SteamVR SS vs. in-game PD

[VC]

I'm going to do a bit more experimenting with this but wanted some opinions as well.

 

 

Are the two methods of supersampling the same, as in do they use the same algorithms behind the scenes etc. or do they produce slightly different results?

 

 

For example:

 

 

SteamVR recommends 200% resolution based on my PC specs.

 

 

Now, SteamVR is measuring number of pixels but in-game PD is linear edge length, so 200% in SVR is equivalent to a PD of 1.42 (approx. square root 2).

 

 

So, I can leave SteamVR as default (200%) and set PD in game to 1.0, or I can drag SteamVR back down to 100% and use PD 1.4 in game, and I should be rendering (almost) the same number of pixels.

 

 

But am I actually seeing the same thing? Does it downsample the same way? Are they actually all equivalent or is one "better" in your experience?