Good livestacking practices with dual/tri narrowband filters

[clouzot]

Hi all,

This is a continuation of a CN post, where I showed the issues I'm currently having with dual narrowband filters, OSC, Sharpcap and livestacking.

These filters let pass little light, and the relative sensitivity of my two cameras (ASI183 and Altair 294C) causes histogram peaks to be separate, as can be seen in the following screenshot taken with an IDAS-LPS-D2 (red peak all the way to the left, followed by the blue and green peaks (here the latter are aligned, but it's not always the case, especially under moonlight). The Optolong L-Enhance has the same behavior.

SharpCap Pro (v3.2.6121) - ZWO ASI183MC (via USB3) - C__Users_admin_Desktop_SharpCap Captures 26_10_2019 00_20_05.png (401.1 KiB) Viewed 1511 times

Colour balancing is then really difficult, as I have to set an pretty high gain to red, while trying to align peaks in the mini Display Histogram on the right. The black level setting is then really sensitive as well.

While ZWOs have white balance controls (under Image Controls) that could probably be used for this purpose (assuming darks and flats are taken with the very same settings), my Altair doesn't provide them in RAW, only in RGB. I don't know if this is a known issue though.

What could be the best way to get things under control? An option to be able to set different black/mid/white levels depending on the channel would be a solution, albeit a bit overkill for real-time purposes.

[admin]

Hi,

An interesting question. You are right that the ZWO colour balance options could be used but would require you to take all your calibration frames with the same settings. As you know, these controls are not available on Altair cameras in raw mode.

On the other hand, the colour adjustment applied in the live stacking is applied before the image stretch, so it should be able to achieve equivalent results to the colour balance controls that the ZWO cameras make available. I guess the issue is just how to get the adjustment exactly correct. I wonder if the dotted line showing the adjusted histogram in each channel would help?

Cheers, Robin

[clouzot]

On the other hand, the colour adjustment applied in the live stacking is applied before the image stretch, so it should be able to achieve equivalent results to the colour balance controls that the ZWO cameras make available. I guess the issue is just how to get the adjustment exactly correct. I wonder if the dotted line showing the adjusted histogram in each channel would help?

Thanks Robin,

It's a relief to know that the signal path is as you described (ie. colour gains followed by image stretch), that makes things more workable from a mental standpoint

A Sharpcap-side postprocess would of course be better than a camera- or driver-side white balance, especially as regards darks (which one would have to do again for each balance setting. And boy, does the 294 need a lot of darks to achieve a proper master with no unwanted pattern...)

Your idea of an optional, secondary indicator showing how colour sliders affects the RVB histogram would definitely make my life easier (and hopefully help other OSC users as well), if it's not too much work. The added value would be an easier white balance, as the Mini Display Histogram is definitely a bit too narrow for the task, especially on retina-like displays, and as the auto white balance is often set off by such extreme narrowband filters.

[admin]

Hi,

Just an update on this – I tried adding the second dotted line to show where the histogram is after the adjustments and it made the whole histogram graph much too confusing and muddled to work with. As a second attempt, I went back to having a single line but I made the histogram in the live stack reflect colour adjustments made using the live stacking colour controls and that seems to work quite nicely, so watch out for that in the next update.

Cheers, Robin

[clouzot]

Thanks Robin, you rock! Looking forward to the update

(the histogram is already pretty crowded, so I totally understand that showing both pre- and post-balance curves would be way too confusing. It could also be an explicit option to either show the post-balance histogram)