Maybe the first thing is to determine why you got “grainy” footage. Is it pixelated, or footage where the little square pixels are visible, or is it noisy which has tiny bits of color most visible in the medium and dark parts of the picture or is the graininess sort of a halo surrounding sharply defined and/or moving objects?
Pixelated footage is the result of too few pixels. It often is caused by excessive up scaling of the original footage. I am not aware of any good fix other than being more careful the next time you shoot. The problem is due to the fact that 1080×1920 HD produces a picture with only two megapixels. That isn’t much for upscaling.
If your graininess is really fringing or haloing of moving and/or high contrast objects in your final product, that is usually due to the codec. Setting a higher bandwidth or resolution may fix the problem. Otherwise, find a better codec.
Finally, if the graininess is due to noise (technically, video can’t have grain, that is a characteristic of film), there are some software packages that offer some help. I have used the Neat package for stills with good results and currently am using Red Giant Denoiser for video. But, the Red Giant application runs in After Effects and takes a very long time to render. For example, using the application’s default settings and running nothing but After Effects on a six-core Wondows 7 machine it takes thirteen times run time to render HDV mpeg video.
As is almost always the case, shooting for lower noise is a better option. You can do this by using a lower gain setting on your camera. Lower gain means less noise just like in film, lower ISO means less grain, but also less light sensitivity. Consequently, it is necessary to use a lower shutter speed and/or a larger lens opening.
On my Sony Z5u, grain starts to be noticeable at 9dB gain. So, if possible, I manually lock it at 6dB or lower. Auto exposure still works because the camera can adjust shutter speed and iris. I have also found that locking the shutter speed to 1/30 of a second allows me to reduce the gain by 3dB. Going slower than 1/30 of a second makes the video jerky. Of course, neutral density filters should be off.
An often overlooked trick is simply to stand closer to the subject. This allows you to use less zoom and since most zoom lenses have an effectively larger iris opening at short focal lengths you get more light into the camera which means you can use less gain.
If none of these shooting techniques works for you, you can always leave the noise. Some people like it because they say it gives a “film look”. Otherwise, there is the software route which as one small advantage – it teaches patience :o)
.
