I prefer blur-o-vision to "shimmer-o-vision" and "pixelcrawl-o-vision", and "jaggies-o-vision".
I wonder why Ubi in its infinite wisdom decided that users will not be allowed to disable SSAO completely, or to toggle FXAA. The review could mention that it is possible to disable AO, DoF (and supposedly enable FXAA) by tweaking a config file. Is PC gaming so consolized, that tweaking a config a file is no longer a valid option?
Personally I have to disable AA and dial down some graphical settings like PostFX or shadows (impact of which was sadly omitted in this review) in order to play in 1080p. Why? Because whenever my framerate dips below my refresh rate (which I already lowered on purpose to 50Hz) unbearable stuttering begins - that is why I am forced to play on lower settings because if I want to avoid stuttering, I need to have a rock solid 50fps - so effectively my AVERAGE fps need to be at or above 50 to avoid stuttering.
These benchmarks are supposed to reflect a real life gaming scenario and that is what I love [H] reviews for - that they find optimal playable setting for a given card at a given resolution and then test. This methodology has been very relevant and helpful, so far.
But in case of Far Cry 3 it stops being relevant because the stuttering is not present when vsync is disabled - and [H] did all the benchmarks with vsync off. I know this was done so that the FPS numbers are not artificially capped below cards capabilities, but in this case it is not a good reflection of real life conditions, because lack of vsync results in a very distracting screen tearing, so disabling it is not an option for me. Did you internally test with vsync on? Probably, because then the stuttering issue would be mentioned (assuming this issue affect all, not just me).
Can someone with a similar rig as mine can confirm or deny having stuttering issues when framerate dips below refresh rate, when vsync is on? Perhaps it is some kind of glitch on my side and can be fixed? That would be great as it would allow me to play on higher settings, because without stuttering, the dips below 50fps would be much more tolerable.
Also, I think that it is a pity that the 9 graphical options which can be customized to a large extent were ignored in this analysis - especially PostFX which has a large impact on framerate but almost unnoticeable impact on how the game looks (at least without switching between still shots).
Since MSAA has significant visual and performance impact on the game, I think this review would be much more helpful if these 9 options were customized to enable MSAA more often - for example it could have been tested if dialing down PostFX/shadows/geometry, etc would allow to enable 2 or 4x MSSA without noticeably compromising image quality. For example, I noticed that dialing down PostFX, shadows and geometry would improve performance without compromising visual quality.
I wonder why Ubi in its infinite wisdom decided that users will not be allowed to disable SSAO completely, or to toggle FXAA. The review could mention that it is possible to disable AO, DoF (and supposedly enable FXAA) by tweaking a config file. Is PC gaming so consolized, that tweaking a config a file is no longer a valid option?
Personally I have to disable AA and dial down some graphical settings like PostFX or shadows (impact of which was sadly omitted in this review) in order to play in 1080p. Why? Because whenever my framerate dips below my refresh rate (which I already lowered on purpose to 50Hz) unbearable stuttering begins - that is why I am forced to play on lower settings because if I want to avoid stuttering, I need to have a rock solid 50fps - so effectively my AVERAGE fps need to be at or above 50 to avoid stuttering.
These benchmarks are supposed to reflect a real life gaming scenario and that is what I love [H] reviews for - that they find optimal playable setting for a given card at a given resolution and then test. This methodology has been very relevant and helpful, so far.
But in case of Far Cry 3 it stops being relevant because the stuttering is not present when vsync is disabled - and [H] did all the benchmarks with vsync off. I know this was done so that the FPS numbers are not artificially capped below cards capabilities, but in this case it is not a good reflection of real life conditions, because lack of vsync results in a very distracting screen tearing, so disabling it is not an option for me. Did you internally test with vsync on? Probably, because then the stuttering issue would be mentioned (assuming this issue affect all, not just me).
Can someone with a similar rig as mine can confirm or deny having stuttering issues when framerate dips below refresh rate, when vsync is on? Perhaps it is some kind of glitch on my side and can be fixed? That would be great as it would allow me to play on higher settings, because without stuttering, the dips below 50fps would be much more tolerable.
Also, I think that it is a pity that the 9 graphical options which can be customized to a large extent were ignored in this analysis - especially PostFX which has a large impact on framerate but almost unnoticeable impact on how the game looks (at least without switching between still shots).
Since MSAA has significant visual and performance impact on the game, I think this review would be much more helpful if these 9 options were customized to enable MSAA more often - for example it could have been tested if dialing down PostFX/shadows/geometry, etc would allow to enable 2 or 4x MSSA without noticeably compromising image quality. For example, I noticed that dialing down PostFX, shadows and geometry would improve performance without compromising visual quality.