TCL 65R617 Image Quality Review

The 2016 TCL P-Series was a TV that we felt performed much better than its price would indicate. At a time when true HDR TVs were selling for well over $1,000, the $650 TCL made for a massive bargain. A year later, TCL has changed the naming from the P-Series to the 6-Series but also come back with some improvements. They’ve added a 65” screen size, which we tested, and more local dimming zones than before. So let’s look at the image quality for the 2018 TCL 65R617.

TCL 65R617 SDR

Out of the box, the Movie mode is the most accurate option for the TCL 65R617. To do any advanced calibration, you’ll need to install the iOS or Android app, as the controls for white balance, CMS, and gamma are only available there and not in the set. I’d like to see Gamma be available inside the TV, but putting the other controls in the app is fine and makes calibration easier without menus to interfere with measurements.

The white balance in warm has a red push that you can see on the chart. The gamma, when set to 2.4, also tracks away from our target of BT.1886. If I change the target to 2.4, or even 2.2 and change the TV gamma to 2.2 (the default), then it tracks perfectly. For a living room situation, you probably want to use 2.2 anyway, but we use BT.1886 as a reference point for all displays. The native contrast ratio measures in at 3852:1 and the grayscale dE2000 average is below the visible average of 3.0, but some readings at 70% and above do move past 3.0 as you can see in the data chart.

Colors are generally good, with Rec.709 gamut coverage coming in at 98% and saturation errors almost entirely below a dE2000 error level of 3.0. Some of the luminances for white are off, but we expected that with the grayscale numbers. Color checker errors move right up to the error line at 3.0 but most of them stay under. Overall the numbers out of the box are fine, but we have seen more accurate displays. The biggest issue with measurements here is that you don’t have a BT.1886 gamma option.

Post-calibration we’ve improved the grayscale, so the average is close to half of what it was. We can’t fully fix it with the controls we are given, and the gamma is still not perfect, but it’s better and good enough that no one should complain. The highest error level is 2.07, which is below the visible threshold.

Colors have mostly improved, but you still see some luminance issues below 60% which we will talk about more later. The contrast ratio after adjustments and without local dimming is still 3916:1, about what we expect for a VA panel display. The color checker still has an average error level close to 2.5, but more samples are below 3.0 now while some still rise above it. The major issues look to be with orange tones, where the amount of red saturation is now off after trying to correct the red color point. It’s a balancing act in calibration, but this result was better than leaving red alone.

With local dimming set to low, the contrast ratio improves to 8319:1. Medium moves it up to 14000:1, and high to 46,560:1. However, the higher you set it, the more likely you are to notice it in action. I recommend testing it with a scene with light and dark elements (read the section below on Harry Potter) and changing it to see how aggressive you can get without seeing it in action. The other option is to set it to high and take it down a level everytime you notice it in action until you find a setting where it is invisible to your eyes.

Overall for SDR the TCL 65R617 does a fine job in movie mode and will do even better if you leave the gamma at 2.2 for a living room since it targets that better than BT.1886.

TCL 65R617 HDR

Somehow I managed to not save my HDR calibration session in CalMAN, so I’m going to have to look at the HDR Analysis workflow and work from memory here, so my apologies. I did manage to pull in all of the post-cal data into our standard chart, but with a couple of caveats.

The TCL 65R617 shares a white balance with the SDR mode, so that has been copied over here. We see that the RGB balance is quite good up to 60%, then it starts to lose red as it tone maps. Since we initially had too much red in SDR, this is likely then causing HDR to not have enough, but only past 60%. Since this means the entire SDR portion of the HDR image, which is still most of the image, is balanced, this is a trade-off we would always recommend. Some very bright HDR highlights might have a small color shift to them, but these are likely to be fleeting and far less likely to be noticed than if the entire SDR area had the shift.

We also see that peak nits came in at 782 cd/m2, with a gentle roll-off. Additionally, the EOTF tracks low below 10%, which is likely causing a bit of shadow crushing that we see and discuss later. The grayscale without luminance errors is quite good, with the error level below 1.0 even using the tougher dEICtCp method. Once you add in luminance you see that errors are high in the dark shadows, and in the highlights when it runs out of headroom and can’t hit the 1000 nits target. Some people would use relative nits here, in which case the TCL would be almost perfect, but we don’t because the content isn’t relative and it would make the TCL score the same as a display that does 1000+ nits, even though they’d look different.

Colors are overall good, with gamut coverage in the 89-94% range of the DCI/P3 gamut. The errors are highest in Red and Green (and Yellow, which is a mixture of those two) as the saturation falls short of 100% for them. Red is also off at lower saturation levels and rises above the 3.0 error level target. With luminance, the errors are similar, except for White as it tries to hit 1000 nits and can’t.

The color checker here is a much smaller one than usual but still provides good insight. I failed to correct the numbers, but the average error (dEICtCp) with luminance included is 3.50 while without luminance it is 2.01. Again the largest errors are with Red and Green that cannot hit their saturation targets, but most colors fare well as they aren’t 100% saturations and the TCL does a fine job of mixing them together.

Color Volume for the TCL 65R617 comes in at 345 MDC using the Dolby method, 63.47% of the space using the DCI gamut and 1000 nits as a target, and 42.9% using Rec.2020 and 1000 nits. The chart below shows this compared to other displays and it notably comes in behind the 2018 Vizio P-Series, the main competitor this year IMO. These are all far better numbers than the 2017 TCL P-Series, but most other TVs made a bigger leap this year than the TCL did.

One final note is that people I trust that tested the 55” TCL R617 managed to get 1000 nits or more from their samples. Usually a larger screen has brighter highlights, but in this case, it might be a smaller one that does better. I’d likely still choose the larger screen for the immersion factor, but if you go smaller, you might get superior HDR performance.

TCL 65R617 Image Quality Notes

As I mentioned in the SDR section, the luminances for the image below 60% had more visible errors than above. Some of this is because of the gamma selection, but also the 65R617 has some shadow crushing that you can see at times. Watching the opening of Black Panther, in the jungle, you miss some of the fine details that I can see on displays like the Sony X900F or LG C8 OLED. It isn’t awful but it is certainly there and not something you can fix.

Another issue is vignetting that is clearly visible on the sample I used. In most cases it didn’t bother me while actually watching content, but in scenes that were bright or had a more uniform palette, the dimmer corners stood out. Most lay viewers didn’t notice it, but fellow writer Stephen Hornbrook did almost instantly. My assumption is that the backlights in the corners have no overlap with another backlight in that area, which causes the relative dimness compared to the rest of the screen, where there would be overlap, but I can’t be certain.

Our sample had no banding, but we know some people are receiving displays with banding from TCL and the only remedy is to exchange it.

Motion on the TCL is generally fine, but not up to the level of a 120Hz panel. The panel typically runs at 60Hz but can run at 48Hz when needed for 24Hz content. Watching the first episode of The White Rabbit Project on Netflix, when the schematics pan around 40 minutes in, it is jumpy and juddery on the TCL, which much smoother on the 120Hz panels we tested this year. No 60Hz content like basketball we didn’t notice this, but 24Hz content with pans you will see these issues possibly. The Netflix example is one that is very easy to see compared to most but be warned based on how sensitive you are to this.

The full array backlight algorithm isn’t as good as those from Vizio, Sony, or Samsung. During the panning scene in Chapter 12 of the final Harry Potter, with Voldemort on the hill outside Hogwarts, you can see zones fire on and off as it pans around. It isn’t as invisible and clean as it is from other displays. You might not notice it depending on how much you pick up on this, but it just isn’t as refined.

TCL 65R617 Conclusions

For the price, the image on the TCL 65R617 can’t be beaten. It offers accurate HDR and SDR modes, good calibration tools, and looks good with most content. Dolby Vision support is also good to see as more and more discs support it. Compared to last year, though, the TCL is in a tougher position. Spending 20% more gets you a Vizio P-Series (analysis forthcoming) that offers improved motion thanks to a 120Hz panel, improved black frame insertion, and brighter HDR highlights with better color volume numbers. Strictly on an image-quality standpoint, the Vizio offers superior performance for not much more.

The TCL is still a good TV. The numbers are good, but there are some issues that don’t come up in the numbers as you can see. I recommend seeing it in person if possible (Best Buy stocks the R615 model, which is identical aside from the remote) to see if the vignetting and motion are issues for you or not. It’s a perfectly good TV, just spending a little bit more can fix the issues that it has.