screens or scenes with sky at dusk (not daytime sky). In detailed scenes, especially with a lot of things going on, I can play for an hour without noticing a single flicker.

It’s worth noting that I do take some steps that reduce VRR flicker – I drop down graphical details to make sure that the framerate stays above 90 Hz (not because of VRR, but because anything below 90 doesn’t feel responsive to me) and, in the case of Jagged Alliance 3, I set a framerate cap at 144 Hz, because that game has wild framerate swings.

It’s likely that I will eventually come across a game that has large enough framerate fluctuations to make VRR flicker annoying for me. In that case, I will probably just disable both VRR and Vsync, as I don’t find tearing to be very noticeable at a 180 Hz refresh rate.

Screen coating

I would classify the coating on the AOC Q27G3XMN as “semi-glossy”, because, unlike my matte screens, it not only diffuses light, but also reflects a faint mirror-like image. You can see the difference in these two photos – on the AOC, you can see the edges of the flashlight in the reflection.

I did not have any expectations with regard to screen coating on AOC, so I have to say I was pleasantly surprised by the quality of the antireflective layer. I can easily see grain on my Dell IPS screen, and even on my high-end-ish Eizo, but – to my 45-year-old eyes – the AOC is basically grain-free. I may be able to barely make out some grain, but only when I’m slowly moving a solid-colored window around the screen and my attention is completely focused on detecting grain. In other situations, including staring at uniformly colored backgrounds while coding or editing text, everything looks clear and pure. During gaming or watching video content, the benefits are less obvious than in desktop use, but even there, there is a definite bump in clarity. When I put the AOC alongside my old Dell, the Dell looks a bit like a textured painting, whereas the AOC looks more like a window to another world. Furthermore, if I compare the picture on my glossy iPad and the AOC, I don’t really see a difference in clarity.

The only bad things I can say about the coating on the AOC is that reflections are more defined than on traditional matte screens and that the screen reflects a bit more light in general, so I have to control my room lighting a bit more whenever I want to see deep blacks.

Text clarity

I use the AOC with 125% Windows display scaling. This makes text slightly larger than on a 24″ 1080p monitor at native scaling, so I place monitor slightly further away from my eyes, which I think helps with eyestrain (it’s easier to focus on objects that are further away). I was worried that 125% scaling would make some things look blurry, but the only time I notice blurriness is in applications that I use occasionally and for brief periods of time, like the Windows Device Manager (come on, Microsoft – you can’t even update your own system components for hi-res monitors?). All the apps that I stare at for hours a day (Google Chrome, Visual Studio Code, Sublime Text) look nice and sharp.

The switch from 1080p to 1440p – or, more accurately, 90 ppi to 109 ppi – is a nice upgrade. It’s not quite like 4K, but everything does look significantly smoother and smaller text is more legible. The subpixel layout is totally standard RGB, so there’s no ClearType blurriness like you get on an OLED.

Coil whine

My AOC Q27G3XMN has easily audible coil whine when the brightness is set to between 60 and 70 percent. I can also hear it in some other situations, depending on the image that is displayed on the screen. Sometimes moving a window to the “wrong” spot can produce a slight buzz or whine. All in all, the AOC is not much different from my old Dell when it comes to electrical noise. I do believe that hardware should be silent, but I have little confidence that another display would fare any better, so I guess I just have to put up with it. (Another unit I’ve tested has softer coil whine that appears between 50 and 60 percent brightness, so it looks like it’s a lottery.)

Bad pixels

If I hadn’t methodically examined the screen with a magnifying glass, I would not have realized that my unit has 5 bad subpixels, all near the edges:

  • two stuck blue subpixels (always on, but low brightness)
  • one broken blue subpixel (always off)
  • one broken red+blue subpixel (does not turn on fully)

Fortunately, none of them are brightly glowing pixels and all of them are extremely hard to detect at a normal viewing distance – even if I know where they’re supposed to be.

HDR

None of the games that I’m playing right now support HDR well. I’ve tried Cyberpunk 2077, but turning on HDR in that game completely ruins black levels and my attempts at calibration didn’t bear any fruit. So far, my favorite HDR content is this Sony demo video showing the crafting of a Japanese sword.

I’ve noticed that in HDR mode, the AOC has a greenish tint, which cannot be fixed with OSD settings because the whole color setup section is disabled. Another thing I didn’t like is that HDR is simply too bright for playing late at night. The monitor only supports VESA HDR1000, which means the maximum brightness of a screen element is 1000 nits and there is no way to turn down the brightness because the brightness control is disabled in HDR mode (which is typical of HDR monitors). Furthermore, unlike on some other displays, you cannot switch to a less ambitious HDR mode like HDR400 (400 nits). It’s either HDR1000 in all its eye-searing glory or nothing.

For all those reasons, I only turn on HDR when I want to show off my new display to a friend. If I find new exciting uses for HDR, I’ll be sure to update this review.

Annoyances

The controls on the AOC Q27G3XMN were designed by raging dimwits who haven’t got the slightest idea how people use their monitors. You get five identical buttons with invisible, black-on-black labels and I guarantee you will accidentally turn off the monitor many times, because the power button is right next to the menu button. I recommend carefully feeling for the rightmost button (power) with your finger and then moving one button to the left to access the OSD menu. Changing the brightness requires 6 button presses (menu, right, menu, right, menu, left/right), while gimmicks like “game mode” and FPS crosshairs are super-easy to access with their own dedicated buttons. Let me also add that the buttons feel cheap and require quite a bit of force to press – I got pain in my forearm after one day.

Given the dreadful physical controls, changing the brightness with the OSD is obviously a no-go, so the only way out is DDC/CI. (For those unfamiliar with it, this is a technology which lets you send commands to your display via DisplayPort or HDMI. You can install an app to control things like brightness and contrast with your mouse or keyboard.) Of the four DDC/CI apps that I’ve tried, only two – TwinkleTray and Win10_BrightnessSlider – even detect the monitor, and TwinkleTray does only if you put it in legacy mode. Neither works reliably – about 50% of all commands are randomly ignored by the display, which means a lot of repeated keystrokes if you want to set the brightness just right. Even this, however, is less hellish than using the OSD, so this is what I do.

I’ve found this issue bothersome enough that I actually ordered a newer unit of this monitor (manufactured in March 2024) based on reports that the bug had been fixed in later batches, but that turned out to be a mistake, as DDC/CI support is broken in exactly the same way on the new unit. (Note: DDC/CI works fine if you connect the monitor over HDMI, but that means no VRR and a 144 Hz refresh limit, so I won’t do that.)

When you turn the AOC off with the power button, it disconnects from the computer. You’ll hear a “device disconnected” sound and Windows will rearrange all your windows and icons to your other display. This behavior is slightly annoying, but I understand it is common on modern displays. To avoid it, I use a keyboard shortcut (assigned in TwinkleTray) that puts all my monitors to sleep.

When the monitor turns on, it shows a bright AOC logo, which can be quite jarring if it happens at night. Fortunately, this can be disabled in the service menu. To access the service menu, while the monitor is on, unplug the power cable. Then press and hold the menu button while plugging in the power cable. Then open the OSD menu – you will see an extra option labeled “F”.

The monitor has an sRGB mode (exception: early European units manufactured in October 2023 or earlier), but if you select it, all the color controls are disabled, which makes it pretty useless, as you cannot correct the red tint. The AOC has, by default, quite a wide color gamut, so if you don’t do anything, standard video content will appear oversaturated. I wouldn’t mind a little extra “pop“ to the colors, but unfortunately this causes human skin to take on an unnatural reddish hue. Once again, software comes to the rescue. If you own an NVIDIA card, you can install an open-source utility called novideo_srgb, which will emulate sRGB just fine. If you own an AMD card, you can do it directly in the driver control panel. (More on sRGB emulation)

Given all the bugs that I’ve just listed, I suspect you won’t be pleased to learn that the AOC Q27G3XMN has no firmware updates of any kind. Even if AOC fixes some of the issues down the line, the only thing I can do is sell my old monitor and buy a newer unit, which is wasteful and inconvenient.

Price and availability

The Q27G3XMN is often billed as a budget mini-LED display. This is accurate, because the price (or at least the MSRP) is very attractive, but also somewhat misleading, because it suggests you can get something better if you pay more money. In reality, there simply are no other options if you want a 27″ non-curved screen with good contrast and good text clarity. You can get a Samsung Odyssey VA display, but that will be curved. You can get a flat OLED, but that will have either poor text clarity or it will be larger than 27″.

Abysmal availability has become a pattern with monitor releases in recent years. The AOC Q27G3XMN followed a typical timeline from the perspective of the European consumer:

  • December: you begin to see reviews
  • January-February: everybody talks about it in forums, but nobody actually has it
  • March/April: a small early batch (with multiple firmware bugs) is shipped to stores – this sells out in a week
  • June: you begin to see decent availability in stores

The situation has been much better in the US – which, incidentally, proves that the free market doesn’t work. For months, the Q27G3XMN has been listed in the US at $280 and $400-$500 in Poland. You would think that if a company can sell the same product for more money in another country, they will redirect their stock to that country, but apparently AOC prefers to make less money so long as it can sell to Americans.

Final words

I am happy with the AOC Q27G3XMN, because it improves on my old Dell IPS display across all the dimensions that matter. The upgrade in my day-to-day experience is so radical that I can even forgive the numerous annoyances – something I’m usually not very inclined to do. Even at the $400 price point, I feel the Q27G3XMN is a good deal, especially if you consider that it can be reasonably compared with a $1,100 OLED. (Yes, an OLED will have much better microcontrast, better motion clarity, less VRR flicker, and much better viewing angles, but the AOC has better text clarity, higher brightness, and doesn’t have to be coddled to avoid burn-in.) As of May 2024, I’m reasonably sure there is no better mixed-use 27″ 1440p display on the market.

→ 4 Comments

Full-screen videos and games look oversharpened with shimmering edges – possible solution

March 19th, 2024 · Technology

About a month ago, I noticed a weird thing on my PC. Whenever I played a high-resolution video, it would appear oversharpened, as if someone went overboard with a sharpening filter. You know what that looks like – excessive contrast in details and brightened, jagged edges that shimmer in motion. Strangely, this sharpening effect would only manifest itself when the video was playing full-screen. In a window, everything looked nice and smooth.

At first, I put it down to some weirdness with my video player. I fiddled with the player settings for a bit, but that didn’t help, so I just gave up and forgot about it, since the problem wasn’t visible in typical movie/TV content.

Then, one day, I launched a new game (fittingly titled Jagged Alliance 3) and thought to myself, “Wow, they really went crazy with the sharpening here!”. I blamed it on the developers and enabled DLDSR, which made everything beautifully smooth. Problem solved! But then I tried playing some of those high-resolution drone videos on YouTube, and I noticed distractingly jagged edges on roads, fences and houses. And then I launched another game and I noticed excessive contrast and aliased edges.

Interestingly, all of those issues were specific to “exclusive fullscreen” situations. Just like before, as long as the video or game was in a window, the issue would go away. Even more weirdly, when going into YouTube fullscreen mode, the video would look fine at first, but as soon as the YouTube status bar disappeared, the sharpening problem would come back. Go figure.

As an experiment, I plugged my monitor into another video adapter (integrated Intel graphics). The issues went away. This seemed to point at my NVIDIA card as the culprit.

Solution:

The NVIDIA Control Panel has a 3D setting called “Image Scaling”. I had it turned on, with “sharpen” set to 50%. I thought a feature called “image scaling” would only apply to situations where a game outputs a lower-than-native resolution and the NVIDIA drivers upscale it. Apparently, it will process all full-screen output, even in your display’s native resolution!

Screenshot showing the Image scaling option in nVidia Control Panel

→ 4 Comments

Steelcase Leap review – crooked back update (also, video is hard!)

November 22nd, 2021 · Office, Seating

I have updated my review of the Steelcase Leap office chair with information about the “crooked back” issue, which has been reported by users here and on Reddit. I even recorded a video that demonstrates what the issue is and – I think – pretty conclusively shows that it is caused by a design flaw, not some kind of random manufacturing defect.

Making the video was a surprisingly fun project, even though it did confirm what I already knew about video as a medium: it takes an awful lot of work to say anything with video, especially if you’re aiming for semi-professional quality. Just filming the shots took me two days – finding a decent, uncluttered spot in my apartment, setting up the lighting, experimenting with framing, fighting technical issues with my smartphone cameras, etc. Then there was audio – recorded in small pieces, with multiple takes needed due to my lack of skill as a narrator (for example, I tend to speak in a monotone, which sounds very dull, and when I try to jazz up my speech, I often end up sounding unnatural). Editing the video and audio together – choosing the best video takes, making sure the source videos (from different devices) play well together, deciding which parts of the videos to use, strategically using slowdowns, speedups and still frames, drawing the on-screen graphics, cutting the audio and making sure it corresponds to what’s happening on the screen – all that while learning to use Hitfilm Express – took me a day or two as well. And then I had the idea to add a soundtrack to cover up the noise on the voice track, and instead of using a stock track, I decided to emulate 3Blue1Brown and make my own music for the video. That little side job took another 2-3 days, including 1 day to synchronize the music track with the voice track and the video.

In the time it took me to make a 2-minute video, I could have written several blog posts. At the same time, the work I’ve created is very hard to edit. If I want to correct a post – for example, because I made a mistake or because I’ve come up with a better way to explain something – I can simply rewrite a couple sentences. But I can’t do that with a video – I would have to re-record the audio, likely add more video scenes to show while the new audio is playing, figure out how to make the music track fit the new content (record extra music?), re-edit the whole thing – it’s just way too much work. What’s more, you may not know that YouTube explicitly disallows updating a published video – you have to delete the old video and upload a new one. But if you do that, you lose all the likes and comments, which influence your ranking in YouTube search, so you’d basically have to be an idiot to do it. Video may be easier to absorb for the viewers (at least for most topics), but yeah – it is HARD to work with.

Anyway, as I’ve said, I had a lot of fun working on the video. As an introverted programmer type, I had the most fun writing the soundtrack and editing the video. In another life, I could see myself working as a composer or a film editor. The least fun part was putting my voice out there – I’m just not very comfortable doing that (even though it’s still much easier than putting my face in front of the camera!).

By the way, I’ve uploaded the soundtrack to SoundCloud in case some mom-and-pop candy store wants to reuse it in a promotional video.

→ 4 Comments

“Account locked out” error when connecting to a Windows network share – possible solution

October 5th, 2021 · Technology

You try to connect to a computer on your Windows network (using what is now called “Windows File Sharing”). Instead of a login prompt, you are greeted with an error message that reads: “The referenced account is currently locked out and may not be logged on to“.

"Account locked out" error message

This error can only occur if you’ve set an account lockout policy, which locks your account after several failed login attempts (this is recommended by Microsoft). But how could there be a failed login attempt if you were never asked for your user name and password in the first place?

Here’s how. When you connect to a network share, before you get the chance to type in your user name and password, Windows will try to log on with your credentials from your current computer.

Let’s say your local user name is “Clara”. If the target computer also has a “Clara” account, but with a different password, the autologin fails. Furthermore, Windows appears to retry the operation a few times (God knows why) and this may be enough to trip up the account lockout mechanism. (As a quick confirmatory experiment, I just set my account lockout to 5 failed attempts and tried to connect to it – I was locked out after the first failed autologin.)

Of course, a lockout means you’re screwed. You cannot try again with the correct password. Worse, you cannot even log on to the other computer locally. You have to wait whatever the lockout duration is set to – typically something like 15 minutes. On the plus side, you can use the break to do anything you like. I used mine to estimate what yield of a nuclear weapon would be enough to wipe out Redmond, WA.

Bad solution:

As you know from the previous post, renaming your account on one of the computers is not a good idea. Not only will it not solve the autologin issue, but it will also add enormous confusion.

Good solution:

Go to the Credential Manager and add the user name and password for the target computer. Now Windows will use these credentials automatically whenever you connect to that machine.

→ 2 Comments