I have stacked 3 exposures of the same shot in Photoshop to fuse it manually, so that I can make a better HDR image of the scene.
But I am not quite sure whether I am following the right method for fusing these images, so let me explain:
I am placing the darkest exposure layer on top then mid level and the brightest on the bottom and revealing required portions from these three layers using a layer mask. Is this the right way? or does anybody have any better suggestions?
That's one way to do it, and it's perfectly fine. If it's working for you, there's probably no reason to do it any other way.
Other techniques can include:
Simply shooting RAW, and adjusting highlight/shadow exposure on the single image.
Using HDR software (more recent versions of Photoshop/Lightroom have an HDR merge feature), such as Photomatix, to perform HDR processing and tonemapping.
Exposure fusion software, such as GUI frontends (Lr/Enfuse, Photomatix, Hugin, etc.) to the open source enfuse command line tool, which basically does what you're doing manually on a per-pixel basis with algorithms for selecting/blending pixel values.
Lighting. (see: the Strobist, and Scott Hargis).
There is no right way to do this. Whatever gets the results you're after, is the right way.
I typically pick a base Exposure which will be the bulk of the image, this goes on bottom.
The rest doesn't make a difference because you're going to mask anyways. Say I have a foreground, a building, and a sky. I need 3 exposures one of each. Well the building as my base goes on bottom. The others don't matter because the foreground will be masked and not interfere with the sky, like the sky will be masked and not interfere with the foreground. That's the whole point of masks.
What you might however want to look at is Luminance masks.
Basically go into your channels. Duplicate the RGB channel, then intersect it with itself, duplicate and repeat a few times. That will give you progressive light channels.
Then select the base and do invert selection Ctrl/Cmd+Shift+i and repeat the same steps. Now you'll have progressively darker channels.
Your base exposure still goes on the bottom but then you can save some time by using an appropriate luminance mask on however many additional exposure layers you have, then manually masking out / in parts that the luminance mask didn't get.
