I’m soon venturing into unknown territory of Mono imaging with a filter wheel ( previously OSC imaging for the past 3 to 4 years and prior to that DSLR for 3 years )
For example, if I capture a Ha , Oiii and Sii data set and want to process a typical SHO image what is the most appropriate compositing Luminance , Color option should I choose in Compose ?
Same question for an SHO or HOO bi Color image ?
(Previously using my OSC and Dualband data set was easy, I always used -
L + Synthetic L from R ( 2 x G ) B, R ( GB ) ( GB ) for OSC / DSLR data sets to create Bi Color images).
If I capture an LRGB data set with my Mono camera I’m assuming I can use the following compositing Luminance , Color schemes -
RGB Mono ( for Mono RGB data sets where output result will be Mono )
L RGB ( for Mono RGB data sets where output result will be colour )
L + Synthetic L from RGB , RGB ( for Mono data sets where a Synthetic Luminance is derived from RGB channels and added to Luminance channel, output result will be colour )
L + Synthetic L from RGB , Mono ( for Mono data sets where a Synthetic Luminance is derived from RGB channels and added to Luminance channel, output result will be Mono ) Also known as LLRGB
L + Synthetic L from RGB , R ( GB) ( GB ) Bi Colour ( for Mono data sets to create Bi Colour images )
Appreciate advice regarding most appropriate compositing schemes in Compose for both Mono broadband and Mono Narrowband Data sets
NB: I’m currently using ST v1.8 ( not v1.9 )
Thanks
Martin
Compose, Mono Data Set (Ha , Oiii , Sii )
-
Mike in Rancho
- Posts: 1166
- Joined: Sun Jun 20, 2021 10:05 pm
- Location: Alta Loma, CA
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Just mash buttons, Martin, all will work out. 
But really though, just a matter of keeping in mind two main things. The left side options for luminance, color dictate the framework of the composition. Anything that ends in "mono" is really a specialty situation and you are unlikely to need that now. Or maybe ever.
And then the right side is the exposure weighting and only applies to the synth L (or L + synth L as if you have a separate file in the L slot, you can work that into the relative weighting as well).
Routine SHO would load into R, G, and B and you would keep the default setting of L + Synth L from RGB (no 2xG). In 1.9 you can further choose identifiers which will get you to purple cap (rather than green cap) in Color. In 1.8 if you want to purple cap a SHO, you'd have to do an inverted green cap.
For HOO, you would choose the non-2xG bicolor for R(GB)(GB), and, zero out one of the G or B exposure times so that the GB (OIII) doesn't get double-weighted.
For broadband, if you captured L and R, G, B, standard is again the default L+Synth L option (and of course color). This is aka LLRGB, not the one ending in mono. Again, anything ending in mono gets you a grayscale (no color!), not generally used unless you like grayscale, or are getting creative in some kind of multi-step process.
If you don't like the SNR of your discrete RGB filters and things aren't blending well, you can forego synth and try L, RGB, for old school LRGB (not LLRGB).
The same goes for narrowband, you can alter the weights if you have to. For example last night (mostly as a guiding test after a little mount tuning) I captured about 30 minutes each of S, H, and O on the Elephant Trunk. Not much integration. It's all faint, but the OIII and SII are especially faint. So I felt that the overall composition was too noisy, and not properly giving the far better SNR of the Ha its due, if the channels were weighted about equally. So for a quick test I just doubled the Ha exposure time (the times are all relative btw). That processed better. Still noisy of course, but the Synth L was just cleaner that way.
You'll get the hang of it and how this all works for your discrete filters in no time.
But really though, just a matter of keeping in mind two main things. The left side options for luminance, color dictate the framework of the composition. Anything that ends in "mono" is really a specialty situation and you are unlikely to need that now. Or maybe ever.
And then the right side is the exposure weighting and only applies to the synth L (or L + synth L as if you have a separate file in the L slot, you can work that into the relative weighting as well).
Routine SHO would load into R, G, and B and you would keep the default setting of L + Synth L from RGB (no 2xG). In 1.9 you can further choose identifiers which will get you to purple cap (rather than green cap) in Color. In 1.8 if you want to purple cap a SHO, you'd have to do an inverted green cap.
For HOO, you would choose the non-2xG bicolor for R(GB)(GB), and, zero out one of the G or B exposure times so that the GB (OIII) doesn't get double-weighted.
For broadband, if you captured L and R, G, B, standard is again the default L+Synth L option (and of course color). This is aka LLRGB, not the one ending in mono. Again, anything ending in mono gets you a grayscale (no color!), not generally used unless you like grayscale, or are getting creative in some kind of multi-step process.
If you don't like the SNR of your discrete RGB filters and things aren't blending well, you can forego synth and try L, RGB, for old school LRGB (not LLRGB).
The same goes for narrowband, you can alter the weights if you have to. For example last night (mostly as a guiding test after a little mount tuning) I captured about 30 minutes each of S, H, and O on the Elephant Trunk. Not much integration. It's all faint, but the OIII and SII are especially faint. So I felt that the overall composition was too noisy, and not properly giving the far better SNR of the Ha its due, if the channels were weighted about equally. So for a quick test I just doubled the Ha exposure time (the times are all relative btw). That processed better. Still noisy of course, but the Synth L was just cleaner that way.
You'll get the hang of it and how this all works for your discrete filters in no time.
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Thanks Mike,
Much appreciated
Looks too simple
I did read Guy’s user notes for Compose which made some sense but still can’t quite get my head around the weightings for individual channels. With my OSC for Bi Color narrowband Ivo advised to leave them default or all the same ( 1hr 3600s ) which did make sense as you’re loading the same data set in all three channels.
You example about the weightings when you bring noise into the equation does make sense. I always assumed the weightings was more to do with Color which is wrong I guess , it’s more to do with signal and fidelity.
In Guy’s user notes ( Special Techniques Hubble Palette ) he mentions about an option of Luminance and Narrowband. Why would you capture Luminance and Ha , Oiii and Sii and process that option ? Under my Bortle 8 the luminance file would be a white out under say a full moon ???
It’s probably a good idea to stick to the basics , strictly broadband and strictly narrowband for now
Cheers
Martin
Much appreciated
Looks too simple
I did read Guy’s user notes for Compose which made some sense but still can’t quite get my head around the weightings for individual channels. With my OSC for Bi Color narrowband Ivo advised to leave them default or all the same ( 1hr 3600s ) which did make sense as you’re loading the same data set in all three channels.
You example about the weightings when you bring noise into the equation does make sense. I always assumed the weightings was more to do with Color which is wrong I guess , it’s more to do with signal and fidelity.
In Guy’s user notes ( Special Techniques Hubble Palette ) he mentions about an option of Luminance and Narrowband. Why would you capture Luminance and Ha , Oiii and Sii and process that option ? Under my Bortle 8 the luminance file would be a white out under say a full moon ???
It’s probably a good idea to stick to the basics , strictly broadband and strictly narrowband for now
Cheers
Martin
-
Mike in Rancho
- Posts: 1166
- Joined: Sun Jun 20, 2021 10:05 pm
- Location: Alta Loma, CA
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Hi Martin,
The User Notes for Compose is pretty long. I've read through it, but am not sure exactly which subsection you are referring to. If you could pinpoint it, I'll have a go at seeing what he is trying to describe there.
Luminance for SHO, in the way we normally think of it, generally isn't a thing. You just capture your discrete NB filters and create a full synth L. Theoretically, if one had a tri or duo-band filter which matched the discrete filters well, you could pop that on the mono cam and more efficiently capture a narrowband "luminance," which you would then compose in the same way you would LRGB. I did something like that once, before I had all my filters for a complete EFW. I captured L-eNhance on my D5300 for an HOO amenable target, then put the L-eNhance on my 2600MM, and composed them together. Not quite the same, but the idea holds up.
Of course there are also old-school methods (before NB Accent) of combining broadband and narrowband, such as - famously or infamously - HaRGB, where one would use Ha as the luminance, or pre-combine Ha into red.
Yes exposure weightings are all for the Synth L. In fact if you choose a composition that doesn't use Synth L, you will see all the exposure sliders become grayed out as they are not used. You are not weighting the color with those settings, just the synth L that you will work for detail -- though I suppose in an indirect way, luminance does affect color when the final blending happens in Color.
Color itself is stashed away for use later. I don't know the nuts and bolts of what is going on mathematically, but I'm figuring it's somehow a saved state of all pixels levels, only altered as needed by the Wipe module, and then by linear balancing in the eventual color module. Of course that is exactly what one wants, or pretty close, for broadband. For narrowband, that starts getting close to what Stefan and I were discussing recently. It permits the relative gas emissions throttling, but otherwise, do discrete NB filters really need to be glued to one another that way, and more so for here, are exposure times really a valid measure of SNR for each such filter channel?
That may not be the case, as 2h of Ha might be a whole lot cleaner than 2h of OIII. Thus, you can probably re-weight those exposures manually based on your evaluation of the data quality. Color will still be the saved-off linear levels, relatively balanced, and with hues as one may have chosen in the matrix mappings, to colorize what one has created using the synth L.
The User Notes for Compose is pretty long. I've read through it, but am not sure exactly which subsection you are referring to. If you could pinpoint it, I'll have a go at seeing what he is trying to describe there.
Luminance for SHO, in the way we normally think of it, generally isn't a thing. You just capture your discrete NB filters and create a full synth L. Theoretically, if one had a tri or duo-band filter which matched the discrete filters well, you could pop that on the mono cam and more efficiently capture a narrowband "luminance," which you would then compose in the same way you would LRGB. I did something like that once, before I had all my filters for a complete EFW. I captured L-eNhance on my D5300 for an HOO amenable target, then put the L-eNhance on my 2600MM, and composed them together. Not quite the same, but the idea holds up.
Of course there are also old-school methods (before NB Accent) of combining broadband and narrowband, such as - famously or infamously - HaRGB, where one would use Ha as the luminance, or pre-combine Ha into red.
Yes exposure weightings are all for the Synth L. In fact if you choose a composition that doesn't use Synth L, you will see all the exposure sliders become grayed out as they are not used. You are not weighting the color with those settings, just the synth L that you will work for detail -- though I suppose in an indirect way, luminance does affect color when the final blending happens in Color.
Color itself is stashed away for use later. I don't know the nuts and bolts of what is going on mathematically, but I'm figuring it's somehow a saved state of all pixels levels, only altered as needed by the Wipe module, and then by linear balancing in the eventual color module. Of course that is exactly what one wants, or pretty close, for broadband. For narrowband, that starts getting close to what Stefan and I were discussing recently. It permits the relative gas emissions throttling, but otherwise, do discrete NB filters really need to be glued to one another that way, and more so for here, are exposure times really a valid measure of SNR for each such filter channel?
That may not be the case, as 2h of Ha might be a whole lot cleaner than 2h of OIII. Thus, you can probably re-weight those exposures manually based on your evaluation of the data quality. Color will still be the saved-off linear levels, relatively balanced, and with hues as one may have chosen in the matrix mappings, to colorize what one has created using the synth L.
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Thanks Mike,
Here’s the part where Guy talks about Luminance and NB data
“Loading Narrow Band data using the Hubble Palette
Mapping narrowband data onto RGB channels like this does not produce a documentary result.
Load the Luminance data - if you have it - into the Luminance channel.
Load the SII data into the Red channel.
Load the Ha data into the Green channel.
Load the OIII data into the Blue channel.
If you want to create a weighted synthetic luminance channel set the 'Luminance, Color' to 'L + Synthetic L from RGB, RGB'
Otherwise if you have Luminance data - but don't want to use synthetic luminance (LLRGB) - set the 'Luminance, Color' to 'L, RGB'
Set the Total exposure times for each channel.
Press 'Keep' when done.”
Reading through your notes makes more sense now
I’m just waiting for my TS GPU coma corrector to arrive from Europe ( about 2 weeks away ) then I can build the new Mono set up ( I have my 2600MM camera , ZWO 7x2 EFW and Antlia LRGB and 3nm NB filters )
Thanks again
Martin
Here’s the part where Guy talks about Luminance and NB data
“Loading Narrow Band data using the Hubble Palette
Mapping narrowband data onto RGB channels like this does not produce a documentary result.
Load the Luminance data - if you have it - into the Luminance channel.
Load the SII data into the Red channel.
Load the Ha data into the Green channel.
Load the OIII data into the Blue channel.
If you want to create a weighted synthetic luminance channel set the 'Luminance, Color' to 'L + Synthetic L from RGB, RGB'
Otherwise if you have Luminance data - but don't want to use synthetic luminance (LLRGB) - set the 'Luminance, Color' to 'L, RGB'
Set the Total exposure times for each channel.
Press 'Keep' when done.”
Reading through your notes makes more sense now
I’m just waiting for my TS GPU coma corrector to arrive from Europe ( about 2 weeks away ) then I can build the new Mono set up ( I have my 2600MM camera , ZWO 7x2 EFW and Antlia LRGB and 3nm NB filters )
Thanks again
Martin
-
Mike in Rancho
- Posts: 1166
- Joined: Sun Jun 20, 2021 10:05 pm
- Location: Alta Loma, CA
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
That should be great, Martin! Do you already have the EFW bolted to the 2600MM?
At first I actually used an all-threaded connection, CC on the one side, camera on the other, still with the so-called 5mm ZWO tilt plate on plus another separately-bought adapter. Had to make sure that adapter had a sufficient flange though!
Later I removed the tilt plate, to get the filters 5mm closer to the sensor, and bolted it. Now at the time I got it my 7x2 EFW did not come with those darkening gaskets (they started including them a few months later), so I did order those separately just in case. But (knock on wood) my as-bolted connection has not had any light leaks and I have no uncovered holes either.
On the User Notes. I guess I just glossed over all that before.
I can't say that I wouldn't re-write that section to reduce confusion.
To start, if doing narrowband I think it is documentary enough -- for what it is and what it is understood is being displayed. Even more so using ST, which holds one to a relative balancing in Color for the various emissions. But even without that I think it could still be deemed documentary as narrowband, as long as one doesn't start playing artsy tricks.
The procedure is indeed how ST works in all cases, but as already discussed I don't think there's a need to mention the Luminance slot here for any sort of typical use case. Heck, query if there even needs to be Hubble as a special technique here, unless one also points out the SHO matrix that comes closest to what is generally seen in HST (since straight-up SHO often looks quite a bit different). Also, the SII, Ha, and OIII is already right there on the file slot buttons.
Unless I'm missing something, I just can't think of a good use for the L slot in an all-narrowband workflow. And any kind of actual Luminance filter usage would just dilute everything.
For Hubble I would just load tricolor right into the indicated slots, set the exposures (in most cases this is now automatically filled in when using 1.9, which is oh so handy), adjust those exposures if they don't seem to result in the overall synth L SNR that you like, and then once in Color what comes up by default (matrix 1, purple cap, artistic, and RGB ratio CIE retention) gets quite close to the HST palette appearance.
The User Notes do link to a video in that subsection, on Pillars of Creation, which indeed is a very good video. Watch that one for sure, and if you haven't already, download the actual Hubble data and process side-by-side along with the video. Great practice for then doing your own SHO later.
But below that it links to a forum thread that is titled "LRGB & Hubble Palette" but I actually couldn't find any Hubble Palette discussed in there at all.
At first I actually used an all-threaded connection, CC on the one side, camera on the other, still with the so-called 5mm ZWO tilt plate on plus another separately-bought adapter. Had to make sure that adapter had a sufficient flange though!
Later I removed the tilt plate, to get the filters 5mm closer to the sensor, and bolted it. Now at the time I got it my 7x2 EFW did not come with those darkening gaskets (they started including them a few months later), so I did order those separately just in case. But (knock on wood) my as-bolted connection has not had any light leaks and I have no uncovered holes either.
On the User Notes. I guess I just glossed over all that before.
I can't say that I wouldn't re-write that section to reduce confusion.
To start, if doing narrowband I think it is documentary enough -- for what it is and what it is understood is being displayed. Even more so using ST, which holds one to a relative balancing in Color for the various emissions. But even without that I think it could still be deemed documentary as narrowband, as long as one doesn't start playing artsy tricks.
The procedure is indeed how ST works in all cases, but as already discussed I don't think there's a need to mention the Luminance slot here for any sort of typical use case. Heck, query if there even needs to be Hubble as a special technique here, unless one also points out the SHO matrix that comes closest to what is generally seen in HST (since straight-up SHO often looks quite a bit different). Also, the SII, Ha, and OIII is already right there on the file slot buttons.
Unless I'm missing something, I just can't think of a good use for the L slot in an all-narrowband workflow. And any kind of actual Luminance filter usage would just dilute everything.
For Hubble I would just load tricolor right into the indicated slots, set the exposures (in most cases this is now automatically filled in when using 1.9, which is oh so handy), adjust those exposures if they don't seem to result in the overall synth L SNR that you like, and then once in Color what comes up by default (matrix 1, purple cap, artistic, and RGB ratio CIE retention) gets quite close to the HST palette appearance.
The User Notes do link to a video in that subsection, on Pillars of Creation, which indeed is a very good video. Watch that one for sure, and if you haven't already, download the actual Hubble data and process side-by-side along with the video. Great practice for then doing your own SHO later.
But below that it links to a forum thread that is titled "LRGB & Hubble Palette" but I actually couldn't find any Hubble Palette discussed in there at all.
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Thanks Mike,
My Mono image train will be 2600MM with 5mm tilt plate > ZWO 7x 2” EFW 20mm > 2mm adapter flange > 16.5mm adapter > TS GPU CC = 55mm to 56mm BF
I won’t be using Lum with my narrowband data sets that’s for sure
Thanks for all your comments , experiences and recommendations
Clear Skies
Martin
My Mono image train will be 2600MM with 5mm tilt plate > ZWO 7x 2” EFW 20mm > 2mm adapter flange > 16.5mm adapter > TS GPU CC = 55mm to 56mm BF
I won’t be using Lum with my narrowband data sets that’s for sure
Thanks for all your comments , experiences and recommendations
Clear Skies
Martin
-
Mike in Rancho
- Posts: 1166
- Joined: Sun Jun 20, 2021 10:05 pm
- Location: Alta Loma, CA
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Hi Martin,
Yep, that's the same boat I was in, which had me ordering various new little parts.
But it'll be better once you have everything in front of you. Perhaps what comes with your kits will be sufficient.
In my case, an all-threaded solution required extra stuff. Three possible snags in your described optical train -- the ZWO 16.5 has a 42mm side, and I thought that was getting possibly a little tight that far out from the sensor in the light cone. And I don't see a connection from 2600 to EFW in your train. Finally, whatever BF you end up with can't be shortened if needed.
What I had to do was order a (Blue Fireball?) second M54 adapter, this one to M42, and make sure it has a big enough flange. That adds 2mm, so the 16.5 has to go, to be replaced with another adapter. And for that I got all 48-48 both ends. I forget what length I used for that when I was all-threaded, but I could probably find it. I know I went a little short, in order to use micro adapters. I did get sets of both 42 and 48 micro spacer rings - the 3D printed black ones from Agena. But you can get them in aluminum too by other brands.
Then when I went to bolted, that removed the need for the M45 to M42 flanged adapter, but I had to get new M48 extensions again so I could end up just short and then use micro spacers.
With what came along with my camera and EFW, they would not (reliably) connect between the EFW and the 5mm tilt plate. Oh, maybe you could get it to wedge tight, but again not reliably in my opinion. Elsewise, you'd just keep threading in and could end up protruding into the EFW, impairing it.
Possibly moving the M45 flanged adapter to the other side, with an included ring adapter on the CC side, and then adjusting down the flange that comes with the TS GPU, might do it. Again if it would stay tight. Maybe. And then check your measurements.
Let us know what you come up with!
Also if you are getting the 1.0 CC (I think the TS is just a rebranded Sharpstar), it has quite a long nose, I think about 25mm longer than my 0.95x.
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Mike,
I checked with ZWO and my supplier , I can bolt or screw my EFW directly to the 2600MM tilt plate, 4 screw holes are provided in the EFW which line up with 4 x M2.5 threaded holes in the tilt plate of the 2600MM. A darkening gasket is placed between tilt plate and EFW. ZWO provide M2.5 countersunk fixing screws
Hopefully there will be no micro flex in the EFW or for that matter the whole assembly.
I’m using a M48 male to M48 female 16.5mm extender and M54 to M48 2mm adapter between EFW and TS GPU CC which has an M48 male thread. That achieves my 55 to 56mm BF
I’m sure I’ll get a little bit of Vignetting in the corners but I can live with that ( flats are routine anyway )
Cheers
Martin
I checked with ZWO and my supplier , I can bolt or screw my EFW directly to the 2600MM tilt plate, 4 screw holes are provided in the EFW which line up with 4 x M2.5 threaded holes in the tilt plate of the 2600MM. A darkening gasket is placed between tilt plate and EFW. ZWO provide M2.5 countersunk fixing screws
Hopefully there will be no micro flex in the EFW or for that matter the whole assembly.
I’m using a M48 male to M48 female 16.5mm extender and M54 to M48 2mm adapter between EFW and TS GPU CC which has an M48 male thread. That achieves my 55 to 56mm BF
I’m sure I’ll get a little bit of Vignetting in the corners but I can live with that ( flats are routine anyway )
Cheers
Martin
-
Mike in Rancho
- Posts: 1166
- Joined: Sun Jun 20, 2021 10:05 pm
- Location: Alta Loma, CA
Re: Compose, Mono Data Set (Ha , Oiii , Sii )
Hey Martin,
Awesome, so you will be bolted up. That must be the inner set of 4 holes, I forgot about that.
I think it all holds together quite well, it sure doesn't feel like it would be susceptible to flexing. Now, the weight of the whole contraption - the 7x2 + 2600 is big - if the camera is out to the side, pulling down on the focuser and maybe even torquing the OTA - that seems more likely but not a lot I can do there. If you are on your carbon scope, things ought to be better in that regard.
Awesome, so you will be bolted up. That must be the inner set of 4 holes, I forgot about that.
I think it all holds together quite well, it sure doesn't feel like it would be susceptible to flexing. Now, the weight of the whole contraption - the 7x2 + 2600 is big - if the camera is out to the side, pulling down on the focuser and maybe even torquing the OTA - that seems more likely but not a lot I can do there. If you are on your carbon scope, things ought to be better in that regard.