Looking at that “screenshot”, you might wonder what I’m on about. It seems perfectly normal. It might even leave you scratching your head as to why it could ever be a problem.

It’s not just that button: The entire Roku app is awful. Among other things, it now forces your home screen to present Plex-corporate-specific-content, and buries your own libraries in a very irritating menu at the top of the screen.

The library selection menu. Imagine if there were 10 libraries here. I pared it down to the two I use most to cut down on button presses.

From a UX perspective, it’s horribly optimized (or not optimized at all) for quick and easy navigation. The prior version of their Roku app wasn’t perfect, but Plex is clearly aiming for a use case other than consuming one’s own content.

So let’s look at that button again, and compare it to the button on the Plex web interface:

Watch versus Play.

One of those words is the app telling you what to do (watch), while the other is you telling the app what to do (play).

Yep, that’s it. That’s the reason for my annoyance. It emphasizes their philosophy: Plex will now order you around, rather than being ordered around by the user.

You want to watch your own media libraries? Buried. Watch Plex-branded monetized content instead.

Play means “commence presenting the media”. Yet, Plex feels confident enough to redefine it as watch.

However, I often don’t watch visual media. Sometimes I just have it on in the background and I listen. But I’ve always starting it playing by pressing the play button.

Sometimes I leave the room while something is playing, even though I’m no longer watching.

So not only is the button labeling instructive rather than receptive, it’s also just plain wrong in common use cases.

On top of that, it’s unnecessary.

Obviously this is a “play” icon, identifiable even in a logo.

The sideways equilateral triangle pointing to the right is universally understood to mean “play”, across just about all languages and cultures of Earth. Why Plex bothers to label the button is strange, as perhaps their time could be better spent on other areas.

Sure, it doesn’t take a lot of time to label a button, but probably more than you’d think.

Aside from the fact that their codebase needs to contain what are ultimately pointless conditionals to decide whether “watch” or “play” should be shown depending on the type of media, there’s also localization. A simple play icon by itself needs no localization.

But the text accompanying it must be made available in every language supported by the Plex app. Adding text to a universal symbol is just wheel spinning. Moreso when that text actively redefines the meaning of the symbol.

Bizarrely, in a music library, the button isn’t labeled “Listen”. No, it simply says “Play”.

Not only is Plex wrong (IMO) in their button labeling, they’re also inconsistent.

Getting back to iconography, there’s these gems on the left-hand side of the interface. If you’ve never seen them in this context, try and conjure in your mind what they might do in reference to your media library before you scroll any farther:

That top one? I thought it was an eyedropper icon at first, viewing a 2160p 42″ screen from perhaps 3 meters away as I was.

Nope. That’s intended to be a magic wand.

Or rather, I think it’s meant to be a magic wand. It’s a bit oddly proportioned, the top segment isn’t filled in with white as with a stereotypical magic wand, and the spacing of the “magic radiance dots” is too uniform. (Perhaps 4 dots would be better?)

The bottom icon is clearly a grid, which I suppose is at least recognizable. Though, a grid of what? And magic for what?

I polled a few people around the office, showing them an image of the entire screen. No one could correctly identify the purpose of either icon, even in context.

Fortunately, hovering over the icons reveals their purposes:

My question here is: Why persistently label something obvious like the play icon, yet create two extremely vague icons and leave them unlabeled by default?

What’s even more baffling is that the Plex web interface uses a very similar icon to the “Browse” above, yet for a completely different purpose. The grid is used to represent — get this — a grid view. That makes sense.

All design issues aside (and there are more — I’m still on the home screen at this point!) there are major bugs in the Plex Roku app.

For example, twice today the interface froze while I was listening to music. (I use the Roku when working out.) In both instances, it occurred when I was navigating between albums under a couple of different artists. Not selecting anything, mind you. Simply using the arrow buttons on the remote was enough to freeze the UI.

And it was just the UI; Music was still playing, and hitting the remote’s Home button brought me back to the Roku home screen to re-launch the Plex app, indicating that the Roku OS hadn’t seized up. (I was able to reproduce that problem on a completely different type of Roku device, so it’s almost certainly an app bug and not specific to my one device.)

Tonight, when launching the Plex app to get the above screenshots, it greeted me with a black screen. Oh, the volume keys on the remote still worked and the volume indicator popped up on screen. The Roku home button worked. The Plex app did not.

I never had those kind of problems on the old, better Plex app.

In conclusion, my take is this:

Plex initially went in with a very consumer-friendly business model, offering lifetime subscriptions at a reasonable price. This, though great for me and many others, was clearly unsustainable. There’s a certain addressable market for personal media server subscriptions, and they probably addressed all of it.

Now, they want to become some kind of half-assed media company, pulling in revenue from wherever they can, forcing their existing customers deeper and deeper into the Plex content ecosystem. A place none of their users want to be. This is also an unsustainable business model, but meanwhile they’re spending significant money on re-developing all their beloved apps, removing valuable features and rendering others inconvenient.

Why, oh why, do I need 2 different apps on my phone for music and video? Often when listening to music, I’ll play the music video when that’s available and otherwise listen to the MP3/FLAC. It was great when that could be done within the same app. Now, not so good.

I’d move to Jellyfin, but Plex’s web interface is by far the best in the industry. And by “industry”, I don’t just mean that of personal media servers. Plex’s web interface is superior to any video streaming service out there, paid or otherwise*. They had a top design team at one point. Maybe they still do, but that team is being crippled by management. I have no way of knowing.

Jellyfin is (for now) very immature by comparison (though philosophically much, much better).

*Though I wager Plex’s web interface won’t be aces for long. It might even be enshittified by the time you read this. I’d promise to update this post to reflect that, but by the time they kill off this generation of the web interface, I won’t care about this anymore because I’ll be using something else. Probably Jellyfin.)

Of course, I only use Plex for my personal consumption of content which I purchased and ripped from my own hard media.

I’ve has this sitting around my house for years now. I was mostly interested in it because it was a piece of broadcast gear which clearly had an ATX computer hiding within.

I happened to accumulate 3 Beelink computers at once: Two are going to be PtP VPN nodes at undisclosed mysterious and super interesting locations (some of that is a lie), and one will be a portable Plex Media Server for travel.

This video is an unboxing and disassembly of all three, comparing them and showcasing their design advantages and disadvantages. If you’re looking for a video about their performance, well, I kinda didn’t even get to powering them on yet. This is 100% about the physical hardware.

The SER and SEi series are incredibly similar, the former having an AMD CPU and the latter an Intel. The cases are different (metal vs. plastic and different venting), but the motherboards are interchangeable. Both are incredibly easy to disassemble (and hence reassemble), having the same screws throughout. Even the NVMe SSD hold down screw is the same as the motherboard and case screws!

The EQR was a pain, however. I lost count of how many different screw heads, thread pitches, and lengths. Hex standoffs/posts of varying sizes. Weird multi-layered construction, a seemingly unnecessary (due to poor optimization) daughter board, and worst of all the line voltage inlet connector is half supported by the motherboard itself. I really don’t like this one, just from a design and serviceability standpoint. For example, the BIOS/RTC battery is underneath the motherboard. It would require removing something like 25 screws and posts just to extract it!

As usual, this video is NOT sponsored, endorsed, nor in any way related to Beelink. I paid for all three computers out of my own pocket for my own purposes.

The other thing I didn’t address in the video is about the EQR series’ internal power supply. There’s no airflow whatsoever over the PSU, and it’ll be radiating its heat into the case. Time will tell, but that doesn’t speak well to the reliability of the PSU, and puts more of a burden on the CPU fan (the only fan) to handle the removal of that additional heat.

I recently got a Star Trek TOS reproduction tricorder from The Wand Company to accompany my cosplay at the upcoming STLV: Trek to Vegas 2025 convention. But, more important to me right now than getting a costume ready is the contents of this device!

So in this video I do a full teardown; Though this isn’t a tutorial necessarily, as I shot it whilst figuring things out for myself the first time.

My apologies to anyone that would rather I have preserved this NIB. But remember, Lore was disassembled and he turned out fine! Seriously though, as much as I’d love to preserve this in perfect condition, I purchased it to bring it to a con. It’s gonna get banged up and scuffed at least a little bit as it hangs on my hip. So long as I can put it back together again, it should be no worse off in the end.

Sorry for the ruse in the beginning, but I wanted to show my other costume prop option: An AM/FM radio that seems to have been included with a Japan-exclusive TOS DVD box set! That was an eBay find, as I can’t imagine anyone is still handing out analog radios. I couldn’t find much in the way of details on it, so if you know anything about the device please lmk in the comments!

As usual, this video was *not* sponsored nor endorsed by The Wand Company, nor whatever company was behind the AM/FM radio in the 2000s. :)

PCB Images

I did my best to get a nice rectilinear view of all the PCBs at a reasonably high resolution and with even, sharp focus. I was mostly successful. The lighting might seem to be too shadowy, but I had to use a somewhat oblique light to get most of the component markings to appear legible.

However, I didn’t realize until after all the photos were taken and resized that the markings on both Winbond chips were illegible. But both of those appeared in the video, and I included screencaps below, which should hopefully suffice.

All images are in 2 sizes: Medium and Max Resolution. The Medium size is to make it less cumbersome if you just want to take a quick peek at about 1-3MB; The largest of the Max Res photos is about 12MB.

Medium Resolution

Max Resolution

I needed a USB PSU capable of supplying about 85W @ 20V on three separate ports simultaneously (so 255W total). The UGREEN Nexode 300W GaN “Charger” seemed like a good option, as it would have a decent amount of headroom and has a port labeled “140W”, and two “100W”.

However, IMO it’s pure marketing nonsense. First of all, it’s very particular as to what loads you can place on which ports in combination. The only way it can output its rated 300W is if one of those loads is @ 28V. That’s fine, I guess. They can spec it however they want.

These are the various permitted combinations of loads

However, in my testing while loading it with 250W across 4 ports, it started thermally throttling after about 45 minutes. In other words, shutting ports down to save itself from melting down. IMO if it can’t even provide 250W consistently, it should not be sold with a 300W rating. (The testing was performed at about 24°C [75°F], which I think is reasonable.)

In its defense, it does have active PFC and feels incredibly solid. It seemed to be able to provide 200W consistently. But, I would not recommend it unless they change the markings to 200W total and changed the architechure so 200W could be drawn from any combination of ports up to their individual maximum labeling.

While I was searching for a suitable PSU and settling on the UGREEN, I ran across a listing for a “700W” USB-C PSU. Right off the bat, it seemed like an absurd claim. So I purchased it just to take a look and see how much of a BS device it was. It is total crap as one would expect.

Well, not “total” crap. It was able to provide 360W of output for about as long as the UGREEN was able to provide 250W. And it did either thermally throttle outputs (or begin failing intermittently — I’m not sure which) rather than having a meltdown. That being said, the quality of the output was far worse than the UGREEN’s, and it got significantly hotter. (Peaking at about 107°C in one area.) It was untouchably hot after about an hour of loading. Compare that to the UGREEN throttling closer to a 75°C hotspot, and a 65°C average-ish case temperature. (The UGREEN is likely chock full of thermal compound, whereas the 700W no-name had some sparsely strewn about.)

Oh, and the architechture of the “700W” PSU was such that it was essentially 2 completely independent power supplies from the mains to the outputs. So essentially you could only load it to its actual maximum (~360W) by distributing the load across both.

Links to products featured, but this is (obviously) not sponsored, the links are not affiliated, and I don’t recommend you buy either:

UGREEN: https://www.amazon.com/dp/B0DBZY57ZF
700W: https://www.amazon.com/dp/B0CMNQMJ9G

REUPLOAD – In the previous version I made the erroneous claim that the PYXIS Monitor couldn’t be used with the BMMSC 4K G2 and a USB hub to allow simultaneous monitoring and recording. The issue was that the monitor uses HDMI over USB-C, not USB-C data. So it will only work with hubs that support USB-C HDMI alt mode, none of which I had. Apologies if this misled anyone; It was definitely not my intent, as I very much wanted the monitor + SSD to be usable! Though a bit embarrassed, ultimately I’m glad I was wrong about it.

Here’s the USB-C hub from the intro that *does* work with an SSD + PYXIS: https://www.amazon.com/dp/B0DRG9K3BJ (Not affiliated nor sponsored.)

The original video is still available in unlisted mode for posterity, to preserve the comments and etc: https://www.youtube.com/watch?v=-cGBQrV702o

The title isn’t clickbait. After messing around with the monitor, the BMMSC 4K G2 and various USB accessories, I found a bog standard USB-C touchscreen monitor from Amazon that provides the same functionality as the PYXIS. To be fair, I tried two other 1080p touchscreens and neither worked, so it’s not universal compatibility.

Here’s the product, but it’s *not* an affiliated link nor am I shilling for the brand (which has a silly name anyhow): https://www.amazon.com/dp/B0DF1P542S (PeakDo 7 inch Portable Touchscreen etc..)

All that being said, the PYXIS monitor does have a much better look and feel than a cheap alternative, and advantages include a locking USB port, 1/4″ mounting threads on 2 sides, a sunshade, and 3 function buttons on the back that are extremely flexible (you can assign them to do most anything).

IMO if budget isn’t an option and a 5″ monitor is good enough for you, go with the PYXIS. It’ll probably be less pain in the log run. Then again, maybe go with an HDMI or SDI monitor, because the monitor won’t work via a USB hub (I tested 3 of them and none worked, anyways), and if you occupy the BMMSC 4K G2’s single USB-C port with the monitor *you can’t record to external media*.

So you’d either need the PYXIS as well as an HDMI/SDI recorder, or something like a Video Assist 5 or Atomos Ninja to monitor and record.

When used in a studio, I’d think a cameraperson would prefer a monitor larger than 5″.

If Blackmagic Design added another USB port to the camera, it would solve many problems. Or enable the Video Assists to work with USB for camera menus. IDK, either way it seems like a hole in BMD’s product line that they probably want you to fill with a PYXIS camera (or URSA) rather than making the BMMSC capable of fully competing with those when rigged out.

Part 2 of this video shows one way to add WiFi connectivity to the camera. It’s slightly convoluted in that it uses a WiFi bridge device which essentially passes networking from the wireless network to an Ethernet port. That Ethernet cable can then be connected to a USB-C Ethernet adapter attached to the camera. (There may be other WiFi bridges that have a USB Ethernet adapter built in. IDK, this is the one I found.)

The reason for using a bridge is that you can’t just connect a USB WiFi adapter directly to the camera. There’s no way to configure SSID and password (or etc.) in the camera’s interface. However, the camera does natively recognize a USB Ethernet adapter and will (by default) use DHCP to grab an IP address.

Micro Studio Camera 4K G2 video.

I take a look at the Clicks Keyboard phone case (or, as they call it, the “Precision Typing Instrument”) with a Pixel 9 Pro and a Motorola Razr 2024.

It’s a well-made product, and works without the need to install an app (and the app that is available does *not* require you to create an account for no reason). However, it’s not something I’ll be using, both because I don’t want to switch to a Razr but also because I’m probably overly sentimental about a tactile keyboard. Also, the form factor is cumbersome on a slab phone like the Pixel (and I’d have the same issue when paired with an iPhone).

Note: As usual, this video is not sponsored nor am I associated with Clicks in any way. I purchased these phone cases with my own money for my own purposes.

Once again, I’m trying to cram a consumer graphics card into an “enterprise” Workstation. This time it’s a Dell Precision 7960 Tower from eBay, with a w9-3495x CPU, 128 GB RAM, and two 1TB SSDs (one of which I’m upgrading to a 4TB Samsung 9100 Pro). As best I can tell, the MSI Inspire 3X is one of the few consumer GPUs that will clear the height constraints of the case (an RTX 5070 Ti here, but you can also get a 5080 in the same form factor).

Also, I put on the tin foil hat to talk about how consumer graphics cards are probably bigger than they have to be to protect Big GPU.

Storage Driver

To download the driver for the on-board disk controller, go to Dell’s support site here:

https://www.dell.com/support/product-details/en-us/product/precision-t7960-workstation/drivers

(I don’t want to link directly to the actual driver, as it may have been updated to a later version by the time you read this.)

• Select Storage Controller from the Category dropdown.
• This should bring back a result for Intel Virtual RAID on CPU Driver. (This is correct even if you’re not using RAID mode.)
• Download that EXE and run it.
• Select the option to Extract, not install.
• Extract it to a USB stick.
• Attach that USB drive to the system on which you’re installing Windows 11.
• Install the driver called “” in the folder “\F6\IntelVROC_f6_iaStorE_win10_64\”. (That worked on my system, but if it doesn’t work on yours, try the other driver listed.)
• Note: It may depend upon whether you have RAID enabled in BIOS. I probably shouldn’t have.)
• You’ll know it worked if/when you see your NVMe drive(s) listed as options on which to install Windows.

I’m sprucing things up in the studio (basement), which includes removing most light stands and tripods. They’ll be attached to the ceiling instead, and with that comes re-wiring everything. Also, upgrade of a couple of full-size Blackmagic Studio Cameras (gen 1) to the much smaller and lightweight Micro Studio Camera 4K G2. Unfortunately, whereas the old BMSC devices have SFP slots for fiber connectivity, the BMMSC G2 does not.

So I’m also using a couple of Blackmagic Mini Converter Optical Fiber 12G modules with the cameras, but as with the cameras they require power which can’t be provided by fiber. (I did look at structured cable, but it wasn’t worth it.) To keep things somewhat neat and tidy, I decided to provide power remotely using 4-pin XLR cables going into custom splitter boxes with 3 barrel jack outputs. That’ll run the camera, converter, and a small LCD monitor if necessary.

Mostly this video ended up being about the adapter box, and I didn’t get into the fiber converter as much as I intended to from the outset. That’s because the whole project got a bit sidetracked by the power supply not working out of the box. I’d picked up a Core SWX Cube 200 Power Supply from B&H, which is supposed to put out 14.3V constantly, but was putting out 16V occasionally when I banged on it. Just a faulty unit, I suppose. Might return it or investigate further.

In a subsequent video I’ll go over the actual setup of the cameras using the adapter boxes, as well as the fiber converter (and hopefully a working high-amp PSU).

As usual, nothing in this video was sponsored and I’m not promoting anything. I paid for all this stuff out of my own pocket, for my own reasons.

I got this Sandisk Pro-Dock 4 card reader holder and laptop dock a while back for a bargain, hoping it would be somewhat slapdash in its electronics such that I could “convert” it to be used as a simple USB 3 card reader apparatus. Turns out, it’s so Thunderbolt that Zeus himself is scared.

The engineers at Sandisk (and/or Western Digital) did an absolutely masterful job with this hardware, though the product appears to have been let down by poor software/driver support. It seems to be discontinued now, which is a shame.

In this video, I tear down the thing and check out its components. Relatively high-res photos of the two internal circuit boards:

I got this Schaffner NSG 200E Interference Simulator mainframe with an NSG 223A module a while back to test a device featured in a previous video. That was quite a while ago, but I finally got around to taking a look at it.

In this video we’ll check out the device’s internals and power it on for the first time in years. Then test some USB power supplies for interference (spike) pass-through.

The USB PSUs were the cheapest I could find for fast delivery from Amazon, at about 4 bucks each. I wanted them to be of questionable quality, but they seemed surprisingly well-made. Though I’m not an expert, I am a fan of ⁨@DiodeGoneWild⁩ and would love to send these over to him for a more thorough analysis if he’d be interested. (Pre-opened and pre-abused, of course.)

Pictures of the Internals

A.I. generated video content presents plenty of problems in the short and medium term: Job loss in the commercial, TV and film industries. False or misleading information. Dystopian idolatry of fake celebrities.

But IMO people are missing the really frightening aspect of all of this. The road leads to bespoke content generated for each viewer in any given moment. Once it becomes cheap enough, why would a company spend 0.5 cents to create a commercial to show to 100 million people when they could create 100 million commercials to show to each person individually? And why wouldn’t streaming platforms create content on-the-fly to cater to each individual’s preferences? It’ll lead to cultural isolation wherein all of us consume only our own media, in an infinite feedback loop. We each become an echo chamber of one, disconnected from our fellow humans.

I gave up on the HPE Microserver in the previous video on this subject and decided to go with a custom build. That went.. well.

It’s an iStarUSA D-300 case, ASRock Rack motherboard, Xeon Silver 4509Y, 64GB RAM and a Samsung 990 Pro NVMe. And, of course the Dell/IBM LTO-8 tape drive. Finding the right power supplies and fans was a bit of an issue due to my poor planning.

The PSU originally was a Thermaltake Toughpower PF3 1050W, but that full-size ATX form factor didn’t 100% fit so I ended up using a Corsair SF1000 SFX-sized PSU which was technically too small but whatever.

The fans were almost completely replaced between the beginning and end. I went with Noctuas for quietness and reliability, but the small ones in the back of the case didn’t move enough air. (As found out in the “smoke test” — not what you think). Swapped those out.

Of the two CPU coolers I tried, one suffered from extremely bad vibration at medium and high RPMs. The other also vibrated excessively (but less so).

Could I have made this a little easier on myself? Sure. But this was also my first ever custom rackmount build. And though I’ve probably built scores of computers over the last 30+ years, they were all desktops, and lately I’ve been buying so much used enterprise hardware that I lost the knack. Hence the super-long video.

BTW: Absolutely NOTHING in this video is sponsored, endorsed, nor otherwise shilled. I paid for everything out of my own pocket, bought it for my own reasons, and have no relationship whatsoever with any of the companies I’m about to mention below. There are no affiliate links or anything similar. You do you as far as these products go.

I compare the superficial differences between the SmallRig VB99 and the Neweer Mini V-Mount batteries. I was looking for inexpensive V-Mounts and came across both of these on Amazon, and from the basic descriptions and product photos they seemed to only have cosmetic differences.

They’re both ostensibly 99Wh battery packs with D-Tap out, USB A and C out (with charging on the C connector), as well as 8V and 12V barrel plug outputs (and, of course, the V-Mount connector).

And they are in fact pretty much the same as far as features, weight, and capacity goes. Though, neither one was able to deliver the promised 99Wh (they got close at around 92Wh).

The SmallRig was almost twice the price, so as far as value-for-money is concerned: Buy the Neweer (or something entirely else).

This video is NOT sponsored nor affiliated with the companies featured. As usual, I just fancied comparing these two products for my own purposes, and purchased them with my own money just like anyone else would.

I examine the SmokeNinja fog machine from the company of PMI (which stands for Practical Magic & Innovation, though the logo is largely illegible and they only mention what it means in a tiny font in one place for some baffling reason). Despite them being embarrassed by their own company name, it seems to be a very well-constructed, well-designed and well-implemented device. (That’s 3 out of 3 “wells”, my highest score yet on this new scoring system I’ll never use again.)

It puts out a decent amount of smoke/fog, is easy to use, remote controllable, and has a user-swappable common type of lithium ion battery so you could use it extensively.

They pitch it for photography/videography effects or theatrics, and I guess it’d almost certainly be good for that. Now that I think about it, it would also be good for Halloween effects on the front porch or something. Though, I purchased it for the purposes of showing the airflow in computer cases, as well as the beam dispersion patterns of lighting fixtures.

As usual, this video is NOT sponsored nor did I receive any kind of compensation whatsoever. All opinions are mine and mine alone, and I paid for this thing out of my own pocket for my own reasons.

I’m not a terribly good gamer. I play Dark Souls and Elden Ring with a mouse and keyboard. It’s been my dream to actually be able to use a modern day controller, so I figured I’d force myself to by putting a “gaming” HTPC in the spare bedroom. Some kind of mini PC. That didn’t go to plan because I ended up getting distracted from this task by playing through Dark Souls 3 on my main PC.

But regardless, in this video I open up an Intel NUC Enthusiast 11, stick in some RAM and an SSD, then other RAM, then yet other RAM… then eventually play some games to see what the performance is like. It’s not too bad considering the thing has an RTX2060, which isn’t awful for 1920×1080 60Hz.

Pairing that with some Logitech keyboard (w/trackpad) and a PlayStation 5 controller, I will eventually learn to master gaming from a couch or bed. Because that’s living.

If you’re looking for a high-power charger/power supply that can also make toast and confections, the SlimQ 330W laptop power supply might be for you! I actually purchased this to use with a Starlink Mini and a USB-C laptop, and it should suffice for that. It seems like a really high quality unit, except that they’re trying to dissipate about 22W of power from this relatively small plastic case (at full load). Though it is around 93% efficient, that extra 7% has to go somewhere.

I make a really ugly cable/adapter thing to test battery charge capacity using a BK Precision 5335B (the actual battery tests will be in a different video, though). All it does is interrupt the VBus current path and provide a couple of voltage monitoring lines on a pair of USB-C connectors that are otherwise in parallel.

Taking a look at a Dell-branded LTO Ultrium 8 external SAS tape drive that I got new in box from eBay. (It’s actually an IBM drive and enclosure, as it turns out.) Paired the drive with an HPE MicroServer G10+ v2 that was also a NIB eBay find. Turns out it’s not the right machine with which to use this, but at least was enough to get the drive tested.

I laboriously go through the process of unboxing and testing the new-ish Blackmagic Micro Studio Camera 4K G2, or BMMSC G2 as it’s slightly more easily said. It’s kinda a combo of the old BMMSC G1 and the Micro Cinema Camera. Also got two Mini Converters to apply LUTs to both cameras over SDI, also from Blackmagic Design, of course.