Why on earth would anyone buy a 55-inch TV for $20,000 when they can get one for 250 bucks? It doesn’t even have features. No high refresh rate, no variable refresh rate. There’s branding and buttons everywhere. And these bezels look straight out of my grandma’s basement. It doesn’t even have any goddamn HDMI ports or a remote control. But what it does have is features and capabilities that are coveted by creators. And not this kind, no offense. This kind. One does not simply see Lord of the Rings or Star Wars until they’ve watched it on a display like this one from Flanders Scientific, who happened to be in the neighborhood and were nice enough to swing by with their latest and greatest so we could take a gander. It boasts near-perfect color accuracy, tetrahedral interpolation, 12-bit 444 signaling, zero bullsh** image processing nonsense, meaning that for the first time in my life, I will finally see content the way the colorist saw it, the way the director saw it. What content, you might ask? Well, to start with, we’re gonna be checking out the remastered segue Trilogy to our sponsor. Hetzner, if I know you, you’re looking for a company with GDPR-compliant, high-performance cloud servers that offer on-demand cloud computing platforms and APIs on a metered pay-as-you-go basis. Classic you. Don’t miss out and use code LTT24 at the link below to get 20 euros off. The XMP550 Mastering Monitor is likely the most accurate display I have ever laid my two eyes on. It’s QD OLED, so we’ve got incredible viewing angles and color that pops. It’s rated for 2000 nits peak brightness in a 4% window, and while 55 inches might seem small by today’s TV standards, it’s actually big for what it is, with most of the predecessors to this product being in the 30-inch range. Now, I’m gonna tell you a lot more about it with some help from the Labs team, but first, I’ve gotta experience it. Now, we didn’t do a TV roundup for 2023, so for comparison, we’ve got two outstanding TVs from 2022, one with a first-generation QD OLED panel from Samsung and the other with the competing WOLED panel from LG. Now, neither of these is gonna do 2000 nits, more like half of that, but they still might hold their own okay because that often-boasted-about number, that’s peak brightness in a tiny window. Once we get to bigger window sizes, the gap should narrow, and in filmmaker mode, hey, theoretically, these displays are super color accurate too, but maybe I just don’t know what I’m missing. To find out, I’ve asked Brandon from the Lab to join me, and he’s gonna be taking us on a little tour of some content that might either tell us, well, this Flanders thing is amazing, or hey, don’t bother, don’t worry about it. Which will be awkward with the Flanders guys there, but what can I say? Whoa, you guys are gonna notice that the LG G2 looks wildly different from the two QD OLED displays. That is due to the different spectral power distribution on the G2 and the way that cameras capture TV output. Now, in some scenes, it’s gonna be a subtle difference, but in others, this difference is going to stand out a lot, but I can assure you that in person, the G2 was relatively accurate to our human eyes. It was a lot more subtle. Also, please note, there are some really dark scenes here that our cameras didn’t capture very well, so for the best experience, you’re going to want to watch in HDR. Man, the other displays almost look like they have cheesecloth over them. Is that right, or did we screw something up? Their ABL’s kicking in. That makes sense. Consumer displays have ABL, or auto brightness limiting, which is to protect them so that they’re not issuing warranty replacements every six flippin’ months. Also, because I guess they would have to comply with energy standards that a professional display wouldn’t. So they’re just dimmer. For now. For now. It should be noted that for a brief period of time, ABL’s not gonna be nearly as much of a factor. We just unplugged and replugged these TVs, and you can see, now they’re a lot closer, even if, you can tell already, shadow detail. I actually find the green saturation is a little bit better than the W OLED. It’s not much. Even the trees themselves are popping a bit more here. Yeah. They kind of blend into the background there on that one. But to get a better idea, we’re gonna cut the lights, and we’re gonna have a look at some test footage. We’re gonna start off with this house scene here. Right away, look at the Samsung compared to your reference. It’s not even close, dude. It’s purple. Yeah. I know that what HDR is supposed to do is it’s supposed to make things look more true to life. But what blows me away is that I feel like I’m looking at a dollhouse with an actual light in an actual window on the Flander Scientific. Yeah. I think it’s just the consistency of the color. It just makes things look way more realistic. It’s not white. It’s light. Ah, this is gonna ruin everything for me. And look at the, oh my God, I didn’t even notice. The Samsung doesn’t have a fence. Nope. The burglars could get right into the backyard. You know, LG is doing a much better job at retaining detail in the shadow compared to Samsung. I think the LG is still raising the blacks a little bit though. Do you see where the garage is? Mm-hmm. To the right of it, there’s like some dark detail, but it’s pretty dark. Yeah. If you look at the LG, it’s raised a bit, right? Like even the sky behind it. So they’re retaining that detail in the shadow, but they’re cheating a little. Yes. Where Samsung’s going the opposite way, they’re giving it like that punchier look, but you’re losing, you’re clipping all your black detail. No question. I would take the Flanders. Then I would take the LG actually. And then the Samsung for this. The Samsung is not handling skin tones the way that I’d like. Yeah. I’ll try to posit when it goes back to her face. Yeah. Particularly the Asian skin tone. Yeah. Samsung tends to oversaturate. Even in its filmmaker mode, I’ve noticed that things look a bit rosier, more saturated on people. Even the dirt. Like look at the dirt compared to- It’s orange. Yeah. Dirt’s not orange. LG doesn’t look too bad. No, no. To their credit, it’s more accurate, even if the capability is not quite as good. Yeah. Like look at his hair. He’s red. Yeah, and his skin. He’s red. And back to what we said about realism, right? True to life, what color is a late night street lamp? Yellow. It’s yellow. It’s not orange. That’s what it’s supposed to look like. And that seat is a blue leather. Not whatever that color is. It’s got pink in it. Yeah. I can’t believe how bad this Samsung looks. This is the TV I have at home and I’m shocked. Well, right? And guys, you gotta remember, Brandon test displays for a living. Like how long have you been doing it now? Four, over four years. Yeah. And that’s how easily, like even we’re bamboozled. There’s no way that that was the director’s intent. Looking at this, we’re supposed to be able to see everything that’s going on. Here, look at that. She’s wearing a purple shirt. She’s wearing a black shirt with purple, I don’t know, something along here. This doesn’t look too bad. No, none of this bothers me. And that reminds me, part of why I had a hard time differentiating there was because I was kind of busy looking at the content. And no matter which of these TVs you have, that’s what you’re gonna be doing most of the time. So these are still hair splitting differences that we’re picking at. It’s just fun if you’re a giant nerd. These strawberries look really good. They look super red. I could eat those. Yes. They look pretty washed out on the LG. Yup. They look okay on the Samsung, but it’s still not as red. It’s not, no, it’s not as accurate. Oh, and look at the shadow detail on the reflection. It’s like all gone on the Samsung. Oh yeah. So on the LG, but you’re desaturated. But it’s, yeah, it looks like a ghost of what a strawberry is supposed to look like in the reflection, the top looks all right. But those don’t look as delicious. Those are more sour. And those are super sweet. As long as it’s not a super saturated color, the LG does okay. And I believe the Mario movie, the mastering display primaries are P3, which all these displays should be able to do. Now my understanding is the Mario movie was mastered for a peak of 800 nits. Is that correct? According to the metadata. So theoretically, any one of these displays can reach the brightness that we need to see. Even the Samsung, it’s oversaturating Bowser’s eyes a little bit. They’re a bit too orange. It looks nice. Like my eyes like that, but that’s what it should look like. So LG, pretty close. Honestly, the LG is very color accurate. I’m really glad we’re looking at real content now though, because looking at test patterns or the ASC footage made the difference seem a lot differenter. Looking at a real movie, which is going to be mastered with home displays in mind a little bit more, I guess. I don’t find the difference nearly as noticeable. Can you pause it right here? I can. What a great still. Mario’s hat does look a bit oversaturated or a bit like maybe pink or purplish compared to here. The true version looks almost right in between the almost too orangey Samsung and the almost too, I don’t know, blue purple. Yeah, blue purplish. That’s going on on the LG. I think especially with animated content, you have less like memory colors to reference, right? Like you don’t have realistic skin tones and all those kinds of things. So I think it’s just a lot easier to be fooled in thinking this looks correct. That’s true. And like back to that night scene where I was like, oh yeah, no, I know what an ugly yellow street lamp is supposed to look like. Well, this rainbow burning fire could be anything. He looks almost pallid and dead on the Flanders. Which- He’s rosy on the Samsung. Was clearly the intent. They’re not supposed to look life full. Brown uniform, green uniform. This is probably the starkest difference I’ve seen between the right color and completely the wrong color. That looks like it could have been made out of leather. Also, you can tell that this is daylight lit because on the Flanders, there’s a bit of a bluey cast to everything. Whereas we’re losing that with the oversaturation of the greens on the other displays. Here’s what I want to know. Maybe this is my first question for the Flanders folks. Is the problem that it’s more work to just not do any bull processing? Yeah. Fundamentally, I always tell people it’s more about what we don’t do almost rather than what we do. Is that really hard to not do that? The problem is most of the chips that I use are off the shelf chips from companies like MediaTek. So we don’t do that. We do everything on FPGAs instead. That’s why they cost so much. So it’s a cost savings thing. Slash an unwillingness to build your own processor from the ground up to not do any of this BS. It’s another dark scene with oranges. Oh my gosh. The Samsung is doing so bad. Yeah. And not that Brandon has a horse in this race or anything, but that’s his TV. Incredible. Okay. Dang it. I mean, even here, that’s an orange fire. Those are red fires. Yeah. Oh, that looks so very metallic. According to HGTV test, the scene coming up should be 2000 nits coming off of Gandalf. So the maximum capability of our Flanders and well beyond the maximum capabilities of our consumer TVs here. Yes. Although it is a bright scene, very high APL. Not sure how much that’s going to come into play here. Ooh. Pause it. Everyone’s handling this pretty well. There’s clipping going on the Samsung though. I don’t know if it’s because of the color stuff, but you can’t make out nearly as much. Losing the trees in the, whatever, the light source that’s behind Gandalf. LG pretty good. Oh yeah, yeah, yeah. You can barely make out the form of his body through the robes on the Samsung. And it’s also just not as bright. I think the Samsung is tone mapping a lot here. I’m also losing the folds in Legolas’ cloak. Yeah, so these two are about the same. They’re both hitting around 1000 nits off of this little beam. The Samsung is down to about 700. And I guess that’s probably because of tone mapping. One thing we didn’t focus on last time around is an undesirable behavior of the Flanders if you were using it for consuming content. And that is that if there is a signal that is outside of its maximum brightness range, rather than just tone mapping it down and trying to smooth out the transition, it’ll just peak, which is to say it will completely lose any detail that was supposed to be there. And you can really see it on the bird’s left cheek here on the right side of the frame here. There’s nothing left of the feather detail there. Whereas, actually the LG doesn’t do a very good job. It’s not doing great. Samsung, though, retains some of the texture that’s supposed to be there. Oh yeah, LG actually manages to peak even worse than Flanders. Yeah, LG does have dynamic tone mapping on right now, but that is the default behavior in filmmaker mode. Yeah, and it’s just not working very well. Huh. Oh yeah, wow. This one’s really interesting. Hey look, there’s green there. Yeah, first off, the colors are completely different. Yeah. But here, this is a great quantization banding test pattern. So just look at the Samsung. You can clearly see that white circle ends right there. Yeah, and then another circle. Yeah, and then look at the LG. The LG is much worse. You can really see, it’s like looking at the rings of Saturn around the outer perimeter. Whereas the Flanders, I guess that’s what a 5D LUT or whatever the crap it is that they’re using is gonna do for you. How do you guys describe the LUT? A tetrahedral LUT interpolation. It looks so smooth here. It’s falling apart. It looks so yucky on the rest of them. This is not a bit rate issue. This is the ability of the processor to interpret all the different little micro steps of color that might exist. And you can see the consumer displays are falling flat on their faces. There’s more to this thing than just a pretty face though. Build quality is solid with an aluminum chassis and a graphite heat sink. And it operates fanless in spite of the fact that 2000 nits, even in a small area, can generate some serious heat. And this is cool. Check this out. Part of why this display is so expensive is because it can offer up information that might not be relevant to you, but is relevant to filmmakers, like showing false color or a live readout of luminance and waveforms so you can make sure that what you’re looking at is as you intended. And that’s the reason why you or I, budget aside, might not actually want to use this as a home display. Also the lack of HDMI, et cetera, et cetera. The XMP 550 is a tool for professionals and it offers no post-processing whatsoever. That includes, or rather doesn’t include, tone mapping. This TV, or more accurately, display, shows the footage in its unprocessed form with only a custom 3D LUT or a lookup table to make sure the colors are displayed accurately. And because of that, when we show content that was mastered at a higher brightness value rather than roll off pleasingly and retain detail, it’ll simply clip. That’s intended because a colorist doesn’t want to be fooled into thinking, this is what a 4,000-nit scene looks like. So it’s kind of a safety mechanism to ensure that they know, hey, you’ve gone beyond the capabilities of this display and you can do that, but you’re kind of grading blind at this point. But wait, hold on a second. Grading blind, isn’t this thing the be-all and end-all? All right, we’re working on a tech quickie that’s gonna help further explain color spaces and terms like chromaticity in the not too distant future. But when it comes to HDR and Rec. 2020, that’s a volumetric color space that can go all the way to 10,000 nits beyond the capabilities of even this display. I mean, QD OLED has gotten us close in terms of color saturation, but the brightness just isn’t there yet. Now, the primary competitor to the XMP550 is Sony’s new BVM-HX3110, a dual-layer LCD mastering monitor that uses a technology similar to the Hisense U9DG that we covered almost two years ago. It can reach a whopping 4,000 nits, but the color saturation is narrower. To hit 100% of the entire volume, you’d need basically pure color wavelengths like frickin’ lasers displaying your image, and that’s gonna be a while. Also, despite the Sony unit offering extra brightness, there are a few reasons why you might prefer the Flanders Scientific Display. First and foremost, they’re really nice people who are entirely non-threatening. No, I’m kidding, but seriously, the near-perfect viewing angles, thanks to the Quantum Dot OLED panel, mean that you can use a single display for both working and showing references to your client or boss, someone who’s likely to be a big deal, like a studio head or a creative director. To illustrate why this matters, we had an older dual-layer LCD mastering display at LTX, and Brandon was constantly instructing people to, no, no, you have to stand directly in front of it, or it really doesn’t look right. Second, OLED’s lightning-fast self-emissive pixels give us zero rise or fall time motion blur, not to mention inky blacks without any halation or blooming. Right, and also the price. Yeah, I know, right? At only $20,000, it’s really well-positioned for a brand new mastering monitor. Our quote for the BVM-HX3110 was almost $30,000, and it’s under a third of the size. Of course, though, your choice might be dictated by creative requirements rather than budgets. If your film contains scenes like this, for example, the ABL, or Auto Brightness Limiter, that protects our QD OLED panel might preclude its use, even if it doesn’t operate as aggressively as a consumer display. However, if I was working on something like the Mario movie, I’d be a fool to go dual-layer LCD at any price, just like I’d be a fool if I didn’t sign up for a notification for our new precision screwdriver. The Flanders is also, finally, a viable alternative to the venerable BVM-X300, which is still sought after today despite being discontinued. That one used a true RGB-stripe OLED panel, which avoids washing out color with a white sub-pixel, but being older, it could only reach 1,000 nits in a 10% window. They’re also not getting any younger, so burn-in is becoming a concern with these displays. There are other options that are still in production with the BVM-E251 and 171, but those are 1080p at 24 1⁄2 and 16 1⁄2 inches, respectively. If anyone watching this video actually needs any of this stuff, hey, we’ll have links in the description below. For now, I wanna throw over to Brandon from the lab to talk about our measured results from the Flanders scientific model. Due to our limited time with the XMP550, we were only able to test its color accuracy with HDR signals. For our testing, we used the Delta E ITP metric for our color error evaluation, since it’s designed for high luminance values and a wide color gamut, two integral parts of HDR. As we saw earlier, the S95B is skewing a bit red, and on top of that, it oversaturates, but it’s still reasonably accurate, as is the G2. However, the XMP550 is a clear winner, hitting a Delta E ITP average of less than two and a max of less than five. This is by far the best color accuracy I’ve ever measured. What’s even more impressive is the lack of any major change when adding luminance error to the equation. See how the values are nearly identical? The same can’t be said for our other sets, especially the G2. That’s because the XMP outputs light and color basically perfectly to give us accurate HDR content. Its PQ EOTF curve is tracking nearly identically to the HDR standard. Meanwhile, the G2 and S95B both have minor inaccuracies, as well as a worse white balance by comparison. And even with the XMP550 capped out at 1800 nits with a D65 white, it still handily beats our two other TVs. Overall, a very impressive piece of technology for the people who need it, even if it does also kind of highlight that there’s still room for these displays to improve. There’s no such thing as a perfect solution right now. Now you might’ve noticed, we’ve been toying with HDR recently and you’re probably wondering, hey, Linus, I’m watching this video in HDR right now. Do you guys need one of those? And depending who you ask, production team’s not here, that’s good. I don’t think we need one. No, I don’t think we need one. At least not until the check clears from our sponsor. Squarespace, creating your own website doesn’t have to be difficult. Luckily with Squarespace, it isn’t. Their all-in-one platform makes it easy to get your website up and running quickly. Designing with their Fluid Engine Site Builder is easy. Start with a template and customize every detail imaginable with drag and drop technology for desktop or mobile. You can also use their asset library to manage all your files from one central hub and use them across the Squarespace platform. With Squarespace’s analytic insights, you can see what’s working well and what needs a little TLC. And if you need help, Squarespace has got your back with helpful guides and a 24-7 support team. They don’t sleep. Head to squarespace.com forward slash LTT and get 10% off today. If you guys enjoyed this video, why don’t you go watch the one where we checked out Alienware’s monitors that are also using Samsung QD OLED technology. They look incredible.

source