Performance Analysis: The Witcher 3: Wild Hunt
The first distinction is that PS4 caps its frame-rate at 30fps (with v-sync), while Xbox One runs with no cap at all to allow a variance between 30-40fps during play. This means Sony's machine is in theory capable of a smoother experience - with the cadence of frame delivery forced to a consistent rate. However there are issues on the Sony side side as well, and the net result is that neither feels truly smooth, though for very different reasons.
Actual gameplay paints a slightly different picture. A trip through the busy Novigrad city really puts both consoles through their paces, and Xbox One is essentially reduced to a 29-31fps range of fluctuations that produces stutter. By comparison, PS4 doesn't go over this line, but nor is it especially well optimised for large areas like this, with 25fps lurches seen at its worst. This is perhaps an issue of background streaming lots of new geometry and NPCs at each turn - as made evident by the heavy texture pop-in for each console. Sadly, each turn prompts Sony's platform to hiccup to 29fps as we canter forward, meaning it rarely feels much smoother than Microsoft's platform in a synced test, despite the frame-rate cap.
Overall, the fact Xbox One tends to holds a higher frame-rate is misleading in terms of the quality of the end experience; a capped 30fps is by far the preferred option in this case and we still hope to see it implemented in a future update. The PS4 release has the right tactic, but it lets itself down by being unable to stream in world geometry or render effects at a perfect 30fps to take advantage of this. The perceived effect is a judder to motion on both consoles, for cut-scenes and gameplay, though PS4 is capable of smoother passages of play at times. When it comes to performance, and factoring in the game's superior 1920x1080 output, Sony's hardware is a preference - but it's not the clear choice we had expected.