33 replies to this topic

[jimthompson]

Greetings All,

 

I have recently procured a sample of the Radian Triad Ultra (used) and the Antlia ALP-T, and have been working on some testing to compare how these filters perform relative to each other.  To fill out the comparison I have also included some other well established filters:

 

- Baader Planetarium UV/IR Cut ($127)

- IDAS NB1 ($199)

- IDAS NBZ ($299)

- Optolong L-eXtreme ($309)

- Radian Triad Ultra ($1075)

- Antlia ALP-T ($380)

 

My testing included both imaging and bench top spectrometer measurements.  I intend to put a test report together eventually, but figured I would share my test results so far.  You can view my image captures at the Flickr link below.  Also posted to the album are my measured filter spectra, plotted assuming a few different telescope f-ratios.

 

https://flic.kr/s/aHBqjzAbzR

 

My initial take-away is that the difference in performance between the NBZ, L-eXtreme, Triad Ultra, and ALP-T is pretty subtle.  Based on what I am seeing, I don't understand why anyone would pay $1075 for the Triad Ultra when it does not perform significantly different from filters 1/3 the price.

 

I will update the thread as I get more of my data reduction completed.

 

Best Regards,

 

Jim T.

[OregonAstronomer]

Hello Jim,

 

Thanks so much for performing such a valuable service for us all.    I, for one, will look forward to your completed results.

 

Sincerely,

 

Arnie

[SilverLitz]

To me:

 

NBZ looks clearly better than NB1

Antilia ALP-T has best halo control

On M42, Triad Ultra has a nicer color diversity (more blues, not as red dominant; due to Hb??) 

[bobzeq25]

The imaging results (actual, not theoretical) will vary a lot depending on the optical speed of the setup.  The notable advantage of something like the NBZ is how much better it works than a non-high speed filter on really fast optics.  For people with those, engineered high speed filters are a clear choice.  For other people, it's a minor thing.

Edited by bobzeq25, 27 January 2022 - 06:14 PM.

[slepage]

When you scroll over each image you have the number of subs and the exposure times.  I noticed that for each filter they are different from other filters.  I'm not sure how you can compare the results?  I would propose that you try to control all variables other then the filter.

[bobzeq25]

When you scroll over each image you have the number of subs and the exposure times.  I noticed that for each filter they are different from other filters.  I'm not sure how you can compare the results?  I would propose that you try to control all variables other then the filter.

That's risky.  DSO AP has a great many factors involved.  Artificial tests that freeze parameters may not give meaningful results.  What counts is how "X" affects the final image.

[jimthompson]

When you scroll over each image you have the number of subs and the exposure times.  I noticed that for each filter they are different from other filters.  I'm not sure how you can compare the results?  I would propose that you try to control all variables other then the filter.

This topic has been a challenge for a long time:  how do you collect image data in such a way that the results can be compared equally?  I have selected to compare filters with quite a large range of total transmission rate, something I call "Luminous Transmissivity".  As a result, if I tried to use a single sub-exposure time for all the filter configurations, the brightnesses would be all over the place.  The reference image with just the UV/IR filter would be totally over saturated (white end of histogram) if I used the same sub-exposure as I did for the ALP-T for example.  Conversely if I used only 3.5sec subs so the UV/IR cut image was not over saturated, the image with the ALP-T would be woefully under saturated (black end of histogram).  The compromise I have selected is to vary the sub-exposure time from filter to filter in order to give the same apparent scene exposure.  In practice I did this by matching visually the extent of the over exposure in the center of M42 ... how many stars are visible, how many are obscured by the saturated part of the image, etc.  I then stacked subs to get the same total image exposure time of around 5 minutes.  Finally, I did my best to white balance and stretch each stacked image exactly the same.  I think the result is images that are visually comparable, but it is certainly not a perfect method.  That is why I normally put more weight on the filter spectrum measurement results.

 

Best Regards,

 

Jim T.

[DogPaws]

I ran a similar test last summer... to be able to compare I was (given the same optics and caméra) targeting a specific ADU by adjusting sub exposure time.... so for instance to reach 1500 ADU on a lum filter I needed 15 sec subs.  To reach the same ADU of 1500 on a l-extreme I needed subs to be 90 seconds .... etc...

.

Then i would stack the same amount of subs to eventually compare.  For instance I choose the best 90 subs of a 150 image set for each filters...

 

my my logic was that normalizing ADU across filters would then make all the camera parameters that could influence the result similar (gain, offset, QE, saturation etc...). That gave me images with similar dynamic ranges.  

 

Now i have no clue if that was a good approach but I found out (at least for my usage). The slights difference to choose between the filters I examined.

 

hopefully we can debate on how to make things comparable with a test protocol.

 

good luck

Claude

 

btw:  I found out that my most expensive multi band was not necessarily the best and that some of the more economical where actually very decent.

[dx_ron]

The imaging results (actual, not theoretical) will vary a lot depending on the optical speed of the setup.  The notable advantage of something like the NBZ is how much better it works than a non-high speed filter on really fast optics.  For people with those, engineered high speed filters are a clear choice.  For other people, it's a minor thing.

The NBZ clearly is the only viable option for f/2, among the filters in the test. The way it accomplishes this, while still being usable at slower speeds is by having a much wider passband. Even at f/3, the performance of the L-eXtreme is only slightly degraded and should be quite usable. The NBZ does have a 10-15% absolute advantage in transmission at slower speeds, along with the wider passband. Is it also typical for mono filters to trade off max transmission for selectivity when going from 12nm to 6nm?

[jimthompson]

I have finished my test report and it is ready to download for those interested.  In it I compare five different multi-narrowband filters:

 

- IDAS NB-1

- IDAS NBZ

- Optolong L-eXtreme

- Antlia ALP-T

- Radian Triad Ultra

 

Comparisons are performed using my benchtop spectrometer measurements, and one-shot colour images.  The bottom line is that the new ALP-T performs pretty well; slightly better than all the other filters tested.  I also determined that the performance of the Triad Ultra is not significantly different than the L-eXtreme, and that it is the most sensitive to f-ratio out of all the filters tested.

 

http://karmalimbo.co...est_Feb2022.pdf

 

Cheers,

 

Jim T.

[AarondeVries]

Another fine read from you, Jim.
What surprises me is the only slightly better contrast gains of the 5nm vs the 12nm bandpasses? I have the NBZ filters and am quite pleased with them. Based on these findings it looks like I should stick with them as going to the current batch of duo filters with narrower bandpasses does not give much of an improvement.
Halo control on the Antlias is impressive, though.

Edited by AarondeVries, 17 February 2022 - 04:41 AM.

[jimthompson]

Another fine read from you, Jim.
What surprises me is the only slightly better contrast gains of the 5nm vs the 12nm bandpasses? I have the NBZ filters and am quite pleased with them. Based on these findings it looks like I should stick with them as going to the current batch of duo filters with narrower bandpasses does not give much of an improvement.
Halo control on the Antlias is impressive, though.

Hi Aaron,

 

I suspect that if I used fainter targets and/or longer total exposure times (i.e. hours instead of minutes), the difference between filters would be easier to see in my images.  In theory imaging results closer to the predictions of contrast improvement presented in Figure 9 should be achievable with more time and effort applied.  Different levels of light pollution probably also plays a role.  I guess the question is how much of a theoretical contrast improvement is really necessary when you can compensate with image post processing tools?

 

Best Regards,

 

Jim T.

[DogPaws]

I have finished my test report and it is ready to download for those interested.  In it I compare five different multi-narrowband filters:

 

- IDAS NB-1

- IDAS NBZ

- Optolong L-eXtreme

- Antlia ALP-T

- Radian Triad Ultra

 

Comparisons are performed using my benchtop spectrometer measurements, and one-shot colour images.  The bottom line is that the new ALP-T performs pretty well; slightly better than all the other filters tested.  I also determined that the performance of the Triad Ultra is not significantly different than the L-eXtreme, and that it is the most sensitive to f-ratio out of all the filters tested.

 

http://karmalimbo.co...est_Feb2022.pdf

 

Cheers,

 

Jim T.

A fine study that put science into perception.  As I said above when I ran my qualitative study on a bunch of multi band filters I could not justify the cost of some of them...but I do use Antlia filters and appreciate the performance.  As for the radian quad.... well I reached similar conclusion as you.... 

 

thanks for the work, it is not easy to be methodical but greatly appreciated

 

Claude

[UKalwayscloudy]

Great work here. I've got an NBZ, NB2 and Altair 7nm dual band. It would be great to see both the standard Altair 7nm and newer 4nm duos added to this comparison. This is the one I have right now:

 

https://www.altairas...r-2-10457-p.asp

 

and this is the 4nm: 

 

https://www.altairas...ort-11468-p.asp

 

Subjectively so far I think the 7nm AA filter is comparable to the NBZ, but it would be good to get a more rigorous comparison. The 4nm vs the Antlia is intriguing. 

[DirtyRod]

I thought it was a good test as well so Thanks to Jim for doing it. 
 

One additional factor to consider is that the Ultra is the only one to pass SII. I have both the L-Extreme and L-Enhance and the only reason I’m considering the Ultra is the additional SII band pass. 
 

Perhaps one day I can do a comparison on a target with SII and post the comparison.

[jimthompson]

Great work here. I've got an NBZ, NB2 and Altair 7nm dual band. It would be great to see both the standard Altair 7nm and newer 4nm duos added to this comparison. This is the one I have right now:

 

https://www.altairas...r-2-10457-p.asp

 

and this is the 4nm: 

 

https://www.altairas...ort-11468-p.asp

 

Subjectively so far I think the 7nm AA filter is comparable to the NBZ, but it would be good to get a more rigorous comparison. The 4nm vs the Antlia is intriguing. 

Thanks for bringing the new 4nm Altair filter to my attention, I was not aware of it.  I have no experience with Altair brand filters as they are not easy to get in Canada.  It makes me suspicious that Altair does not publish their filter spectra like other manufacturers do.  I assume they have their filters made by some Chinese OEM, but I have no idea which one.

 

Best Regards,

 

Jim T.

[jimthompson]

I thought it was a good test as well so Thanks to Jim for doing it. 
 

One additional factor to consider is that the Ultra is the only one to pass SII. I have both the L-Extreme and L-Enhance and the only reason I’m considering the Ultra is the additional SII band pass. 
 

Perhaps one day I can do a comparison on a target with SII and post the comparison.

FYI, you will never be able to see the SII emission that this filter passes as it will be lost in with the much stronger Halpha signal.  Both of these emissions will be picked up on your camera's red channel, but since the Halpha emission is so much stronger than SII, there is no advantage to the SII pass band.  The same is true with the Hbeta pass band which will be mixed in with the O-III band on your camera's green and blue channels.  If you want to have proper pass band separation so you have more freedom to process the image, you should a dual-band filter.

 

Regards,

 

Jim T.

[UKalwayscloudy]

Jim - Altair produce individual spectrograph analysis for the individual 4nm filters. Some examples are published here.

https://www.backyard...JIWuc&fs=e&s=cl

[DirtyRod]

FYI, you will never be able to see the SII emission that this filter passes as it will be lost in with the much stronger Halpha signal.  Both of these emissions will be picked up on your camera's red channel, but since the Halpha emission is so much stronger than SII, there is no advantage to the SII pass band.  The same is true with the Hbeta pass band which will be mixed in with the O-III band on your camera's green and blue channels.  If you want to have proper pass band separation so you have more freedom to process the image, you should a dual-band filter.

 

Regards,

 

Jim T.

Not sure I understand that last statement but what you are saying is that the Ultra will never show you SII data so the only way to actually pull SII is to go mono and/or use an SII filter. Is that correct?

Edited by DirtyRod, 03 July 2022 - 08:21 PM.

[jimthompson]

Halpha and SII are both emissions in the red part of the spectrum. If you are using a one shot colour camera, both of these emissions will be picked up together on the sensor's red pixels. The camera can't distinguish between Halpha and SII, they are both just "red" to the camera. Halpha is the main emission we see from nebulae. It is a much stronger emission than SII, so if your camera is seeing Halpha you will not see SII because it is drowned out. The only way to capture the SII is with a single narrow band pass filter - SII only, no Halpha.

Best Regards,

Jim T.

[UKalwayscloudy]

If you get an IDAS NB3 filter that will pass Oiii and Sii and you can combine that with data from any of the Ha Oiii dual bands. I agree that you cannot separate two signals on an OSC cam that are both in the red channel, but it is not true that the only way to get Sii is with a single band pass filter. 

 

I've done a bit this way but Roger Nichol is expert at it. He uses an L Extreme and an NB3 with a 2600MC. You get some Oiii from both duos which is a helpful boost to an often faint signal. 

 

https://www.astrobin.com/tv44gx/D/

 

https://www.astrobin.com/lketio/

[jimthompson]

If you get an IDAS NB3 filter that will pass Oiii and Sii and you can combine that with data from any of the Ha Oiii dual bands. I agree that you cannot separate two signals on an OSC cam that are both in the red channel, but it is not true that the only way to get Sii is with a single band pass filter. 

 

I've done a bit this way but Roger Nichol is expert at it. He uses an L Extreme and an NB3 with a 2600MC. You get some Oiii from both duos which is a helpful boost to an often faint signal. 

 

https://www.astrobin.com/tv44gx/D/

 

https://www.astrobin.com/lketio/

Sorry, you are correct.  The IDAS NB-3 will allow one to capture S-II emissions by themselves on the red channel.  I have tried this technique myself in a test report I prepared on IDAS filters, but totally forgot about it.  My previous post was assuming that one was trying to capture S-II at the same time as H-alpha - in that situation the two channels would be captured together on the same sensor colour channel.

 

Best Regards,

 

Jim T.

[UKalwayscloudy]

I keep hoping that the NB3 will get some development or that other players will step in. Lots of progress, as your report shows, on the Ha Oiii combos. The market might be quite small maybe, as folks (including me) tend to drift to mono if they are going to do this a lot. 

[JethroXP]

I image from a Bortle 5/6 location with both an EdgeHD 8" and RASA 8.  I own the L-Pro, L-eNhance, L-eXtreme, and both the standard NBZ and the UHS NBZ.  I don't really use the L-eNhance anymore.  The L-Pro is my filter of choice on both the EdgeHD and RASA for broadband targets such as star clusters and galaxies.  For everything else I'm using the NBZ (UHS version on the RASA).  I actually prefer the tone of the reds from the L-eXtreme but the halos around bright stars is what drives me to use the NBZ instead.  And based on your analysis I just ordered the Antlia ALP-T, so I'm excited to give that a try.

 

I posted a comparison of the NBZ to the L-eXtreme, nothing as detailed as your analysis, but it does show the richer red of the L-eXtreme but also the better ability of the NBZ to pick up blue, and of course no halos like the L-eXtreme.

https://www.flickr.c...th/51907119627/

Thanks for effort, much appreciated!

Edited by JethroXP, 08 July 2022 - 09:38 AM.

[mikenoname]

I have finished my test report and it is ready to download for those interested.  In it I compare five different multi-narrowband filters:

 

- IDAS NB-1

- IDAS NBZ

- Optolong L-eXtreme

- Antlia ALP-T

- Radian Triad Ultra

 

Comparisons are performed using my benchtop spectrometer measurements, and one-shot colour images.  The bottom line is that the new ALP-T performs pretty well; slightly better than all the other filters tested.  I also determined that the performance of the Triad Ultra is not significantly different than the L-eXtreme, and that it is the most sensitive to f-ratio out of all the filters tested.

 

http://karmalimbo.co...est_Feb2022.pdf

 

Cheers,

 

Jim T.

Can anyone else reach this link? Just times out for me.