This is a 10stop Lee Big Stopper plus 5stop Singh-Ray filter. The Big Stopper hold true to calculations of exposure reduction whereas the Singh Ray tested to about 50%. I have multiple images at 2minutes with the Lee, so I added another 21/2 stops which worked out to 9 minutes.
Most of the "white" you are seeing is not beach but is the long exposure effect of high tide and motion blur on the beach. I have some crops which look like "surface of the moon". If you want, I will post it.
Sat 19-Oct-13 09:24 PM | edited Sat 19-Oct-13 09:59 PM by NanoMeter
That is very interesting. I remember this question came up before, and it caught my interest, and since then I tested some Schott NG3 2mm thick glass. NG3 is not the darkest neutral density glass that Schott makes, and 2mm is not as dark as 3mm, obviously, but I did have good results stacking it with RG850 (850nm IR longpass). It worked out to be 4-1/3 stops difference, a 1 to 20 exposure ratio between the 850nm alone and the stack. So I plan to test again with darker NG glass, and perhaps thicker. I will post some tests when I get them done.
Very nice photo too. Shooting with LENR (long exposure noise reduction) turned on (in the D7000) will help clear up specks, but LENR will add the same 9 minutes processing the exposure before you can make another shot and even before you are suppose to move the camera... So 9 minutes becomes 18 minutes waiting to just see the results!
Also, I compared RG850 to RG1000 (1000nm IR longpass, @50% transmission). RG1000 produces a much longer exposure time than RG850, 3 stops difference, 1 to 8 exposure ratio. Although, results with the 850nm glass proved to be somewhat sharper on close inspection. However, if needed, stacking neutral density glass with 1000nm glass can add even more exposure time.
What is the nm of the IR glass you are stacking with?
I have tested LENR at 2 minute exposures and found minimal difference with and without it on. The 9 minute image was denoised in NIK Dfine which is fair for this. At 9 minutes I expected the noise to be much worse, especially around noon. At 100% on raw image, it was acceptable in the shadows and other areas.
This next image is a 2 minute exposure with similar conditions to above except LENR is on, but NIK was also used as above. Surprisingly the rocks on the beach are pretty sharp.
Wed 30-Oct-13 03:58 AM | edited Wed 30-Oct-13 04:04 AM by NanoMeter
Paul, I tested Schott RG850 2mm stacked with NG9 2.2mm, and got a 50s exposure that looked almost the same as the RG850 at 1/5s. I also tried RG850 2mm stacked with NG1 2.2mm and got a fairly dark image at 10 minutes. I didn't have time to test that stack longer (yet), but I think something between 10 and 15 minutes would work for that stack, depending on the light.
D7000, same settings as you, ISO 100, f/20, Different lens (18-55mm VR) @18mm.
From those tests, I decided it would be good to have a thicker version of NG9 on hand (perhaps 2.5mm and/or 3mm), as well as a thinner version of NG1 (perhaps 1mm).
I didn't use LENR for the 10 minute NG1 test, because I just didn't have the time to wait that long, but it looked terrible, lots of specks and noise. The 50 second NG9 using LENR looked perfect however.
What was the histogram like when you stacked the RG850 2mm + NG1 2.2 mm for 10"?
I have found that the only ND filter I have tested that gave a true 100% reduction in exposure time was the Lee Big Stopper, whereas others calculated to about 50%. What have you found the relationship to be so that you can accurately predict exposure times?
The beach is a good place for me to test as there is opportunity to evaluate noise reduction and to experiment with different software. Using NIK skin softening for noise reduction has some promise and appears better than other noise reduction software I have tried. Skin softening is a filter in Color Efex.
Paul, I didn't look at the histogram. This was my first test using the darker NG9 and NG1 glass. At those camera/lens settings, I had to use manual exposure for those two darker glasses, so I only came close to a target exposure in one of them, because I just ran out of time. I will let you know what I come up with when I have tested them some more.
Where is the best place for me to purchase the glass you are testing? I was on Schott website and it was difficult for me to purchase. I found the downloads for all the glass but no order form.
With the Lee Big Stopper(10 stops of ND reduction), at f20, I get a 2 minute exposure between 11-1:00. I would like to test some of your glass stacked with this to see if I can do better than 50% exposure time with other glass stacked that I have tested.
At f11 with LBS on filter holder, not "screwed" on, exposure time about 30 seconds. When I stop down to f20, time increased to 2 minutes. This filter is added in holder to my 830 conversion.
Filter holder has 3 slots, so additional filters are added and tested with LBS in slot closest to lens. Problem with LBS is that it is very light sensitive and often in combo with other filters , produces a "mirror" like effect resulting in significant flare on image. This is despite covering filter and eyepiece and trying to block out light as best as I can.
What I would like is to be able to replace the LBS when stacking it with other filters as results are difficult to edit often in post. Alone it us fine, but I want to be able to produce images with more that 2minute exposure. The benefit of the LBS is that exposure time predictions are accurate with no fall off as I have seen with other filters.
Here are the specs for the Lee Big Stopper (LBS) which I would like to equal with another glass for reasons above.
1. 4x4" (100x100mm), 2mm thick 2. Z Pro Cokin filter holder, (4x4", slotted) 3. Manufactures Exposure Guide which I have tested and confirmed to be as stated with no reduction in times as with other ND filters(50%). Normal shutter speed LBS 1000th 1 sec 250th 4 sec 125th 8sec 30th 30sec 15th 1min 1/4 4min 1/2 8min 1sec 16min
If we can reduplicate this, I would then stack other glass in holder to further times.
Paul, interesting, 2mm thick. I am not understanding the list you have. Does the list mean for example, that any photo with any filter shot at 30th of a second will take 30 seconds with the addition of LBS?
The list is for the LBS only. It means as you say that with the LBS with a shutter speed of 1/30sec without it, your exposure will be 30 sec. with it. Other ND filters would reduce exposure time to about 50% of above, that is, for the same shutter speed of 1/30sec., exposure would be 15sec as opposed to 30 sec with the LBS.
OK, most of those work out to be 10 stops, except these: 15th 1min 1/4 4min 1/2 8min Those three work out to 4 stops, unless I am calculating those wrong. But I will test out the same idea with the NG9 and NG1.
Sun 03-Nov-13 07:11 AM | edited Sun 03-Nov-13 07:11 AM by NanoMeter
Paul, I am guessing the NG1 2.2mm filter I have is about 15 stops with an acceptable exposure. The tests I did calculate to be about 12 stops, but they are way too dark. So my NG9 is the next lighter in my set, but is not dark enough to be close to 10 stops.
I am going to give a try using NG9 stacked with NG3, which looks like it might produce an optimal exposure closer to your exposure time.
When you sat that the tests with NG1 2.2mm are dark, what do you mean? Does this mean a 15 minute exposure hypothetically would produce an image? LBS is black, can't even see any image in live view at any aperture.
Sun 03-Nov-13 02:07 PM | edited Sun 03-Nov-13 02:24 PM by NanoMeter
Paul, Yes hypothetically, I think 15 minutes given how dark 10 minutes was. When I say dark I mean the final image, NG1 in live view is black, yes. When I say dark I mean you can see the image, but it needs to be exposed much longer. I will show you what I mean when I test it again. A picture is worth a thousand words they say. You have to get up pretty early in the morning on a nice day to test 15 minute exposures. It is laboriously time consuming doing hit and miss long exposure testing in manual mode, guessing what amount of time might yield the optimum exposure, which is not as hard when you are hunting in the 1-30 second range using aperture priority, but 15 minutes means I am going to have to take up knitting or something. We have rain here for the week also, so if I get an umbrella, then I can learn to knit in the rain. :-/
NG1 is VERY strong, even at 1mm to 2mm, but looking at the graphs, NG1 is not as flat of a curve as the other NG glass. I like the curve flatness of NG9, but I am wondering how thick I can go with NG9 without loosing sharpness. I don't know how much difference those two curves will make in the IR range however. NG1 looks like a slightly different glass recipe than the other NG glass, this is even apparent from Schott's catalog showing 1mm NG graph comparisons. So these are things I am also thinking about.
If you test at 2 minutes at f5.6 and get an image, then at f20, it should be about 16 minutes (3 stops). I would then test dark glass at 5.6 to get an image with even histogram at left, not clipped .On the images I have shown at 81/2 and 9 minutes histograms were in shadows but detail was recoverable.
Cheers, Klaus ------------------------------------------------------------------------------------------------------------------------ Link(s) to the lens database (macro, UV, IR) as well as to my UV BLOG, plus many UV images here: http://photographyoftheinvisibleworld.blogspot.de/ ------------------------------------------------------------------------------------------------------------------------
Wed 06-Nov-13 10:58 AM | edited Wed 06-Nov-13 10:58 AM by NanoMeter
>Working with those filters can be confusing if the term >"optical density" is unknown, which Schott and >others are using (Helioplan for instance) in naming their >filters.
Klaus, The names are unrelated to optical density. I see no relationship or order between the numbering-names of Schoot's NG filters and their optical density or transmittance. NG9 is out of sequence, and other numbers are in reversed order of optical density.
I am using 1mm for these first two graphs, because that is the thickness Schott uses for their NG graphs in their catalog.
Internal Transmittance at 1mm each:
Optical Density at 1mm each:
Optical Density for NG glass mentioned in above posts:
I've been studying those curves from the standpoint of which to test. The Lee Big Stopper is 2mm thick and gives me 10stops of ND reduction. Looking at the curves for Optical Density, I would test first the NG9 at eithe 2.2 or 3mm thick at f5.6 and see exposure time to get histogram to left, not clipped. We can then calculate added exposure times by stopping down.
I already get 9 minute exposures by shooting at 11-1:00. If I shoot later in day, I should have no problem getting at least 1 if not 2 added stops which would easily double exposure time.
Paul, Sounds good, but I think we should stack it with an IR filter for the tests, given that is the intended purpose, and also that the curves change right around 700nm, so stacking will avoid lower NM influences. What do you think?
Sorry. I thought that it already was going to be stacked from prior correspondence. I have an 830 conversion. For future use, which converted camera would you use for long exposure work? if deep IR, then testing is more reliable for both of us for future use. But, whatever filter you prefer.
Unrelated, do you have any interest in Infrared HDR. That is a 3 image IR composit that I posted below. Not much work in IR that I can find for HDR, blending or focus stacking. Plenty to do.