Digital Foundry Article Technical Discussion Archive [2010]

Status
Not open for further replies.
So when can we expect a DF feature on kinect's lag?

It seems that on setup you can get a view on kinects vision + skeletal tracking, how about comparing the delay there to in game measures?

Personally I would hope for a feature on something other than just lag. We know there is lag there, but the gaming experience doesn't live or die by it. Plus, it's hard to measure what feels right.

Tommy McClain
 
Personally I would hope for a feature on something other than just lag. We know there is lag there, but the gaming experience doesn't live or die by it. Plus, it's hard to measure what feels right.

Tommy McClain

Reviews don't really seem to agree with you on that.

But for DF analysis, IMO, it should be pure technical without accounting for personal perception...so yes, I would like to see the input lag measured for the retail version, especially how it varies in different games/implementation. I would also like to see how accurate and divers Kinect tracking is.

The other stuff, fun of games, personal feeling and influence of lag during actual gameplay is something everyone has to test and judge for himself, IMO of course.
 
Personally I would hope for a feature on something other than just lag. We know there is lag there, but the gaming experience doesn't live or die by it. Plus, it's hard to measure what feels right.
DF's articles are more about the technical aspects. The actual game reviews are for the likes of Eurogamer and IGN.
 
Oh I agree that is should be solely about the technical aspects. I just don't think it would be fair to focus solely on lag. As long as there's no direct comparison to lag in regular controller games, I won't have a problem. The 2 can't be directly compared(apples to oranges). Not even sure comparing to Wii or Move is appropriate, but understand it would be natural to do so. The problem with DF's analysis is that it will show the lag that we all know is there, but people will latch onto the rather high numbers compared to regular games and say it doesn't work when it does.

Tommy McClain
 
The problem with DF's analysis is that it will show the lag that we all know is there, but people will latch onto the rather high numbers compared to regular games and say it doesn't work when it does.
That is, sadly, their fault for misunderstanding the numbers and not DF's for finding out what they are. In the same way framerate and resolution don't really define the gameplay, but people look to them when deciding games to buy. To be fair on Richard, he normally includes a sense of what the game is like to play, such as if lag is noticeable in game or not, and I expect him to do justice to his Kinect experience while keeping it relevant to us tech-nuts.
 
So when can we expect a DF feature on kinect's lag?

It seems that on setup you can get a view on kinects vision + skeletal tracking, how about comparing the delay there to in game measures?

And, I would like to see how important fps is to motion gaming. How much smoothness is need for a good casual motion gaming experience? Are there amounts of smoothness needed based on the type of casual game? If so, what are the categories? There are so many places to go with this. Frames per second for gaming was a big topic in other threads. This would be a great opportunity to tackle LightHeaven's suggestion and mine in one big feature. It would be great knowledge to have in the gaming community.
 
After using Kinect, I have a feeling it's going to vary hugely on player.

For example, if I focus on the lag, I can see it reflected around me. But if I'm actually playing the game and not actively looking for lag, I'll notice virtually no lag at all. It's a bit analgous to everytime I pick up a console controller and play a platformer that requires somewhat precise jump timing. If I look for it, the lag between pressing a button to jump and jumping is sometimes huge. But if I just play the game and don't look for it, my brain just automatically adjust and I don't ever notice the jump lag.

My non-gaming friend that I played with today was absolutely oblivious to the presence or non-presence of lag and their play never suffered due to it. They happily waved their arms around, kicked their legs around, acted goofy, and whatever. And in game happily proceeded to beat me at one of the competitive games in Kinect Adventures. :p It was one of the most "fun" experiences they've ever had. And as I've mentioned in other threads, she hates consoles (including the Wii) due to the fact that have to "fight" the controller. Er, well scratch that, she hated consoles until last night. :p Kinect changed her mind.

In other words, I don't think there's any way to measure for "smoothness" needed for casual gaming experience.

Those measurements appear, at least at first glance for my circle of friends, to be only of concern to core gamers. In that respect though, it would be interesting to see what the delta is between Button press -> action displayed on screen (including TV display lag) between a standard controller and Kinect. But at least with regards to what I've tried so far, the lag is almost completely unnoticeable when playing the game and not actively looking for lag. It is, however, quite noticeable if you are looking for it and know what you are looking for.

Regards,
SB
 
Sure lag is an easily measured phenomena. "Smoothness" however is a perceptual quality. Even with something as simple as frames per second in video games, what one person perceives as smooth may not appear smooth to another person. With lag it's a simular situation. Most people may not notice lag from enabling Triple Buffer + Vsync for example. Some do. In the first case, the game is smooth. In the second case, not so much.

Start throwing casuals into the mix that aren't into twitch gaming and just how much tolerance for lag is there going to be for a "smooth" gaming experience? It's not really a quantifiable number.

That doesn't make me less curious about how much lag there is with Kinnect compared to a standard controller. However, it remains just a curiousity and not some sort of benchmark on smoothness or playability of a game or system.

Heck, even with a theortical casual FPS (weird concept) you could always throw in lots of aim assist in game to counteract the lag of aiming. Since it's a casual title, it's not like you'd absolutely need precision twitch aiming as long as you could make the game fun for casuals.

Regards,
SB
 
At the party I had on the weekend, only the core gamers even noticed any lag, everone else just took to it like it was natural for them. I pretty much sold 3 Kinects to people who until now have never had a game console.
 
I would pretty much doubt that the majority of gamers out there both core and casual, even potential-gamers wouldn't notice Kinect's lag. I doubt many people notice input lag in most games anyway, and i've always held that position.

I think that the majority of those who do even notice some lag, and aren't nerds that frequent forums like this, will subconciously adjust and not think twice about it at all.

I think a significant benefit that Kinect games have is that its gameplay experiences are completely unique and therefore player will have little to compare them to.

Lag perceptability to me has always been relative. If i play a game like street fighter ex plus alpha (on PS1) for instance then i subconciously register some input lag. If i then switch to Street Fighter alpha i instantly recognise that the game is much more responsive, as it requires much faster input from me (or can accept much faster input) to perform my super moves. Again, when i played KZ2 i hadn't played any FPS for a while prior to that so i didn't conciously register any input lag. If i switched to MW after playing KZ2 and then went back to KZ2 i would have been amongst those who were complaining "the games isn't as responsive!".

I think a device like Kinect benefits from having no meaningful gameplay experiences readily available to players that can be meaningfully compared. I think that as time goes on however, and if MS starts updating the FW of the device to reduce lag then players will start noticing a difference in their gameplay experiences. How meaningful that difference will be and what impact it may have remains to be seen, and most likely will be rendered unimportant given the target demographic for the device.
 
I think a significant benefit that Kinect games have is that its gameplay experiences are completely unique and therefore player will have little to compare them to.

I think a device like Kinect benefits from having no meaningful gameplay experiences readily available to players that can be meaningfully compared.
Well, Eyetoy gameplay was similar, but it also had significant lag that didn't interfere with the enjoyment of the games. I think Kinect works best because the players are aware they aren't 'in' the game, but are direction another person, the avatar, telling them what to do. So it's pretty natural to act and have them respond with a delay.

Still, all this talk about Kinect in DF's thread should probably wait until DF's analysis!
 
Well, Eyetoy gameplay was similar, but it also had significant lag that didn't interfere with the enjoyment of the games. I think Kinect works best because the players are aware they aren't 'in' the game, but are direction another person, the avatar, telling them what to do. So it's pretty natural to act and have them respond with a delay.
Still, all this talk about Kinect in DF's thread should probably wait until DF's analysis!

No. It's quite the opposite. When you're playing, you are the avatar and that's how your mind thinks and reacts. I and others, dont' go "i'll make my avatar do this!" You just do it because the avatar is you and you are the avatar.

The fact that your avatar does everything you do and nothing is canned animation blurs the line between "you" and "avatar" greatly. Esp when you're in an intense game.
 
just read the BO face off:

http://www.eurogamer.net/articles/digitalfoundry-call-of-duty-black-ops-faceoff?page=1

man, this is a brutal one. must be one of the worst MP conversions this gen...and this for such a high profile game. Reminds me of GTA 4 desaster, but in this case, what is even worse: even the Xbox360 version does not look that good, IMO - engine is getting kind of old now...

what did they improve compared to MW2 that could cause such difficulties on ol PS3? If I understand it right, creation of assets, i.e. how to handle available resources, is an important task in game deving and game performance - maybe this is the main problem this time?

I decided to at least wait until it is cheap, before buying.
 
just read the BO face off:

http://www.eurogamer.net/articles/digitalfoundry-call-of-duty-black-ops-faceoff?page=1

man, this is a brutal one. must be one of the worst MP conversions this gen...and this for such a high profile game. Reminds me of GTA 4 desaster, but in this case, what is even worse: even the Xbox360 version does not look that good, IMO - engine is getting kind of old now...

what did they improve compared to MW2 that could cause such difficulties on ol PS3? If I understand it right, creation of assets, i.e. how to handle available resources, is an important task in game deving and game performance - maybe this is the main problem this time?

I decided to at least wait until it is cheap, before buying.

It definitely seems like a step back from MW2. They say there's better lighting but it looks worse to me. And there is stuttering sometimes for no apparent reason. Hair/beards look terrible. Fog/smoke on some of the multiplayer levels seems to flicker or be applied inconsistently.

And the sound of the guns is very weak.

All this is on 360.
 
just read the BO face off:

http://www.eurogamer.net/articles/digitalfoundry-call-of-duty-black-ops-faceoff?page=1

man, this is a brutal one. must be one of the worst MP conversions this gen...and this for such a high profile game. Reminds me of GTA 4 desaster, but in this case, what is even worse: even the Xbox360 version does not look that good, IMO - engine is getting kind of old now...

what did they improve compared to MW2 that could cause such difficulties on ol PS3? If I understand it right, creation of assets, i.e. how to handle available resources, is an important task in game deving and game performance - maybe this is the main problem this time?

I decided to at least wait until it is cheap, before buying.

Wow, that's pretty bad.

Doesn't Treyarch use the latest version of the CoD engine, so in this case the MW2 iteration? If so, how can the engine struggle on the ps3 so much that it needs to be scaled back so much? I can't read the entire article yet as I'm at work, but I hope they explain why the engine needed to be scaled back on this game.
 
Wow, that's pretty bad.

Doesn't Treyarch use the latest version of the CoD engine, so in this case the MW2 iteration? If so, how can the engine struggle on the ps3 so much that it needs to be scaled back so much? I can't read the entire article yet as I'm at work, but I hope they explain why the engine needed to be scaled back on this game.

I believe they said they used a modified W@W engine which was a modified Cod 4 engine.

Going to be interesting next year sledgehammer games has a cryengine 3 license and they are making the next cod. But the question is will they drop to 30 fps or still try to bring 60 fps.

If they go 30 then it's a generic shooter for me.
 
Doesn't Treyarch use the latest version of the CoD engine, so in this case the MW2 iteration?

Guys, you have to understand that before the "IW fiasco" Treyarch didn't have access to anything from IW - no help, no assets, no MW2 engine, nothing.
W@W was developed from a MW1 code drop - meaning, "here's bunch of code and assets, go have fun".
So when asking "why is this a step back from MW2" you have to understand from Treyarch's perspective it never was MW2, it was W@W.
By the time MW2 came out (and later the "fiasco" when there was more access to MW2 source code), COD:BO was basically wrapping up.
 
If they go 30 then it's a generic shooter for me.

That's funny, I've wanted them to drop the frame rate to 30fps ever since MW2 was released. :p

I'm sure it isn't easy, or even possible, but adding in a deferred lighting engine and some additional post-proc effects would make a killer looking game IMO.

Guys, you have to understand that before the "IW fiasco" Treyarch didn't have access to anything from IW - no help, no assets, no MW2 engine, nothing.
W@W was developed from a MW1 code drop - meaning, "here's bunch of code and assets, go have fun".
So when asking "why is this a step back from MW2" you have to understand from Treyarch's perspective it never was MW2, it was W@W.
By the time MW2 came out (and later the "fiasco" when there was more access to MW2 source code), COD:BO was basically wrapping up.

My mistake, I admit I assumed Treyarch received updated versions of the engine ever since W@W.
 
Status
Not open for further replies.
Back
Top