Jump to content
UltravioletPhotography

NP365. Is this bandpass filter worthy?


BruceG

Recommended Posts

Recently I found this bandpass filter with following specifications from some Chinese manufacturer, it seems to rely on interference design only.

 

Name: NP365

Peak: 365±5nm;

Half band width:about 60nm;

Peak transmission:>85%

Rejection wavelength:200-800

Optical density:OD3

 

post-127-0-05750900-1529683334.png

post-127-0-07178500-1529683354.jpg

 

Anyway I'm getting a 20mm sample to test its performance on some stackings, it will arrive in one week's time. For this diameter even it turns out leaky on IR, I can always use it for my UV torch. 20mm costs little less than 30USD, a 30mm piece would be double of the price.

Link to comment
Andy Perrin
It leaks IR according to its own graph even. And they only claim OD3 anyhow. So it needs to be stacked with S8612. And this is assuming the sample you get matches the specs given.
Link to comment

You can give it a try. But the Ir pass through looks to be really bad. Nearly same cost and slightly larger are the 330WB70, 330wb80 and 330wb80 improved that you can find on ebay by bjomejag. The ones that look good are about 24 to 25mm and block ir. Take a look, they are under $60.

I have the 330wb80 improved, and in process of monting it into a 25mm holder. He also sells 28mm holders for $12 which may be better though.

Link to comment
Andy Perrin
If you plan to use a filter without stacking for UV photos in sunlight, then the filter needs to block from 400nm to 1100nm to at least OD3. Our cameras are ridiculously sensitive to IR (partly because the bayer array becomes transparent, so more light is let through, but also just because that's how silicon sensors behave).
Link to comment

What Andy said, OD3 is not enough, border line, you would get some warm blacks with that, and it transmits above OD3 in the 800nm range, that's not good at all, that would definitely need extra stacked suppression.

Once you stack it with whatever, it then reduces the UV, so the same as using plain U glass.

Also looks like it would have a slight blue edge, but not that important.

 

What is the white disk pic of? PTFE? Please explain. Thanks.

Link to comment

What Andy said, OD3 is not enough, also looks like it would have a slight blue edge, but not that important, just that once you stack it with whatever, it then reduces the UV some, so the same as using plain U glass.

What is the white disk pic of? PTFE? Please explain. Thanks.

I assume it to be the filter itself.
Link to comment

You can give it a try. But the Ir pass through looks to be really bad. Nearly same cost and slightly larger are the 330WB70, 330wb80 and 330wb80 improved that you can find on ebay by bjomejag. The ones that look good are about 24 to 25mm and block ir. Take a look, they are under $60.

I have the 330wb80 improved, and in process of monting it into a 25mm holder. He also sells 28mm holders for $12 which may be better though.

Looks interesting, do you use 330WB70 for reflected UV work or UV fluoresence?
Link to comment
SteveCampbell
I think this filter combined with a BG38 or BG40 would be quite good - even outperforming the S8612 combination. Stacking with BG38 might carry the risk of a very very very small IR leak, but chances are it wouldn't have any impact.
Link to comment
Andy Perrin

Pretty much nothing outperforms S8612.

 

Seconding the 330WB70/330WB80 etc suggestion. It doesn’t need to be stacked.

Link to comment

It partly depends on how much red/IR you need to suppress, which is impossible to know from the linear graph of that filter shown above.

 

post-87-0-80450900-1529776816.jpg

Link to comment
Andy Perrin
Cadmium, I still don't see why you'd choose anything other than S8612 from those graphs you just posted? It looks like if you are willing to tolerate just OD3 suppression up to 1100nm then S8612 1mm beats or equals everything else in the UV side. (The assumption is that there is not a VISIBLE leak, I suppose, in the NP365.)
Link to comment

S8612 is the best, and usually the only BG type glass to use for UV stacking.

However, reading what Steve said, the only way you might 'over perform' with another glass would be to suppress enough of the visual/IR range, and still transmit more of the UV range.

So given the top graph, it shows IR above OD3 starting above 800nm. It also shows quite strong IR in the 1000-1100nm range.

Given that, we can't use BG38 2mm thick, because it doesn't suppress the 1000nm range.

However, we might be able to use BG40 2mm or S8612 1mm, which both suppress the 1000nm range, also 800nm, even 740nm, and transmit UV a bit stronger than S8612 2mm.

Those do have a little IR around 1200nm, but I don't know if that would show or not...

But, come to think of it, you are actually correct... even at that 1mm thickness, the S8612 is still the best, only because it transmits UV just a hair more than the BG40 at 2mm.

So S8612 still wins, but either of those might work well, and at least better for the UV range, at those thicknesses, however, we still don't know how much red/IR we need to suppress in the visual range and up to 800nm.

It says OD3, and that implies not the best suppression, but we don't know where that point is that we need to concern ourselves with exactly.

It would help to know the 'diabatic' view of the graph... but without that, best would be to play with different thicknesses of S8612, 1mm, 1.25mm, 1.5mm...2mm,

to see where we loose all visual/IR pollution in the photos, and still keep the strongest UV.

But if you had that figured out, then using the correct thickness of BG40, maybe around 2.5mm thick (which is the same as using S8612 1.25mm thick), should work practically the same as S8612,

but I don't see anything over performing S8612 at any thickness.

Link to comment
Andy Perrin

I don’t think anyone has shown the sensors to respond to anything past 1100nm. Jonathan hasn’t tested the IR side (I don’t think) but I’ve done my own test using a Thorlabs Premium Longpass filter that blocks all of the light from 200-1100nm to OD5 and I could not get anything but noise. This was for my SWIR setup, so I could be sure I was seeing SWIR.

 

FELH1100 here:

https://www.thorlabs...ctgroup_id=6082

Link to comment
SteveCampbell
Maybe I didn't understand how these filters work in conjunction with their relative thicknesses. Is a 1mm BG40 or BG38 not possible? My comments were based on comparing a 2mm S8612 with a 2mm BG38/40, but from Cadmium's comments I'm inferring that it's a false equivalency on my part?
Link to comment

Steve, Yes, 1mm thick BG38 and BG40 are possible, but in this situation you would want to go the other direction, thicker, because you would need thicker than 2mm of the BG38 (especially) to make that work,

and doing that would cut the UV transmission of the BG38.

S8612 has the best UV transmission to red/IR suppression ratio of all BG glass types.

So whatever thickness of whatever BG glass type you need to adequately suppress Red/IR to an acceptable level, S8612 retains the highest UV transmission compared to the others.

 

This graph illustrates the difference between BG glass types when they are all equalized to have OD3 at 700nm.

S8612 and BG39 need only 1.5mm thick, and S8612 retains the strongest UV transmission of all BG glass types, and BG39 has poor UV transmission (like BG38 at the same suppression).

BG40 needs 3mm thick, but has a fairly close UV transmission profile to S8612, but twice as thick.

BG38 needs to be 4.5mm thick, which is very thick, and raw sheets are only normally made in 4mm thick maximum, and the UV transmission is reduced to that of BG39.

 

post-87-0-62996100-1529806945.jpg

Link to comment

I would mainly focus on getting a good IR-attenuation.

The S8612, 2mm is safe for almost all filter stack combinations.

 

The 10% gain in transmission for UV when using a S8612 will 1mm decrease the exposure time by only 10%.

That is often a not very important improvement.

 

With my camera I only see one problem with the S8612, 2mm.

I had problems when I used it alone for VIS-images as it cut off too much of the red spectrum.

 

For VIS I use a BG38, 2mm instead as that seams to agree better with the Bayer response.

The BG38 has a rather similar response to the filter removed during camera conversion.

I think the BG40 might be an even better alternative for my VIS, but the BG38 works without any problems and I see no reason to change.

Link to comment

I agree, BG38 and BG40 work best for visual, S8612 is not as good for visual, too blue.

The graph above is just an example, not about thickness, more about equalized suppression comparison.

To determine the S8612 thickness needed, it would be best to know the diabatic curve of the U filter first, without that, it would be trial and error for the thickness of S8612 needed.

All I was doing with the graph was illustrating how all the BG filters compare when the suppression is made the same.

Link to comment

Looks interesting, do you use 330WB70 for reflected UV work or UV fluoresence?

I am going to mainly use it for reflected uv and compare to my ZWB1 filter. For fluorescence work I am now using two Convoy S2+ flashlights with u340 filters. This end up being much cheaper than my first attempt with 199a flashes. Sadly the flashes didn't work as planed and very expensive filters to isolate the light.

Link to comment

Just wanted to add to the second conversation here. The only advantage of going with the BG40 over the S8612, is that it does not get oxidized over time. But if your using a UG5 or UG11, then you have to deal with oxidation anyway and cleaning one more filter is no big deal.

Thus why I went S8612.

 

Link to comment

Actually, either can get the oxidation, but S8612 has more of that tendency.

Cerium oxide, great stuff, get some, works. I would not use toothpaste for that job myself, you don't know what all is in it.

Link to comment

Toothpaste? Really, I never heard of using that for optical filters. I know people tried it on DVDs if scratched, and then surprised it didn't work. But must depend on brand I guess. That Thomas toothpaste high in Cerium oxide for a whiter look, might work.

I didn't know the BG40s were also still sensitive to oxidation. Ok so no point going down that road.

Link to comment
SteveCampbell

Steve, Yes, 1mm thick BG38 and BG40 are possible, but in this situation you would want to go the other direction, thicker, because you would need thicker than 2mm of the BG38 (especially) to make that work,

and doing that would cut the UV transmission of the BG38.

S8612 has the best UV transmission to red/IR suppression ratio of all BG glass types.

So whatever thickness of whatever BG glass type you need to adequately suppress Red/IR to an acceptable level, S8612 retains the highest UV transmission compared to the others.

 

This graph illustrates the difference between BG glass types when they are all equalized to have OD3 at 700nm.

S8612 and BG39 need only 1.5mm thick, and S8612 retains the strongest UV transmission of all BG glass types, and BG39 has poor UV transmission (like BG38 at the same suppression).

BG40 needs 3mm thick, but has a fairly close UV transmission profile to S8612, but twice as thick.

BG38 needs to be 4.5mm thick, which is very thick, and raw sheets are only normally made in 4mm thick maximum, and the UV transmission is reduced to that of BG39.

 

post-87-0-62996100-1529806945.jpg

 

It seems like this new chinese filter cuts IR until around 825nm, so IR leaks around 700nm shouldn't be a problem. Wouldn't this give you leeway if you wanted to use a BG40 at 1mm? With this new chinese filter you would be more worried an OD3 at 825nm I would think, and cutting the S8612 any thinner than 1mm might be difficult. I have scanty knowledge in this field of course, so I'm sure I'm missing something

Link to comment
Andy Perrin
The trouble is that you can’t tell how much blocking there really is at 700nm from the linear scale graph. You need diabatic or log scale to be able to tell how bad the leak is there. Same with visible light region. We don’t know where the weak spots are.
Link to comment

Steve, You can use BG40, it works like S8612, but at twice the thickness.

BG40 1mm is like S8612 0.5mm thick.

As I showed above in the graphs.

Andy is correct again, but even if we assume that the visual range up to 700nm is all suppressed by the dichroic filter, then we would need S8612 1.5mm or BG40 3mm thick to suppress the rest of the leaked range above 700nm.

NOT 1mm BG40.

But, if any part of the visual range is not suppressed below OD3 with the dichroic filter, then you would need even thicker S8612 or BG40 to suppress it.

And if any part of the lower visual range is not below OD3, then basically toss out this entire discussion, because no BG type glass is going to suppress lower visual range.

Please explain to me how BG40 1mm is going to suppress the 70-80% peak in the 900nm+ IR range?

My opinion, buying such a diachroic filter is a huge wast of time and money, it has a very STRONG built in IR leak, and it needs to be suppressed, and yet is has no visual range data telling us how much that range needs to be suppressed.

Waste of time.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...