Home Theater Forum and Systems banner

Silver Fire v2 Comparisons

16K views 36 replies 7 participants last post by  Harpmaker 
#1 ·
The purpose of this thread is to compare the latest silver fire formulation to other diy methods to see how it holds up. If you do a comparison of your own, feel free to add it. :T We intend to compare it to several, cheaper alternatives. This first post will be a place keeper for the data on SFv2 3oz.

Spectro analysis



Gain



Grayscale and Gamut readings

Have yet to be performed. But when they are I'll put it here.

The Contenders

The first contender will be the neutral gray tint Gray Screen (Sherwin Williams SW7071) in an eggshell finish. We've done a little bit of work with the eggshell finish and have found that the myth (hot spotting) surrounding it seem to be false. Paint used was PPG Grand Distinction.​

The second contender will be a sfv2-3oz color match in satin finish. We selected the color based upon the numbers from the spectro analysis and identified a Sherwin William color that seems to exhibit the same characteristics as sfv2-3oz. Paint used was PPG Grand Distinction.​



We'll fill in more as we get to them. I know of at least two possibly three other things that I am planning on trying. And I believe Harp has a thing or two that he's gonna test out. :T
 
See less See more
2
#2 ·
SFv2 3oz vs Eggshell

This comparison will be to a PPG Grand Distinction Eggshell finish paint tinted to SW7071 (Sherwin Williams Gray Screen)

Spectro Analysis

Coming soon

Gain



Grayscale and Gamut

Coming soon



Pictures

In the pictures there are three separate samples. The small sample is SFv1 Lite. It was made by jinjuku. The large sample on the left is the Eggshell finish gray. SFv2-3oz is on the right side. A couple of things to note is that the SFv1 Lite sample is the darkest of the three, then SFv2-3oz and finally the Gray Screen panel. There are 8 photos all from 4 different angles and each angle having 2 shots - one with a moderate amount of ambient light and one with no ambient light at all. The angles are on axis, ~15 degrees, ~40 degrees, and ~65 degrees.


















Conclusions

In the photos above it is clear that SFv2-3oz has a serious problem with viewing cone. While the eggshell panel does not. On axis, SFv2-3oz is very bright. But moving the camera over about 4 feet clearly shows the change in the luminance of the panel. The Eggshell sample is brighter from there on out. Comparing two angles, 0 and any of the others, almost makes it seem like the eggshell panel is a much darker shade of gray in the on axis shots. But it is the same all the way across.
 
#3 · (Edited)
SFv2 3oz vs Satin

Spectro Analysis



RGB 170 176 178
LAB 71.3 -1.77 -1.83

Pretty nasty blue push on this sample. The sample used was SW7072 Sherwin Williams Online. Number wise it should have been a close match. But one never knows what the paint mixers are going to come up with. This was a PPG Grand Distinction brand of paint mixed at Menards.

Gain



Grayscale and Gamut



Pictures

The Satin panel takes up roughly two thirds of the left side. Silver fire version 2 is on the right. First picture is with ambient lighting. Second image is darkness. First two images are on axis, second set is ~15 degrees off axis, third is ~30 degrees off axis, fourth is ~45 degrees off axis and the last set is ~60 degrees off axis.

***EDIT*** There are no 45 degree shots. I think my brain hiccuped for a moment. ***EDIT***



















Conclusions

The satin sample is pretty much on par with the sf sample on axis for brightness... in the center of the image. Looking at the left side you can see that the whites are a bit dimmer. This is a classic example of hot spotting. And this is clearly evident as you move off axis on the satin panel. The satin panel hot spots slightly more than the sf panel. One thing to keep in mind though is that the satin panel is darker than the sf panel. The L value is 73.8 for sf and 71.3 for the satin sample. It should look a tad bit darker off angle

Looking at the blacks in ambient lighting, the winner is an easy to distinguish - the satin panel.
 
#4 · (Edited)
SFv2-3oz vs GSEgg:LBS 1:1

Spectro Analysis



RGB 198 201 199
LAB 80.7 -1.38 0.53

The paint used for this is PPG Grand Distinction Eggshell finish tinted SW7071 Sherwin Williams Gray Screen. This was mixed at a 1:1 ratio with Liquitex Basics Silver.

Gain



Grayscale and Gamut



Pictures

For this set, the Eggshell:LBS 1:1 sample panel is on the right two thirds, with sf taking up the left third or so of the test area.


















Conclusions

This test panel is what I would consider a sf-killer. Whites on axis are slightly better than sf. Blacks in ambient lighting are slightly better on the sf panel. So why take the test panel over sf? Because the test panel would cost about 1/4 of the price of a sf screen. Quart of paint would be $10 and about $15 for the Silver paint. Plus, while I didn't roll this panel, I'm certain it could be rolled.
 
#5 · (Edited)
SFv2-3oz vs Satin:LBS 1:1

Spectro Analysis



RGB 173 178 179
LAB 72.3 -1.66 -1.12

This panel consists of one part of the satin paint above (SW7072 - Online) and one part Liquitex Basics Silver. Again, this panel is darker than silver fire so keep that in mind as you are looking at the pictures.

Gain



Grayscale and Gamut



Pictures

The Satin:LBS panel is again on the right side with the silver fire panel on the left two thirds.

















Conclusions

From the photos above it appears that the Liq Silver seems to have an easing effect on the hot spotting witnessed with the straight satin paint only panel. The whites are brighter on axis with the panel getting slightly dimmer near the right edges. Blacks are definitely blacker on the sample panel than sf. As you move off axis the sample panel still retains the white edge over sf except on the first off axis shot. I think that one may be another camera anomaly though due to the WhiBal card. If you look at the card it's reflecting quite a bit of glare from the projector.
 
#6 · (Edited)
SFv2 vs matched colorant

SF v2.0 versus a matched color component

This post is meant to debunk some of the most ridiculous things you've ever read about screen paint. Here's a quote someone sent me about this particular topic:

the colorant is really individualized color components and each mutually exclusive of each other. suspended side by side in a translucent solution, the totality of the mix appears to grey, but the individual components (RGBY) each react differently to the light being projected. very much unlike how white reflects ALL colors the same and black/grey both reflect and absorb all colors the same dependent on shade.

so to make a long story short... what you have proposed is essentially what RS-MAXXMUDD already is... minus the additional LAMP BLACK that would be added by having the store try to color match the components.

and that in fact, would make substitute SF mix less than RS-MaxxMudd.
This post will contain the proof that the above statement and many others similar to it are utter nonsense. Panel will be sprayed tomorrow. Color matched color component cost a whopping $3 as opposed to $20+.

Measures



As you can see, the color match has a Delta Error of less than 1. Most paint stores would consider a anything with a DeltaE of less than 2 a good match. About the only thing I am worried about is that this will fix SF2.0's color issues. :whistling:

Spectro Analysis

I didn't get to this yet. I did a preliminary measure of the new panel last week and I recall it had a slight red push to it. It was about on par with the blue push from sf v2 3oz. I'll have it done by the weekend though and I'll post it here.

Gain

Grayscale

Pictures

There are 4 separate panels in these shots. Two of them are just for more comparison. One of the extras is the 1:1 VUP Eggshell:Liquitex Basics Silver and the other is a test I did using an N6 paint with polyurethane and Liquitex Basics Silver. The first image shows the placement of the panels.











Roughly 15 degrees off axis:











Roughly 45 degrees off axis:











Roughly 60 degrees off axis:













Some high ambient light shots:







Misc shots:



















While taking these pictures, I was laying on the floor in front of the screen with multiple remotes laying around me. From my viewpoint, the N6 conglomeration was performing better than both sf v2 and the color matched sf v2. In order to show that I took a few pictures from where my head during this test. While the color balance is off on the N6, it's amazing to me that it has an image that is on par, brightness-wise, with sf v2. I think that is further proof of the terrible viewing cone and horrendous hot spotting of the silver fire mixes.





 
#32 ·
Re: SFv2 vs matched colorant

SF v2.0 versus a matched color component

This post will contain the proof that the above statement and many others similar to it are utter nonsense. Panel will be sprayed tomorrow. Color matched color component cost a whopping $3 as opposed to $20+.

Measures



As you can see, the color match has a Delta Error of less than 1. Most paint stores would consider a anything with a DeltaE of less than 2 a good match. About the only thing I am worried about is that this will fix SF2.0's color issues. :whistling:
Don't worry Mech, it won't fix any color imbalances the SF mixes have since neither the SF base nor the SF colorant (hand mixed or color-matched) are neutral. If any given combo of base and colorant are in fact neutral it is more by luck than design. What having a colorant that can be had by giving a specific formula to a paint-person at Lowe's or Home Depot will do is greatly reduce the color variations people are seeing between their hand-made component mixes and the SF mixes made from them. This isn't about "fixing" SF (which it's designers say isn't broken), it's about making it easier to mix SF and disproving a very faulty theory on how the colorant works.
 
#8 ·
I'm curious as to why the SFv2 "looks" brighter in the 0° shot over the SW gray. Yet they have nearly identical gain readings at the 0° position.

Are the gain readings a misrepresentation (not intentional by the measurer I must add:T) when dealing with a DIY paint that is heavily laden with metallics or mica?
 
#9 ·
That is something that I wondered myself. And I have to admit that I was a bit flummoxed at first. :dontknow: But I figured that it had to do with the camera focusing on the center of the shot and the amount of light it let in from each panel. I think it has more to due with the pictures being a misrepresentation of the gain. :T
 
#10 ·
Just a general FYI, proper gain measurements are taken using a white reference target (sometimes called a Unity target) that is totally diffusive in reflection (has NO gloss whatsoever) and is as close to a pure N10 white as possible. There are several materials used for this purpose, magnesium carbonate being one of them, and that is what we use.

In practice, for every angle tested a reading is taken of the reference target with a 1° spot meter and the reflectance written down. Next the screen under test is swapped for the reference target and another reflectance reading is taken. The ratio of the two readings is the gain of the screen and is given in percentage.

If a white screen has a gain over 1.0 then it must be achieving the additional brightness on-axis by taking brightness from off-axis. The same applies if a screen has a gain of 1.0, but is darker than pure white.

Any time a gray screen is said to have a gain over 1.0 be skeptical that either the gain figure is wrong or the screen will have viewing cone problems. You can Rob-Peter-To-Pay-Paul only so much before viewing cone and/or hot spotting become real issues.

All this applies to commercial screens as well as DIY.
 
#11 ·
I updated the titles to reflect some of the tests I'm going to run. You'll notice the second to the last one states SFv2.0 vs SFv2.1 because as has been the case since we've run these tests, the formula did, in fact, change. Fortunately, I have enough supplies to mix up a few more batches if need be. :T
 
#13 ·
Yes the camera was set to 'auto-everything' with the flash shut off. That's why I suspect the camera. But who knows? :huh: When I do get time, I'll put it all back up with the MgCO block and snap a picture. The key words though are 'when I do get time'. It seems like something is always coming up lately. :foottap:
 
#14 · (Edited)
I did a little experimenting with exposure.

f10 - 1/2 sec - ISO 500



f4.6 - 1/3 sec - ISO 500



The shot in the eggshell post was shot at f4 - 1/3 sec - ISO 100

This is all I could get done at the moment. I'm going to try and get a few more shots later this evening.

Also, I have sprayed all the panels for comparisons. :T Should be able to get all the photos and data this upcoming week.
 
#15 · (Edited)
I think you are getting closer with the pics, but you have to be careful of highlight clipping. The white areas are all perfectly white 255 in the 2nd picture. I think the first one is closer to an accurate exposure, probably only 1/3 to 2/3 of an f-stop underexposed.

That is an Elite Gray screen underneath right? The eggshell is really close to that in appearance.
 
#18 ·
#19 ·
I'll be interested in the gain readings of the LBS samples. Especially that satin one. It does appear to hot spot, BUT, the off axis shots seem to be consistent, even though they are getting darker. To me the 15° looks the same as the 60°:dontknow:

How grainy is that LBS anyway? Does it color shift the paint much?

What would happen if one were to use a neutral white and use the LBS in a 2:1 ratio?:sneeky:
 
#20 ·
I'll be interested in the gain readings of the LBS samples. Especially that satin one. It does appear to hot spot, BUT, the off axis shots seem to be consistent, even though they are getting darker. To me the 15° looks the same as the 60°:dontknow:

How grainy is that LBS anyway? Does it color shift the paint much?
Mech does all the gain stuff.

LBS has a very fine grain since the mica particles are very small, but it is mica and it does refract light like it's larger grained cousins in the craft paint aisle. The secret to making a screen mix with mica that doesn't color-shift is to use enough opaque paint in the mix to control most of that refracted light.

What would happen if one were to use a neutral white and use the LBS in a 2:1 ratio?:sneeky:
It pushed blue, but not as much as other silver paints. I did have to develop a base color to correct it enough to comfortably fit within our neutrality standards when developing the new C&S™ Ultra mix (which is tinted Valspar white enamel and LBS mixed 1:1).
 
#22 ·
Mechman, would you mind letting me know what software you're using to test the different screen paints? I have an i1 Pro and have been using HCFR for screen calibrations. I have a pretty good handle on that software, but have never tested a reflective target. Any tips you could give me would be much appreciated!

Also, what would be involved in testing gain properly? I've been calculating it using the i1 Pro, but after reading your stuff here, I gather that's not the right way to do it.

Thanks!
 
#23 ·
The cheapest option would be to use the X-Rite software. It's also the biggest pain to use. Harp would be a bigger help with it than I but basically it gives you the data for the spectrum chart and then you have to export it as an excel file. Then you open that file up in excel and create a chart like this:



It also gives you the data numbers. But, IIRC, they are either done in C or D50 so you would need to convert them.


The next option would be to use something called BabelColor. The only problem with BabelColor is that the spectrums don't look as nice. Here's what they look like:



Outside of the spectrum chart, I like BabelColor.

Finally, the option that I use is CalMAN version 3.7. I don't know if this version is for sale/available anymore. The latest version does seem like it would be able to do what I do, but I haven't had the time that would be needed to set it up properly. So I keep using the old version. :huh:

Gain cannot be determined using an i1pro, at least not that I'm aware of, for that I use a Sekonic L758C digital spot meter. There are several papers written by John Ptylak from Kodak Labs floating around on the web regarding measuring screen luminance/gain. I don't have copies of them but they were the basis for how we measure gain. :T
 
#26 ·
I have the color component match - in satin since that's what the Basics acrylics are. Harp had to do the leg work on this one as most, if not all, paint store spectro's have issues matching dark colors.

But I have it and I'm planning on spraying yet another panel this Thursday. :T
 
#27 ·
So explain again what the procedure is?

Are you trying to match the color component and then mix that with the base and viscous? Or trying to match the color component AND base, then mix that grayish slew with the viscous?

My guess is the former so that you can still try to experiment with the silver and pearl?
 
#30 ·
Actually we are trying to show/prove a couple of things.

First is that a color match to the end mix that has a high enough sheen to match the mica and poly in SF will be virtually indistinguishable between the two. Of course they both will hot spot as we already know, but the point is to show that it's not alchemy, but sound color science as the reason behind what is going on.

The second thing is to demonstrate that if the colorant is matched, it could be added to the rest of the mix to the formulas exact specifications... minus of course the complex colorant. I guarantee even the developer would not be able to tell the difference between the screens. The true difference though would be that with the color matched 'colorant' the results would re reproducible every time, and right now that is not the case with SF. If you were to examine a dozen screens, you would get at least 8-10 wildly varying screens.

The colorant is just too difficult to reproduce exactly the same every time. I've seen it said that because artist paints are used it's a better quality and that store mixes vary too much. Not true. They can match paints to a very low DE and make it reproducible. If not then contractors would not be able to paint houses and entire interiors without having one wall not match the others. Private individuals don't have the same type of precision and everyone does things just a bit different, which could be enough to alter the colorant enough to throw the entire mix off. Am I making that up just to take a cheap shot? Not at all. Just read through all the people that had problems and needed 'on the fly' adjustments. Even the dev has run into some problems with a mix gone wrong when out in the field doing a job. However if the mix was neutral to begin with as well as being reproducible, there wouldn't be any need for 'tweaks' and constantly changing formulas.

Again... it's been proven that it doesn't matter if you use lamp black and white, red blue and green, or any other color combination to produce a neutral gray. If it is neutral it is neutral- period. The only difference is the more colors that are used, the harder it becomes to make make it neutral, or consistant. That and it has never been proven that the 'colorant' does what it is said to do. All you'll get is basically a 'because I said so'. For some people that's good enough. I like to know why things work and do what they do.
 
#28 ·
Color component only - the rest stays the same. It's been some time but I recall Roland and Pete saying something along the lines of that the paints 'don't mix - they stay independent of each other on the molecular level' and that it could never be color matched. etc.

So far this thread has already found a cheaper solution - Sherwin Williams Online Satin:LBS in a 1:1 ratio. The color component match will make it cheaper as well as disproving the 'molecular level' statements.
 
#29 ·
That is pure bunk. I proved that it does not matter how a neutral gray is made, if it spectro's neutral then it is neutral. On the other hand nobody has ever shown one shred of proof that color components react with a projector's light to produce a better image. It sounds good, but simply doesn't work that way. It's advertising hype to try and sell a product and that's all it is.

Honestly it took me three years to replace a simple OTS neutral gray with something else, and I tried just about all of them out there, including commercial screen paints.
 
#34 ·
Hi,

am about to build a homecinema in my basement and just planning to paint the screen... and stumbled into this forum pretty quick. But I still have no idea of what Silver Fire is. Could someone please explain that to me?

Is it just a color code for some paint or a certain paint?
 
#35 ·
Hi, welcome to HTS! :wave:

We have several threads on SF that should tell you everything you need to know about those mixes, but to put it all in a nutshell SF is a series of screen mixes developed and recommended on another home theater forum. I say a series of mixes since they have a base mix that doesn't change much from one variation to another to which is added varying amounts of a dark "color component", the more of this is added then the darker the final SF mix will be. There are many problems with these mixes, the most glaring is it is a real pain to make the "color component" because one must mix 4 different color paints together and even the slightest error will affect the color of the final mix. The developers of SF say that making this "color component" this way is very important to making SF perform properly. They are simply wrong since proven color science says that how a color is derived doesn't matter, all that matters is the spectral reflectivity of the final color. Another problem with these mixes is that they contain an overabundance of mica as the reflective element in the mix; this causes visual problems such as hot spotting or shimmering in large bright areas of the projected image. The next problem is that the SF mixes contain a fair amount of polyurethane; not only isn't this needed in the mix but it WILL yellow as the screen ages changing the color of the screen.

Some facts about SF:
No documentable shade values are given for any of the mixes (such as N9, N8, etc.).
No documentable gain values are given for any of the mixes (though lots of gain values are stated they are literally all just guesses).
It takes 8 different paints, plus the polyurethane, to make any of the mixes.
All SF mixes MUST be sprayed and cannot be rolled.
 
#37 ·
Thanks for chiming in here Bill! Your voice has been missed!

The developers of SF also like to play up the popularity of SF saying that virtually everyone that has made a SF screen likes it, but if you read their forum you will find out differently. I can't count how many SF users we have helped here to get a better performing screen.

Something I forgot to mention is that SF is THE MOST EXPENSIVE DIY screen to make I have seen. I think the cost is now upwards of $100 to even try the mix! :spend:

Most users have to ask how to color-correct their batch of SF and between mixing it up and color-correcting it spend hours, if not days, getting the mix to even approach being a neutral gray. After spending all that money and putting in all that time very few are willing to admit the screen doesn't perform as stated by the developers. I mean the screen is so expensive and time consuming it just HAS to be better than a simple gray screen of the proper shade right? :doh:

And just so folks know, my first DIY screen was made with SF so I am personally aware of it's short-comings.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top