246 replies to this topic

[bratislav]

Jim,

check the last picture in post #28. This test was done by Optolong themselves, so it seems your filter was done using HWB630 substrate, not B270. 

[yuriy oseyev]

 

So interest test. How about the same test for Optolong-OIII?

Hi Yuriy,

I have tested both Optolong 6.5nm OIII and Baader 8.5nm OIII. No significant off-band failure has been detected. But Mike did another test and showed there is some red leaking through https://www.cloudyni...lter/?p=8282207.

He are my tests:
 

Regards,

 

Thanks for test. I see the same problem. Optolong OIII filter, has little leaking through the red.

[Fongky]

I wonder Fongky if somehow your friend simply didn't get the filter he was supposed to at all.  Maybe the Halpha narrow band filter performs like the Optolong measured data but by accident some other filter glass got put in the aluminum cell and shipped to your friend.  When I look at your colour pattern with my mirrorless camera through my Optolong Halpha 7nm I get the attached image below.  Clearly different than what your friend has.  It is like he got entirely the wrong filter.

 

Best Regards,

 

Jim T.

That was my initial thought as well but Optolong has denied such mistake at their end. They sent me a a document with the picture of the test certificate of my friend's Ha filter, with the screen capture of the file folder and the transmission curve generated when this filter was tested for QC before shipping. Here is the document, https://drive.google...ew?usp=sharing. From the document, it seems they are able to trace the QC records of all the filters produced by them.

Is there a way to know that the filter you have tested is made of HWB630 or B270 substrate?

Regards,

[GA-HAMAL]

At one time I also found the interesting anomaly.

 

https://www.cloudyni...der-sii-filter/

[vdb]

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

Edit: anyway you will always have some signal in Blue and Green, so make sure your debayer algo does not use this as it will inject noise into your image. (APP has a special Ha debayer algo to take care of this)

 

/Yves

Edited by vdb, 21 December 2017 - 06:13 AM.

[rockstarbill]

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

/Yves

You should read the thread. You are way off of the mark of what this is all about. 

 

Case in point -- this has nothing at all to do with Bayer Matrices. 

[vdb]

 

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

/Yves

You should read the thread. You are way off of the mark of what this is all about. 

 

Case in point -- this has nothing at all to do with Bayer Matrices. 

 

I'm not 100% sure, there is interaction with the bayer matrix/filters and do you know for sure that interaction does not influence the result, are you a specialist ... if so please explain as otherwise I find this remark quite sharp ...    

 

 

/Yves

[vdb]

To be complete I tested this myself under a real sky with a Baader Ha 3.5nm filter and did find my Ha object in green and to a lesser extend in blue, which is quite normal as you have 2x green pixels ...

 

/Yves

[Fongky]

To be complete I tested this myself under a real sky with a Baader Ha 3.5nm filter and did find my Ha object in green and to a lesser extend in blue, which is quite normal as you have 2x green pixels ...

 

/Yves

I have a Baader Ha 7nm filter. I don't get anything in green or blue other than noise in these colors, when imaging the sky with a modified Pentax K-5 DSLR through a telescope. I suspect the green that you get may be injected by the camera software for the purpose of white balancing.

Edited by Fongky, 21 December 2017 - 08:22 AM.

[vdb]

 

To be complete I tested this myself under a real sky with a Baader Ha 3.5nm filter and did find my Ha object in green and to a lesser extend in blue, which is quite normal as you have 2x green pixels ...

 

/Yves

I have a Baader Ha 7nm filter. I don't get anything in green or blue other than noise in these colors, when imaging the sky with a modified Pentax K-5 DSLR through a telescope. I suspect the green that you get may be injected by the camera software for the purpose of white balancing.

 

I record pure raw and split channels before debayer ...

I have been thinking and maybe an explanation could be reflection of the cover glass to filter and back ...

 

Could also be an issue ...

 

/Yves

Edited by vdb, 21 December 2017 - 09:23 AM.

[Bart Declercq]

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

Edit: anyway you will always have some signal in Blue and Green, so make sure your debayer algo does not use this as it will inject noise into your image. (APP has a special Ha debayer algo to take care of this)

 

/Yves

An observation: even if there's light bleed in the bayer mask, that's not what we're seeing here - if the bayer mask lets red light through to the blue & green pixels, that would turn the red&white parts of the test image pale red (R + some G & B) - the blue & green parts of the test image should remain black (as that light is supposed to be blocked by the H-alpha filter, not the bayer mask)

 

That is not what we're seeing here, we're seeing actual Blue & Green light get through the H-alpha filter to be detected by the color sensor behind it.

 

So like others said, this issue has nothing to do with light bleed on a Bayer mask - in fact, you can safely do the same test using a monochrome camera, though the effect wouldn't be quite as obvious or easy to interpret - in the Optolong filter, the red circle and white background would be "white" and the blue & green sections would be grey, while in the Baader filter, the blue & green sections would be black.

[vdb]

 

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

Edit: anyway you will always have some signal in Blue and Green, so make sure your debayer algo does not use this as it will inject noise into your image. (APP has a special Ha debayer algo to take care of this)

 

/Yves

An observation: even if there's light bleed in the bayer mask, that's not what we're seeing here - if the bayer mask lets red light through to the blue & green pixels, that would turn the red&white parts of the test image pale red (R + some G & B) - the blue & green parts of the test image should remain black (as that light is supposed to be blocked by the H-alpha filter, not the bayer mask)

 

That is not what we're seeing here, we're seeing actual Blue & Green light get through the H-alpha filter to be detected by the color sensor behind it.

 

So like others said, this issue has nothing to do with light bleed on a Bayer mask - in fact, you can safely do the same test using a monochrome camera, though the effect wouldn't be quite as obvious or easy to interpret - in the Optolong filter, the red circle and white background would be "white" and the blue & green sections would be grey, while in the Baader filter, the blue & green sections would be black.

 

You clearly missed my last remark ;-)

What about refections, look I'm not saying there is an issue with this filter, but I find it hard to believe Optolong would have such a bad behaviour, or they really have burned them selves ... as they even can trace back to the individual tests and claim nothing is wrong. That would be serious ...

 

/Yves

 

 

/Yves

Edited by vdb, 21 December 2017 - 09:48 AM.

[jimthompson]

I did not read trough all the posts so maybe this has already been mentioned, and what I'm about to say does not explain the difference between Optolong and Baader ...

OSC bayer bleed all over the spectrum, I can "See" blue and green even when using Ha filter from baader, even the 3.5nm, this is inherent to 2 problems, bayer filter overlaps, so yeah that explains why you see green ... but not the blue, well the blue is from not parallel light, so "red" photon's end up in Blue baskets ... so use a high F number ... anyway you will always find leakage, I did anyway ;-)

 

So if Optolong and Baader where done with identical setups then of course the above does not explain what you are seeing in the tests. 

 

Edit: anyway you will always have some signal in Blue and Green, so make sure your debayer algo does not use this as it will inject noise into your image. (APP has a special Ha debayer algo to take care of this)

 

/Yves

Hi Yves,

 

The "Bayer leakage" you speak of is entirely in the near infrared band and is a result of the realities of making RGB filters using absorptive type materials.  The cameras are designed for imaging in the visual band so Bayer matrix filters are designed to achieve the desired performance in the visual band.  This results in all three colour channels having some level of response to near infrared.  Your OSC camera interprets this NIR light received on the G and B channels as green and blue light even though it isn't.  This can often cause some unusual appearing images when you let the camera do the white balancing.  Myself I often experience images that have a pinkish or orangish tint as a result.  This is easily dealt with by adding a UV/IR cut filter or by using a band pass filter.  If you are imaging with a narrowband filter you are probably better off manually white balancing your OSC camera so the software does not inflate the off-band colour channels introducing noise.  For example when using an Halpha band pass filter you will probably get better contrast in your images (but poorer resolution) if you manually adjust the G & B channels down to zero.

 

By the way, I am fairly certain that the leakage you speak of is not what is being observed with the Halpha filter of Hongky's friend.  There would be little or no near-infrared emitted from the LED monitor used for the filter test, so the blues and greens we are seeing are in fact blue and green light that is being passed by the Halpha filter.

 

Best Regards,

 

Jim T.

[jimthompson]

Is there a way to know that the filter you have tested is made of HWB630 or B270 substrate?

***************

Hi Fongky,

 

I dug through my files this morning and found the invoice for my Optolong Halpha filter.  It was purchased in early November 2016 from Agena Astro in California.  It is possible I have an older filter that was made with the HWB630 glass depending on how frequently the stock at Agena Astro is getting replaced. 

 

When I hold my Halpha filter at arms length and slowly rotate it, the colour of the filter goes from red to orange and then back to red.  My guess is it is made from the HWB630 glass.  If your friend holds his filter up at arms length in front of a white background and slowly rotates it, it should change colour from red to orange to green to blue to purple as he rotates it, if indeed it is made from B270 glass.  If it does not look dark red to begin with, not rotated, then his filter is not an Halpha filter.  For example Optolong's UHC, CLS, Moon and Sky, and L-Pro filters will all have a light purple appearance when you hold them up and look at a white background.

 

I searched around a bit online on the two glass types mentioned by Optolong.  I found B270 easily, it is a standard type of clear white crown glass from Schott.  My guess is that since it is a commonly used glass it is easier to source in good quality at low cost.  I could not find a direct web reference to HWB630 but my guess is that it is a red absorptive glass, a low pass filtering glass with cut-off wavelength of 630nm.  From Schott this would be called RG630 glass, or from Hoya R-62.

 

Best Regards,

 

Jim T.

[Fongky]

When I hold my Halpha filter at arms length and slowly rotate it, the colour of the filter goes from red to orange and then back to red.  My guess is it is made from the HWB630 glass.  If your friend holds his filter up at arms length in front of a white background and slowly rotates it, it should change colour from red to orange to green to blue to purple as he rotates it, if indeed it is made from B270 glass.  If it does not look dark red to begin with, not rotated, then his filter is not an Halpha filter.  For example Optolong's UHC, CLS, Moon and Sky, and L-Pro filters will all have a light purple appearance when you hold them up and look at a white background.

 

 

Thanks for the information.

Using your method, the color changes from yellow to pink, orange, red, light blue, green (more like teal blue), red, orange, pink, then back to yellow as I rotate his filter. It turns red but not dark red. There is no purple appearance but red against a lit white background. It is at night here. I think I will try it again in the morning.

[jimthompson]

Thanks for the information.

Using your method, the color changes from yellow to pink, orange, red, light blue, green (more like teal blue), red, orange, pink, then back to yellow as I rotate his filter. It turns red but not dark red. There is no purple appearance but red against a lit white background. It is at night here. I think I will try it again in the morning.

 

 

Yikes!

 

To get that kind of response the filter would need multiple pass bands that all shift as you rotate the filter.  It is a little hard to understand but I will try to explain:  The Halpha filters and other narrow band filters we have been talking about are "interference" type.  They are made by many many thin layers of glass-like materials with different refractive indexes.  At each interface between layers you get some light passing through and some reflecting.  By tuning the thickness of the layers you can construct a filter that preferentially passes a particular range of wavelengths.  A filter constructed in this way using a small number of layers has a response that looks like the image below, something called an etalon.  There is light passed at the design wavelength as well as at harmonics of the design wavelength.  By adding more layers of different thicknesses, or by using absorptive band pass glass, you can block the harmonics and leave just the design wavelength.  By your description of what you see when you tilt the filter, it sounds like your friend either got a totally different multi-band filter or an Halpha filter that did not complete its coating process and still has a number of harmonics passing through the filter.

 

By the way, the tilt test works by way of the fact that tilting the filter effectively increases the thickness of the layers, shifting the design wavelength of the filter down (towards blue).  The more you tilt the more towards blue the pass bands are shifted.  Eventually you can shift them past violet and can not see them anymore.  In your case, as one pass band is being pushed into the UV another is coming into view at the red end from infrared.  Note that pink is the result of seeing a pass band at blue and red simultaneously.

 

Best Regards,

 

Jim T.

Attached Thumbnails

• 
• 

[TareqPhoto]

How do I test my LRGB filters with mono camera with that band passing charts?

Also, from this topic, I don't know which nb filters to buy not Astrodon.

[SandyHouTex]

I tested my Orion NB filters with test image back on the 1st page, and the Ha and SII work great.  However the OIII lets a bit of red through and some of the blue and green.  The brightest is the blue/green overlap.

 

Weird.

[TareqPhoto]

Oh, i am sorry and apologize about my post above earlier, it is about NB filters then, not LRGB.

[Fongky]

 

Yikes!

 

To get that kind of response the filter would need multiple pass bands that all shift as you rotate the filter.  It is a little hard to understand but I will try to explain:  The Halpha filters and other narrow band filters we have been talking about are "interference" type.  They are made by many many thin layers of glass-like materials with different refractive indexes.  At each interface between layers you get some light passing through and some reflecting.  By tuning the thickness of the layers you can construct a filter that preferentially passes a particular range of wavelengths.  A filter constructed in this way using a small number of layers has a response that looks like the image below, something called an etalon.  There is light passed at the design wavelength as well as at harmonics of the design wavelength.  By adding more layers of different thicknesses, or by using absorptive band pass glass, you can block the harmonics and leave just the design wavelength.  By your description of what you see when you tilt the filter, it sounds like your friend either got a totally different multi-band filter or an Halpha filter that did not complete its coating process and still has a number of harmonics passing through the filter.

 

By the way, the tilt test works by way of the fact that tilting the filter effectively increases the thickness of the layers, shifting the design wavelength of the filter down (towards blue).  The more you tilt the more towards blue the pass bands are shifted.  Eventually you can shift them past violet and can not see them anymore.  In your case, as one pass band is being pushed into the UV another is coming into view at the red end from infrared.  Note that pink is the result of seeing a pass band at blue and red simultaneously.

 

Best Regards,

 

Jim T.

 

Great explanation! Thanks

I know the physics behind constructive and destructive interference and etalon, also how it is applied to solar filter. However, I have never thought they use it for narrow band filters. I should do more research in this area.

It is difficult for me to hold the camera against a bright white background. Instead, I use the sunlight that is coming through my window to project through the Optolong Ha NB filter as I rotate it. Here is the video that looks exactly what you have described, https://youtu.be/7PwAENhugEs

 

Everything makes more sense now. The older (manufactured before August 2016) Optolong Ha and SII NB filters were using HWB glass which acts as absorptive filter that allows only the infrared passband. Then, it is coated with interference coating to cut off the remaining off-band wavelengths to achieve the desire narrow passband. While the new NB filters rely solely on interference coating to get the job done.
 

The remaining question I have now is whether these new NB filters are really OD3 since Optolong only tested them for sodium and mercury emissions.

Cheers,
 

Edited by Fongky, 21 December 2017 - 09:53 PM.

[Fongky]

Oh, i am sorry and apologize about my post above earlier, it is about NB filters then, not LRGB.

In fact, you can test the RGB filter the same way. The red filter should only see the red color while blue and green appear as black.

[Fongky]

I tested my Orion NB filters with test image back on the 1st page, and the Ha and SII work great.  However the OIII lets a bit of red through and some of the blue and green.  The brightest is the blue/green overlap.

 

Weird.

So long the off-band (such as red for OIII filter) is not excessive, the filter should be fine. You will need special equipment to measure the amount of off-band blocking.

The Doubly ionized oxygen or OIII emission is at 493.1nm, 495.89nm, and 500.69. They are indeed between green and blue. Visually, some describe it as teal blue.
 

[GA-HAMAL]

Leakages aren't a fault of the mask or other secret phenomena, I one's originally I noticed with the naked eye in the spectroscope.

[cuivienor]

Wow, this is terrible. I just got my LRGBSHO filters from Optolong via Robin at Cyclops Optics. If this is correct, I would definitely feel cheated. I based my purchase on sample pictures and reviews that were likely taken/done based on the filters produced using the old manufacturing process. The curves on the site may also be valid for the old manufacturing process and not the new one.

 

I have sent an email from Robin at Cyclops Optics referencing this thread as well as the talks with Optolong - as a reseller, he may be able to get more information (and I feel he'd be angry/disappointed as well to learn that) and force them to at least change their advertising, and/or produce new full spectrum curves for the "v2" of these filters.

 

The change in manufacturing process is really creating a second, different product than the filters that have been reviewed up to now. So in effect we are being sold product 1, but delivered inferior product 2.

Edited by cuivienor, 22 December 2017 - 01:08 AM.

[Fongky]

Wow, this is terrible. I just got my LRGBSHO filters from Optolong via Robin at Cyclops Optics. If this is correct, I would definitely feel cheated. I based my purchase on sample pictures and reviews that were likely taken/done based on the filters produced using the old manufacturing process. The curves on the site may also be valid for the old manufacturing process and not the new one.

 

I have sent an email from Robin at Cyclops Optics referencing this thread as well as the talks with Optolong - as a reseller, he may be able to get more information (and I feel he'd be angry/disappointed as well to learn that) and force them to at least change their advertising, and/or produce new full spectrum curves for the "v2" of these filters.

 

The change in manufacturing process is really creating a second, different product than the filters that have been reviewed up to now. So in effect we are being sold product 1, but delivered inferior product 2.

I have just checked (December 22, 2017, 12:21, GMT+8) the Chinese version of the specification of Optolong Ha 7nm NB filter. Since the manufacturing materials and process were changed in August 2016, the specification is still saying these filters are made with HB610 substrate. Here is the translation of two red boxes I highlighted:

 

1. "....using HB610 color glass that effectively absorbs 200-610nm......"

2. "Glass material: Chinese HB610 color glass (similar to the American RG610 substrate), utilize the absorption property of the glass that absorbs UV and those below 610nm to better block off unwanted light"

You may highlight this to Robin.
 

Attached Thumbnails

• 

Edited by Fongky, 22 December 2017 - 02:08 AM.