Keep in mind that you don’t need to limit yourself to a single particle system, or a single emitter. Mix 'em up! Use different halo specs for different visual effects, and different particle physics for different 'layers" of stellar atmosphere. Maybe combine that with nested spheres of various materials as well, maybe even some Compositor effects – always a good way to fuzz up things that are intrinsically too sharp.
Keep in mind that sun corona / flair images you are trying to depict are taken by an instrument at particular wavelength. This is not what you or I “can see”.
Corona / flair event is more like electrical magnetic discharge and not like particle flying off like fire. How real do you want to make it?
Any way here is the all texture model I built for your reference:
Attachments
sun.blend (427 KB)
This may not be realistic, but I did an animation of solar flare-type activity to go along with the previous sun image. Basically, it’s particles with halo texture set to random emisson across the icosphere. I kept the halo fairly large for a blurred effect.
(For everyone’s information) I’m not concerned much with realism. I’m doing this for an abstract film, and I’m trying to achieve an h-alpha-esque depiction of a red giant whilst corncerning myself more with aesthetic appeal than physical accuracy. I really haven’t taken the time to study the science of stars’ visible spectrums etc. But my latest render gives you a good idea of what I’m shooting for:
I’m actually getting pretty close to being satisfied with it now. I just want the limb to be much brighter than it currently is, with a more yellowish orange color. I’ll continue working on a second particle system to achieve that effect.
Thank you for the .blend ridix. I’ll look at it shortly. And Illusionist, that’s very similar to what I want to achieve, but I want there to be a visible streak tracing the path of the flares. I guess halo trails should do this?
can someone explain how to make these flares?
may be with a sample file
thanks
That last image is looking very Betelguesian, transmoderata! From what I know about such red supergiants, they would have an active photosphere with many large prominences extending far (well, relatively far) off the surface, which could help create the fuzzier limb you mentioned. Your dim red corona jibes very well with some theories about late-evolution reds, as one of the components of their solar wind/mass shedding can be carbon dust, which would tend to glow brightly in infrared and rather dimly but deeply red in visible light. Art & science seem to be working and-in-hand 
RickyBlender: as soon as I get to where I’m satisfied with my result, I’ll upload my .blend for reference.
I’ve found that a good way to achieve a flare effect that hasn’t been mentioned yet is with the “lines” and “star tips” settings of the halo material. I haven’t had near as much luck with particle trails.
Okay folks, I am bumping this thread once again, because stupid lil me still hasn’t been able to get these tendrils looking right. There’s been a lot of good advice given here and I don’t doubt it’s a problem of my personal incompetence. So, we move on to the problems I’m running into now…
I’ve basically been trying two different methods, chipmasque’s method and a method that just uses a static emitter with children particles “clumped” and “shaped” to look like tendrils. Details of my problems with both are below.
chipmasque: I tried setting things up like you suggested, and just as I expected, the particle trails above the surface emitted in awkward directions and most a good distance from the surface, creating free-floating lines of halos around the sun pointed in random directions. If you’ve gotten this to work yourself, please tell me how you did it.
The other method gets decent (while not as good as hoped for) tendrils going. The problem is that they die in the order they were created, so when they start dying, you see the trails separating from the surface and form small blotches of halos remote from the surface before dying. It looks awful. Illusionist: what settings did you use for the video you linked that bypassed this problem?
By the way here’s what I have so far:
the fuzz and the “fresnel color fade” effect around the limb are about how I want them. But…no tendrils obviously.
I wrote a tutorial about this, a long time ago. Not sure it’s entirely what you’re going for, but it might give you some ideas :
http://www.rhysy.net/Resources/Stars.pdf
I guess I’m somewhat confused on what you’re trying to accomplish. Early in the thread you published a vid link that I thought was close to what you wanted, so I described how I might do that, but now it seems you’re going for something much more naturalistic, much closer to “the real thing,” in which case, you’ll probably have to take a number of different approaches depending on what your “tendrils” are depicting.
First are the typical prominences that erupt from the surface, often from sunspot activity. They frequently follow lines of magnetic force and loop back to the surface after reaching a considerable altitude in the corona. They have a very distinctive internal structure as well caused by plasma reacting to the magnetism and flowing in sheets and flame-like tendrils.
When these “typical” prominences are not trapped by magnetism they can simply erupt as blobs of plasma that shoot off the surface and fade. These are usually rather small eruptions.
Then there are prominences and other solar events that are so energetic they actually eject material away from the surface, creating a solar flare. These can have a massive visual component that often has a flame-like structure.
On a much smaller scale, small “fountains” or “jets” of hot gas & plasma called spicules are seen all across the surface of our sun. In some very energetic stars like Wolf-Rayets these are thought to become gigantic.
Since each of these forms of solar surface activity have a different form, you’ll probably have to come up with a different means for producing each, with a lot of experimentation along the way.
Which type of feature do you favor for your “tendrils?”
i’v heard of one method to do the tendrils with particules and animated curves
but never been able to redo it !
an example would be nice here to show how to do these animated sun flares
with closed or not curves like following magnetic lines or not !
don’t know with particules field there is a force field magnetic or spiral can this be used may be ?
thanks
:spin:
You were not confused sir. I am going for a look that emulates that of the video as closely as possible. And I’m telling you the problem I’m having that is keeping your method from turning out that way. If I need to be more clear on what my problem is, please say so and I’ll upload some pictures for visual aid. Basically, the tendrils in the video I linked are all directed outward from the sun, as natural. Your method makes it impossible to achieve that.
Rhysy 2: Thanks, I’ll look in a moment.
Whoa, dude, chill a bit, I was just tryin’ to get a handle on what you’re trying to do, sorry if my ignorance offends. I think your progress so far greatly exceeds the naturalism of the example video, so I wondered how close to reality you wanted to get with the prominences, etc. The vid’s FX are very stylized and don’t really follow stellar atmospheric physics very closely, though it looks pretty good. I think your image deserves better, to be honest.
In any case, what I described is quite possible, as shown by this video:
The basic model is pretty much as I described. The main stellar sphere is a solid using only RGB coloration and a Clouds texture for the surface granulations. No animation there, though it could be done easily. Concentric with that is a deformed sphere with shape keys that make it pulse in and out, breaking through the surface of the main sphere in irregular locations. Surrounding all is another sphere to create the corona.
The video is in three parts, two different particle effects passes and the composite made from them.
Only Halos were used to create the gas/plasma effects. The pulsating sphere is also the particle emitter, and the emitter and particles share the same material. The emitter is set to render. The particle system was rendered with two sets of parameters: in the first of the three parts, the system has more children and no trails, and provides a somewhat chaotic look that changes quickly. The second part uses many fewer children and particle trails, giving a more structured and smoothly-moving effect, more like the “tendrils” of the example vid posted at the head of the thread. The third segment is a blending of the first two with some masking and filtration using the Compositor.
If you look closely you can see places where the distorted pulsing sphere breaks the main sphere’s surface and creates arc-like prominences. This can be emphasized by further shape key manipulations and Compositor effects if desirable.
The coronal effect is simply a relatively large-scale halo material with very low alpha. All the uses of the halo materials were set to various Add values.
I’ve uploaded the file to BlendSwap if you want to snag it. The Compositor and VSE modules won’t work properly without the image sequences available but that’s too much to upload. However you can still see how things were set up. You’ll have to bake the particle system. It’s currently set to produce the look of the second segment of the vid. For the first segment look, just increase the child particle count (to about 10 or 15, iirc) and disable particle trails (set Trail value to zero).
chipmasque: Thanks. Because you uploaded the .blend I came to discover that the reason I was running into particle direction problems was because my shape-keyed sphere’s “prominence” points were WAY more exaggerated than yours. I was shaping the actual mesh to how I wanted the prominences to appear, and giving the particles a very tiny initial velocity so that they more-or-less just form the shape of the mesh. That’s how I interpreted your instructions. But now with an actually scene to look at I can see all the things I’ve been doing wrong. I really appreciate your help. I’ll get back with the results.
I probably wasn’t all that clear on how to do the shape key stuff, 'cause I hadn’t done it yet! 
At least, not for this particular subject. And to some degree the shapes do provide prominence effects, the arcing type that rise and fall. The particles are influenced in vector by the shape deformations, but not to a huge degree, just enough to keep them looking a bit random and not too mechanical.
Hope the .blend helps, it was fun to experiment on, and it’s another effect for my “library.” Some possible elaborations include greater use of shape key deformations for more “rise and fall” prominences, animated materials and textures instead of rather bland halos, using “sprite-like” textured planes as particle objects and giving them some rotation as well as motion vectors, etc. I kept things simple to demo my idea (and do it quickly), but there’s a lot of room for expanding on it. Go nuts!