Screen Noise

[mechman]

Not much is ever said about screen noise. And in researching Da-Lite's JKP HD Progressive screen material I ran across a review at PJCentral. In the review they talk of 'subtle artifacts' and 'sparkles' in the Grayhawk material they compared the HDP to. This can also be seen in the latest 'beta' samples from Elite - the PowerGain 3, 4, and 5. In the diy world, probably the 2 biggest culprits are Silver Fire, and S-I-L-V-E-R. This is because they use a massive amount of mica particles in an attempt to make the image brighter. While finishing up my review of the dnp Supernova lately I did some comparison shots to Silver Fire (one of the makers has always claimed that SF was better ) and the amount of noise generated by the screen was clearly evident. And no, Silver Fire is not better than the Supernova - period.

I have to admit that I'm uncertain if this would be the proper term - noise. In the past I've always used 'sparklies' or 'graininess'. But I believe video noise has always been used for the 'snow' which is the result of a poor signal. It is also referred to from bad tracking on old vcr tapes. Regardless, it's signal which is not from the origin - something that is not supposed to be there. Well the same can be applied to screens I'd guess. Some screens with a lot of reflective materials will generate this noise. The best example is that which I talked of earlier - Silver Fire. These are some shots from the Supernova review that clearly show what I'm talking about.

In these shots there is a matte white reference in the upper left and the SF is upper right. The Supernova is on the bottom with a Neutral Gray Reference Card in the red shot on the bottom right.





What do you think?

The number of pms and emails I get about adding more reflective particles to mixes that we've created is a lot. And the answer is pretty much the same, adding more would result in graininess or sparklies. I think we can now coin that as screen noise. And adding more than what is actually necessary would result in excessive screen noise. Black Widow uses 20% aluminum. The beta N8.5 Scorpion mix use even less.

It's my opinion that in order to preserve the image fidelity as best we can, one should avoid mixes with high concentrations of micas. And this isn't to say that mica is all bad. There are actually good mica paints out there. benven was known to use them I believe. But it all must be used in moderation.

[wbassett]

Good topic on noise.

I'd like to add to your last comment

mech wrote:

It's my opinion that in order to preserve the image fidelity as best we can, one should avoid mixes with high concentrations of micas. And this isn't to say that mica is all bad. There are actually good mica paints out there. benven was known to use them I believe. But it all must be used in moderation.

I'll add... not only in moderation, but more importantly the quality. Some products that are listed as 'mica' are actually ground up fish scales!

My personal stand is that we have much better products available to use.

[Prof.]

wbassett wrote:

Good topic on noise.

I'd like to add to your last comment

I'll add... not only in moderation, but more importantly the quality. Some products that are listed as 'mica' are actually ground up fish scales!

My personal stand is that we have much better products available to use.

WOW!..They really use fish scales!?..
I had heard that before, but I thought it was a fishy story (pun intended!!)..

I catch fish on my property that have really big colourful scales..Maybe I should make up my own silver paint!!!

[wbassett]

Yes some actually do, but that's typically the much lower grad craft paints, and even then some do, some don't. Even seashells have been used. Basically anything that causes that iridescent rainbow effect.

Some mica is coated with TI02, which basically makes it a big flake of TI02 and isn't really mica any longer. Why do that though? TI02 is very very fine. It's like extremely fine grain talc, and never stick your finger in a container of it! That stuff seems to stick at a molecular level! I had a white index finger for a day or two!

I digressed though from the topic if the thread... now back to our previously scheduled program.

[lcaillo]

I think the only problem with noise as a term for it is the ambiguity it may create in associating it with the physical property of the screen as opposed to the signal chain. Perhaps "screen graininess" is another option. The look that you get is more similar to film grain (and the source of the problem is similar...partical size). I would avoid the term screen "grain" because it may get easily misunderstood as "gain."

[mechman]

lcaillo wrote:

I think the only problem with noise as a term for it is the ambiguity it may create in associating it with the physical property of the screen as opposed to the signal chain. Perhaps "screen graininess" is another option. The look that you get is more similar to film grain (and the source of the problem is similar...partical size). I would avoid the term screen "grain" because it may get easily misunderstood as "gain."

Yeah that was my main reservation for using 'noise' - confusion with video noise from the signal. I guess we'll stick with graininess.