Review of Baader’s New UHC-L Filter: The Ultimate Weapon Against Light Pollution?

In this video, I test Baader's new UHC-L filter.
Alpine Astronomical product page for this filter:
alpineastro.com/products/baad...
Music credits:
* Scott Holmes Music: Feeling Home
* The 126ers: See You On The Other Side

[@ScrapYardDog64]

Wonderful, thats most appreciated 👌

[@mikegore7669]

Thanks for publishing these results and math. It's clear they are blocking too much of the spectrum. Ever since you first showed the led spectrum in your other video I have been searching for a led blocking filter. There is a company called stcoptics that has a multispectra filter however it looks like it may also pass too much blue. It would be nice if you could try tracking down other led specific blocking filters and perhaps review them - it would be a great service to the astrophotography community who live in badly light polluted area with led lights. Thanks for your work on this!

[@darkskygeek]

The IDAS GNB has potential but I don’t agree with their cutting a large chunk of the deep red/near IR. If they had not done that, and with a mono camera that is decently sensitive in that part of the spectrum, I think we would be ok. But in the end, there is no silver bullet, unfortunately. CS!

[@MattVentureNL]

Really enjoyed your video and immediately subscribed to your channel ofcourse. I really appreciate the experimental exploration and validation you bring to the "filter graveyard of broken promises". I am somewhat confused by the problem statement; as both the galaxy and iris (reflection) nebula are broad band targets. Using a filter to suppress broad band spectra (like LED is broadband) would not be a typical application, but a rather obvious tradeoff. I assume your point was that this should have been clarified on the suppliers website. Now in the case of emission nebula - with only OIII and Halpha emissions; I wonder if you can show that a UHC filter does make sense as a better than nothing (ofcourse a dual narrow band 3nm being the best solution for OSC). My hypothesis is that as long as you pass the target wavelengths, surpressing anything else will help a bit in the right direction (given finite well depth). Because some UHC filters are much cheaper (50$ vs 300$) than dual narrow bands, a UHC could be attractive to beginners in AP shooting emission nebula only. Would you agree? I could not reconstruct your SNR calculation to support this hypothesis because I did not understand how you derived Lp in the Bader case (Python script)? I would expect that a UHC SNR with an smaller and smaller bandpass would eventually approximate a dual narrow band (emission nebula). Again, really enjoyed this excellent exploration, looking forward to your next video!

[@darkskygeek]

I absolutely agree with you that a UHC filter is better than nothing when imaging emission nebulae, that is a good point. Thank you!

[@SeptentrionGlace]

Bonjour Julien. Merci beaucoup pour cette video et ce test. La question que je me pose dans le cadre du protocole de test de deux filtres (ou plus) est : ne faudrait il pas comparer les deux résultats à SNR équivalent ? Je sais que communément on compare des temps de poses équivalents mais si l'on considère, comme vous le précisez, que le filtre Baader va nécessité plus de temps de pose pour un SNR équivalent (quasiment 2 fois plus si mes calculs sont bons), et que c'est le prix à payer, en retirons nous profit au final ?

[@darkskygeek]

Si tu vas exposer deux fois plus longtemps, alors il sera mieux de le faire avec un simple filtre L car le SNR sera encore plus élevé.

[@Calzune]

Interesting! Can you have a look at Antlia filters? Both narrowband and LRGB?

[@darkskygeek]

Hi! I already did, in a previous video. Check out my channel. Thank you for watching!

[@scottrk4930]

Hello . I was hoping you might have scanned the Filter from both sides since you did mention that it has a Telescope-Lead-side Indicator . Would be interesting to see if they measured differently . Also , with real DSO objects , taking a few images again through both sides to see what the differences would be . Halos I imagine but still interesting . Thanks for your video ./SRK

[@darkskygeek]

Hi! Yes, you are correct, I could have done that. I don’t know if that would have an impact on the transmission graph. However, based on evidence from other filter testing videos, halos would be the biggest issue with a filter mounted in the reverse direction. CS!

[@galactus012345]

Quelle probité ! Avec une démarche méthodique en plus. Admiratif. Finalement la seule méthode c'est une longue durée d'intégration avec filtre luminance ou rien ? J'en suis à mon 4ème filtre broadband.

[@darkskygeek]

Correct. Comme certains l’ont noté, l’IDAS GNB peut aussi être intéressant, mais je pense qu’on peut faire encore mieux que ça (voir les commentaires sur cette vidéo) Sinon, rien de mieux qu’un ciel noir, bien sûr ! Merci 🙂

[@aradani3]

Hi, Thanks for the very instructive video. I'm not sure I understand the SNR formula you use, I would expect 0 LP to produce a value of 1 (f.e. that would be the result with squared values under the root) When I add the squares I get that the filter would help only when the LP is about 2.5 time the Signal or more. The formula I used is SNR = S / SQRT( TS^2 + LP^2) where TS is the total signal and doesn't change when you put the filter. so f.e. when using the filter with LP = 0, leeds to an SNR of 40%; when S = LP = 100 then the SNR0 is of 71% without the filter and SNR = 39% with the filter; when S=100 and LP = 300 then the SNR0 is of 32% without the filter while the SNR is 34% with the filter. Please let me know if these numbers make any sense. It would be interesting to do this test with targets at different elevations and see if the SNR is better with the filter when looking at targets that are closer to horizon. I also don't think this is the intended use for this filter; what I understand from what they write in their website is that it is intended to replace the luminance filter in an LRGB setup, and I have no idea how that would affect the picture. All that been said I agree with you that this filter will not improve the results on a broadband targets (as you have proved) but maybe it can help in light polluted areas, by allowing you to take longer exposures without clipping the image. And this might have advantages depending on your setup / processing power.

[@darkskygeek]

Why would you expect no light pollution to give a SNR of 1? A Poisson signal has a SNR equal to S/sqrt(S). Next, what is TS in your formula? The signal that reaches the sensor is affected by the filter. I think something may be confusing you, but I’m not sure what it is. Also, your conclusion (“[…] allowing you to take longer exposures […]”) is simply flawed. The goal is to compare the SNR for the same total overall exposure time, and increasing the sub exposure won’t help at all. Hope this helps. CS!

[@aradani3]

​@@darkskygeek Ok, indeed I was indeed a bit confused, well actually I was more like three towns off. LOL I think what is confusing is that SNR refers to the noise caused by the nature of the signal itself but LP is in fact another signal that also adds noise. Thanks for explaining

[@garybrown1127]

I thinks it's unreasonable to expect that a narrower filter will produce better results at the same exposure. One of the costs of living in a light polluted area is exposure time. A more reasonable test would be to increase the exposure for the LP filter by the ratio of the bandwidths. That may in fact produce superior results since the in-band S/N should be better. Excellent work none the less. I always enjoy your videos.

[@darkskygeek]

I respectfully disagree. If you are going to increase the total exposure time, then you might as well do that with a standard luminance filter, and your SNR will be even higher! If you don't use the same total exposure time, you are not comparing apples to apples. Hopefully, that makes sense. Thanks for the kind words.

[@Robservatory]

Isn’t the point of a light pollution filter to cut down the light pollution and allow for longer exposures? I think the test would be better if you are comparing how much signal you can get from each filter at its optimal exposure length. Shooting them at the same duration is handicapping the the light pollution filter.

[@garybrown1127]

I see Julien's point if you're imaging a narrowband target (nebula) where all of the emitted light falls in the passband of the filter. The filter would reduce the light polution (noise) without reducing the target signal. The same exposre time should produce superior results with the filter. If you're imaging a whitelight broadband target (star, galaxy or globular cluster) the filter would reduce both the noise and signal by approximately a factor of four. If you compensate by increasing the exposure time, it seems that you should still get an overall slightly better signal to noise ratio with the filter.

[@jlecomte0719]

@@Robservatory > Isn’t the point of a light pollution filter to cut down the light pollution and allow for longer exposures? Not at all. The point of a light pollution filter is to increase the SNR. That it may allow for longer single exposures is almost irrelevant in the era of extremely low read noise CMOS cameras. In the CCD days, when the read noise was quite high, it would have been a possible benefit, but not anymore.

[@jlecomte0719]

@@garybrown1127 > If you compensate by increasing the exposure time, it seems that you should still get an overall slightly better signal to noise ratio with the filter. No. If you are referring to the sub-exposure time, it is not relevant (see my response to @Robservatory) and if you are referring to the total exposure time, you will get better results if you increase the total exposure time on a standard L filter.

[@meibergstrmandersen9181]

Hi Julien, do you have planned any test / comparisons of triband filters, for us novices using color cameras and on the hunt for good filters. Could be like Anitllas RGB Triband or Idas GNB filters.

[@darkskygeek]

Only if a manufacturer sends me one of these free of charge. Otherwise, no. These reviews are extremely time consuming, so I want to focus on other things.

[@meibergstrmandersen9181]

Makes sense 🙂

[@khoatran9804]

The use case I would want to try with this filter is to use it as the L-channel for RGB imaging of emission nebula targets, especially those with reflection components. There, a clear filter for L would be too low-contrast, and narrowband/dual-narrowband as L might be too aggressive. Clearly, as your #s show, reflection nebula targets are not the use case for this filter, and I agree that it'll almost certainly be a good visual filter for emission nebula targets.

[@darkskygeek]

You can certainly try and compare. If you do, please share your results here. Thanks!

[@licho52]

One of the problems of filters having a case use is that the capacity of filter wheels is limited. I can't simply put a limited use filter in my 7 position EFW.

[@desbarry8414]

Would this be a suitable filter for broadband targets using a osc camera and a fast f2 Hyperstar setup? I use a pre bandpass shifted IDAS UHC NBZ for narrowband emission objects but looking for a filter for broadband targets but with a fast hyperstar setup. Thanks

[@darkskygeek]

A simple luminance filter will give you similar or better results. That is the conclusion of this video.

[@desbarry8414]

@@darkskygeek yes thankyou for reply

[@deep_space_dave]

I agree this filter should not be used as a broadband LP filter but it does work well for narrowband. I use this filter and it is great for bright and planetary nebulae.

[@darkskygeek]

Sure. It will not perform as well as a narrowband filter, however. I think that was my conclusion in this video. Basically, use a standard (and cheap) UV/IR cut filter for broadband targets, and a narrowband filter for emission nebulae.

[@deep_space_dave]

@@darkskygeek yep the only way to combat LED lights is more integration time. BTW thanks for the video!

[@JasonFennec]

This filter appears to be designed for narrow band targets, not broadband targets. You will need to integrate luminance data from a broad band light pollution filter, or an IR pass filter, to recover your stars. Keep in mind that refractors may not be apochromatic for IR light.

[@darkskygeek]

No, absolutely not. Their marketing material, which I showed in the video, states that this filter was designed as a replacement for a luminance filter in LED light-polluted areas. The point of this video is that for imaging, this filter does not work well for that.

[@JasonFennec]

@@darkskygeek Hence, why it looks meant for nebulas, not galaxies.

[@kecimalah]

Isnt problem that UHC filters are not suitable for broadband targets like galaxies as it is blocking big part of its spectrum? If you try it on suitable targets like emission nebulas it should work much better.

[@darkskygeek]

The UHC-L filter is SPECIFICALLY marketed as a luminance filter replacement in areas affected by LED light pollution. This video proves that it does not work. Additionally, for narrowband targets, a narrowband filter will work much better. That is the conclusion of this video.

[@kecimalah]

OK, but 2 main usages as written on their web are: -Visual and photographic UHC-L Nebula filter with highest transmission -Ultra High Contrast L Filter blocks city lights and increases contrast of nebula and comets (C2 lines) So it would be good also to test that filter also on theese objects to see how it works. But you are right that statements you mentioned are not true..

[@khoatran9804]

@@darkskygeek and @kecimalah : I think it's generally-understood that a UHC filter is only good for emission nebula targets, but I can see how the wording on their website can be confusing. It says: Can be used as LED-opimized Luminance filter for RGB-Imaging wtih skyglow suppression if they simply inserted [of emission nebulae] after "RGB-Imaging" there wouldn't be a problem. I think it might be useful for a target such as the Lagoon and Trifid nebulae complex, for instance, where there are enough reflection components that using a duo-narrowband filter as luminance would be too restrictive, and the UHC-L would have higher contrast than a clear L filter.

[@darkskygeek]

@@khoatran9804 Right, but again, a narrowband filter will give you much better results on emission nebulae, which is my conclusion. For astrophotography, this filter serves no purpose. But visually, it is probably quite good.

[@khoatran9804]

@@darkskygeek I disagree. Not all emission nebulae need to be imaged in narrowband all the time, and I would rather use this as an L channel over a dual narrowband filter for some objects where there’s more going on than just Ha and OIII. I did post a comparison image of this filter vs the ZWO duo band under my previous comment, but I can’t find it and it mustn’t have gone through the first time. I’ll post it again later when I get back to a desk. Cheers.

[@UnderDriven17]

If S = 100 and LP = 50 then the SNR = 8.165, so it is possible to get a better SNR with this filter--but you have to have more signal. So it seems the question is really whether you are willing to increase exposure time sufficiently to increase the SNR when using this filter. If the number of electrons captured is proportional to exposure time, then 2.5x exposure increases SNR by ~1. 5x longer exposure increases SNR to ~11.5, which is more significant. 7.5x longer exposure increases the SNR to ~14.1, doubling the SNR without the filter. I don't know if taking 8x subs and stacking them is equivalent to a single 8x exposure time increase, though, as I am not an experienced imager. However, I thought I should point out that this filter can be effective if exposure time is increased sufficiently, based on the math in your video. You do mention increasing exposure time in an offhand way, but it might have been helpful to give an example (or show a graph) of how increasing exposure time with a light pollution filter can achieve a better SNR. Whether this one is better than the other similar filters on the market is another issue--I don't know if your mathematical analysis is sufficient to compare different filters precisely. It might be better to just do imaging tests and look at the resulting photos, as other people do on KZbin. That said, I appreciate the spectral testing of this filter--it's nice to see if the bandpass graphs are accurate.

[@darkskygeek]

If you increase exposure time with a regular luminance filter by the same factor, you will get higher SNR with the luminance filter than with the Baader UHC-L. That’s the whole point of this video. Maybe I wasn’t clear enough about that. So, for astrophotography, this filter serves no purpose in my opinion. Visually, however, it can be quite nice by darkening the background, but don’t expect a miracle. Nothing replaces dark skies. CS!

[@UnderDriven17]

@@darkskygeek Thanks for your reply. Yes, using your formula if you increase the exposure by a factor of 4 with the L3 (which is only a UV-IR cut filter) you get the same SNR as with the 7.5x longer exposure with the Baader filter. However, you must be able to capture 800 total electrons with the L3, as opposed to only 450 electrons with the Baader filter. If you have no limit on the number of electrons your sensor can capture, then the L3 is the winner. For a single exposure the well capacity of the sensor could come into play, and thus the Baader could be better (there has to be some advantage in filtering out light you don't want to capture). If stacking shorter subs is equivalent to a single long exposure, then there is effectively no well capacity limit, and there is no point using the Baader filter (or any light pollution filter). However, claiming that all LP filters are useless for AP seems extreme, given that so many are sold and used by photographers. Maybe it's true, but I would think that experienced imagers would have demonstrated this long ago, and thus no one would be manufacturing LP filters. Am I missing something?

[@darkskygeek]

There are lots of threads on this topic on Cloudy Nights, and the point that broadband LP filters are useless for astrophotography with the modern form of light pollution caused by broadband LEDs is now well established.

[@UnderDriven17]

@@darkskygeek If the LP is uniform in intensity, I agree. However, your LP graph was not flat. It seemed there was a significant amount of energy in the blue part of the spectrum that would have been rejected by the Baader filter, as well as some in other wavelengths. I agree that things were much easier when city lighting was LPS and HPS--those narrow sources could easily be filtered out. Nebula Photos' review of LP filters (kzbin.info/www/bejne/jqDVoWecZZqXnrc) is an interesting watch--he concludes that LP filters can be useful in some circumstances, primarily with unmodified DSLR and mirrorless cameras in Bortle 9 skies. My house is Bortle 5, and a 30-minute drive can get me to Bortle 4. In those conditions it probably isn't worth buying an LP filter. Thanks for the conversation...

[@testboga5991]

The problem isn't the filter, the problem is the LEDs continuous spectra

[@darkskygeek]

The marketing materials for that filter indicated that it was able to boost SNR in areas with light pollution caused by LEDs. This video proved that this was a misleading statement, full stop.

[@anata5127]

I cannot understand why manufacturers don’t include 750-800nm in such filters. This band is very effective for galaxies.

[@darkskygeek]

I agree. This filter does not let enough red/infrared through.

[@MrWacha]

Isn't IDAS GNB supposed to do just that?

[@darkskygeek]

@@MrWacha Yes, you are absolutely right! It does just that. But even in the case of the IDAS GNB, I don't understand why they cut between ~680nm and ~780nm...

[@anata5127]

@@MrWacha Yes. I take pictures only with mono camera. Galaxies are often LRGB + Sloan-I filter. L+Sloan-I is Lum.

[@licho52]

The reason is that most refractors or Newtonians with CC aren't corrected in this band and star bloat would occur. That said the latest Antila Quad does just that.

[@gillesleroy734]

Vidéo très instructive comme d'accoutumé. Par contre, il me semble que les cibles choisies pour les tests photographiques ne sont pas pertinentes. Les filtres UHC sont généralement dédiés à l'observation de certains types de nébuleuses et plus particulièrement les planétaires. Même en visuel, ce type de filtre n'est pas conseillé pour les galaxies. Il faudrait donc faire un test photo sur M57 par exemple où la majorité de ses raies d'émission se trouvent dans la bande passante de ce filtre UHC-L. Le spectre d'une galaxie s'étendant sur l'ensemble du spectre de la lumière visible, il n'est pas surprenant que le filtre L3 s'en sorte mieux que le UHC sur ce genre de cible...

[@darkskygeek]

Salut! En fait, si vous regardez la vidéo attentivement, vous verrez que j'ai intentionnellement choisi des cibles qui émettent sur une large partie du spectre visible car le fabricant prétend que ce filtre peut être utilisé en remplacement d'un filtre luminance dans des zones polluées par des lumières d'origine LED. Mes calculs et mes images montrent que ce n'est pas vrai. Et pour les nébuleuses diffuses ou planétaires, des filtres à bandes étroites sont bien meilleurs. Voila, j'espère que c'est plus clair. A+

[@gillesleroy734]

@@darkskygeek J'avoue que cette subtilité en langue de Shakespeare m'avait échappée... Comme j'avais l'intention de me procurer ce filtre aussi bien pour du visuel que pour de l'imagerie des nébuleuses planétaire par soucis d'économie, un complément d'informations avec essais à la clé allant dans ce sens serait bien cool 🙂. Peut-être que ce filtre peut être utilisé en luminance pour des objets bien précis...