[H]OCP does Radeon9700 I.Q....

Reverend said:
Nah, PNG isn't even required... just be gentle on the compression ratio for JPGs !

Saving a BMP or TGA screenshot as PNG provides better quality than a JPG even when no compression is used though.
 
Completely *FALSE*.... The gama correction is going to be different on *Each* system. therefore getting the *best* results. I cant belive none of you called the guy on a statement like this.

We didn't "call him" on it, because I believe he is correct. You misunderstood his point. It is precisely because gamma profiles can be differnt on each system, that this "screenshot" issue can be real. (He's talking about viewing SCREENSHOTS.) If a person takes a screen shot on system 1 (which has a specific gamma correction profile), and that screenshot is viewed on System 2 (which has a DIFFERENT gamma correction profile), then the shot that was taken on system 1, but VIEWED on system 2, might look worse than a shot taken on system 3 (without gamma correction) and viewed on system 2.

But the shot taken on System 1, and viewed on System 1 should always look better than that shot taken on System 3, viewed on system 3. (Though "look better" is probably not a good word to use: "be more accurate" would be more appropriate.)

There is a clear difference between people who are havign problems and are objectively, and constructively posting about it.. and those with an *agenda*.

I agree. However, there is also something to be said of people who seem to read "negativity" into other people's posts where it really doesn't exist. One might say they have an agenda as well.
 
Doomtrooper said:
Maybe its hard for people to understand here what I believe I'm trying to point out..the 2X AA 4600 shot is not even close, the 2X 9700 and 4X 4600 shots are actually very close on the vertical pillars and is certainly a MUCH better comparison in IQ than the 2X 4600.

I will remove the 4X shot to point out the obvious.

Um, Doom, the "2x" shot on the 4600 is using no FSAA at all.

In addition, I seriously doubt that the Radeon's 2x is better than the GF4's 2x, unless ATI finally fixed their "adaptive" AA sampling algorithm. I haven't seen any evidence of this yet.
 
I have 40mb of webspace at the University, I would be happy to host the uncompressed images if that helps anyone.

Good stuff, thanks Brent.
 
Chalnoth said:
In addition, I seriously doubt that the Radeon's 2x is better than the GF4's 2x, unless ATI finally fixed their "adaptive" AA sampling algorithm.
The gamma correction can make a huge difference in some cases.
 
Bigus Dickus said:
Well, 9700's 4x looks better to me than the GF4's 4x, so what reason is there to believe that 2x isn't similar?
The Radeon 9700 uses a rotated grid for 2x and 4x AA modes, but the GeForce 4 only uses a rotated grid in 2x mode. As I said above, the difference between the GeForce 4 and Radeon 9700 in 2x AA is the gamma correction, which can make a huge difference. Our 4x AA with the rotated grid and gamma correction looks very nice as well. 6x AA with gamma correction... you get the picture :)
 
OpenGL guy said:
The gamma correction can make a huge difference in some cases.

What, precisely, does gamma correction have to do with this?

That is, are you claiming that the Radeon 9700 has higher-precision gamma correction than the GeForce4? And what does gamma correction have to do with FSAA?
 
How bright an image is has absolutely nothing to do with image quality, and brighter is most certainly not always better.

Color separation is the key.

And, most importantly, the brightness of an image doesn't affect edge AA in the least...nor should it have anything to do with any anti-aliasing.
 
Chalnoth said:
OpenGL guy said:
The gamma correction can make a huge difference in some cases.

What, precisely, does gamma correction have to do with this?

That is, are you claiming that the Radeon 9700 has higher-precision gamma correction than the GeForce4? And what does gamma correction have to do with FSAA?

I think the gamma correction referred to is in the blended pixels from anti-aliasing. I think the 9700 has a new feature wherein it weights the pixels according to gamma correction before it blends them, rather than just blending them linearly. This is supposed to make the intermediate colors from AA more natural, visually.

Although Chalnoth's skepticism is, of course, valuable.
 
gah. It was a facetious remark.

TV manufacturers ship their TVs with the brightness and saturation pushed, because the human psyche responds to bright and saturated in the showroom with their dollars.

Regardless of whether its 'correct' or not, brightness sells.

And that was the point of my stupid little remark. Put a winky face next to it and maybe it makes more sense.
 
Chalnoth said:
In addition, I seriously doubt that the Radeon's 2x is better than the GF4's 2x, unless ATI finally fixed their "adaptive" AA sampling algorithm. I haven't seen any evidence of this yet.

Ummm, I'm pretty sure they have actually.

As for the gamma stuff, read the review.
 
antlers4 said:
I thing the gamme correction referred to is in the blended pixels from anti-aliasing. I think the 9700 has a new feature wherein it weights the pixels according to gamma correction before it blends them, rather than just blending them linearly. This is supposed to make the intermediate colors from AA more natural, visually.
This is exactly it.
Although Chalnoth's skepticism is, of course, valuable.
Skepticism to the point of idiocy is not.
 
I'm not sure if anyone else has noticed this, but why in Microsoft Flight sim does the Radeon 9700 has lots of shadowing that's not in either 8500 or GF4 pictures? Was a setting changed?
 
I think the reaction to my post is just what I expected. :LOL:

Both sides' fans has a hard time supporting half of my post while calling me a liar on the other half of it.

I have to have a few clarifications though.

Xmas said:
Then obviously we're not looking at the same pictures.

Actually we're not.
I based my comment on my own experiences, mostly with our own engine.
I just looked at the pictures at [H] but they are all way blurry and compressed so hard they are not suitable for IQ comparsion at all.

I didn't say 2x AA on R9700 is better than 4x AA on GF4 in all respects.
Having 4 samples has obvious advantages at a lot of cases.
On the other hand there are lot of cases where 4x AA of GF4 doesn't look better than 2xAA GF4.
And 2x AA of R9700 is beating 2x AA of GF4.

But I probably shouldn't said it this way, or at all.

Hellbinder[CE said:
]Hyp-X..

Dont even go there...... You may be having issues with your specific hardware config.. maybe you are trying to OC... or some other issue.... but the Drivers are Stable and have few issues for 99% of the people. I do *NOT* like the insinuation that the drivers are crap or they are to blame.

It is quite clear from reading several forums that there are a number of people out there going out of their way to exagerate, overstate..... and imo... intentionally trying to view the 9700 in a bad light. There is a clear difference between people who are havign problems and are objectively, and constructively posting about it.. and those with an *agenda*.

Hmm.

1. I do not overclock.

2. The computer is stable with a GF4Ti4200

3. The computer locks hard in our engine within a minute
I read (and constantly do) the official support forum at rage3d, and a lot of people has the same problem on different configs with different games.
Support people admitted that the problem is in the drivers, and will be fixed in a future release of it.

4. I can't work if I can't test out program, and I can't test it if it locks the system.

I don't think I have stated anything untrue.
I have also stated I can't wait to put the card back.

btw, do you have an R9700? According to your sig, you run a Ti4200 ... :LOL:
 
McElvis said:
I'm not sure if anyone else has noticed this, but why in Microsoft Flight sim does the Radeon 9700 has lots of shadowing that's not in either 8500 or GF4 pictures? Was a setting changed?

I saw that too. It looks stupid.
 
Back
Top