PC Buying Guide - July 2011

[Sarath]

Whoa! How much time did you take to write that up.
Ok that clears a lot of things

I knew it doesnt need double the proc but I didnt know exactly how much extra for 3D.

Movies and designing have no place in my PC. Its used purely for gaming. Although it excels in that domain I though better colours would look better in game. But I should tell you games do look significantly better than in this panel.

But I got your point. Games dont make full use of an IPS panel so a TN would be fine.

I have been through 5 generations of consoles and all I saw was improvement in visuals and smoother gameplay. I dont understand why the developers would ditch realism for shoddy colours.

 

[Skud]

@Cilus: That was excellent!!!

 

[Tenida]

@Cilus bhai-You're great Thanks again for giving us 3D rig

 

[Cilus]

Thanks a lot guys for taking your valuable time to read my super lengthy post and share you valuable opinions...it means a lot to me.

Sarath, I am from Computer Engineering background and a very old reader of Digit magazine, especially their in depth articles. So in my college life I used to study all those things and used to try really hard to understand those things by consulting available computer science books on Microprocessor, Vector processors, Operating system etc. Thankfully, I still have the habit...trying to understand everything and go as much deeper as possible.
I think all the computer science students should go through those books man, they are really interesting.

Guys if you like my post, I have posted another article regarding the new Architecture of AMD graphics cards in AFDS: The architecture of future AMD GPUs! thread. Some of the members like Jas, SKud has already shared their opinion. SO if you guys spend a little time on it, it will be a great favor to me.

Warning: That post was far bigger than this one.

And one thing about duplicate Frames in 3D... Suppose you have 1 Frame having all the sahders processed and the 2nd Frame which is need to be created is just a little different in terms of Viewing angle, color depth and intensity of light. Now here the 2nd frame does not get created from the scratch. The game program can actually keep track of the changes between those two frames and it duplicate the 1st frame 1st and then apply the changed Shader/Pixel matrix on it to generate the 2nd frame. So the GPU overload for the 2nd frame is just to apply the Difference Matrix on the 1st frame, not to build it from scratch.

 

[Sarath]

I thought the picture moves at 120hz and the active shutter glasses give you 60hz per eye for 3D. Thats why I though the frame rate needs to be doubled.

 

[Cilus]

What you know is correct in some extent but I was talking about how the double frame generation doesn't hurt the GPU performance that much. I was explaining how the 2 Frames which are required to generate a single 3D frame are getting generated by GPU.

 

[Jaskanwar Singh]

cilus superb man! just awesome info.

 

[Extreme Gamer]

While it is true that they are only slightly different images in 3D, you need to remember Cilus that the viewing angle changes a bit by ~1-2 Degrees max, that means two images have to be generated. While you dont get 2x load, you do get almost 80% higher load on average. Of course I am refering to Nvidia 3D vision here, not some other 3rd party vendor.
Each image is separately rendered one by one in stereo 3D.

Put your thumb between your eyes and close your eyes alternatingly and see the difference.

 

[Sarath]

That makes sense actually. I mean if the two offset images are not similar then the brain would have difficulties patching up the images from the two eyes. So every 2 consecutive frames are similar for 3D generation giving a net effect of 60hz while the processing power increase is marginal. Is 20% increase safe to say?

Will my HD 6970 do 3D? on a 120hz monitor ofcourse. (and kit)

Also I wanted to make sure you guys add a disclaimer for 3D rigs that:
~General: Long usage even for 3hours causes headache and dizziness in many. Needless to mention eye fatigue.
~Some people fail to perceive 3D for some unknown medical reason. So ask them to check 3D beforehand at a Sony store or friends place. For them 3D appears as 2D albeit with a lot of blur.

 

[Jaskanwar Singh]

cilus so final config like this?

Sub 110k (Extreme 3D Gaming)

Component | Make | Price Processor |Intel Core i5 2500k|10500
CPU Cooler |CM Hyper N620|2500
Motherboard |Gigabyte GA-Z68X-UD3H-B3|11500
RAM |G.Skill Ripjaws X F3-12800CL8D-8GBXM|5400
Graphic Card |MSI GTX580 Twin Frozr II|25900
3D Vision Kit |Nvidia 3D Vision Kit|8100
HDD |Seagate 1TB 7200.12|2700
DVD Writer |LG 22X SATA DVD|900
PSU |Glacialtech GP AX 950 AA|6000
Case |Lian Li Lancool K58W|4300
Monitor |Asus VG236HE BK|18500
Mouse |Razer Deathadder|2000
Keyboard |RAzer Arctosa|2000
Mouse Pad |Razer Goliathus Dragon Age II Edition|900
UPS |APC 1.1KVA|4300
Speakers |Logitech Z-506|5000
| Total |110500

Optional upgrades for 110k -

Component | Make | Price SSD |OCZ 60GB AGILITY 3 SATA III|7500
Headset |Roccat Kave|5000

 

[Cilus]

Great Jas, I think our work is done here....for the 1st 3D config.

Sarath, Yes all the AMD 6000 series card support 3D, but they have different implementation than nVidia Stereoscopic 3D.

AMD 3D is supported in any 3D Television but not in most of the 3D monitors available in market. Currently Viewsonic and some other comanies are working for AMD based 3D monitors and I think some are available in newegg. But their number are very low currently. Hopefully this condition will be increased in couple of months as AMD is working seriously to bring full phased 3D support.
The Viewsonic V3D241wm supports both AMD and nVidia 3D. For all supported 3D TVs and LCD monitors Check the Official Supported hardware page.

And ya, we are gonna put the disclaimer as u suggested. But the best thing is if you feel this kinda problem, just switch to the 2D mode for some times.

Extreme games, thanks for the info. I think I have mentioned the viewing angle point in my previous post.

 

[Extreme Gamer]

The brain is only able to patch up the objects in focus. You will see that the other background objects appear dubled and/or blurred.

When you put a finger in front of your eyse and focus beyond it, you see that the fingers become doubled.

Have you noticed, that when you play GTA4 in 3D, only upto the player the image is truly 3D? Beyond that it appears bifurcated

Moral of the story is that we need parralax 3D to go mainstream before eye stran can be reduced. Active stereo is the most expensive and the most strenuous form out there. We are trying to force our brains into thinking we are looking through glass when we are actually looking at opaque light emitting devices right in front of us.

Depth also differs person to person btw.

Cilus, you probably have mentioned, but you need to know that when viewing angle changes, howsoever small it may be, there will be a new image altogether and not just some width extension
To get a proper image, two separate images have to be rendered. The frames may share calculations but during the culling and clipping portion of the calculation, separate calculation has to be done.

For example, the calculation of smoke in case of grenade explosion might be shared, but the portions culled from the image render will have to be separately calculated.

 

[Jaskanwar Singh]

That makes sense actually. I mean if the two offset images are not similar then the brain would have difficulties patching up the images from the two eyes. So every 2 consecutive frames are similar for 3D generation giving a net effect of 60hz while the processing power increase is marginal. Is 20% increase safe to say?

Will my HD 6970 do 3D? on a 120hz monitor ofcourse. (and kit)

Also I wanted to make sure you guys add a disclaimer for 3D rigs that:
~General: Long usage even for 3hours causes headache and dizziness in many. Needless to mention eye fatigue.
~Some people fail to perceive 3D for some unknown medical reason. So ask them to check 3D beforehand at a Sony store or friends place. For them 3D appears as 2D albeit with a lot of blur.

thanks for that info sarath.

 

[tkin]

I thought the picture moves at 120hz and the active shutter glasses give you 60hz per eye for 3D. Thats why I though the frame rate needs to be doubled.

No, not necessarily, frame rate does not need to be doubled. Going by your theory most games would be unplayable at even 1080p by cards like 570(I mean less that 60FPS, some people find it "Distracting").

 

[Sarath]

3D needs a few more years I guess. But if people stop buying then it will never be developed further.

Its nice to see however that a 3D rig has come up. When I joined TDF a month back I was wondering as to why no 3D rigs were mentioned, but since I wasnt looking for one, I never bothered.

Now this above is a high end system or was if it was a 2D one. So where does it stand for 3D. I mean owing to the requirements of the 3D and the performance hit is this still a rig that can max out all settings in 3D? Simply is this a high end 3D rig or a good 3D rig?

Good work Jas and Cilius. You guys should rest your fingers a little bit now.

Oh you are welcome. Always a pleasure to help troll
[I missed that post as it came while I was typing]

 

[tkin]

The brain is only able to patch up the objects in focus. You will see that the other background objects appear dubled and/or blurred.

When you put a finger in front of your eyse and focus beyond it, you see that the fingers become doubled.

Have you noticed, that when you play GTA4 in 3D, only upto the player the image is truly 3D? Beyond that it appears bifurcated

Moral of the story is that we need parralax 3D to go mainstream before eye stran can be reduced. Active stereo is the most expensive and the most strenuous form out there. We are trying to force our brains into thinking we are looking through glass when we are actually looking at opaque light emitting devices right in front of us.

Depth also differs person to person btw.

Cilus, you probably have mentioned, but you need to know that when viewing angle changes, howsoever small it may be, there will be a new image altogether and not just some width extension
To get a proper image, two separate images have to be rendered. The frames may share calculations but during the culling and clipping portion of the calculation, separate calculation has to be done.

For example, the calculation of smoke in case of grenade explosion might be shared, but the portions culled from the image render will have to be separately calculated.

And that would make it computationally impossible, very well explained on your part.

And 3d hurts my eyes, hurts like hell, one reason to get 3d monitors is due to decrease in effective brightness 3d monitors need high luminance levels to start with, 400cd/m2 is minimum, that will look very bright, and not to mention: NO FLICKERING upto 120FPS, so I don't need to turn on vsync(god, the lag with that on), only reasons I would get a 120hz monitor and not use 3d.

Great Jas, I think our work is done here....for the 1st 3D config.

Sarath, Yes all the AMD 6000 series card support 3D, but they have different implementation than nVidia Stereoscopic 3D.

AMD 3D is supported in any 3D Television but not in most of the 3D monitors available in market. Currently Viewsonic and some other comanies are working for AMD based 3D monitors and I think some are available in newegg. But their number are very low currently. Hopefully this condition will be increased in couple of months as AMD is working seriously to bring full phased 3D support.
The Viewsonic V3D241wm supports both AMD and nVidia 3D. For all supported 3D TVs and LCD monitors Check the Official Supported hardware page.

And ya, we are gonna put the disclaimer as u suggested. But the best thing is if you feel this kinda problem, just switch to the 2D mode for some times.

Extreme games, thanks for the info. I think I have mentioned the viewing angle point in my previous post.

AMD HD3D is for Blueray 3d only, no games support 3d with amd cards like nvidia, amd does not have any active program for 3d like 3d vision, 3d gamers will need to stick with nvidia.

 

[Cilus]

Tkin, why don't you check others' post before posting your misleading info? The AMD official link I've given, check it 1st before posting because I guess your level of knowledge is very limited in this field. In that list it is clearly mentioned that all the 5000 series and 6000 series cards are capable of
3D Gaming
Bluray 3D
2D to 3D Video Conversion
Photos

Here is the review of the Viewsonic V3D241wm. Read it , you will find that they tested it with both AMD and nvidia based 3D solutions, although nVidia Stereoscopic 3D was proven far mature than AMD's current solution in some games.

Extreme Gamer, please do care to read my previous posts first. Where did I mention that you don't need 2 images for 3D images. What I was stating that although you need two frames to generate a single 3D frame, that does not cost GPU 2 X t times where t is the processing time for the 1st frame. Reason is the 2nd frame does not need to be generated from the scratch, it can be created by applying the Difference Matrix over the 1st Frame.

Quote from my older post:

And one thing about duplicate Frames in 3D... Suppose you have 1 Frame having all the sahders processed and the 2nd Frame which is need to be created is just a little different in terms of Viewing angle, color depth and intensity of light. Now here the 2nd frame does not get created from the scratch. The game program can actually keep track of the changes between those two frames and it duplicate the 1st frame 1st and then apply the changed Shader/Pixel matrix on it to generate the 2nd frame. So the GPU overload for the 2nd frame is just to apply the Difference Matrix on the 1st frame, not to build it from scratch.

This technique is also used in Video compression methodologies. MPEG or Moving Picture Expert Group extensively uses these techniques where although the whole movie is consisting of say 10000 frames, the video itself does not need have all the frames separately because some of the frames are just having minor differences with their previous/Next frame. So instead of keeping the whole information of the frame, the compression technique actually keeps the reference frame and just records the changed info about the next frames. In decoding time, the dependent frame is created by Reference Frame + Changed Information.

Now what part of it you don't understand?

 

[vickybat]

The brain is only able to patch up the objects in focus. You will see that the other background objects appear dubled and/or blurred.

When you put a finger in front of your eyse and focus beyond it, you see that the fingers become doubled.

Have you noticed, that when you play GTA4 in 3D, only upto the player the image is truly 3D? Beyond that it appears bifurcated

Does GTA4 even supports stereoscopic 3d? Afaik, it doesn't have native support. And ofcourse titles not having native support will look bad in 3d mode.

Moral of the story is that we need parralex 3D to go mainstream before eye stran can be reduced. Active stereo is the most expensive and the most strenuous form out there. We are trying to force our brains into thinking we are looking through glass when we are actually looking at opaque light emitting devices right in front of us.

Parallex barrier has a lot of shortcomings. For instance, the viewer has to position himself/herself in a specific point to view correct depth. Since a parallex barrier has a number of slits to allow light to pass at different angles, the position of the eyes has to be such that two images are foccussed simultaneously in order to view depth.

Another shortcoming is that the horizontal pixel count is reduced by one half.
Its kind of a cheap implementation of 3d it hardly will be mainstream.

Nintendo 3ds is a great example of a having a display with a parralex barrier enabling glass free 3d.

Real glass free 3d has to be a lenticular display. Here , instead of a barrier having a no. of slits, it has a refractive layer on top of the panel. This method relies on a display coated with a lenticular film. Lenticules are tiny lenses on the base side of a special film. The screen displays two sets of the same image. The lenses direct the light from the images to your eyes each eye sees only one image. Your brain puts the images together and you interpret it as a three-dimensional image.

This technology requires content providers to create special images for the effect to work. They must interlace the two sets of images together. If you were to try and view the video feed on a normal screen, you would see a blurry double image.

Here's a comparison between parralex barrier and lenticular method.

*img31.imageshack.us/img31/7618/220pxparallaxbarriervsl.png

You can clearly see that that the lenticular display has a wide source coming to each eye than the parralex method.This allows the user to switch positions more freely than the parralex based display.

But lenticular displays are far expensive to make and will take sometime to get mainstream. So lenticular displays are the future cause they also provide more sense of depth than their parralex counterparts.

 

[tkin]

Tkin, why don't you check others' post before posting your misleading info? The AMD official link I've given, check it 1st before posting because I guess your level of knowledge is very limited in this field. In that list it is clearly mentioned that all the 5000 series and 6000 series cards are capable of
3D Gaming
Bluray 3D
2D to 3D Video Conversion
Photos

Here is the review of the Viewsonic V3D241wm. Read it , you will find that they tested it with both AMD and nvidia based 3D solutions, although nVidia Stereoscopic 3D was proven far mature than AMD's current solution in some games.

Extreme Gamer, please do care to read my previous posts first. Where did I mention that you don't need 2 images for 3D images. What I was stating that although you need two frames to generate a single 3D frame, that does not cost GPU 2 X t times where t is the processing time for the 1st frame. Reason is the 2nd frame does not need to be generated from the scratch, it can be created by applying the Difference Matrix over the 1st Frame.

Quote from my older post:


This technique is also used in Video compression methodologies. MPEG or Moving Picture Expert Group extensively uses these techniques where although the whole movie is consisting of say 10000 frames, the video itself does not need have all the frames separately because some of the frames are just having minor differences with their previous/Next frame. So instead of keeping the whole information of the frame, the compression technique actually keeps the reference frame and just records the changed info about the next frames. In decoding time, the dependent frame is created by Reference Frame + Changed Information.

Now what part of it you don't understand?

You should try to read bit more, start here perhaps? Come back when you read this article:
ASUS VG236H 120Hz LCD Monitor 3D-Vision Set | ASUS VG236H,Review,3D Vision,120Hz,Widescreen,LCD,Monitor,ASUS VG236H BK/2MS/NA 120Hz Widescreen TN-Panel LCD Monitor with 3D-Vision Set Performance Review

3d vision cuts performance by a lot, about 40% in most games.

 

[Cilus]

^^ What are you trying to say? First of all it does not justify anything about your misleading info posting that AMD does not support 3D gaming. It is actually strengthening each and every point I've said.
It uses the same series of Asus monitor I've suggested and the review is pretty good about it.
2ndly it also specified that frame rate drop in 3D for the 3D optimized games are just 30% in case of single 480 and just 11% in case of SLI. The games which were performing poorer, it is mentioned on the review that the reason is that the titles are older and not at all optimized for 3D vision. MAFIA II and Lost Planet II, the two new titles with 3D Vision ready actually performed far better than the older title.

Here is the quote from the review itself:

However, when 3D Vision is enabled the video frame rate usually decrease by about 50%. This is no longer the hardfast rule, thanks to '3D Vision Ready' game titles that offer performance optimizations. Mafia II proved that the 3D Vision performance penalty can be as little as 30% with a single GeForce GTX 480 video card, or a mere 11% in SLI configuration

So in case of a GTX 580, the loss will be lesser than 30% due to his higher processing power.

Again and again you are just posting garbage data, irrelevant of anything. I think we need to design a garbage collector for DIGIT forum. Buddy, at least read what you are posting. For processing the 2 images there is a overhead but it does not drop the performance by 50% or half. It is like 1/3 or 1/4 th at maximum. Did I say that it is free of performance hit in any of my post? If I said then please point me. All I said is the performance hit will not be such will will reduce the performance to 50% of the original and your review is basically proving every bit of my word.

You need to understand a bit more.