59 replies to this topic

[Starman1]

OK, lab reports are back:

 

Lumicon Gen.3 O-III filter:

 

FWHM bandwidth--11nm

Transmission 495.9nm--95.1%

Transmission 500.7nm--94.7%

no red output.

This confirms what I saw--a tad narrower than other O-III filters, but with as much nebula showing, but details a bit more like the narrowest filters.

An excellent filter, in every regard.

 

However, the sheet that came with the filter bore no relationship to the measurements of the filter.

I'm not sure what the graph is or was.  The filter is a lot better than the graph provided.

However, this filter joins the ranks of the state-of-the-art filters out there.

[CrazyPanda]

OK, lab reports are back:

 

Lumicon Gen.3 O-III filter:

 

FWHM bandwidth--11nm

Transmission 495.9nm--95.1%

Transmission 500.7nm--94.7%

no red output.

This confirms what I saw--a tad narrower than other O-III filters, but with as much nebula showing, but details a bit more like the narrowest filters.

An excellent filter, in every regard.

 

However, the sheet that came with the filter bore no relationship to the measurements of the filter.

I'm not sure what the graph is or was.  The filter is a lot better than the graph provided.

However, this filter joins the ranks of the state-of-the-art filters out there.

Any thoughts as to how this compares to the new TV filters built by Astronomik? (transmission & bandwidth-wise)

Edited by CrazyPanda, 24 July 2018 - 09:45 AM.

[Starman1]

https://www.cloudyni...uses/?p=8723874

[CrazyPanda]

Thanks for the link. Very interesting.

 

Looks like you choose between contrast and transmission when deciding between the Lumicon Gen3 and the TeleVue. 

 

In your estimate, is there a substantial contrast difference between 11nm and 12nm? 

 

It would seem that the narrower bandwidth would be preferable at larger exit pupils where the slightly lower transmission will not be that important, but the higher transmission of the TV would be preferable at smaller exit pupils. 

Edited by CrazyPanda, 24 July 2018 - 11:27 AM.

[Starman1]

As I understand the making of filters, it is difficult to achieve high transmission and narrow bandwidth.

It's especially difficult to achieve a steep side to the bandwidth, a high transmission, and pick up all the lines desired.

As the bandwidth is narrowed, the transmission usually drops.  Many H-ß filters, for instance, have fairly low transmission because

the bandwidth is narrowed for contrast.  DGM's NPB has a lower transmission than their much-wider VHT, for example.

Sacrificing 10% to gain a narrower bandwidth is quite acceptable, which is why whether it's 88 or 98% transmission isn't as important as controlling the bandwidth.

You won't see the difference between 99% and 95% in a filter any more than you would in an eyepiece or a mirror coating.

I've tested O-III filters with bandwidths from 7 to 30nm, and the narrower ones definitely display better contrast, albeit with a dimmer field overall, so aesthetics are involved for many viewers.

 

None of the O-III filters tested under 11nm bandwidth were visual-specific filters.  A 7-10nm O-III photographic filter (concentrating on passing primarily the 500.7nm O-III line for imaging purposes)

can, of course, be used for visual use, and what you see will have fantastic contrast, but you won't see as extensive a nebula as you would with a slightly wider filter.

So I'd put the best compromise between large size of the nebula and contrast for small features within the nebula at 11-15nm for an O-III filter and 21-27nm for a UHC-type.

 

You're right that, in general, narrower bandwidths go with larger exit pupils.  If the image field is already quite bright, restricting it is less injurious to overall brightness.

There is no appreciable difference visible between 11nm and 12nm, though.

 

But what if you have a small aperture?  Well, as long as the exit pupil is large, the same thing applies.  I've used a 12nm O-III filter on a 50mm finder to look at the Veil and it was quite well seen (the whole nebula at once, too), so I agree with David Knisely that a smaller aperture doesn't require a wider filter.   A slightly wider filter of its type might pass more light (see the Lumicon UHC figures, for example--incredible), but the difference isn't likely to be visible if the difference is only a few percent.

 

What I see in some filters is misplaced bandwidths, where the FWHM figures clip one of the spectral lines trying to pass.

I don't see why a 30nm wide UHC type filter would clip either the H-ß line or the higher O-III line, for instance, yet some do.

One O-III filter actually extended up in the spectrum far enough to catch the C2 spectral lines in comet's tails, but, unfortunately not at a high enough percentage to be recommended as a comet filter.  And its

transmission at the lower O-III line wasn't high enough either.  It was simply a poorly-designed filter.

 

But when we're talking Lumicon, Astronomik, or TeleVue, we're talking high-end filters and the results in the field will be excellent with any or all of them.

My problem is that I have them all, so which ones do I sell?

Edited by Starman1, 24 July 2018 - 01:21 PM.

[Lumicon]

 OIII Gen3 Data Sheet.pdf   44.59KB   132 downloads

 

Here is a Lumicon Gen3 OIII Filter Scan Report:

 

Doug

[Starman1]

Very consistent.

This scan is from mine (the 2 green lines are O-III lines, and the purple line is H-ß; H-α on the far right):

Attached Thumbnails

• 

Edited by Starman1, 25 July 2018 - 05:36 PM.

[MrJones]

I see some of these filter threads need attention. Too busy as usual. Anyway, these filters have an issue. The Lumicon scan from Doug is showing blocking of ~0.5% in the out-of-band areas?! The old Lumicon specs are standard OD4 spec out-of-band blocking, T < 0.01% at I think 440-475 nm and 520-630 nm, so not even close. There is always the question of instrument and calibration for measuring these values but from what I've seen of other Lumicon scans with their Ocean Optics system it really needs to be showing at most 0.1% T out-of-band. In the scan from Pensack, apparently actually from LunarFox? (it's ok to credit people responsible and post experimental conditions including instruments), some of the out-of-band baseline seems to be below the zero line which shouldn't happen but it's quite obvious how rough it is when at this scale it should be perfectly flat on the zero line. Feel free to check out other Lumicon and LunarFox spectra in this forum for comparison.

 

Contrast is of course S/N , in this case S ~ T and N ~ area of transmission over the visual range. So all of that area contributes to lower contrast. This is maybe the worst blocking performance I've seen from a "premium" narrowband filter. As good blocking can cut transmission a little, one could speculate that the designer decided to compromise blocking to eek out a little more T but it could also be just to cut costs. Regardless this would be an automatic rejection with professional contract narrowband filters. Very disappointing.

[Starman1]

Here are the actual measurements (300-1100nm):

1) don't forget the red shift with angle in faster scopes (like f/3)

2) 11nm is the narrowest FWHM I've seen in an O-III filter that captured both the 495.9 and 500.7nm lines.  I've only seen narrower in single line filters.

Attached Files

•  Lumicon OIII Gen3.xlsx   65.94KB   67 downloads

Edited by Starman1, 26 July 2018 - 01:44 PM.

[Starman1]

I can provide similar scans for Orion O-III, DGM O-III, an older Lumicon O-III, The TeleVue O-III (new and old versions),

The Astronomik O-III (visual), StarGuy O-III, Explore Scientific O-III, and Baader O-III upon request.

[MrJones]

Here are the actual measurements (300-1100nm):

1) don't forget the red shift with angle in faster scopes (like f/3)

2) 11nm is the narrowest FWHM I've seen in an O-III filter that captured both the 495.9 and 500.7nm lines.  I've only seen narrower in single line filters.

Maybe not Hitachi. Do you not know? Anyway, your data supports what I said - bad oob blocking far from the OD4 standard.

 

1) Yw. Need to measure to know though.

 

2) I get 12nm FWHM, 12.1nm actually from this. I measure 11.4nm with my SVOTek OIII. Do you sell SVOTek filters?

[MrJones]

Read that PM. I have signed NDAs that are occasionally relevant so don't talk about some things. But otherwise I have a minute and might as well post here.

 

FWHM is always a little controversial. Vendors want whatever gives the best number, scientists/customers want the number that best reflects the performance.

 

With this data and filter, assuming everything else is good, the highest measured T is 95.32 so FWHM could be measured at T=47.66. Many would argue, especially with this fairly low resolution and complicated passband shape, that some kind of average or interpolation value should be used to define the top. Let's say half height is T=47.5.

 

It's less controversial with peak width. With the low resolution you have to use an interpolated value, typically measured from a plot directly if you cannot use a good line shape fit such as Gaussian, Lorentzian, mixed, etc. Blowing this up on my screen and using a ruler I get 200nm = 361mm for the scale (bigger spread than I used the first time, so more accurate). At T=47.5 I get a width of 21.7mm. 21.7 x 200/361 = FWHM = 12.02nm. You can fudge this around a little but the FWHM from this data is closer to 12nm than 11nm.

 

12nm at T=~95% is very good but not obviously better than the other top tier options from Astronomik, TeleVue/Astronomik, SVOTek direct and the current Lumicon OIII (I don't know what Gen to call this) and the oob blocking is a little disturbing. This is all based on the evidence at hand and I would speculate that the actual performance and quality of these are about the same, but dependent on how bad the actual blocking of this filter is and ~0.5 (from their scan) is pretty bad. I've attached the DeepSky scan they did for me showing ~0.1 oob blocking.

 

I have no vested interest in any of this other than I like to see great $200 narrowband filters and don't mind sharing some of my expertise to make sure CN'ers know what they are buying. All of these filters are compromises in cost vs. performance and are not really state-of-the-art. Alluxa for example could make a >90%T, OD4 or better, 8-9nm OIII filter for a couple thousand $$$. The best amateur astronomy filters are really state-of-the-art compromises or maybe state-of-the-art values. This looks like an ok OIII filter with a big ? on the blocking. I have a very good SVOTek that is even better than my 2013 Lumicon, so me, I'm waiting for Gen 4.

[MrJones]

DeepSky scan:

Attached Thumbnails

• 

[Starman1]

 

12nm at T=~95% is very good but not obviously better than the other top tier options from Astronomik, TeleVue/Astronomik, SVOTek direct and the current Lumicon OIII (I don't know what Gen to call this) and the oob blocking is a little disturbing. This is all based on the evidence at hand and I would speculate that the actual performance and quality of these are about the same, but dependent on how bad the actual blocking of this filter is and ~0.5 (from their scan) is pretty bad. I've attached the DeepSky scan they did for me showing ~0.1 oob blocking.

 

I have no vested interest in any of this other than I like to see great $200 narrowband filters and don't mind sharing some of my expertise to make sure CN'ers know what they are buying. All of these filters are compromises in cost vs. performance and are not really state-of-the-art. Alluxa for example could make a >90%T, OD4 or better, 8-9nm OIII filter for a couple thousand $$$. The best amateur astronomy filters are really state-of-the-art compromises or maybe state-of-the-art values. This looks like an ok OIII filter with a big ? on the blocking. I have a very good SVOTek that is even better than my 2013 Lumicon, so me, I'm waiting for Gen 4.

I agree this isn't really any better than the offerings from Astronomik, TeleVue, or SVOTek.

You're right that it is a compromise.

However:

--it isn't a $2000 filter

--it performs excellently in the field  (contrast and sharpness was as good as my older Lumicon O-III and the Astronomik and TeleVue O-IIIs

and better than several others).  So it's in the realm of an affordable, excellent O-III filter for amateur observers.

--It's as good as you're going to get at an inexpensive price

--it's close to being the equal of my Lumicon O-III from Barr (now Materion) (about 10 years old), but it's been many years and even with inflation, this one is less $, and yet, it's very close.

The older filter had faster drop offs on both sides (more of a square wave).

 

Still, a lot of that is nitpicking.  It works great in the field, and is the equal of my reference standard (the Astronomik/TeleVue O-III).

Compared to a lot of modern "O-III" filters", it's in the top 5 or 6 consumer-level products.

[Redbetter]

I find it somewhat misleading to claim horrible blocking when the data indicate a rather tiny effect overall.  I summed the out of band transmittance for those 1nm increments from 400 to 490nm, and 510 to 700nm.  The total is 43.5%.  So worst case one is talking about less than 0.5nm of effective extra band width.  And visually over half of that is in the red from 640 to 700 (total was 23.3%).  That is where it is going to do the least damage as a visual filter, as at that level it will be effectively imperceptible.  Keep in mind the red leak in the old Lumicons (of "rust" fame) made them pink when held up to light...just looking at my 1990's 93% Hbeta and 94-95% OIII.  They passed orders of magnitude more light in this range.

 

One could instead look at the 440 to 475 and 520 to 630 bands that were specifically mentioned...again the sums are 2.4 and 9.1% transmittance respectively, for a total of 11.5% out of band, equivalent to no more than 0.115 nm wider 100% transmission in the full peak-to-base bandwidth...about 1% increase in noise compared to the sum of 1129% of the full 19nm width (peak to base) of the pass band.  Good luck being able to see that difference visually when the FWHM is in the 11-12nm range and the transmittance of the target lines is 95%. 

 

I don't have one of the newer Gen 3 Lumicon OIII's.  I do have one of the newer Tele Vue OIII's.  I would need to see a scan performed on the same instrument to compare them, but the Gen3 Don linked above looks a little narrower than Tele Vue's published curve both at FWHM and base of the curve.   That difference is likely to overwhelm the small out of band values I tabulated above. 

Edited by Redbetter, 28 July 2018 - 04:02 PM.

[MrJones]

"From passband to thread tolerances, the filter industry seems like the “wild west” of astronomical products." - TeleVue website

 

Indeed. Interesting that the designing and making of these filters is a fairly evolved and refined science at this point. Courses and textbooks even. Much of the sales is professional to professional where the customers know a lot about the filters and want certain specs. There's some of that with astronomy filters behind the scenes obviously.

 

But up front it is the Wild West. Customers mostly know nothing about these products, learn about them from forums and such from other hobbyists and consequently shady producers with shady products pop up and there are Snake oil salespeople to promote them. "You need 5 types of OIII magic filters to fully experience the wonders of the OIII universe!!" One for this nebula, one for that nebula and so on. Meanwhile Science says if you want to best see the OIII in a nebula, what you need is one filter with highest OIII transmission and the best blocking of everything else. How do we define "best"?

 

I remember when I started pushing the novel idea in these forums several years ago that you need to look at an independent transmission spectrum of a filter from a reliable spectrometer to begin evaluating a filter because that can tell you a lot about them and even most of what you need to know. There was actually disagreement on that. Now we have Cary and crew at OSI/Farpoint/Lumicon/Astrodon providing scans with all their filters and for customers who sends theirs in. Awesome.

 

So now you know there is a blocking spec that, at least in the past, some astronomy vendors have sought to guarantee, if not make public. Some of these are publicly available now. This filter likely does not meet the commonly accepted blocking specs based on data above. The people that made it seem professional, probably already know this and in the future might try to make their version of OIII filter with better blocking if there is any demand. Is there such demand? Regardless, you are now a more informed public.

 

I actually just ran across a nice, relevant PDF presentation by Don Goldman, founder of Astrodon, now owned by this company. Enjoy: http://www.aicccd.co...IC2007Talk2.pdf

[25585]

I have one and used it in the field.

I'm waiting to get test lab reports back to see if they confirm what I saw.

I have two other "reference standard" O-III filters to compare it with, and I used it on several objects this last weekend.

I'll report later.

My reference standard OIII is my original Lumicon (1.25") OIII. Its UHC casemate is very good too - closer to an Astronomik OIII than the Lumicon. Both Lumicons were the versions that came in dark blue translucent boxes, late 80s or early 90s.

[Starman1]

One thing to remember is that the requirements of visual users and photographic users are different.

For imaging, a narrower filter with as close to zero out of band transmission as possible is better.

For visual use, however, the tiny amounts of out of band transmission aren't particularly important because of the low sensitivity of the eye as a receptor

outside of a fairly narrow scotopic bandwidth of 450-550nm.

The shape of the transmission curve is fairly important, however, as this can admit more light pollution into the filter and rob the desired contrast from the image.

A Gaussian curve is not as desirable as a perfect square wave.  More cavities of filter coatings can improve the transmission curve but at an increased cost and a possibly reduced peak transmission.

 

In the field, sometimes the narrower filter reveals details not seen in a wider filter, but also shows less nebulosity than a slightly wider filter (and I don't mean ridiculously wide to

the point where the filter is simply bad for its intended purpose).  So which is the better filter--the one that shows the small details or the one that shows more nebula?

Photographically, the choice is obvious.  Visually, it's not as clear.

 

The Lumicon O-III filter in question goes from less than 0.3% transmission to over 50% in less than 3nm, and from 50% on the way down to less than 0.3% transmission in about 4nm.

One could argue the filter has too wide a curve under the 50% level, but there is little difference between it and other good O-III visual filters, and, in the field, it proved to yield a nice high

contrast image with lots of O-III details.  If this filter does not meet the "commonly accepted blocking specs", then this seems to be a universal condition for commercial-grade filters

sold to amateurs.  And at least 3/4 of the filters made and marketed don't even meet the minimum standards to be considered for purchase.

[Miranda2525]

I am quite skeptical with these "new" filters because of what went down in the recent past.

[nicoledoula]

 It wouldn't hurt to keep reminding ourselves QUI BONO when reading these comments. Does someone want the new incarnation of Lumicon to fail? I can think of quite a few companies who wouldn't shed any tears.... Thanks Don, what we see does matter way more than specs, I agree.  And if someone thinks $3000 filters will sell than they should make them.   I'm an old sheepdog Miranda,That's a cute puppy you have there.  Now I'm definitely buying Lumicon. Lumicon is my TV. I'm going to give Lumicon $400. And I'm going to buy them from you Don. It'll take a while to save up but you've got a promise from me. 

Edited by nicoledoula, 29 July 2018 - 07:50 PM.

[Redbetter]

My reference standard OIII is my original Lumicon (1.25") OIII. Its UHC casemate is very good too - closer to an Astronomik OIII than the Lumicon. Both Lumicons were the versions that came in dark blue translucent boxes, late 80s or early 90s.

With the rusting problem inherent to the blue box Lumicons and the red leak (and violet/blue on the other end) I don't consider these a good reference standard.  Because of the laminated build their performance is only going to degrade over time.  Mine are effectively retired. 

[Miranda2525]

 It wouldn't hurt to keep reminding ourselves QUI BONO when reading these comments. Does someone want the new incarnation of Lumicon to fail? I can think of quite a few companies who wouldn't shed any tears.... Thanks Don, what we see does matter way more than specs, I agree.  And if someone thinks $3000 filters will sell than they should make them.   I'm an old sheepdog Miranda,That's a cute puppy you have there.  Now I'm definitely buying Lumicon. Lumicon is my TV. I'm going to give Lumicon $400. And I'm going to buy them from you Don. It'll take a while to save up but you've got a promise from me. 

You also seem to be Don's follower, which is very cool. Seems that's all you go by on here, lol.  Let's not forget, he is a "vendor". 

 

PS: I already have two of the "good" Lumicon filters with hard dielectric coatings. I don't need to save up for anything. PS: I only paid 1/4 of what you'll be paying for less performance.

[Starman1]

Perhaps this post might help:

https://www.cloudyni...lter/?p=8735637

I cannot say that every filter from each company will match those results.  Some will be worse, and some better.

But those are the measurements obtained on my filters and they pretty much matched what I saw in the field.

 

Miranda,

If you have a Lumicon filter from the 2005-2011 era, then I wouldn't bother looking at newer ones.

[Redbetter]

 

PS: I already have two of the "good" Lumicon filters with hard dielectric coatings. I don't need to save up for anything. PS: I only paid 1/4 of what you'll be paying for less performance.

Unless you have had them scanned with the same instrument, you don't really know for sure what you have relative to the Gen3, new TV's, etc.  Your samples could be better, or they very well could be worse.  The former ownership of Lumicon was doing some odd stuff with regards to their filters in the later years, so there is little reason to have confidence in a used one from that era. 

[Udderly Abducted]