Jump to content
UltravioletPhotography

[Filter Test SEU Gen2 #9A] Longpass Stack Wandering Discussion. See #9B for results.


Andrea B.

Recommended Posts

Update Tuesday 24 July 2018: This topic originally was going to show some SEU2+Longpass stacked filter tests. We all got sidetracked discussing some nuances of such tests! And I lost some of the tests I had made due to a corrupted card. So I will start a new topic for SEU2+Longpass tests when the weather permits.


Update: Saturday 21 July 2018: There is going to be a delay in posting these results. My SD card has gotten corrupted!! I think I might have to have to reshoot the series. My apologies even though this was out of my control. I don't think this SD card was very old, but I suppose corrupted files can happen in various ways.

 

I will leave what I wrote here and the reference scenes, but I removed the example photos until I can be sure their companions are OK or until I reshoot, however it works out.

 

Thank you for your patience !!! I know you all are eager to see these results.


[Filter Test SEU Gen2 #1] Introduction to the SEU Gen2

[Filter Test SEU Gen2 #2] White Balance, Raw Histogram & Andrea's "White Signature"

[Filter Test SEU Gen2 #3] Filter Speed & A Windy Bull's-eye

[Filter Test SEU Gen2 #4] Dealing with the Usual Dichroic Effects

[Filter Test SEU Gen2 #5] Landscape Interlude

[Filter Test SEU Gen2 #6] Monochrome Museum Comparison

[Filter Test SEU Gen2 #7] Measured Filter Transmission

[Filter Test SEU Gen2 #8] Dichroic Reflection Detour

[Filter Test SEU Gen2 #9A] Longpass Stack Wandering Discussion. See #9B for results.

[Filter Test SEU Gen2 #9B] Longpass Stack Results

[Filter Test SEU Gen2 #10] What good is a filter test without a Rudbeckia?

[Filter Test SEU Gen2 #11] A Fascinating New Rudbeckian View

[Filter Test SEU Gen2 #12] Summary


 

Longpass Test #1:

Stack BaaderU and SEU Gen2 with 2 mm thick Schott GG longpass filters and see what happens. For this first attempt, I'm looking only at violet, blue or cyan cut-in.

 

Zen Question:

Are we testing what longpass light leaks through the UV-pass filter?

Or, are we testing what UV light leaks through the longpass filter?

 

List of Schott GG Filters, 2mm:

  • GG 395
  • GG 400
  • GG 420
  • GG 435
  • GG 455
  • GG 475
  • GG 495

Reference Scenes

D610 full spectrum, full frame + UV-Nikkor 105/4.5

 

During the time I was making the stacked filter shots after these initial three reference photos, the sun was strong and steady. I checked the exposure times twice more for the BaaderU and SEU Gen2 (f/11, ISO-400 for both) and the 1/5" and 1/8" exposure lengths used in these UV photos still held.

 

 

Visible Scene: Baader UV/IR-Cut

As shot

d610_uvNikkor_vis_sun_scene_20180721swhME_12782.jpg

 

 

UV Scene: SEU Gen2

As shot

d610_uvNikkor_uvSeuGen2_sun_scene_20180721swhME_12808.jpg

 

 

UV Scene: BaaderU

As shot

d610_uvNikkor_uvBaader_sun_scene_20180721swhME_12789.jpg

 

 

Rebooting the shooting....... "-)

 

See Post #10 for one example.

Link to comment

Longpass Test #2: Stack SEU Gen2 and BaaderU with 2mm thick Schott RG filters and see what happens. For this first attempt I'm looking at the area around 700 nm.

 

Schott RG Filters, 2mm

  • RG 665
  • RG 695
  • RG 715
  • RG 780
  • RG 9, IR bandpass

This test was performed in [Filter Test SEU Gen2 #9B].

Link to comment
Thank you for working hard on these tests Andrea. These results will explain what has been mentioned in other posts.
Link to comment

Andrea, Something seems askew here.

If you are stacking the Baader U with any longpass filter, that is above 400nm, then you should either get a black image or else some image, even a faint image (depending on exposure time),

and not any kind of gray or neutral toned image like you are showing above.

Here is an old simple example (sorry, but I don't know off hand what the ISO or aperture was for these, but probably 200 to 640, and f/8?).

Regardless, if you can shoot some similar example as a starting point, say 610nm at 20-40 seconds, and your shot should show some kind of image, like my example below.

Before you jump into umpteen tests, try to recreate my test below first.

If you are not getting a black image with the stack at shorter exposure times, then something is amiss.

 

post-87-0-85007200-1532220356.jpg

Link to comment

And as it happens I had a dangrabbity corrupt card!! Whether that had anything to do with anything I had previously posted (and now removed), I don't know. I haven't had time yet to run a recovery on the card. I think I'm probably going to have to reshoot. Some of the frames are only half there and with weird colors and pixel-y streaks. So there's no recovery on those anyway. And I was missing a whole raft of IR stack work. Like it just disappeared !! Dangnabitty rabbity.

 

I did confirm it is not the camera which is working just fine with another card.

 

I don't get why the review shots looked OK on the LCD while I was shooting but then I started finding corrupted files while editing???? Maybe this happened in the transfer from card to PC??

 

Well, enough griping from me. I'll figure it out. Easy enough to get a new SD card or two.

Link to comment

Now when I put the card in the SD slot on the MacBook, the Mac doesn't seem to see it. So I guess I can't try to find my missing IR stack shots. But the Mac can see my other SD card, so I don't thnk it is the Mac SD shot slot.

 

Arggggghhhhhhh.......

Link to comment

Here is a very small sample using the Schott GG 420 nm longpass filter stacked with the SEU Gen2. I think I've gotten the "green bump" which has been discussed in other SEU Gen2 test topics. It is more of a cyan bump. It seems reasonably obvious to me that is does not play much of a role in an SEU Gen2 photograph. As always, please let me know what you think about these results -- so far. It is totally OK to disagree with me.

 


ADDED 2018.07.22: Later comments have focused on the forced exposure aspect of this green bump experiment. I tried to make the point in this topic that I was looking to find out what kind of light might be passed once violet was blocked. To do that I needed to gather enough of the light to be able to see the result in the raw composite. Now that I know there is a possibility of greenish-cyan light, I will be better able to judge whether or not it is a contaminant in "ordinary" exposures. [[ADDED 2018.07.25. To verify, there is no green contamination.]] As mentioned immediately in the intro to this post #10, I gave my opinion that the green bump is not a contaminant from what I have seen so far. I didn't think anyone would be disagreeing about how I tried to trap the green bump with the forcing. I was expecting disagreement about whether or not it is a contaminant!! :D


 

The weather turned against me (fog, rain), so I could not capture any more examples this morning using other Longpassers or with the BaaderU.

 

(BTW, kindly do not throw any shade anywhere onto anyone in this topic.

Shade is as ruinous to good UV photography as Infrared is. :D :D :D

Discuss this like ladies & gentlemen who love tech & science....

......and who enjoy preserving my good nature as a forum admin.)

 

 

First, here is a well-exposed SEU Gen2 UV photo, both as shot and in raw composite form.

Raw Digger reports no over- or under-exposure.

 

SEU Gen2 Reference Photo, As Shot

Sorry, not much of a scene. I just wanted to get one good sample up for you all.

610_5888.jpg

 

SEU Gen2 Reference, Raw Composite

Raw composites do not have finished contrast adjustments, so tend to look dull.

Or it could be the gloomy day we are having.

610_5888_rawComp.jpg

 

 



 

The difference in exposure lengths is important in this example. Please note that for f/5.6 at ISO-400, the UV reference photo above required 1/30" but the UV+Longpass photo below required 17/3 = 5.66 more stops at 1.6" to be "well exposed".

 



 

Second, here is a well-exposed stack shot, SEU Gen2 + GG 420x2.00, both as shot and in raw composite form.

Raw Digger reports no over- or under-exposure.

This stack shot can be considered "forced", I will agree. But we must gather enough photons to figure out what light we are capturing in a stack shot test. (I will later add back the explanation in the first post about how I "well expose" a stack. Basically, I try to center the white histogram.)

 

Stacked SEU Gen2 + GG 420 Longpass Photo, As Shot

Remember, this stack shot needed 5.66 more stops to capture a "centered" brightness histogram.

Forced but not hammered.

610_5892.jpg

 

Stacked SEU Gen2 + GG 420 Longpass, Raw Composite

Wheee.....forced cyan!!!

610_5892_rawComp.jpg

 

 



 

Third, here are the raw histograms for both photos.

 

SEU Gen2 Reference Photo, Raw Histogram

The two Green channels are averaged. Green channels seem to be the same for the D610.

610_5888_rawHisto.jpg

 

Stacked SEU Gen2 + GG 420 Longpass, Raw Histogram

Note that this chart has EV0 in a different location because there were different amounts of data to present. Also, unfortunately, Raw Digger rounds up the exposure time in the histo title. To reiterate, exposure length was 1.6 seconds.

[ADDED 2018.07.25: Let me emphasize that this is a forced stacked filter exposure. There is no green leakage in an SEU Gen2 photo.]

610_5892_rawHisto.jpg

Link to comment
In Post #10 I did not discuss the fluorescence of GG glass. I'll cover that later when I am able to make the complete post after this card delay and weather delay.
Link to comment

Forced and non-forced.

Generally, to compare any out-of-band leaks with the actual intended bandpass range wanted in a photo,

I would simply shoot a photo with a satisfactory exposure, then use those same setting and exposure time with the stacked test, and that would be a non-forced test,

and it would show you how much if any out-of-band leak pollution is mixing into the desired bandpass shot.

Forcing, or pushing, simply uses a much longer exposure, and it can be useful in several ways, but it doesn't show the true ratio of in band and out-of-band mixture.

For example, if say a rudbeckia black center had a browner color, or lighter color, then you could test to see if that flower center was polluted by out-of-band light,

in the case of a UV-only filter, the out-of-band light would be anything from 400nm and above (visual...IR), so up to 1100/1200nm.

Using strong UV lighting, like direct sun on a summer day, if you can't see any image, even a faint image,

and if the frame is entirely black with a stacked test, and using the same exposure time and settings as the satisfactory UV-only exposure,

then you have no out-of-band leaks to worry about.

In short, you need a UV shot with optimal exposure, and a stacked shot with the same exposure/setting to evaluate realistic leak pollution.

If the stacked test is black, then you are good to go.

 

However, on the other hand, if you compare the unstacked image and the stacked image, and the stacked image has any hint of even a faint image, then you have a leak, you have out-of-band pollution,

and it is up to you (the user) if that is pollution you need to worry about, but the ideal is to see a totally black stacked image.

If your pollution is from the visual range, usually in the high red (up to 700/720nm, and even a little above), then expect some brown pollution in the UV-only black areas,

and if the pollution is in the higher IR range (above 780/850nm), then expect the UV-only black areas to be less black, some white/gray mixed into the UV-only black areas.

 

I am not saying all this for you Andrea, I know you know all this, and the same goes for a lot of other people on here, but I am saying it for a few people who might be reading, who might not understand all of that.

 

PS: I find it a little odd that your stack shot above basically has invisible foliage.

Link to comment

Hold on there....light transmission decreases through two filters. How would a two-filter stack trap anything at the same exposure length used for the single filter?

 


 

I'm not sure you really want to know where the leaves went, but I ran a quick contrast enhancement to show you that they are still there. The wind caused motion blur which naturally makes the foliage less distinct. However, there is another phenomenon at play which I do not have a name for, but that I am currently calling "dichroic stutter". It would seem that enough UV or maybe violet light passed through the GG 420 to hit the SEU Gen2 and get broken up into this repeated staggered pattern. But of course we didn't get enough of any kind of light to make a real image.

 

Filters do strange things. Stacked filters do even stranger things. We cannot and must not judge contamination solely by an informal stacked filter test. Such a test certainly does alert us about what to look for. But a black stacked filter frame does not necessarily indicate no contamination is present. It's only good for that exposure with that gear and that time of ambient lighting. This is why it is so important to use good reflective standards and known UV-reflective/absorbing patterns when testing. And even those tools have their problems.

 

Oh well. Enough !! :lol:

 

Before we all get off into the technical forest examining tiny details and missing the tree trunks*, let's just note that for the non-laboratory UV-reflective photography we all do, the commercially available filters and filter stacks work quite well with the only notable exception so far being the un-standardized ZWB glass. (If you get a good pour from a Z manufacturer with good quality standards, it's ok. But too many bad examples are still cropping up.)

 

 

*...of which I'm so often guilty meself.....

 

 

 

The hammered version of the forced photo.

610_589201.jpg

 

Optical density of 2 mm GG 420

Screen Shot 2018-07-22 at 8.20.17 PM.jpg

Link to comment

You do the math :)

The longpass filters are about 100% Ti and 90% T, so if that 10% difference is a factor for your equation, then by all means increase the exposure if you want,

but I doubt you need to do that, because essentially the 90% T of the longpass filters are wide open, and they don't suppress the OD3/4 bumps that Ulf has shown.

http://www.ultraviol...dpost__p__22642

So personally yes, I would use the same exposure for the stacked tests as was satisfactory for the U filter shot alone (un-stacked).

Link to comment
It's not what I want, rather it is what is the correct thing to do with the exposure of a two-filter stack. I don't have any equations....maybe Andy can help me out on this one.
Link to comment

The line you present is not reality. The Schott data for the longpass filters is limited to 1E-05 for 3mm, that is the 'bottom' of the data, but the ocean is deeper than that, that is only the length of their chain.

But that would be a separate question. almost an opposite question from 'how much added expose time is needed' to make up for the 90% longpass peak amplitude.

Personally, I would just use the same exposure time for each, it is more real than unreal, and if you want to push it a little, then fine with me, but I would present the same exposure as a starting reference.

My example above is a pushed/forced example, of course, otherwise you would see nothing but black using the same exposure of either the UV stack or the Baader U, because neither of those leak, unless pushed,

neither have any out-of-band pollution.

My opinion is that pushing the exposure is not really a fair result.

You might want to use a 495nm or 515nm longpass if you want to test for the 500nm range bump that Ulf was showing.

Just keep in mind that the bottom of the longpass filters goes deeper than the bottom line data in the graphs will show.

 

post-87-0-36416700-1532306502.jpg

Link to comment

Here's a UV-pass photo together with a UV-pass + GG 420 photo made at the same exposure which seems black but which is also full of information. I'm not sure this information should be ignored and I'm also not sure what that information is telling us. Is UV getting thru the GG 420? Or is Visible light getting through the UV-pass filter? Clearly some kind of light is getting through the stack made at the same exposure as the single filter.

 

But does it matter? (no) Is my UV photo contaminated? (no) How do we measure whether suspected contamination is at an acceptable level or not? Perhaps Andy's subtraction analysis routine would be helpful -- that thing where the "bad" light is stripped out of the file. Would that help in deciding this stuff?

 

The transmittance charts do not show "reality" any more so than the stacked photos. Transmittance charts only show an ideal never attained in actual shooting.

 

I'm off to watch a movie now.

 

610_5799.jpg

 

Looks black, but has plenty of information inside. Probably irrelavent to actual shooting though.

610_5800.jpg

 

 

Yes, I should probably be embarassed by the sensor blobs. The good old D610 does this. I do clean it. I just do not clean it every day. :D

BTW, this dichroic stutter is clearly not happening in the UV-only photo. Color is false, so please do not judge anything at all by the color.

Is this black frame telling us anything when we look inside?

610_5800pn.jpg

Link to comment

If you want a stronger longpass filter OD than trusting my assumption that the OD actually already is deeper than the OD3+ line calculated from the 1E-05 maximum data they include in the program,

then you can stack several GG filters together for a deeper OD, but I am sure the OD is deeper than what is calculated, because it uses a flat maximum of 1E-05.

Anyway, here is an example you could try for the 500 range if the OD of the longpass is any concern to you.

You could also stack that stack with your Baader UV/IR-Cut filter if you wanted to limit the test range.

 

post-87-0-12148800-1532308404.jpg

Link to comment

What the bandpass test is telling us is how much out-of-band 'light' is mixed with the UV shot.

UV filter shot = UV + Out-of-Band mixed.

UV filter stacked with longpass filter = Out-of-Band alone.

The longpass stack separates the out-of-band transmission from the UV shot, leaving us with only the out-of-band transmission in the photo,

which we can then compare with the UV shot to see if there are any areas that might contain some unwanted mix of UV and out-of-band light pollution.

Such test comparisons should really be done outside on a sunny day for best assessment.

I don't see anything in the original black frame, so based on that I would say it has adequate out-of-band suppression.

Ulf said is has 3.5 OD? I prefer 4+ or 5 if possible, and 3 OD is the absolute danger zone, but 3.5 should work usually.

Like I said, a sunny day is best.

 

What movie?

Link to comment

We cannot conclude that there is no leakage simply because we've gotten a "black frame" from a stacked filter test. We must at least look at the raw file and see what was actually recorded.

 

 

Here is a BaaderU photograph, as shot.

d600_co60_uvBaader_sun_20160202wf_42166XX.jpg

 

 

Here is a BaaderU + B+W092 stacked filter photograph of the same subject at the exact same exposure. Looks black, doesn't it? But if we concluded from this that the BaaderU has no IR leakage, then we would be wrong.

d600_co60_uvBaaderIr092_sun_20160202wf_42168XX.jpg

 

 

Here is the preceding stack shot converted in Raw Digger. No pushes, no levels. Just lots of IR contamination. There is information in the black frames that should be evaluated.

d600_co60_uvBaaderIr092_sun_20160202wf_42168rawCompXX.jpg

 

 

 

Here is a U-360 + S8612 photograph, as shot.

testFilter_u360+s8612_sun_20160202wf_42148YY.jpg

 

 

Here is the U-360 + S8612 + B+W092 stacked filter photograph of the same subject at the exact same exposure.

testFilter_u360+s8612+092ir_sun_2.5sec_20160202wf_4215001YY.jpg

 

 

Here is the preceding stack shot converted in Raw Digger. No pushes, no levels. Much less contamination.

testFilter_u360+s8612+092ir_sun_2.5sec_20160202wf_42150rawCompXYZ.jpg

Link to comment

Sorry to sound contrary, but that is over thinking, and over processing.

You don't need or want to process the black frame, it is what it is, and that is all we want it to be.

 

If you don't see an image using in the stack, using the same exposure as the UV shot, then whatever is there via further processing is just not seen in the UV pic.

This is a simple test. Don't do anything to the stacked shot.

Link to comment
Andy Perrin

Yeah, I don’t understand why we care about things that might lurk in the basement of the black frame. You should make only one change at a time when you do a comparison. You add the long pass and see if light still is getting through under the same conditions except for that one difference.

 

That RAW digger test does seem to show the stack has less contamination than the Baader, but under normal shooting neither have any significant contamination.

Link to comment

Sorry to sound contrary, but that is over thinking, and over processing.

 

I'm really not sure why looking at the actual raw file data is "over thinking"?

And it's the "black frame" JPG which is processed data (processed by your camera JPG engine).

 

 

********

 

You should make only one change at a time when you do a comparison. You add the long pass and see if light still is getting through under the same conditions except for that one difference.

 

OK, I just gave you two examples of that up in Post #21.

But can you tell me why it makes sense to use the same exposure for two filters that was used for the single filter?

 

****

 

I think everyone totally missed the point of my example in Post #10. I was trying to find whether there was a green bump or not. I said back up there that we have to gather enough photons in order to see what kind of light is getting through. Now we know what kind of light it is (indeed, greenish-cyan) we can figure out whether or not it is contaminating in the normal exposure quantities. I don't happen to think the green bump is contaminating, but it would be nice to make further experiments on that to try to "prove" it.

 

 

.

Link to comment

I felt like I had to put up a note in the original Post #10 green bump experiment to stress the point that I was trying to trap the green bump. I didn't expect controversy over how I trapped it. I was expecting controversy over whether or not the green bump is a contaminant. :D It is not.

 


ADDED 2018.07.22 to POST #10: Later comments have focused on the forced exposure aspect of this green bump experiment. I tried to make the point in this topic that I was looking to find out what kind of light might be passed once violet was blocked. To do that I needed to gather enough of the light to be able to see the result in the raw composite. Now that I know there is a possibility of greenish-cyan light, I will be better able to judge whether or not it is a contaminant in "ordinary" exposures. (It is not.) As mentioned immediately in the intro to this post #10, I gave my opinion that the green bump is not a contaminant from what I have seen so far. I didn't think anyone would be disagreeing about how I tried to trap the green bump with the forcing. I was expecting disagreement about whether or not it is a contaminant!! :D


 

I try so hard to write carefully, but somehow I manage to goof it up all the time....Or maybe I write so carefully that it is utterly boring and stuff gets skipped over.

 

"Well, tomorrow is another day," said Scarlett.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...