Back to forums

Ha/O3 and S2/O3 data acquisition.... kind of confused.. I think

Acquisition Duo-Narrowband Deep Sky
14 replies•427 views
peng155 avatar
peng155 avatar
Hi Folks,

So I'm diddling around with something new… after getting my feet wet with the ZWO duo-band filter… I thought I'd try something new for a change of pace…

So I've been collecting data with the Askar C1/C2 filter set… my current plan is to collect the same amount of data for each filter.. something like 20hrs each.. O3 is in the blue-green spectrum and Ha in Red, and S2 in Deep Red

20 or so hrs in each filter… Simple logic says I have 20hr of Ha/O3 and 20hrs of S2/O3… does that mean my final processed image will be 40hrs of integration?? When I see before and after images, of pictures that had additional data integrated into them… more data will result in a better detailed pictures…. Am I thinking correctly on the additive integration time of 20hr + 20hrs = 40hr… or is it 20hr +20hrs = 20hrs, since I am integrating Ha/O3 and S2/O3 seperately..

I diidn't know if this should have gone into the DSO processing sub-group, but since I'm asking about acqusition, here we are… And yeah.. I've got a bunch of questions about processing too, and that another can of worms I'll need to figure out

Thanks
ScottF avatar
ScottF avatar
I would assume that you have 20hrs of HA, 20 hours of S2, and 40 hours of OIII.
Miguel T. avatar
Miguel T. avatar
It's 20hours of the Ha/Oiii filter and 20hours of the Sii/Oiii filter plain and simple. Now keep in mind that you're using a color camera where for each 4 pixels, 1 is dedicated to blue, 2 to green and 1 to red. The Oiii fall somewhere between blue and green. Both the blue and green pixels are receiving a little bit of Oiii but the red ones are doing absolutely nothing about the Oiii. You have only 1 out of 4 pixels receiving signal for the Ha and Sii. Calculate that as you want but the true signal ratio between your filters is not as simple as assuming full sensor coverage.
Helpful Insightful Concise
peng155 avatar
peng155 avatar
I would assume that you have 20hrs of HA, 20 hours of S2, and 40 hours of OIII.

Yes, that would be my assumption
Jan Erik Vallestad avatar
Jan Erik Vallestad avatar
Think of it this way, if you expose a color camera for 20 hours you don't have 20 hours each of R, G and B. It's probably more like 2/4 G and 1/4 each of R and B. However it would still be 20 hours of total exposure.

Either way if you expose both filters for 20 hours each, you will have 40 hours of data and thus 40 hours of total integration. You cannot get 60 hours integration by imaging for 40 hours, obviously. When you enter the data in AB for your image you do so per filter, not per color channel.
Helpful Concise
peng155 avatar
peng155 avatar
Yes, You guy's are right.. I forgot about the OSC bayer filter being RGGB… at this point it might be easier to just speed up my plans of getting a Mono camera with seperate Ha, O3 and S2 filters for extended projects
ScottF avatar
ScottF avatar
Think of it this way, if you expose a color camera for 20 hours you don't have 20 hours each of R, G and B. It's probably more like 2/4 G and 1/4 each of R and B. However it would still be 20 hours of total exposure.

Either way if you expose both filters for 20 hours each, you will have 40 hours of data and thus 40 hours of total integration. You cannot get 60 hours integration by imaging for 40 hours, obviously. When you enter the data in AB for your image you do so per filter, not per color channel.

You're equating a colour to a mono camera; what you say is true about colour sensors though. However, If I use a HOO filter for 20 hours, then I have 20 hours of Ha, and 20 hours of OIII from that run. It is not equivalent to a mono camera, of course. Now, if you expose 20 hours for the Su/OO filter, that run will give you 20 hours of Su and 20 hours of OIII. All told, you have done 40 hours of Integration, with 40 of those hours exposing for OIII and 20 for Su and Ha.  The 40 hours of OIII is not equivalent to 40 hours of OIII on a mono, but the author is not asking that question.
Helpful Insightful Concise
Jan Erik Vallestad avatar
Jan Erik Vallestad avatar
You're equating a colour to a mono camera; what you say is true about colour sensors though. However, If I use a HOO filter for 20 hours, then I have 20 hours of Ha, and 20 hours of OIII from that run. It is not equivalent to a mono camera, of course. Now, if you expose 20 hours for the Su/OO filter, that run will give you 20 hours of Su and 20 hours of OIII. All told, you have done 40 hours of Integration, with 40 of those hours exposing for OIII and 20 for Su and Ha.  The 40 hours of OIII is not equivalent to 40 hours of OIII on a mono, but the author is not asking that question.



OP is using a color camera with dual NB filters, why would anyone use a dual narrowband filter on a mono camera? If you shoot a dual NB filter for 20 hours, you have 20 hours. You don't magically get 40 hours simply because the filter is a dual band one. Those 20 hours would be separated into R/G/B signal on your OSC sensor which would equate to 20 hours alltogether - not 20 hours of Ha and 20 hours of Oiii. The OSC sensor simply isn't capable of collecting each bandwidth equally a the same time. For that you would need a mono camera and separate filters. You clearly cannot get more integration time than the actual exposure time you have, so if you add Oiii/Sii for another 20 hours you get 20+20=40 in total. Out of those 40 hours 3/4 would be G/B signal and the rest would be R. And that's not close to being equal to mono, which is correct - but I did not bring mono into the discussion. 


Again the sensor is not capable of getting 100% Oiii and 100% Sii/Ha at the same time.
Norman Hey avatar
Norman Hey avatar
Miguel T.:
It's 20hours of the Ha/Oiii filter and 20hours of the Sii/Oiii filter plain and simple. Now keep in mind that you're using a color camera where for each 4 pixels, 1 is dedicated to blue, 2 to green and 1 to red. The Oiii fall somewhere between blue and green. Both the blue and green pixels are receiving a little bit of Oiii but the red ones are doing absolutely nothing about the Oiii. You have only 1 out of 4 pixels receiving signal for the Ha and Sii. Calculate that as you want but the true signal ratio between your filters is not as simple as assuming full sensor coverage.

*** Type our reply here ***In reality, all OSC sensors have overlap across the visual spectrum in each channel, so any filter you use will show some signal in each channel. Use an 3nm Ha filter  with an OSC sensor and you will get some signal in the B and G channels. Use an Oiii filter and you will get some signal in the R channel. It's a relatively small amount, but it does "muddy" the image from each channel more than using a mono sensor. Doesn't matter if it's a single bandpass filter or dual band--OSC cameras will have some sgnal spill over by their very nature. So yes, the red pixels don't add to the Oiii signal you wnat, but they do add signal--it's just that what they add is from Ha!
Helpful Insightful Concise
ScottF avatar
ScottF avatar
You're equating a colour to a mono camera; what you say is true about colour sensors though. However, If I use a HOO filter for 20 hours, then I have 20 hours of Ha, and 20 hours of OIII from that run. It is not equivalent to a mono camera, of course. Now, if you expose 20 hours for the Su/OO filter, that run will give you 20 hours of Su and 20 hours of OIII. All told, you have done 40 hours of Integration, with 40 of those hours exposing for OIII and 20 for Su and Ha.  The 40 hours of OIII is not equivalent to 40 hours of OIII on a mono, but the author is not asking that question.



OP is using a color camera with dual NB filters, why would anyone use a dual narrowband filter on a mono camera? If you shoot a dual NB filter for 20 hours, you have 20 hours. You don't magically get 40 hours simply because the filter is a dual band one. Those 20 hours would be separated into R/G/B signal on your OSC sensor which would equate to 20 hours alltogether - not 20 hours of Ha and 20 hours of Oiii. The OSC sensor simply isn't capable of collecting each bandwidth equally a the same time. For that you would need a mono camera and separate filters. You clearly cannot get more integration time than the actual exposure time you have, so if you add Oiii/Sii for another 20 hours you get 20+20=40 in total. Out of those 40 hours 3/4 would be G/B signal and the rest would be R. And that's not close to being equal to mono, which is correct - but I did not bring mono into the discussion. 


Again the sensor is not capable of getting 100% Oiii and 100% Sii/Ha at the same time.

I agree with what you said, but unless I read the OP wrong, he exposed a HOO filter for 20 hours, then a SOO filter for 20 hours. Total exposure time is 40 hours, correct?  He has 20 hours of Ha, and 20 hours of OII, and since the entire 40 hours of exposure has an OIII component, he has 40 hours of OIII does he not. And yes, I realize that he is not getting 20 hours of our Ha, OIII or SII because he is using a OSC. Maybe I'm caught up in splitting hairs , but that's how I interpret it.
Jan Erik Vallestad avatar
Jan Erik Vallestad avatar
I agree with what you said, but unless I read the OP wrong, he exposed a HOO filter for 20 hours, then a SOO filter for 20 hours. Total exposure time is 40 hours, correct?  He has 20 hours of Ha, and 20 hours of OII, and since the entire 40 hours of exposure has an OIII component, he has 40 hours of OIII does he not. And yes, I realize that he is not getting 20 hours of our Ha, OIII or SII because he is using a OSC. Maybe I'm caught up in splitting hairs , but that's how I interpret it.


We may be "talking around" each other, but yeah he has 40 hours of total integration. To split these into how much is Oiii, Ha and Sii would probably not be that straight forward, nor very useful perhaps as you simply add the two filters to the data section of the image, but for the sake of it lets use the RGGB pattern as the basis for this and disregard the overlap and filter specs for simplicity.

3/4 of the image would then be recording G/B signal or Oiii and block the rest, the last 1/4 would be R or Ha/Sii signal. So if you think about it he can't have 40 hours of Oiii signal, given that he only imaged for 40 hours - unless he was shooting pure Oiii with a mono camera for the duration of this time. Using this simplified equation only 75% of the sensor is collecting Oiii signal during those 40 hours.

With each of the used filters the Oii signal was only recorded on a number of pixels, not all of them, which is the key here. Thus, IMO, it will be wrong to claim that the OSC camera with a dual band filter could collect 100% Oiii along with the extra 25% Ha or Sii at the same time. We have to split those 40 hours into what was actually recorded during that period of time.

As mentioned I don't think this would matter in practice since you would just add those filter to your session in the data tab on your image, then add the number of frames and their duration.
Helpful Insightful
andrea tasselli avatar
andrea tasselli avatar
You are not integrating across the whole sensor, just across each pixel, so you forgo full spatial sampling in favor of a reduced (to a degree) sampling but gain in temporal sampling. This is how a OSC operates. So, yes, 20 hours of Ha/OIII + 20 hours of SII/OIII equal to 40 hours of OIII and 20 hours of Ha and SII respectively.
Insightful Respectful Concise
Jan Erik Vallestad avatar
Jan Erik Vallestad avatar
That is a fair point that I hadn't even considered! 

I stand corrected.
Well Written Respectful
peng155 avatar
peng155 avatar
andrea tasselli:
You are not integrating across the whole sensor, just across each pixel, so you forgo full spatial sampling in favor of a reduced (to a degree) sampling but gain in temporal sampling. This is how a OSC operates. So, yes, 20 hours of Ha/OIII + 20 hours of SII/OIII equal to 40 hours of OIII and 20 hours of Ha and SII respectively.

Thanks @andrea tasselli , I hadn't even thought about it this way.... At the time I posted my question, I just took a simple view of the my total frames... and hadn't given the proper thought that each H/O/S wavelength...  Looking back at what I posted and your explanation make sense to me.
D. Jung avatar
D. Jung avatar
There is a good article on dpreview discussing the intricacies of resolution loss due to the Bayer array 

https://m.dpreview.com/articles/3560214217/resolution-aliasing-and-light-loss-why-we-love-bryce-bayers-baby-anyway
Well Written