GT5 shots zoomed in

  • Thread starter Thread starter trk29
  • 239 comments
  • 52,679 views
That's unbelievably false. Gran Turismo 5 uses exactly the same standard of car model whether during gameplay or in Photo Mode. As I've already explained in this post, the apparent difference in quality is down to the level of detail engine, which would be present during both gameplay and Photo Mode.
Exactly. The only difference between photomode and gameplay is anti-aliasing, which is much higher in photomode.
 
That's unbelievably false. Gran Turismo 5 uses exactly the same standard of car model whether during gameplay or in Photo Mode. As I've already explained in this post, the apparent difference in quality is down to the level of detail engine, which would be present during both gameplay and Photo Mode.
Exactly. The only difference between photomode and gameplay is anti-aliasing, which is much higher in photomode.
Is there any evidence for this, or are these just assumptions?
 
Amazing pictures in this thread, that I currently use for my wallpaper on my PS3.

If anyone has an HDTV/PS3 combo, I suggest doing the same. You can zoom the pictures down to 34% (or so) and really get a good look at the textures and modeling job done by PD.
 
There is certainly no visible differences whatsoever, and visible differences are the only differences that matter. :D

Photo Mode: http://i42.tinypic.com/n1xn2a.jpg

Gameplay: http://i216.photobucket.com/albums/cc169/akirozz/f12.jpg

Photomode and Photo Travel are 2 Different things.

The pic you posted is a screen from a replay in-game, Photo Travel (where the Ultra hi-detail Jag was seen) is a whole other method, the car is not in motion and you have a set location.

So theoretically Photo Travel should offer the highest resolution\LOD available.

Which you would HAVE to agree with after seeing the XJ13 Photo Travel and Photomode comparison.
 
Photomode and Photo Travel are 2 Different things.

The pic you posted is a screen from a replay in-game, Photo Travel (where the Ultra hi-detail Jag was seen) is a whole other method, the car is not in motion and you have a set location.

So theoretically Photo Travel should offer the highest resolution\LOD available.

Which you would HAVE to agree with after seeing the XJ13 Photo Travel and Photomode comparison.

I disagree; that does not mean the models are any different. As I've had to explain twice now, if the engine bay was to fill the entire shot during gamelay, the level of detail would be identical to that showcased in the Photo Travel image.
 
I disagree; that does not mean the models are any different. As I've had to explain twice now, if the engine bay was to fill the entire shot during gamelay, the level of detail would be identical to that showcased in the Photo Travel image.
...which is something you assume, but not know.
 
As an avid karter I found this shot of the chassis welds particularly impressive!

27942832.jpg
 
...which is something you assume, but not know.

I assume this on the basis that there is no reason whatsoever to believe that a lower quality level of detail engine is used during gameplay. It would be fairly foolhardy to hold a negative belief about the game when there is no substantial evidence to back this up, for example, a screenshot of the car's engine bay in Photo Travel at a similar distance from which it is seen in the other screenshot.
 
One reason to believe so might be that this was the case in Gran Turismo 4. And it does make sense because computing power is a valuable good during gameplay, so they might lower the level of detail here and there to ensure ideal graphics engine performance.

in the end, there is no substantial evidence for either version, so we actually do not finally know who is correct.
 
there is no substantial evidence to back this up, for example, a screenshot of the car's engine bay in Photo Travel at a similar distance from which it is seen in the other screenshot.

Photomode:
http://www.gtpla.net/wp-content/upl..._JAGUAR_XJ13_66_Ford_GT40markIVRaceCar_01.jpg

Phototravel:
http://www.gtpla.net/wp-content/uploads/2010/08/Circuit-de-la-Sarthe_Jaguar_XJ13-Race-Car.jpg

The distance is similar and the difference in detail is definitely noticeable.
The photomode uses the full detail models for it, gameplay, phototravel and replays use a bit lower detail models.
This is common sense, GT4 did it and even Forza 3 did it.
 
there is no substantial evidence to back this up, for example, a screenshot of the car's engine bay in Photo Travel at a similar distance from which it is seen in the other screenshot.

Photomode:
http://www.gtpla.net/wp-content/upl..._JAGUAR_XJ13_66_Ford_GT40markIVRaceCar_01.jpg

Phototravel:
http://www.gtpla.net/wp-content/uploads/2010/08/Circuit-de-la-Sarthe_Jaguar_XJ13-Race-Car.jpg

The distance is similar and the difference in detail is definitely noticeable.
The photomode uses the full detail models for it, gameplay, phototravel and replays use a bit lower detail models.
This is common sense, GT4 did it and even Forza 3 did it.
Yep well noticed 👍

It's clear some 3D models are just replaced by 2D pictures.
 
there is no substantial evidence to back this up, for example, a screenshot of the car's engine bay in Photo Travel at a similar distance from which it is seen in the other screenshot.

Photomode:
http://www.gtpla.net/wp-content/upl..._JAGUAR_XJ13_66_Ford_GT40markIVRaceCar_01.jpg

Phototravel:
http://www.gtpla.net/wp-content/uploads/2010/08/Circuit-de-la-Sarthe_Jaguar_XJ13-Race-Car.jpg

The distance is similar and the difference in detail is definitely noticeable.
The photomode uses the full detail models for it, gameplay, phototravel and replays use a bit lower detail models.
This is common sense, GT4 did it and even Forza 3 did it.

That is by no means a similar distance.
 
There'll be some sort of progressive lod used in all modes, no point in using a 500,000 poly model to draw something that only covers 100 pixels which might even end up a bit more messy than using a much lower lod and the algorithms for the lod will be tuned specifically for the different rendering modes. There will be a max lod though that if the camera is close enough there won't be any difference in the lod used by different modes but the lighting, shadows, antialiasing and texture filtering should be much better in offline renders.
 
One reason to believe so might be that this was the case in Gran Turismo 4. And it does make sense because computing power is a valuable good during gameplay, so they might lower the level of detail here and there to ensure ideal graphics engine performance.

in the end, there is no substantial evidence for either version, so we actually do not finally know who is correct.

Of course, it's all speculation!
 
As an avid karter I found this shot of the chassis welds particularly impressive!

27942832.jpg

Wow exotic PDI racing tyres. ;)

Joking aside it looks exactly identical to my tonykart (just a different colour).
 
That is by no means a similar distance.

Maybe, to help things along, you could define what "similar" means in this instance.

My guess is is there's a difference of 6" to a foot (15 - 30 cm), accounting for the difference in field of view, among other things.

Anyway, it's close up and there's a marked difference in detail. Signs point to it being scaled according to the environment. No biggie.
 
That is by no means a similar distance.

Take a look at the 2 pictures he posted, on phototravel the engine was a 3D model, in the race photomode the engine was part 2D texture.

Completely different models.
 
Yes, but all he says is that it could be a lower LOD model, of course it looks different.
 
I really do have to wonder why Polyphony would spend so much time creating the premium models that we see in the Photo travel mode when they can't actually be used for driving in the game and only for display purposes. Surely it would have been a far better use of their time to create 400-600 models that reached the in-game model level.

I mean, they are nice feats of modelling but if they were going for pure modelling excellence there are better models out there to be found on the internet. For instance, this image:

http://img119.imageshack.us/img119/7529/koenigsegg0719201zb7.jpg
http://fc09.deviantart.net/fs43/f/2009/090/e/7/Koenigsegg_CCX_in_game_192K_v3_by_ProperGraphics.jpg

Was created by Proper Graphics using the same number of polygons that a GT5 Premium car has. In fact, on their website it's even listed as a low poly car!

http://propergraphics.com/3dlowpoly/Koenigsegg_23_3840.htm

Here's the disclaimer on their website for that car:

In-game model by Gran Turismo 5 standards. Done by blueprints and huge amount of photos. Poly budget: 192 449 triangles. Textures: in total diffuse, normal bump and specular less than 2048x2048 each. Only in-game textures were used. Only normal bump - no displacement. Part of photo reference

Check the detail on this interior shot:

http://propergraphics.com/3dlowpoly/Koenigsegg/Koenigsegg_23_3840.jpg

Maybe I'm just cynical but I could never see the point of it spending so much time on a model that ultimately is superfluous to the experience. Leave that to the design students who are trying to show off their new-found talents.

As a reminder of what's actually possible with 3D modelling, here's Proper Graphic's rendition of a CCX using their normal level of detail...

http://propergraphics.com/vehicles/koenigsegg/CCX_01_3500.jpg
http://propergraphics.com/vehicles/koenigsegg/CCX_10_3500.jpg

*shrug*
 
I really do have to wonder why Polyphony would spend so much time creating the premium models that we see in the Photo travel mode when they can't actually be used for driving in the game and only for display purposes. Surely it would have been a far better use of their time to create 400-600 models that reached the in-game model level.

The premium models are future proof. So they'll be able to plug them into a PS4 GT without anyone complaining about the quality. That'll allow them to focus on other cars since they won't have to worry about rebuilding those premium cars.
 
I really do have to wonder why Polyphony would spend so much time creating the premium models that we see in the Photo travel mode when they can't actually be used for driving in the game and only for display purposes. Surely it would have been a far better use of their time to create 400-600 models that reached the in-game model level.

I mean, they are nice feats of modelling but if they were going for pure modelling excellence there are better models out there to be found on the internet. For instance, this image:

http://img119.imageshack.us/img119/7529/koenigsegg0719201zb7.jpg
http://fc09.deviantart.net/fs43/f/2009/090/e/7/Koenigsegg_CCX_in_game_192K_v3_by_ProperGraphics.jpg

Was created by Proper Graphics using the same number of polygons that a GT5 Premium car has. In fact, on their website it's even listed as a low poly car!

http://propergraphics.com/3dlowpoly/Koenigsegg_23_3840.htm

Here's the disclaimer on their website for that car:



Check the detail on this interior shot:

http://propergraphics.com/3dlowpoly/Koenigsegg/Koenigsegg_23_3840.jpg

Maybe I'm just cynical but I could never see the point of it spending so much time on a model that ultimately is superfluous to the experience. Leave that to the design students who are trying to show off their new-found talents.

As a reminder of what's actually possible with 3D modelling, here's Proper Graphic's rendition of a CCX using their normal level of detail...

http://propergraphics.com/vehicles/koenigsegg/CCX_01_3500.jpg
http://propergraphics.com/vehicles/koenigsegg/CCX_10_3500.jpg

*shrug*



Almost same modeling, its just DIFRENT TEXTURES, LIGHTNING EFFECTS etc etc.. PS3 would not handle such super high res textures to race with. Like the Solid has said, its future proof. By the time GT6 comes out or 7 on the PS4, the textures, lightning etc will be very high resolution with the same polygon counts. Those CCX Pictures probably take long time to render them, GT5 and PS3 needs only 30 seconds to load all the buildings, 16 cars on track, calculate physics, etc... so yeah PS4 will use same GT5's car models but with more high detail.
 
Almost same modeling, its just DIFRENT TEXTURES, LIGHTNING EFFECTS etc etc.. PS3 would not handle such super high res textures to race with. Like the Solid has said, its future proof. By the time GT6 comes out or 7 on the PS4, the textures, lightning etc will be very high resolution with the same polygon counts. Those CCX Pictures probably take long time to render them, GT5 and PS3 needs only 30 seconds to load all the buildings, 16 cars on track, calculate physics, etc... so yeah PS4 will use same GT5's car models but with more high detail.

I do not even see the need for a ps4. Why can't Sony just release a usb ram peripheral. This would allow for better textures and higher AA.
 
Agreed, they look nothing the same. 👍
You know how some PC games, even PS3 games, as they first load you have nothing but smooth objects that just look like shapes, then the textures load and it looks good? That's what it looks like to me. Photomode has all the textures loaded while the in game cars are missing a lot of those detailed textures.
Whether that's different LOD or models or whatnot I don't know cause I'm not pretending to be some 3D artist personman.
 
these shots are missing the reflection of the photographer..
VAMPIRES!!!
 
Last edited:
I do not even see the need for a ps4. Why can't Sony just release a usb ram peripheral. This would allow for better textures and higher AA.

Yes, because such thing exists right? The speed is too slow for a USB. If you want such a thing you'll have to
1. open up the PS3
2. see if the RAM is even possible to remove
3. somehow find an internal DIMM that is compatible with the PS3.
 
We need more ram and everything else to have better graphics in GT6.
So PS4 is needed for GT6. Either way GT6 might be out by 2012, PS4 might be out as well around that date. GT5 alone will have plenty of play time with single play and online, so GT6 can wait till PS4 comes out so it looks, plays, and sounds even more amazing.
 
^Well why don't they change the game engine or tweak it so that its less resource consuming?
 
Back