Can we see Hα?

[PeterC65]

With the absence of clear skies of late I've been considering which filters are useful for visual observing.

I've done the same thing for EAA and for that I created a spreadsheet which allows me to see the overall spectral response of a particular camera sensor with a particular filter on a particular sort of object. So for visual I thought I would add the spectral response of the human eye.

The data I used came from here which gives these response curves ...

The left hand curve is under dark adapted conditions (scotopic vision) and the right hand under daylight conditions (photopic vision).

I was rather surprised by these response curves. Even for the right hand curve, the response to Hα is less than 10% and it is zero when dark adapted.

Does this mean that when my eyes are dark adapted and I'm looking though my UHC filter, all that I see is OIII and Hβ (the same as I would see with an OIII filter)?

 

Edited April 16, 2023 by PeterC65

[ollypenrice]

7 minutes ago, PeterC65 said:

With the absence of clear skies of late I've been considering which filters are useful for visual observing.

I've done the same thing for EAA and for that I created a spreadsheet which allows me to see the overall spectral response of a particular camera sensor with a particular filter on a particular sort of object. So for visual I thought I would add the spectral response of the human eye.

The data I used came from here which gives these response curves ...

The left hand curve is under dark adapted conditions (scotopic vision) and the right hand under daylight conditions (photopic vision).

I was rather surprised by these response curves. Even for the right hand curve, the response to Hα is less than 10% and it is zero when dark adapted.

Does this mean that when my eyes are dark adapted and I'm looking though my UHC filter, all that I see is OIII and Hβ (the same as I would see with an OIII filter)?

 

If you hold an Ha filter up to the light you can certainly see through it and most solar scopes pass Ha only. However, the sun is rather bright! Personally, I've always found visual OIII filters and UHC to give very similar results, though my OIII gives a green tinge which the UHC doesn't.  The fact that DSLR cameras pass very little Ha indicates that it is not a major component to the spectrum we perceive.

Olly

[vlaiv]

9 minutes ago, PeterC65 said:

The data I used came from here which gives these response curves ...

That is not quite response curve of human vision. It is luminosity function which is read like this:

given two colors, bunch of observers will report that their relative brightness for same light intensity is as that curve.

In photopic vision it translated to this:

(do keep in mind that computer screens can't display as saturated color as actual rainbow, so if we compare say 540nm from Baader Solar continuum to above graph - green on it looks rather dull).

If you have uniform light source - like standard illuminant E and you shine it at something white and observe with two filters of same band pass and same peak transmission but one is OIII and other is Ha - relative brightness will be as per above. Ha will seem darker.

But above diagram does not tell if Ha will be detected or not - as this works good in bright scenarios (It is good guide for planetary observations).

On the other hand we have this:

[PeterC65]

Experience suggests that we do see Hα, which implies there is something wrong with the spectral response curve I found. In my search I did come across the separate cone responses, but how do you square these with the luminance response?

My understanding is that for the faint objects we observe in the night sky we see in monochrome, so surely it is the luminance response that is important?

 

[vlaiv]

5 hours ago, PeterC65 said:

Experience suggests that we do see Hα, which implies there is something wrong with the spectral response curve I found. In my search I did come across the separate cone responses, but how do you square these with the luminance response?

My understanding is that for the faint objects we observe in the night sky we see in monochrome, so surely it is the luminance response that is important?

 

Luminance response is given for normal lighting conditions and represents essentially this:

It is the fact that when we present image in black and white without color information - different colors will to us look as different brightness.

Only above is applied to individual wavelengths and not broad spectrum colors (but broad spectrum color brightness is derived from individual wavelength brightnesses of its spectrum).

What you want to look at is this:

Where dotted line is linear sensitivity (we don't perceive in linear but rather in logarithmic - so if in above graph - something is 50% of something else - that does not mean half as bright to us) of Rods and combine this with:

Graph is normalized and does not represent absolute sensitivity of different kinds of receptors. In previous post I posted image of actual blue, red and green cones and their sensitivity with respect to others (non normalized) - where you can see that we are by far least sensitive to blue (or rater blue cones are the least sensitive of the three).

Now mentally add rows 2 (Very light sensitive 6 (stimuli added over time), 7 (Have more pigment than cones, so can detect lower light levels), 9 (x20 more cells in the eye then other type), and you can see that even if graph almost "vanishes" for Ha - we can still see it in the night. Not as good as other wavelengths, that is for sure - but we are not effectively blind to Ha in scotopic vision.

 

[PeterC65]

10 minutes ago, vlaiv said:

Luminance response is given for normal lighting conditions and represents essentially this:

It is the fact that when we present image in black and white without color information - different colors will to us look as different brightness.

Only above is applied to individual wavelengths and not broad spectrum colors (but broad spectrum color brightness is derived from individual wavelength brightnesses of its spectrum).

What you want to look at is this:

Where dotted line is linear sensitivity (we don't perceive in linear but rather in logarithmic - so if in above graph - something is 50% of something else - that does not mean half as bright to us) of Rods and combine this with:

Graph is normalized and does not represent absolute sensitivity of different kinds of receptors. In previous post I posted image of actual blue, red and green cones and their sensitivity with respect to others (non normalized) - where you can see that we are by far least sensitive to blue (or rater blue cones are the least sensitive of the three).

Now mentally add rows 2 (Very light sensitive 6 (stimuli added over time), 7 (Have more pigment than cones, so can detect lower light levels), 9 (x20 more cells in the eye then other type), and you can see that even if graph almost "vanishes" for Ha - we can still see it in the night. Not as good as other wavelengths, that is for sure - but we are not effectively blind to Ha in scotopic vision.

 

So I think what you're saying @vlaiv is that we do use our scotopic vision for visual astronomy (when we are dark adapted) but that while our perception at the Hα wavelength is much less than in the centre of the visible light spectrum (peaking at about 500nm when dark adapted), it is still sensitive enough for us to see Hα?

Is that correct?

If it is, then while light at Hα may be useful (detected), we are much more likely to see faint objects that emit in the 425nm to 575nm range. Planetary Nebulae (mostly OIII) for example should be much easier to see than Emission Nebulae (mostly Hα).

 

 

[vlaiv]

5 minutes ago, PeterC65 said:

Is that correct?

Yes.

6 minutes ago, PeterC65 said:

If it is, then while light at Hα may be useful (detected), we are much more likely to see faint objects that emit in the 425nm to 575nm range. Planetary Nebulae (mostly OIII) for example should be much easier to see than Emission Nebulae (mostly Hα).

That depends.

Ha is the strongest signal in many objects. People that image in narrowband know that they must spend significant amount of time to capture OIII and other wavelengths compared to Ha in most targets (some are exception).

Similarly - it would seem that H beta is much better candidate for observing since we are far more sensitive there - but H beta emission lags behind Ha significantly (if at all present).

Check this resource for example:

https://www.prairieastronomyclub.org/filter-performance-comparisons-for-some-common-nebulae/

Interesting discussion with actual hands on experience with observing thru Ha filter:

https://www.cloudynights.com/topic/785748-visually-observing-halpha-red-nebulae/

[PeterC65]

Reading some more about this subject (here) it seems that when we are observing, our eyes are operating with both rods and cones in a mixed mode called Mesopic. In this mode there is some sensitivity at Hα, but not much.

The post on CN is interesting in that the OP makes the point that they can see very little through their Hα filter and much more through their OIII filter.

 

[PeterW]

Twilight would be mesopic, when you’re observing faint nebulae it’s scotooic and rods only. They will saturate as the light levels increase. I’d use a hydrogen beta filter as most hydrogen nebulae have emission there and the eye is much more sensitive to it. 
https://www.freunde-der-nacht.net/filterexperimente/h-beta/ (use translation) has a review of different h-beta filters, if you want to go after galactic nebulae.

 

Peter

[PeterC65]

1 hour ago, PeterW said:

Twilight would be mesopic, when you’re observing faint nebulae it’s scotopic and rods only.

The general consensus seems to be that mesopic vision occurs between a sky brightness of 5mcd/m² and 5cd/m², so it kicks in at an SQM of 18.3 or less (Bortle 8 / 9). With my Bortle 4 sky I should therefore be using only scotopic vision and will not see any Hα, or for that matter, anything from the upper passband of a UHC filter which I suppose is my concern.

If this is the case then UHC filters make no sense for visual observing and we should be using OIII and Hβ filters.

1 hour ago, PeterW said:

I’d use a hydrogen beta filter as most hydrogen nebulae have emission there and the eye is much more sensitive to it. 
https://www.freunde-der-nacht.net/filterexperimente/h-beta/ (use translation) has a review of different h-beta filters, if you want to go after galactic nebulae.

This website suggests that 6nm Hβ filters are best, even when used with binoculars. This is very interesting, but at odds with the general advice that filters this narrow are only useful with cameras or maybe visually but only with 10"+ aperture scopes.

I thought the comments on the website about using a blue filter on M45 was interesting and I plan to try this.

 

[PeterW]

Christopher Hay has done a lot of testing of these types of filters, not always from good locations as you will have read. I caught a glimpse of the California neb from my streetlight infested SQM 19skies. I did have very good stray light blocking around the eyepieces so I could well dark adapt (these filters let little light through, you’ll be looking at a reflection of your eyes if there is any straylight about) and some long dew shields to cut out the local light pollution. Not got them under better skies yet.

h-beta are your best bet for galactic nebulae and give them the largest exit pupil (brightest view) you can, two eyes giving a calmer view than one.

For reflection nebulae you want a blue CCD filter, ie one that fully transmits upto a given cutoff wavelength, not an older written type that has a Much less distinct transmission spectrum.

Peter

Edited April 17, 2023 by PeterW
Add

[PeterC65]

16 hours ago, PeterW said:

For reflection nebulae you want a blue CCD filter, ie one that fully transmits upto a given cutoff wavelength, not an older written type that has a Much less distinct transmission spectrum.

These are hard to come by individually, I'd need to purchase an LRGB set. I will give it a try with my Baader Light Blue filter for now.