Digital Foundry Article Technical Discussion Archive [2014]

[hesido]

The uncanny valley effect crossed my mind, too. At 24fps, a viewer of a movie is distant and can enjoy a story from an objective position. At 60fps or higher, a viewer feels part of the film. That's great for a FPS, but bad for movies.

I don't think we can categorize it as bad. It's just a matter of getting used to. Currently, most people are *trained* to watch movies at 24fps and anything else feels off. Feeling part of the film is not a bad thing.

Movie lovers will hate me for it, but I now don't like 24 fps for movies, my LG with motion interpolation does a very good job of converting movies to higher framerates. I once objected a lot because of the "soap opera" effect. Now I'm very much distracted by the judder that comes with 24fps.

 

[MrSpiggott]

I don't think we can get used to HFR video just by watching more of it. I think there's more too it than that.

Take the Simpsons. It fails a basic reality check. No matter how badly the sofa or car is drawn we know they're not real and just accept them for what they are. It doesn't bother us and we just get on with enjoying the story.

With a 24FPS film, once we see movement, a higher level reality check fails, and as long as the sets and lighting aren't too artificial, we again accept them.

Now with HFR films the previous reality check passes and we go onto a higher level where we start to notice that the lighting or those materials don't look correct. We've entered the uncanny valley. Indoor scenes look like sets because, well, they are sets with all their plywood, forced perspective and artificial lighting.
Outdoor scenes with natural lighting look totally convincing so I don't think it's a problem with HFR, we just need to change the way we light and dress the sets.

 

[pjbliverpool]

Outdoor scenes with natural lighting look totally convincing so I don't think it's a problem with HFR, we just need to change the way we light and dress the sets.

I totally agree with everything you've said but I think outdoor scenes also have their pitfalls in that HFR, for the reasons you describe will also more easily show flaws in acting, costumes and outdoor props. Hell even camera angles might be at risk. Suspension of disbelief is just so much more difficult when everything looks so goddamn real!

 

[Phil]

I'm not convinced HFR has anything to do with people noticing the fakery of the filmset etc. Hasn't anyone played around with a Hero GoPro camera (capable of HD at 30 or 60fps). I have done various filming when driving on the track or mountain passes and the 60fps content always looks strange compared to what we are used to. It looks a bit like a video game because it's that smooth and the sensation of speed that is portrayed which is not possible when shot at 30fps is fantastic - yet all the same, it kind of feels fake / artificial.

I wish I had a video at hand that I could upload to YouTube, but unfortunately, I don't (and I film exclusively at 60fps, unless I shoot with little to no motion - e.g. everything not driving related).

 

[London Geezer]

People don't like it because people don't like change. Change is bad.
At the same time, same old same old is also bad. No change is bad.

It's human nature. We are destined to be an unhappy, miserable, grumpy species.

 

[HokutoNoKen]

Tomb Raider: Definitive Edition is 60fps on PS4 - report

http://www.eurogamer.net/articles/2014-01-23-tomb-raider-definitive-edition-is-60fps-on-ps4-report

/ Ken

 

[Deleted member 13524]

From the same article:

A third report puts more firm numbers on both - an average of 60fps on PS4 and 30fps on Xbox One, although the latter can climb to 45fps during some simpler sequences.

60 vs. 30 FPS? It doesn't seem like the multiplatform games will look practically the same at all..
Put this together with Assassin's Creed IV doing 720p on xbone and 1080p on ps4 and difference in performance seems bigger and bigger.

 

[AlNom]

Put this together with Assassin's Creed IV doing 720p on xbone.

900p

It's Ghosts that's 720p on XO and 1080p on PS4.

 

[Ceger]

From the same article:



60 vs. 30 FPS? It doesn't seem like the multiplatform games will look practically the same at all..
Put this together with Assassin's Creed IV doing 720p on xbone and 1080p on ps4 and difference in performance seems bigger and bigger.

Third report is questionable as not a verifiable source base. So to make assumptions on that and looping in with launch period cross gen games is not the soundest strategy nor good use of time.

 

[Deleted member 13524]

900p

It's Ghosts that's 720p on XO and 1080p on PS4.

Ok, my bad.

Third report is questionable as not a verifiable source base. So to make assumptions on that and looping in with launch period cross gen games is not the soundest strategy nor good use of time.

That's starting to sound like a bad excuse, since both consoles are equally fresh in the hands of developers.
Nonetheless, we'll address this in half a year.

 

[Phil]

I actually thought the 3rd report made the most sense, especially with the remark that the Xbox One version runs somewhere between 30-45. Isn't that in line with the resolution difference found in some titles so far, only here, both have the same resolution, but the framerate is slightly different?

To me, if anything, it sounds a bit like the lead unit is the PS4 version which is hovering around the 60fps mark most of the time as a result and with this in mind, the Xbox One version is struggling a bit more in the framerate area. Both are above 30fps, so it should be pretty solid overall. Given this is a last gen game being ported to PS4/XbO hardware, I wouldn't draw too many conclusions beyond what we already know.

The more daunting question (which would be more suited in a framerate thread) is who would prefer parity in the framerate area at the expense of some visual compromises. Reading many views on that "1080p30 is enough", one would think this Tomb Raider version has made the right choice (if the info proves to be correct). Or would some prefer less resolution at the same framerate, given it were possible?

 

[Deleted member 11852]

I'm not convinced HFR has anything to do with people noticing the fakery of the filmset etc. Hasn't anyone played around with a Hero GoPro camera (capable of HD at 30 or 60fps). I have done various filming when driving on the track or mountain passes and the 60fps content always looks strange compared to what we are used to.

Seconded. The iPhone 5S has a 120fps 720p video capture mode and it just looks utterly wrong and 'fake' - for want of a better word.

 

[tuna]

Seconded. The iPhone 5S has a 120fps 720p video capture mode and it just looks utterly wrong and 'fake' - for want of a better word.

Can it show 120 fps? Or how do you display it?

 

[MrSpiggott]

Seconded. The iPhone 5S has a 120fps 720p video capture mode and it just looks utterly wrong and 'fake' - for want of a better word.

I think that's because it is fake.
The video has passed the movement reality test so you are beginning to notice that the picture looks wrong somehow, which of course it is. The little camera on your iPhone isn't going to capture the image to the same accuracy as your eye can. It's going to look different when compared with real life.

 

[Deleted member 11852]

Can it show 120 fps? Or how do you display it?

I don't know but I'm pretty sure it can't. It's really for capture slow-motion footage. There's a good "review" here.

I think that's because it is fake.
The video has passed the movement reality test so you are beginning to notice that the picture looks wrong somehow, which of course it is. The little camera on your iPhone isn't going to capture the image to the same accuracy as your eye can. It's going to look different when compared with real life.

It's hard to put my finger on. I don't think it's accuracy or detail because my eyesight isn't great but it's the same kind of weird effect you get when enabling 120hz "motion flow" (Sony's Bravia term) on your TV. It just looks.. wrong

 

[3dilettante]

It's hard to put my finger on. I don't think it's accuracy or detail because my eyesight isn't great but it's the same kind of weird effect you get when enabling 120hz "motion flow" (Sony's Bravia term) on your TV. It just looks.. wrong

If you're enabling image processing in the TV, and I think Sony's scheme involves interpolation, it probably is "wrong" to some extent, and there can be perceptual issues when the effect is applied inconsistently.

 

[pjbliverpool]

Tomb Raider: Definitive Edition is 60fps on PS4 - report

http://www.eurogamer.net/articles/2014-01-23-tomb-raider-definitive-edition-is-60fps-on-ps4-report

/ Ken

I can;t wait for the face off. I hope they include the PC version in there too so we can see how much further the definitive editions goes (and where it may even fall behind). 60fps at 1080p with extra effects is certainly impressive stuff. I'm going to check out what settings I need to run at on my PC to acheive that but I know its not max.

 

[Deleted member 11852]

If you're enabling image processing in the TV, and I think Sony's scheme involves interpolation, it probably is "wrong" to some extent, and there can be perceptual issues when the effect is applied inconsistently.

I know the TV is creating non-existent frames between the broadcast frames but that's the nearest I can describe, what I assume to be fairly accurate 120fps capture, on the iPhone 5S as well.

 

[3dilettante]

I know the TV is creating non-existent frames between the broadcast frames but that's the nearest I can describe, what I assume to be fairly accurate 120fps capture, on the iPhone 5S as well.

My first instinct is to suspect something is not accurate about the capture, encoding, or playback if its results most closely resemble a screen processing effect.

 

[tuna]

I don't know but I'm pretty sure it can't. It's really for capture slow-motion footage.

:

If you have two phones you can test if it does 120 fps playback pretty easily.