Can someone please explain to me what the gain knob at the top of each channel does? Just above ProChannel?
 
When do you use this?
 
I tried to attach an image but it wouldn't let me.
 
Thanks in advance.
 
 

The gain knob, for audio channels, controls the level of the wave file that gets sent to the channel strip.  I use it to make channel's levels approximately equal so I don't have faders way at the bottom or run out of gain on the fader if I need to make something louder.
 
They can also be used to get the signal levels more in line with where you want them for feeding effects such as compressor's or limiters.

It can be used to adjust the input gain, say if a track was recorded too loud or too quiet. They say there is a "sweet spot" for levels to be going into effects processing like EQ's or compressors. The gain knob can be used to level out tracks so the signals hit the effects at their "sweet spot."
 
I usually adjust each track so they are peaking at -10dbFS, not sure if that's the sweet spot per se, but it's likely close and it's easier to calculate the adjustment needed if you're shooting for -10.

Remember that the gain knob cannot affect the level of incoming audio when recording.
This must be set outside of SONAR via your interface.

The gain knob is very helpful for setting up the proper gain staging for all the effects in the effects bin and the Pro Chanel (among other things ) .Once I start populating my tracks with effects , I do not want to be fooled into thinking that louder is better just because I have placed an effect on the track  
I want to hear what the effect is bringing to the table at the same volume as what my track was set to in the mix I have been running ...
There are many other uses for the gain knob ,,,sends, busses and ect ect ...
Some guys even like keeping their Console view track faders set at 0 db and adjusting their gain knob to help with the visual across the console .
I personalty don't  
Another useful thing the gain knob can help with is if you have a very dense mix with a lot of tracks , I have found that there are times when a musical phrase on one track may send the master into the red ...
If you loop that section of the mix you can figure out which phrase may be causing the problem ..
Often times you wont even have to solo the individual tracks to find the offender ...
Once found , I have often used the gain knob on the track that was causing the problem . I will set the volume level down from the gain knob the few db the track needed to keep it from causing a visual over  ....
PS this works best if you don't have the metronome going  that puppy always brings stuff into the red the way I have mine set 
 
all the best,
 
Kenny

As an aside, I often set my tracks to peak quite high (pre fader), perhaps -3dBfs or -6dBfs. It's partly old habit from back when you needed to record hot to get a good SNR but also because most softsynths are so ridiculously loud, there's less of a difference between tracks that way.

It's probably worth shedding this behavior, but what does everyone do with their softsynths? E.g. any patch on a synth like Z3ta+ or Sylenth is basically hard clipping until you yank its GUI master fader down by about 12dB. Do I really need to manually load, adjust and save each patch? Isn't there a "softsynth headroom offset" somewhere or something?

EDIT:
There was actually supposed to be an on topic in here too: on midi tracks the "gain" knob is a velocity offset, and likewise on simple instrument tracks which is what lead to my derailment.