HTS-X2 - an experimental N8 reflective screen mix

[Harpmaker]

We have a number of experimental mixes in development; one that has been designated HTS-X2 has recently been tested by 1canuck2 and he will be adding info and photos to this thread.

HTS-X2 is part of a family of mixes that will run from N9 to ~N7.6. The paints used in these mixes are readily available in many, if not most, areas of the U.S. and Canada. Basically, a N6 gray paint is added to regular Cream&Sugar™ to get darker shades.

HTS-X2 is a N8 reflective screen mix.

Addendum:
This is the mix that will be called Elektra™ N8.

[1canuck2]

My HTS-X2 was a Behr-based mix. Here's the formula Harp gave me:

HTS-X2 is a 8:1 mix of C&S™ and N6 paints. It doesn't matter which C&S™ formula you use. I'm assuming you will use Behr paints from Home Depot.

The N6 tint formula for Behr is:
For 1 quart
Base = Behr #1854
Tint:
      oz.     348 oz.
B -  1          6
C -  0        79
F -  0        22

Depending on how much final mix you need you could always have them make this up in a 8 oz. sampler.

The breakdown of the HTS-X2 mix is:
32 fl. oz. Behr #1850 tinted for C&S™ #3

C&S™ #3 base.
1 quart of Behr #1850 
Tint   oz.   384th oz.
C        0       4
F        0       1

16 fl. oz. of Craft Smart metallic Silver

6 fl. oz. Behr N6 paint.

This totals 54 fl. oz. of final HTS-X2.

The bad news is when I went to Home Depot to pick up the paints, they told me both 1850 and 1854 have been discontinued. I had to go to three Home Depots to find one quart of each base and get the mix I needed.

[1canuck2]

I mixed up the paints and used about 2 fl oz of distilled water to rinse out the various containers. It was still pretty thick looking but rolled nicely. I used the Tiddler single roller method completing each "stripe" with a gentle downroll, overlapping the previous by half a roller.

My first coat scared me because after it had dried for a couple of hours, it was "blotchy" (there was a speckled effect of dark and light, pretty consistently across the screen as if some parts dried thicker and other parts dried thinner). I rolled the second coat at about two hours and after an hour it looked tons better. I decided to roll a third coat since I had the paint to spare.

After the third coat had dried for about 90 minutes I could not resist and fired up the PJ. Without it fully curing and without a greyscale recalibration, it still looked pretty good to my eye! Blacks were definitely better and white still looked good. I watched a few minutes of Bolt, there's a scene where Bolt (white furry dog) and Mittens (black furry cat) are wrestling around and the detail in both the black fur and white fur was pretty good. I also checked out my standard test movie, the Fifth Element and it looked superb. So first impressions are high!

I took a bunch of photos of the process that I'll post here later. I will also try to take some "good" shots of the screen once it has cured (and I learn how to use the white balance feature on my camera...)

[1canuck2]

Here's some pics.

Supplies all laid out:


Two quarts. One is the base for C&S one is the N6 paint Harp came up with.


The C&S base before mixing in the Craft Silver Metallic:


HTS-X2 mix in the can:


HTS-X2 again:


Ready to roll:


The next three shots are the screen (124" diagonal) after rolling three coats of HTS-X2 and a 28 hour cure.

All the lights on, camera flash on:


Partial lights on:


Lights off, just the flash:


I don't really see a difference between the three shots, I guess that's cos the flash fired off for all three. Should I try a shot with the lights all on and no flash? I am not much of a photographer beyond pressing the button on a decent point and shoot, but if there's any particular shots you want to see, let me know and I'll take them. Harp suggested taking a screen shot after calibrating the white balance to a white screen from a calibration disk. I'll try and grab that and any others you want to see tomorrow night after the paint has cured for a full 48 hours.

[Harpmaker]

Sorry about that 1C2, in my PM to you about this I failed to mention I was talking about taking screen photos of movie still frames and not the naked screen - my bad. For taking shots like that you would almost certainly need a camera tripod since the shutter speeds will be quite long. If a tripod is not available, try to put the camera on something that can't move (tall table, table with books stacked to bring the camera up closer to "normal viewing height", even the back of a chair can help).

[1canuck2]

Hey Harp, I understood that your suggestion was for movie stills. I haven't tried the white balance thing yet, I am hoping for a couple of other shot suggestions people would like to see and I'll do them all tomorrow. The shots above were just everything on full auto on my camera at various light levels.

Also, I have a tripod (the three shots above were all taken from about 17 feet back from a tripod) and I'll be sure to use it for the movie still images too. Any specific movies that you want to see images from?

[Harpmaker]

1canuck2 wrote:

Any specific movies that you want to see images from?

Wbassett is the movie guru in these parts so he might have some suggestions for you, but you can never go wrong with the screenies from The Fifth Element.

[1canuck2]

Hi,

I've been a bit tardy about calibrating my PJ after applying the HTS-X2, but today I finally found time to lie on the floor for two hours running through a calibration.

My PJ is an Epson 6500UB with only ~75 hours on the bulb. I only have a Spyder3 colorimeter, so take the results with the appropriate pinch of salt given the gear I have. I used HCFR and followed the procedures in a thread I found on AVS using the 75% Saturation windows. Gamma was a bit of a pain in the bum, but I think the results look good.

Attach is the CHC file (zipped) from HCFR showing the post calibration grayscale results. I am not an expert and don't really know if the attached results are anything to be happy with, but my eye says it looks good.

I plan to take some screenies (hopefully tonight) and will set the white balance of my camera to a 100% White screen from the calibration disk I used (AVSHD709) before taking the shots.

[mechman]

Gonna take a look at it 1C2. Have to load HCFR though.

[mechman]

1C2,

Get readings directly from the pj - the 11 grayscales and the primaries and secondaries. Then get me a second set from the screen.

[1canuck2]

Uh, dumb guy alert. You are going to have to explain a little more clearly what you want me to do... sorry

[1canuck2]

Here's some screen shots. All are from MKV files on my Popcorn Hour as opposed bluray originals. My camera is a Canon G9. I had it mounted on a tripod about 17 feet back zoomed in. Lights were all off. I first did a custom white balance against a 100% white screen from the AVSHD disc.

Fifth Element (1080p) (I was a little too zoomed in for this one). The third shot looks a little "edgy" but it doesn't look so bad i real life:





Bolt (1080p). The fur details are pretty good in this movie:





Finding Nemo (720p)

[Harpmaker]

Those are some pretty amazing shots 1C2! Thanks!

I have to ask, how close are the photos to what the image looks like on the screen with your eyes?

[1canuck2]

I'd say they are pretty close. The only pic that doesn't look quite as good is the Leeloo falling one which in the pic looks more "edgy". Perhaps the freeze frame betrays the special effects more for that shot...

The Nemo and Bolt ones a pretty spot on.

The photos are completely untouched. I did the custom white balance, used a 2 second timer with the camera on a tripod, took the pics on Fine at 4000 x 3000 and then uploaded them straight to PhotoBucket which resizes them with whatever method PB uses.

Oh, and I will pop the sample card I painted for you in the mail tomorrow so you should have it by end of week... I am interested to see what your analysis shows.

[mayhem13]

Vey nice screen shots.....contrast looks fantastic!

[Harpmaker]

mayhem13 wrote:

Vey nice screen shots.....contrast looks fantastic!

Very nice indeed! I'm still blown away by them.

A major component in these photos is that the PJ was calibrated to the screen (which is D65 neutral) and the camera was white balanced to the screen with a white image being projected.

The reason, IIRC, that 1C2 went with the N8 screen was to improve image contrast and black levels over his white screen. I would say it worked!

[custard]

1canuck2 wrote:

Uh, dumb guy alert. You are going to have to explain a little more clearly what you want me to do... sorry

the measurements you have taken are with the colorimeter pointing at the screen.

what Mech has asked for is for a second set of measurements with the colorimeter pointing directly at the PJ. the second set is used as a baseline measurement and helps us see if the screen is causing a color push.

i have an i1 display and it freezes when taking direct readings from the screen. hopefully you dont have the same problem.

[1canuck2]

Harpmaker wrote:

Very nice indeed! I'm still blown away by them.

Well the thanks goes to you... you gave me the formula. I just did the grunt work.

Harpmaker wrote:

A major component in these photos is that the PJ was calibrated to the screen (which is D65 neutral) and the camera was white balanced to the screen with a white image being projected.

The pictures came out better than any other's I have taken, which never seem to do the real world experience justice, I think the white balance idea was the clincher. I would say they are pretty close to what I see with my naked eye. Nemo, especially, looks superb.

Harpmaker wrote:

The reason, IIRC, that 1C2 went with the N8 screen was to improve image contrast and black levels over his white screen. I would say it worked!

This was indeed the reason. I had two goals.
1. For relatively little investment, get a screen that would perform as well as, if not better than, something I'd have to spend $1K+ to achieve if buying a "real" screen. I feel this is achieved. The results have more than exceeded my expectations.
2. Be able to handle a little ambient light for when we play Wii and still have the colours look decent.

custard wrote:

the measurements you have taken are with the colorimeter pointing at the screen.

what Mech has asked for is for a second set of measurements with the colorimeter pointing directly at the PJ. the second set is used as a baseline measurement and helps us see if the screen is causing a color push.

i have an i1 display and it freezes when taking direct readings from the screen. hopefully you dont have the same problem.

Thanks custard. I PMed Mech for some specifics, so now I just need to do it. I am using a Spyder3, so we'll see how it goes. A couple of times when doing 0 IRE readings for the grayscale, the Spyder and/or HCFR froze up, it doesn't do well in low light, but I have never tried blasting it with the light directly from the PJ.

[mechman]

It should be a tad bit quicker getting them from the projector! Last night I did a quick comparison fo HCFR and CalMAN. The numbers are pretty much the same with regards to xyY. For the dE, they were close to the same for gray scale but different for gamut. I think that was due to on e or the other using a different formula and me hurrying through the readings while not paying close enough attention. Regardless, HCFR did very well.

[1canuck2]

Sorry for the tardiness. I was off work today so carved out some time to do the readings Mech requested. Hopefully done correctly!

In the attached zip file should be two files:
ProjReadings.chc - HCFR readings for Greyscale, Primaries and Secondaries with the Spyder3 roughly halfway up the screen, about 5 inches from the screen, angled toward the projector lens. (I took realtime readings and fiddled with the angle to tweak the highest ftL)

ScreenReadings.chc - HCFR readings for Greyscale, Primaries and Secondaries with the Spyder3 just below halfway up the screen, about 12 inches from the screen, angled toward the halfway point of the screen. (I took realtime readings and fiddled with the angle to tweak the highest ftL)

I don't really know what I am looking it in truly analyzing these files, so I'd appreciate some "here's what you discovered" type summary after you've analyzed them, if only for my own learnin'.

[mechman]

1C2,

Basically, all we're doing is comparing the two. The readings from the pj is what is hitting your screen. And the reading from the screen is what is reflected back. If there's a difference between the two, the screen is affecting the image. If it's a big difference, as I've seen with some things, it may be detrimental to the effect of not being a viable solution.

My initial look at the numbers shows me very little difference between the two readings. I see a very slight drop in blue, but nothing significant. Looks good so far!

Also, I disregard all of the readings below 30IRE. I'll have more for you later this evening when I have a bit more time to look at these files.

[1canuck2]

mechman wrote:

Basically, all we're doing is comparing the two. The readings from the pj is what is hitting your screen. And the reading from the screen is what is reflected back. If there's a difference between the two, the screen is affecting the image. If it's a big difference, as I've seen with some things, it may be detrimental to the effect of not being a viable solution.

This part makes sense. Using computer numbers, if I shoot a 255,128,0 and get back a 249,128,0 then the surface I am shooting on is "absorbing" 6 points of red and is thus not completely neutral.

My original question is more understanding the significance of the various graphs available in HCFR.

mechman wrote:

My initial look at the numbers shows me very little difference between the two readings. I see a very slight drop in blue, but nothing significant. Looks good so far!

Good example, in one of the graphs I did see a difference in the blue line from the two different sets of readings.

mechman wrote:

Also, I disregard all of the readings below 30IRE. I'll have more for you later this evening when I have a bit more time to look at these files.

Good, cos I notice the numbers are all over the place at lower IRE. Is that that simply because the Spyder3 is not good at low IRE? I assume the super-spendy colorimeters do better at low light.

Slightly OT: Generally speaking, total neutrality is the ideal goal, but also exceedingly difficult to get perfect. So how big a deal is it? Is not the purpose of a calibration to correct for such differences? I understand a big push in one colour may be too big to correct for, but little ones should be no big deal. Could you actually have a screen surface that is a better overall performer in other respects but exhibits push in one colour. A good calibration can correct for the push, leaving you with a better screen than one that was perfectly neutral but less well performing in other respects?

I understand as screen formula developers that the holy grail is totally neutral, but is there a risk that goal is followed to the detriment of other goals? For example, on other forums it seems that gain or reflectivity is the goal, to the detriment of neutrality. I'm not claiming to understand the complexities of the art/science, just curious...

[mechman]

I just look at the numbers in HCFR, not the graphs. Believe it or not, after a couple years of Bill pounding this stuff into my head, some of it actually stuck! Let us know which graphs your curious about and we can help you out.

Is that that simply because the Spyder3 is not good at low IRE? I assume the super-spendy colorimeters do better at low light.

Yes. And as for the super spendy colorimeters, we may find out in the not so distant future.

Slightly OT: Generally speaking, total neutrality is the ideal goal, but also exceedingly difficult to get perfect. So how big a deal is it? Is not the purpose of a calibration to correct for such differences? I understand a big push in one colour may be too big to correct for, but little ones should be no big deal. Could you actually have a screen surface that is a better overall performer in other respects but exhibits push in one colour. A good calibration can correct for the push, leaving you with a better screen than one that was perfectly neutral but less well performing in other respects?

It is not that big of a deal when everything is close. However, when everything is not close and you're dealing with a projector that does not have a CMS, it's a big deal. Most of the pj's we're dealing with in the diy world are ones without a CMS. Even my BenQ W5000 doesn't have a real CMS. It's kind of a hybrid watered down CMS.

I understand as screen formula developers that the holy grail is totally neutral, but is there a risk that goal is followed to the detriment of other goals? For example, on other forums it seems that gain or reflectivity is the goal, to the detriment of neutrality. I'm not claiming to understand the complexities of the art/science, just curious...

Gain is so yesterday. lol It may be needed in certain cases, such as the house of worship thread, but rarely does any projector need a screen with gain. BW has a tad bit of gain - probably along the lines of a .15 or a .2 gain advantage over a matching neutral gray. That's not bad. The bad is the haphazard dumping of polyurethane and micas in a cauldron of screen goo and calling the result 'pop'. It's more like fizzle. Gain introduces a lot of issues. And those issues are usually coined as 'pop' from the maligned troller of customers (MississippiMan). If you know what I mean... Check out the Silver Fire thread. Check out the readings there. That can tell what gain from heavy doses of mica and polyurethane will get you.

Is Maurice still over at avs? Do they even still have a diy screen forum?

[Harpmaker]

1canuck2 wrote:

Slightly OT: Generally speaking, total neutrality is the ideal goal, but also exceedingly difficult to get perfect. So how big a deal is it? Is not the purpose of a calibration to correct for such differences? I understand a big push in one colour may be too big to correct for, but little ones should be no big deal. Could you actually have a screen surface that is a better overall performer in other respects but exhibits push in one colour. A good calibration can correct for the push, leaving you with a better screen than one that was perfectly neutral but less well performing in other respects?

I'll leave the calibration and colorimetry stuff to Mech and Bill, but I'll weigh in on this. A perfectly color neutral screen is very hard to get, as is a perfect anything in the real world; but luckily we don't need true perfection. Some time ago Bill and Dr. Mark D. Fairchild discussed the matter and came up with a set of neutrality standards for projection screens (which is what we currently use to determine levels of screen neutrality). Most projectors can compensate for small screen neutrality variances, but the more a PJ has to compensate the less chance it can do so properly and the greater chance it will throw other colors off at the same time; thus the desirability for as neutral a screen as possible. Also, the further a screen is from neutral the greater image colors will vary as the projected image drifts away from D65 (an effect called metamerism). Very few screens are perfectly neutral, very few PJ's are perfectly D65 - thus the need for calibration.

Yes, if you need a screen of a certain brightness and that brightness could not be attained with a neutral screen you would have no choice but to use a screen that would achieve that brightness even if it were less neutral. The thing is, screen neutrality (at least with a painted screen) isn't all that hard to achieve if it is approached from a scientific aspect.

The case above where someone hypothetically had to settle for using a non-neutral screen to achieve a viewable image is a case where the wrong projector was being used. Bill has said for quite some time that home theaters should be designed by determining what lighting conditions the screen will be viewed in (no ambient light or lots of it) and then go shopping for a PJ and screen. This makes so much sense! Unfortunately, most people either get whatever PJ is on sale, or buy one that a friend says works well, without taking their viewing conditions into consideration. That leaves them trying to find a screen that will compensate for using the wrong PJ. Many times this can be done, but has definite limitations. No screen can totally compensate for not having enough lumens to start with.

I understand as screen formula developers that the holy grail is totally neutral, but is there a risk that goal is followed to the detriment of other goals? For example, on other forums it seems that gain or reflectivity is the goal, to the detriment of neutrality. I'm not claiming to understand the complexities of the art/science, just curious...

I would say the simple answer is no. Neutrality doesn't affect gain. The color of a screen has nothing to do with it's reflective properties other than how much light it absorbs rather than reflects. The darker the screen, the less light it reflects.

Screen gain is a subject that is almost universally misunderstood. There is a general feeling that more gain is better no matter what. This is just the opposite of what it true. Think of screen gain as the ability of the screen to focus the reflected light toward the central viewer. To do this it needs to redirect light away from somewhere else to point it straight toward the audience. This means that those light rays are not being reflected to areas distant from the central area on-axis with the PJ. This makes for a darker image for those side seats. No screen generates light, it can only reflect what light it receives from the PJ. If a screen has too much gain it will also hotspot, which is when portions of the screen are visibly brighter than other areas of the screen which should be the same brightness.

We only need gain when the projector isn't bright enough for the job. In the past this had been the case more often than not, thus the need for screens with as high a gain as practical. With todays PJ's approaching the 10,000 lumen mark I personally think that the days of high gain screens are just about done.

The other mixes you are probably referring to were designed with gain as the primary goal and neutrality was deemed not to be all that important (and still doesn't seem to be). Back in the day they were designed PJ's had way fewer lumens than they do today.

The perfect screen is one that can't exist in reality. It would be totally black so it would absorb all ambient light hitting it yet it would still reflect 100% of the projected image to a wide audience.

[1canuck2]

Wow! Detailed explanations guys. I warned you I was ignorant to the complexities of the issues at hand. I knew the gain crazy folks were off the mark, more because they didn't seem to care about other factors that would seem as important, if not more. As I said before, that's how I found my way over here from AVS.

I am certainly extremely happy with the results of my screen. I am guilty of buying a PJ without considering my room first, but also think I have ended up with an apporpriate PJ all the same since I have good light control and a powerful PJ (Epson 6500UB). Thankfully it also has a good CMS, so I have been able to tweak out any potential non-neutral quirks from HTS-X2, but the final verdict on that is now down to Mech. Whatever your results, I don't fele compelled to change anything as the screne more than meets my expectations. So unless you have some numbers that show me something is way out of wack, I'll consider myself done.

BTW, I forgot to mention, my PJ bulb only has 90 hours on it. As bulbs age, do they typically change their colour representation, or just get dimmer?

Thanks again. I am looking forward to Mech's assessment.