You're right Jim, it's somewhat subjective. I can get zero grain with 1 f/stop light loss, super sharp etc. with my less diffuse GGs. Problem is when I post something with that GG, I get the "looks like a promist filter" comments. You won't see it in darker footage as much as light stuff, particularly specular highlights.

I'm convinced so far that the 20 degree POC LSD (and more likely the 15) has got the "right" amount of diffusion. Better than anything I've seen yet. It also corresponds both in terms of static grain and diffusion levels to my GG that has produced the greatest positive feedback. The upside is the the POC does it with .7 F/stops relatively less light loss.

Hmm... I've been thinking 60 or greater POC parts would be better suited for the GG as their grain is less visible, but of course the uphill battle with them is severe light loss. To a point, I'm not that hung up on what the bokeh looks like, quite frankly, nearly so much as I am getting grain as supressed as possible while losing as little light as possible.

Jim, if you don't care for bokeh, use a sandwich/veggies/home depot bag. It will give you small grain and the best light loss. Much better than the POC materials. You may even get lucky and find one with circular diffusion and then your bokeh will look great as well.

Here is a compiled table of all materials tested so far:

Beattie Intenscreen
Diffusion angle:~15° Transmission: 97%

Veggies Bag
Diffusion angle:~5x20° Transmission: 96.6%

Thick pouch + Coating on one side
Diffusion angle:~5-7° Transmission: 96.2%

Sandwich bag 2
Diffusion angle:~5x25° Transmission: 95.3%

Thick pouch
Diffusion angle:~15° Transmission: 95.1%

POC LSD5PC10
Diffusion angle:5° Transmission: 94.7%

Sandwich bag 1
Diffusion angle:~5x25° Transmission: 93.5%

Office Depot Magic Tape
Diffusion angle:~15° Transmission: 93.5%

POC LSD20PC10
Diffusion angle: 20° Transmission: 93%

POC LSD30PC10
Difussion angle: 30° Transmission: 91%

Redrock M1 GG
Diffusion angle: ? Transmission: 81.3%

Alain, could post some pictures of the wonderful sandwich, veggies etc bags...I'm looking for one here in Norway....but can't spot any of these...Thanks.

I also wonder which kind of GG the Letus35 Flip uses....

My guess would be hand ground or pre-ground, 1500 grit glass.

I beleive it to be 600 grit, which is an excellent choice for a vibrating adapter.

So replacing the GG grit by a Beattie Intenscreen should per def bring huge improvements to the picture. Right ?

http://www.dvinfo.net/conf/showthread.php?t=58599

Ben's thread on same subject.

Changing the GG to one of these holographic diffusers (POC, Beattie, Bag) will reduce the lightloss, but it may not make your picture better.

For example, here are some quick images of the bokeh produced by a couple materials. They are taken with a 50mm lens with the focus ring pulled all the way towards the near end (excuse the dirt, but was a really quick test swaping the screens):

Thick material (~7°, 95.1%):

http://mentemagica.com/35mmAdapter/BokehTest-Thick1-sm.jpg

Thick material coated (~15°, 96.2%):

http://mentemagica.com/35mmAdapter/BokehTest-ThickCoated1-sm.jpg

Sandwich double layer perpendicular (~15°, 91%)

http://mentemagica.com/35mmAdapter/BokehTest-DoubleSandwichPerp1-sm.jpg

Andy Gordon posted an image from the Beattie bokeh (~15°, 97%):

http://imagedump.filefactory.com/full.php?id=2514

As you can see, light transmission is not the only factor here. The Beattie looks very similar to my thick coated material: Bright hotspot right at the light point and then a fuzzy haze.

But the sandwich bag double layer (unfortunately the lowest transmission of the 3), gives the "nice" circular disks some of use like to see.

At this point I really do not know if there is a way around this. Any thoughts?

Alain,
Tried your software light transmission test on the SG35's GG material, and it reports 97 - 99% depending on where i place the boxes. Im not sure if this is correct or im doing something wrong with the software.

I have now redone the test with your 180RGB (was using pure 225 before) and the test now reads 95.3% - 97% so this sounds more like it.

Question, are you able to convert the %age to F-Stops lost?

According to my testing on the camera, as it correlates to the transmission numbers so far, 95% would correspond to a relative light loss of 1 to 1.5 F/stops with a 50mm F/1.4 under controlled chart conditions. More in the field.

Francois tested one of my early "coarse" GGs at around 90% transmission which is closer to 1.5 to 2 f/stops loss under chart conditions.

I'd postulate 5% being pretty close to 1 f/stop. The beatie at 97% is an approximate relative .7 F/stops loss. So 92% should be around 1.7 stops loss....of course assuming "adequate diffusion" which I suspect hovers from 15 to 20 degrees.

How does this hold up with your on-cam tests?

Thanks Dennis,

I will have to carry out some real world tests, shots with and without the adapter with the camcorder at a fixed iris.

Wayne the reason for using a 180RGB image is to give the digital camera some latitude. If you use a pure white image and take a picture of that, you may get everything blown out to white.

Now, to get a more precise measurement, we have to minimize the illumination variations along the monitor. The first step is to turn off all lights in the room so that the only light captured by the camera comes from the monitor. Second, the screen material should be placed near the center of the monitor so that two very adjacent measurements can be taken either to the side of the material or immediately below. I have found that in my CRT monitor I have less lightitng variation vertically than horizontally, so now I am making my measurements with the test boxes one on top of the other, instead of sideways. That gives me more consistent results independent of placement.

Finally, if you take the picture too close to the monitor the pixles will show up or you may get moire patterns. This will affect the readings as well. Making the test area larger will not improve the reading since the larger it is, the more variation in lighting it will pick up from your monitor. If the variation is not symetric the reading will be innacurate.

I will investigate the light loss correlation with f-stops.

Hey Alain,

Also, the other day I was in the Floral Dept looking at roses. Look at the bags they use. One side is shiny and the other side frosted. Check it out. Also these tests, are they dynamic or static?

Rh

Hey Alain,

Also, the other day I was in the Floral Dept looking at roses. Look at the bags they use. One side is shiny and the other side frosted. Check it out. Also these tests, are they dynamic or static?

Rh

Rich, you may have just found a single sided diffusion material. Have you tried it?

The tests outlined here are independent of the adapter being dynamic or static.

Alain,

Is that good or bad, I have not tried it(adapter in progess) Would you like me to send you some of it. I have a ton of it.

Rich.

Rich, were you at a Floral store or in a floral dept of a K-mart or something? im getting a high resolution network camera shipped to me next week and i need to upgrade from my static adapter that i made with a 1000grit aluminum oxide.

Alain, is there no grit size we can use to simulate this diffusion and minimal light-loss effects of a beattie or the POC materiel? Like, perhaps a 600 or 400 grit on one side of a plastic cd? of course it needs to be a spinner. or is that POC film glued over the cd a better solution for light-loss and bokeh? thanks.

also, for a test, i laid strips of the frosty off-brand scotch tape parrllel on one side of a plastic cd. so each strip overlapped its partner only by a bit. I applied the tape under running water to rid of all bubbles. kinda like how you apply window tint to car windows. anyways, the lightloss wasnt horrible but i would say it wasnt any better than my 1000grit, but it matched it pretty well.

The boquet or roses were at grocery store. I went back to look for some more and they're out right now. I'll see if someone else carries it.

Leave it to us alternative imaging guys to buy things just for their packaging. Thumbs up :)