I thought I would start a thread about this as it can be a difficult area in processing.
I know of a few ways to incorporate Ha into an LRGB image.
Talking in Photoshop terms. One way is to copy the Ha (set it to RGB mode from grayscale first) into 2 new layers on top of your LRGB image. Set them to lighten mode. It assumes the Ha is aligned already with the LRGB and you have not cropped the LRGB so the images are still aligned.
If they are not 100% aligned you can still do it in Photoshop. Just set the opacity slider to 50% then use the move tool and move it around clicking the eye on and off to flick between with Ha and without so you can see if there is still any movement between the Ha and LRGB until they are exactly the same.
Now the first Ha layer will be red. You set the blend mode to lighten. Then delete the green and blue channel. Click on channels, click on green,
hold down the shift key, click on blue then control A then delete. Click on the RGB and now that layer is red. Use opacity to suit (disable the top Ha layer so you can see the colour effect - click off the eye icon). Now the tp layer do the same but this one will be blue. So delete red and green as above. Now usually you set the opacity to only about 15% to represent the HBeta band which is bluish.
Now you can use curves on these lighten modes to adjust more or less as lighten mode lets lighter pixels from lower layers through.
You can add some Ha on top and set it to luminosity but this eats away colour. You can curve that Ha layer down first so its not so overwhelming.
Adding Ha to luminance though will damage subtle blue reflection nebula in your image but sometimes it enhances.
The other method is to combine Ha and red together and add Ha into the luminance layer. I have not really used this method but plan to experiment with it to see if it holds any promise. It seems to lose control of the Ha and your only control would be the % blend of the Ha into the red at that stage.
The other method is the Astrodon method of creating hue/saturation layers and clipping masks (clipping means to limit the effect of a layer to only the layer beneath it and not layers further below it).
There are video tutorials on his site. Just remember to change the mode of the narrowband images to RGB first for that to work.
I have always done what Russ showed in the link - usually the FIRST variant of the procedure - mixing the red channel into the H-a in varying %, then reinserting into the red channel, as well as a layer. I find it works quite well.
It's also the work-flow I intend using with my OSC H-a captures when blending the H-a into the OSC colour.
funnily enough the first crack I tried it on was the running chicken also. my data/conditions/capture was terrible so rather than just delete it - I used it for trialled blending methods (I think this was also my pre RegiStar days so aligned by eye/warp tool).
There are numerous ways of dealing with this and everyone has their own ideas, workflow, etc.
My preference is to use Pixel Math in MaximDL keeping the data in a numerical FITS format until I obtain the aesthetics I'm looking for. Only then will I covert to TIFF for PS work. Sometimes, I'll also import the FITS in to PI for DBE and other stuff before bring it back into CCDStack. FITS is interchangable but you'll need to multiply the PI FITS using Pixel Math to get the data visible in CCDstack again. Anyway, I digress.
In MaximDL, you can blend a percentage of the;
Ha master to red master (50-70% or to taste) to create a Ha+R master
Ha master to luminance master (50-70% or to taste) to create a Ha+L mater
Ha master to blue master (10-15% or to taste) to create a Ha+B master
Once you've got your new masters (Ha+R, Ha+L and Ha+B), you can continue to use your usual routine to create an RGB master with the Ha introduced fits and colour balance them. Rarely will I do a LRGB combine in tools like MaximDL or CCStack, only RGB. Adding the luminance in PS will be provide far greater control.
I've found working with numbers i.e. in FITS format, measuring background levels etc, will give you a more consistent HA blending result. Probably not important for a single image but when conducting a [Ha+R]GB multi panel mosaic, you soon appreciate this.