M_Glenn_M
I guess the main diff was the idea of using one Comp Buss for servicing several tracks thru their sends.
I thought that saved resources and setup.
 

Yes... that's the impression I am under. But scook is right that if the comp has a wet/dry it's easier to use that than keeping track of sends but if you want the same plug across multiple tracks you set the main output of those tracks to the compression bus and then the bus out to the master or whatever other bus you want in the chain before the master. This means your full signal is going to the compression bus instead of only part of it like with the Sends set up.

Just testing the formatting of the forum to see if I can create a flow chart.
                V
               Test > test >
                            V
                           Test

M_Glenn_M
I guess the main diff was the idea of using one Comp Buss for servicing several tracks thru the sends.
I thought that saved resources and setup.
 

That is a completely different question. Parallel compression was never about saving on computer resources. The technique predates the PC.
 
On the other hand, it is true using one instance of a plug-in should uses less resources that multiple instances.
 
Running a DAW on an older or less powerful machine is always a balancing act. Freeze is your friend until your disk becomes your enemy.

Okay, this is how I'd do it.
 
Using Wet/Dry knob. No sends, Just set the main output of the track to output to the compression bus.
 
Tracks 1, 2, 3, etc > Compression bus > Master
 
Using sends. Main output goes Master. Sends go to compression bus. Send knobs blend in compression.
 
Tracks 1, 2, 3 > Master
           V
       Sends > Compression bus > Master
 

 
Routing is another one of those things; there are usually several ways to achieve the same result.

If someone wants the effect of multiple tracks thru a single compressor there's nothing wrong with that, but due to the way compressors work, that is different than a single compressor per track or NY compression on a single track. I don't think you're saving resources, you're creating a different effect. A compressor has a rise time so if a rhythm track has a chord sounding the attack gets compressed, but if a lead track plays once the rhythm track has opened the compressor the lead signal is just limited so what happens to the lead, compression or limit, could be some what random. The same is true for any of the multiple tracks going into the compressor. My 2 cents.

rsinger
If someone wants the effect of multiple tracks thru a single compressor there's nothing wrong with that, but due to the way compressors work, that is different than a single compressor per track or NY compression on a single track. I don't think you're saving resources, you're creating a different effect. A compressor has a rise time so if a rhythm track has a chord sounding the attack gets compressed, but if a lead track plays once the rhythm track has opened the compressor the lead signal is just limited so what happens to the lead, compression or limit, could be some what random. The same is true for any of the multiple tracks going into the compressor. My 2 cents.

Good point.
 
Edit: And yes... I screwed up. I guess true NY style compression should be applied on a track by track basis meaning have the compressor with the wet/dry on each track or have a separate buss for each. Otherwise the compressor isn't working on that specific track (it's working on them all) which I guess is the point. My apologies.

This is all bringing up some interesting options. A lot of flex available here. I'm learning.
My original concept was to emulate numerous track sends to a single reverb buss which gave individual control over how much reverb was added to each track (thru the send level) while adding a consistency which helps blend the tracks without losing the dry track qualities.
I envisioned the benefit with a similar approach to comp was to be able to raise the quieter passages and sibilance while keeping the dynamics of the original.
I then wondered what comp I would use and how to set it. (In the case of the Reverb, it's used to extreme-so my thought was "what would that mean in compression settings?")
The point is well taken that it's different due to the nature of the need for individual compression settings on different instruments, not just volume/blend.
In this case it does seem better to add one normal comp for the general levelling and another, with the side chain switch on, and use wet/dry to bring up the quieter parts. 
Am I getting it?

Forget the side chain switch. That's (afaik) for making one track effect another like a ducking bass or putting a voiceover over top of music for a radio commercial.
 
You could use a compressor on each track with a wet/dry knob (if there is no knob create a specific buss for each track and use sends) and there is your NY compression (I'm sorry I was screwing that concept up as rsinger pointed out). Then use another buss with a compressor to send all the relevant tracks to for overall compression. You could do that second one either by sending the main out of the tracks directly to the compression buss which will effect the whole signal or use sends and just blend it in.
 
Tons of other ways to do it but I think that seems to be what you are trying to do.

M_Glenn_M
 
In this case it does seem better to add one normal comp for the general levelling and another, with the side chain switch on, and use wet/dry to bring up the quieter parts. 
Am I getting it?

I see several posts where there is a mention of the side-chain switch on. What signal is feeding the side-chain? If you turn the side-chain switch on and do not provide a signal to the side-chain, the compressor is effectively turned off.