All measurements are performed using CalMAN 5.4.0 software from SpectraCal. The workflow is one custom designed for home theater testing to better measure performance. We target the Rec. 709 HDTV color space, a D65 white point, and use the BT. 1886 gamma curve. Patterns are generated using a DVDO AVLabTPG pattern generator and measurements use an i1Pro2 spectrometer and C6 colorimeter. With projectors all measurements are made from the lens to remove the screen as a variable. Before actually watching content it is then calibrated again using the screen to make it as accurate as possible.
One feature to be aware of with the Epson Home Cinema 3500 is that it treats YCbCr and RGB signals differently. If you calibrate it using one and a source (Blu-ray player, cable box, game system) uses the other, your settings will not stick. So you need to either calibrate both, or make sure you can set your devices to only use one mode. I stuck with YCbCr 4:4:4 for all calibration and testing.
Pre-calibration the best picture mode for the Epson Home Cinema 3500 is the Cinema mode. The default gamma tracks close to 2.2 but we target the BT. 1886 gamma curve so I used a setting of -1. The RGB balance is good, but the gamma exhibits what some people would call an S-curve. It makes the shadow details a bit darker and the highlights a bit flatter. Colors also exhibit a bit of extra saturation and some tint issues that get worse as the saturation approaches 100%. The worst offender is green and cyan as you can see that those errors on the Color Checker get very high. Cyan comes in a bit under-saturated while green has tint issues that are causing these errors.
The Epson Home Cinema 3500 has a two-point white balance control and a 3D CMS to correct for these issues. The white balance controls work well and let me get the RGB balance almost perfect. The CMS cannot fix the over-saturation and making large changes causes the intermediate saturations to show problems. I focus on correcting the luminance levels and leaving the Hue and Saturation controls alone as much as possible. This results in the best overall final results.
Overall the results are much better in the end, with the skin tones showing the largest errors. Trying to use the Epson controls to fix this causes the overall errors to rise while those errors fall, so I left them alone. The overall color checker average error level of 3.47 is fairly high and means you will see color issues on the screen compared to a better projector. The largest issue remains the gamma, which cannot be made flat to match the BT. 1886 standard. If you select a gamma of 2.2 it will be closer, but we aren’t choosing to use that.
The contrast ratio of the Epson Home Cinema 3500 only measures in at around 700:1. Enabling the dynamic iris will bump this up to 1600:1 or 5700:1, but with an iris action that is noticeable. If I measure the contrast ratio using APL patterns and not full field, the numbers drop to 220:1 without the iris and 270:1 with it. It makes a difference, but not a massive one. 700:1 isn’t uncommon for this price range, but this is where stepping up makes the biggest difference. Here the $2,500 Sony VPL-HW40ES offers a native contrast ratio over 5,000:1 and more accurate color. It isn’t as bright (1600 lumens vs. 2500) but the better blacks show up in a dark room. In a living room room, as I said before, ambient light will render this moot and you’ll probably want to have the extra lumens. I can improve the contrast ratio by further reducing the brightness control, but it starts to clip shadows so no one would do it in the real world.
Overall the Epson performs fine for a living room setup. The grayscale is good out of the box but colors can be improved. The gamma is not that great, and using the controls I cannot improve it as much as I would like to. For watching TV, family movie nights, and video gaming these flaws are going to be almost impossible to notice. There are no major flaws in the Epson Home Cinema 3500 that would keep me from recommending it to someone.
|Average Grayscale dE2000:||2.77||2.69|
|Average Saturations dE2000:||3.73||2.70|
|Average Color Checker dE2000:||4.17||3.47|
|Summary:||The Epson Home Cinema 3500 projector in Cinema Mode is reasonably accurate without a calibration. Calibrating it bring the RGB balance more in line, which helps with the grayscale. The main issue is the gamma curve which cannot track the BT.1886 spec well. In a room with some ambient lighting, the kind of situation the 3500 is designed for, this won't matter as much as it would in a pitch black dedicated theater room. Projectors are still more important to calibrate than a TV because of the interaction with the screen material, so having one one on the Epson 3500 isn't a bad idea, just not an essential one.|