Hisense PX3-Pro Ultra Short Throw Triple Laser UST Projector Review

About The Hisense PX3-Pro

Hisense’s $3,499 MSRP latest entry into the Ultra Short Throw Projector market is a trichroma (red, green and blue) laser, 0.47” DMD DLP projector rated for 25,000+ hours of laser life. They claim 3000 lumens of brightness with a native contrast ratio of 3000:1. In my measurements the PX3-Pro exceeds both of those claims with a highest native contrast of 6350:1 and over 3400 lumens in brightest picture mode, and 2580 after calibration to D65 (UST lumen measurements have a wide +/-25% accuracy range due to angle of the projector, my measurements are average across three angles from 9 positions)

The PX3-Pro is feature rich supporting all HDR formats : HLG, HDR10, HDR10+, and Dolby Vision. Hisense claims 110% of BT2020 color space, and while the measurements didn’t match that, it still impressively has near full BT2020 coverage (98% measured). Measurements also showed quite good out of the box color accuracy overall.

Two front firing 50w Harman Kardon “Dolby Atmos” enabled speakers provide great sound quality and volume for built in speakers.

Screen sizes from 80” to 150” can be had with a throw ratio of just 0.22:1

3D is also supported, with glasses being required and sold separately.

Low input lag and all major current gaming features are supported, along with a partnership with Xbox for “The World's First Designed for XBOX UST Projector”.

Unboxing & Aesthetics

The Hisense PX3-Pro comes well packed in this box, which shows the Xbox branding for it is “The World's First Designed for XBOX UST Projector - With the HSR 240 high refresh rate and low-latency mode, get ready for smooth, lag-free gaming.” It also comes with a cleaning kit, and two sets of soft white gloves.

The gray buttons on the remote are backlit, however something must be pressed first for the lighting to activate. The projector itself is fairly slim and nicely designed overall.

There is a recessed power button on the right side of the projector. It also includes sensors for an eye protection feature that will dim then turn off the image if it detects an object within a few feet. This is great to have with children and pets.

Connectivity

There are three HDMI ports. 2 are high speed HDMI 2.1 ports and port 3 is an HDMI 2.0 eARC port. That means you have two open/available ports for an Xbox Series console and a PS5 or PC. It also means you can’t extend beyond two if using the eARC port to an AVR or soundbar, unless you have a dedicated 2.1 switch on port 1 or 2. You will also find an ethernet, headphone jack, optical audio output, a USB 2.0 port and a service port on the back. The PX3-Pro also has Wifi 6E and bluetooth 5.3 for wireless connections.

Projector Setup

During the initial setup it will prompt you to align the projector and use an automated or manual keystone correction. The automated feature would have you link your phone to it to take a picture, however I would suggest ignoring this and do your best to align as much as possible without using keystone correction. Using keystone can result in a softer or worse image overall. Also in game mode any keystone will be disabled for faster input latency. After this, you can adjust focus if needed, however for me it wasn’t.

When aligning the projector, there are four feet on the bottom that can be adjusted to raise or lower each corner. Using these and placing the projector the proper distance from the screen, I was able to align the projector to the screen well enough without using any keystone correction. Of note, one of the four feet on the unit I tested felt as if it was cross threaded from the factory and was more difficult to turn than it should have been. That was a small annoyance that was likely a QC concern for this one unit, and doesn’t matter after I finished aligning it.

GoogleTV is the OS and software for the PX3-Pro, which is typical for Hisense. The main home screen is easy to navigate and apps are easy to get to. The remote has hotkeys for Netflix, YouTube, Prime Video, Disney+, Tubi, and a customizable hotkey for an app of your choice.

If you have any connection issues with devices, check the HDMI format settings and set appropriately to your device. Game consoles and streaming boxes should be on “enhanced” format while some cable boxes may be older and work better with standard.

Image Quality

The initial picture mode out of the box is Standard which will have a very blue and over processed image. I will cover recommended settings, measurements, and more in the calibration section below. For now, switching this to Filmmaker and setting “content detection” and “automatic light sensor” to OFF would be a good start for most.

The Hisense PX3-Pro hits where it counts. It offers loads of features with excellent overall image quality. Between high contrast levels, great color accuracy, and nearly full coverage of the standard HDR color gamut BT2020, the PX3-Pro brings the best overall experience I have seen in a UST projector.

Post calibration to D65 with a 0.6 gain ALR screen I measured 106 nits of brightness, which combined with the Ambient Light Rejection screen provided a bright and highly viewable image even with a moderate level of ambient light in the room. With projection there is also no glare or reflections from lights facing the screen as well to distract you from what you are watching.

The Spears & Munsil UHD benchmark disc was used after calibration (covered below) for initial evaluation, followed by various movies, TV shows and games.

Colorimetry

The color “wave” image shows each primary and secondary color from black to white in BT2020. It gives a visual representation of the gamut coverage ability of the display.

The red cactus in the Spears & Munsil demo is graded to go beyond the smaller HDR gamut of P3 and extends deep into BT2020. This is an example where a display with lower gamut coverage would not be able to show the deep saturation of this image accurately. As well as the purple flowers in the “squirrel” scene as shown here.

Colors do desaturate somewhat in dark scenes, as seen here in the loss of red saturation in the robes from The Long Night episode of Game Of Thrones. However most displays either desaturate or over saturate in the lowest stimulus levels, and is something that most would not notice or care about.

However color gradients in dark scenes are handled well with only very mild banding or posterization around the glow from the flames in this very challenging scene from 1917.

Image Sharpness and Precision

The level of sharpness and detail is also excellent. The PX3-Pro uses pixel-shifting to achieve a full 3840x2160 image. The lens quality is very good overall, with nice sharpness in all corners. The “Super Resolution” setting may appeal to some, but does add artificial sharpening artifacts.

Look for the light gray rectangles appearing on the sides of these photos with “Super Resolution” or increased sharpness slider.

The sharpness slider on “0” is the neutral setting, however up to “10” should be fine for most with minimal artifacts.

Reaching the halfway point of the slider the artifacts and ringing around edges are very noticeable.

You can see the effects of the Super Resolution in real content with the fur in this image being over processed.

The smooth gradient setting can also remove sharpness and details the higher it is set, and I would recommend just leaving it off.

Contrast & Shadow Detail

The camera exposure is increased in this image to show the dynamic range low pattern from the Spears & Munsil disc. After calibration the PX3-Pro exhibits very good shadow detail, and had a measured black level of 0.0337 nits and 0.0168 in its best HDR mode (and full off of 0.0131). This combined with 106 nits of peak brightness gave an excellent native contrast ratio of 3200:1 and full on full off of 8200:1.

The tree line in this high contrast scene is visible with the expected amount of detail. Other displays may display the tree line too dark to make out individual trees, or over brighten the darker parts of the scene reducing the contrast and visible depth.

Motion Handling

Any motion setting other than “Off” or “Film” exhibits artifacting in complex images. Motion interpolation is a preference that varies from one person to the next. I recommend the “Film” setting and used that for my testing, which did not exhibit any Soap Opera Effect (SOE) or visible artifacting. Overall I didn’t have any issues with the motion handling of the PX3-Pro with the “Film” setting active. A setting of “clear” may be preferable if you’re sensitive to visible judder on panning scenes.

Gaming

Gaming performance is excellent on the PX3-Pro and while they partnered with Xbox, it does well with PlayStation/PC/Switch as well. Input lag ranges from as low as around 5ms in 240hz (1080P PC) to 17ms at 4k 60hz. 4k 120Hz signals are accepted, however they will be displayed as 4k 60Hz. I noticed when setting up the Xbox HDR settings and not in an actual game, but with the PX3-Pro in game picture mode, Dynamic Tone Mapping was available. However once in an actual game, that setting among others disappear from the menus of game mode. However tone mapping still occurs and does a good job keeping games looking bright and punchy without blown out highlight details using the settings below. Shadow detail and color performance were also very good in game mode. Overall the PX3-Pro offers a great large screen gaming experience.

I settled on setting the HDR settings to 500 nits on the console and games with HDR sliders. Lower would reduce the APL (average picture level) luminance too much, and higher can blow out highlight details too much.

The PX3-Pro has pretty much all green check marks on the Xbox Series X console, with the one exception being there is no Dolby Vision game picture mode. Which is not a negative, as Dolby Vision gaming is not something to be concerned with or recommended.

Measurements and Calibration

The PX3-Pro has decent color accuracy in modes such as Filmmaker, IMAX, and Theater, however a calibration is recommended and can elevate it to excellent. There are a few notes to keep in mind through this section:

First, a few minor or basic adjustments can make a lot of difference in getting better accuracy before venturing into the advanced calibration menu. That menu should only be used on a per unit basis with measurement tools and know how. Copying those advanced settings are just as likely to result in pushing your particular unit and screen combination further away from the accuracy standards as it is to making it any better. I would only recommend the more basic user menu settings be copied.

Second, the black level measurement will be lower in full screen black than it is with any pixel activity. In a full fade to black scene, the black level of the image will be darker than if even one pixel is active. In my measurements I measured a full off black floor 0.0131 nits, which raises to 0.0337 nits in almost all picture modes. The exception to this is in the “IMAX Mode” picture mode which is only available in HDR10, it had a better black floor of 0.0168 nits with added active pixels. Also in Dolby Vision Dark and Game Mode the full off black level doesn’t drop like the other modes and maintains 0.0337 nits.

Third, when measuring HDR formats on a projector it is common practice to use a luminance multiplier when measuring how different settings are affecting the image. This is due to the lower luminance levels of projection compared to TVs and their ability to follow the EOTF curve. In the HDR measurement screenshots they have a 2.5x luminance multiplier applied to see how different settings such as dynamic tone mapping, active contrast etc effect the EOTF curve. So while the actual peak brightness is around 106 nits, with the multiplier you would see it reported as over 280.

Fourth, some settings either didn’t work or had odd functionality. For example color saturation in the CMS controls would have no effect when trying to decrease saturation. To get around that I had to reduce the overall color setting then increase other colors saturation individually in the CMS controls. Some of the 20pt settings either didn’t have an effect or could cause posterization in some steps of the grayscale of certain modes, and with just the 2pt I could attain good results so I opted not to use multipoint. It is possible that this will be corrected in a future firmware update as this unit reviewed was one of the first retail models available in the United States.

SDR

Starting off with SDR Filmmaker we can see it is targeting 2.3 gamma and has too much red through the grayscale. Resulting in an average delta error of 5 (around 3 is more acceptable with below 2.3 being imperceptible by eye for most people, and under 1 considered reference or near reference accurate). However that error is mostly from the gamma being brighter than the target of 2.4, and also the PX3-Pro has a drift period while it warms up. After letting it run for a while the grayscale got tighter as seen in the second image, which compares the first to a second sweep done later. So after warming up the average error dropped to just 2.3.

The ColorChecker out of the box in Filmmaker gave an average dE2000 error of 3.56, a pretty good result that can become very good with some minor adjustments.

Before making any changes I also checked for clipping. The legal range for video is 16-235 while full range is 0-255. However some real world content can exceed 235 so having some amount of headroom without clipping can be beneficial for when that occurs. Out of the box the PX3-Pro does well not clipping until 107-108 stimulus (109 is peak white / level 255). However with projection some clipping can be more acceptable than on TVs when you are trying to attain a certain luminance output. In this example with the 0.6 gain screen for testing, it wasn’t reaching 100 nits so increasing the contrast setting can increase the luminance while also lowering the clipping point. To have no clipping through the full range a contrast setting of 35 was needed which lowered the light output to 90 nits. In the end I found a contrast setting of 45 had an acceptable level of clipping and raised luminance to a peak of 108 nits.

The PX3-Pro has two “Brightness Enhancer” settings that I would not recommend ever be used. One makes the image overly cyan, and the other extremely green.

While the luminance would increase with those settings, the resulting image is what I would consider unwatchable.

There is also an “HDR Enhancer” setting in SDR that raises the brightness of red and green and doesn’t appear to do anything else. First image is with it off, the second image is on.

Next is the “Dark Detail” setting which just increases the luminance in darker parts of the image by adjusting the gamma at the lower end of the grayscale.

After testing the available settings and picture modes, using Filmmaker picture mode for SDR with a gamma setting of 2.4 and contrast of 45 (laser luminance at max 10) all processing settings turned off, and a color temperature of “Warm 1” (the “Warm 2” setting has way too much blue) gave the best results before going in to the calibration menu. This lowered the grayscale dE2000 to an average of just 1.4 and tracked 2.4 gamma very well. The color saturation sweep was also very good at this point before using the available CMS controls.

After using the 2pt and CMS controls in the calibration menu things tightened up even more. Luminance dropped slightly after the gain adjustment to 106 nits, while the black level remained at a full off of 0.0131 nits and with active pixels 0.0337 nits. If any of the bias controls (the darker end control of the 2pt) are adjusted to a positive setting the black floor would rise significantly. The bias control should either not be used, or only used sparingly in a negative setting.

Post calibration grayscale resulted in an average grayscale dE2000 error of 0.6, and was very close to 2.4 gamma target. Color saturation sweep came in at 0.8 average and the large ColorChecker SG had a 1.18 average error. Overall an excellent outcome.

In SDR game mode, using the same settings as Filmmaker resulted in approximately the same results. Aside from the full off black floor which doesn’t drop from the typical black floor of 0.0337 nits. Still an excellent 3400:1 native contrast ratio in game mode.

Moving on to HDR10, after testing Theater, Filmmaker and IMAXmodes, I would recommend using the IMAX Mode. There are some misconceptions out there that believe this is required or related to the “IMAX Enhanced” content on the streaming app Disney+. Those are two different things. On the content side (Disney+) it is an aspect ratio change of the content and an enhanced DTS audio format. On the display side (also on other brands such as Sony) it is just another picture preset and is not a Dolby Vision or HDR10+ format competitor as some think it is. It can have additional picture processing “behind the scenes” that are not available settings in the user controls, and in this case the IMAX Mode picture mode has the best native black floor of all the picture modes. Otherwise, it measures the same as Filmmaker and Theater does if the user menu settings are set the same.

With 1000 nit metadata the grayscale is nicely balanced out of the box. Remember this is with a 2.5x luminance multiplier to get a baseline for the EOTF tracking when testing how other settings will change it.

Changing the metadata to 4000 nits we can see a nice roll off that will allow it to maintain plenty of highlight detail in brighter content. However the dynamic tone mapping setting is recommended to be on, so that darker content with 4000 nit metadata doesn’t become too dark.

The active contrast setting on low will darken the lower end and raise the midrange.

On medium it has a more significant rise through the whole mid range.

And on high it has a large increase in mid range luminance.

ColorChecker out of the box is again quite good. In HDR it is more appropriate to look at delta errors in the “ITP” formula, but I will show both ITP and 2000. ITP is more aggressive and will generally report higher errors. In dE2000 there is an average of 2.18, and in ITP 8.52.

Using IMAX picture mode with laser luminance of 10, contrast of 50, black level -1, dynamic tone mapping on, and color temperature of warm 1 with all other processing or “enhancement” features turned off again gave excellent results. Very little adjustment was then needed in the calibration menu for the 2pt and CMS controls. Saturation sweeps of both BT2020 and P3 within BT2020 have low errors of around 1 average in dE2000.

The ColorChecker SG was fantastic for HDR with averages of 0.86 dE2000 and 2.79 in ITP.

Gamut coverage was only slightly reduced during the calibration to 97% of P3 and 96% of BT2020.

Game mode in HDR10 was also excellent with the same settings.

Moving on to Dolby Vision, I used the Dolby Vision Dark picture mode. DV Bright lifted the EOTF and APL of the image too much and is not accurate, but in a room with ambient light could make for a good choice in that scenario. DV Custom is the same as Bright.

DV Dark with the Warm 1 color temperature has a different out of the box white point than SDR and HDR10 Warm 1 does. It is a little cooler and less red compared to the others being a bit too red. Warm 2 is still much too blue though and still would recommend using warm 1. Also using the same 2.5x luminance multiplier we can see DV Dark is a little brighter in the mid range compared to HDR10.

Before calibration the ColorChecker in DV Dark was really good with an average dE200 of 0.92 and ITP of 2.29. P3 in BT2020 saturation sweep shows nice tracking as well.

Post calibration of DV Dark showed results inline with SDR and HDR10, excellent for a UST projector of any price.

ColorChecker dE2000 now has an average of 0.58 and ITP of 1.74.

Final Conclusions

With the full range of HDR format support, excellent gaming chops, decent built in audio, and even 3D support there is very little more that could have been expected or wanted out of the PX3-Pro. However there is always room for improvement. The few minor negatives such as some visible RBE and speckle effects with the ALR screen I tested with, or some loss of saturation in lower luminance levels, these would not be much of an issue for most users if they noticed them at all.

The Hisense PX3-Pro easily earns my top recommendation for a UST in its class, and frankly among all USTs should be at the top of your list. The balance it strikes for feature set, image quality, and value is simply top notch. Only a few other UST projectors can compete in the contrast performance which is the most important aspect of image quality. The luminance output while not top of the charts is very strong, and none of the brighter USTs offer this level of image quality. Even in a dark room this would challenge many traditional long throw projectors, even ones costing more. The performance of the PX3-Pro could be taken even further with the inclusion of an external video processor, considering the excellent BT2020 gamut coverage it has in addition to its strong contrast.

Tested and Recommended Settings

SDR

HDR