Thursday, December 31, 2020

Sharpness of the CHiQ C8UT – Preliminary notes

The CHiQ C8UT is one of the first triple-laser ultra-short-throw (UST) projectors with UHD support. It is not true 4K, as it uses the Texas Instruments 0.66” DMD chip, with a native resolution of 2716×1528. It uses a technique called XPR to emulate the 3840×2160 of UHD, and although it’s a compromise, for regular viewing it is still incredibly difficult to differentiate the image with that produced by a projector with native UHD resolution. Mainly commercialised in Asia, the only real test of the projector so far is that done by Grégory on his blog
(Grégory’s review)

I am working on my own review, from a user perspective, but I wanted to give some preliminary notes, in particular because Samsung are releasing their own triple-laser projector, the LS9PT, but also because I’m not sure I agree with Grégory on one of the main criticisms he has of the C8UT, which is about its sharpness.

My context

I’ve long been a fan of Grégory’s reviews, and I equipped my bat cave with an Acer M550 after a pretty good review from him. Beaming onto a wall painted with Screen Goo, I was more than satisfied with the results.

However, life moves on and this year I sold my house and moved into an apartment. Instead of a dedicated bat cave, I had to compromise with my family (i.e., give in) and find a solution for a different type of viewing.


Closing the blinds gives a darker environment, but my family didn’t want to see a projector hanging from the ceiling in the main room. The solution is an ultra-short-throw projector that sits below, and close to the screen. I lost on the argument of doing a new Screen Goo painted screen as well. Luckily, Grégory had done a nice review of the Vividstorm motorized screen that rises (as opposed to other screens that you pull down – something else the family would not like to see handing from the ceiling) and it made for UST projectors. So, I ordered the Vividstorm S Pro 120, which I will be reviewing separately.

Focus and sharpness of the C8UT

Grégory had some criticisms of the C8UT. One is the size and form, which is indeed a problem. The second is that colorimetry which takes some work to get right. I can’t comment on that right now, as my own calibration equipment is too old, and unless someone is willing to lend me some, it will be a while before I can get new equipment.
A third is that he found the projector was not as sharp as would be expected from a 0.66” DMD based projector. This is where I’m wondering if he may have had other issues or a lack of time to setup the projector to get the best from it. I’d like to show the results of some of my first tests that, I believe, show that the C8UT is capable of sharpness up with the best of the 0.66 chipped projectors.
First, though, a word about setup. Getting good focus on a motorized screen with a UST projector is not an easy thing. One issue is that of tension. A hanging screen can have a heavy weight at the bottom to help with achieving a high tension. It’s far harder for a rising screen to get high levels of tension. Vividstorm themselves recommend not projecting to the edges, as they curve.
Second, and one of my major criticisms of the C8UT, it is very difficult to get positioning and focus right. The projector is round, with nothing to guide you as to whether it is parallel to the screen, or if the lens is centred on the screen. I’m sure many users would end up using keystone correction, but that degrades the image and I want to avoid it like the plague.
The focus is motorized, which I find is far better than having mechanical wheels or buttons. Sounds ideal? Project your favourite calibration image and use the remote to focus. If you could do that it would be fantastic. For some reason, however, the CHiQ only allows you to use a firmware image to focus, which would be ok if it was a good image, but it is not. This is the image, after about 3 hours of positioning and doing my best to get good focus.

(NOTE: Right-click on images to open them in full size)


There are no fine details, and everything in the image seems to be antialiased!



I ended up playing a little with the menu, going out of it to my reference image, judging the result, then going back in again iteratively until I got close enough. A frustratingly long situation!
Anyway, on to the question of the precision of the image. In Grégory’s images, effectively the detail is missing in some of the finer parts of the test image, away from the centre. My experience seems to be different.
The full image:


Here is a close up of the lefthand side.


And, here is the right:


Here’s a few more images often used to judge sharpness and focus





Some caveats. I was hand holding my camera and forgot to manually focus it onto the screen using some reference point. Obviously, I also had to use a slow shutter speed, or I’d risk getting colour bias by capturing the moment when only one or two of the lasers had fired. In addition, it’s not easy to really judge by photos, as there are undoubtedly some artifacts produced by my camera or ambient environment (the photos were taken at midday, with some of the window shutters closed).
Still, comments welcome. Should I be expecting a better performance from a 0.66 DMD based projector?