PDA

View Full Version : CATALYST™ v3.8 available! Feedbacks here please!


Pages : 1 2 3 4 [5] 6 7

Sotos
10-09-03, 07:23 PM
Edit

Probably the reason that most users in Rage3d were disatisfied was not because the drivers were bad...

But because they just were OVERHYPED and expected a lot more

oldsk00l
10-09-03, 07:25 PM
I haven't seen anyone contest my point that a large part of the hype came from the community, so if you want to blame hype blame the users themselves. ATI didn't mislead in any way....but when ATI or nVidia say "this is our best software release to date" people expect more than what that actually means.

Sotos
10-09-03, 07:27 PM
Originally posted by oldsk00l
Sotos

I can answer some of that for you

Not the 1st question but 2 and 3 for sure.

2) Because there is new hardware for the R360 that allows dynamic clock changes by a thermal diode, this circuitry is not present on prior cards.

3) The 3.8's are only 1MB bigger than the 3.7's...are you on dialup? To me that's only %5 bigger......and it's very doubtful smartshader accounts for all that.

[edit]

Just a personal opinion, and a message to Terry,
If you guys had to add 15MB to the current size, to add really cool new features, by all means, PLEASE DO IT!!!!

PILE IT ON BABY!!! I CAN TAKE ALL YOU GOT!!!!!!!!!

For 2) Oww OK :)

For 3) Yes i am on dialup and i have a <only> ISDN connection and takes some time to download 25MB :)

oldsk00l
10-09-03, 07:30 PM
Ouch man

I could never go back....maybe that's a good suggestion....to have seperate downloads available?

I know that they do for JUST the drivers, and then the CP in a seperate DL.

I wonder, would it warrant seperating the drivers even further if the other features got much bigger?

Sotos
10-09-03, 07:33 PM
Originally posted by oldsk00l
I haven't seen anyone contest my point that a large part of the hype came from the community, so if you want to blame hype blame the users themselves. ATI didn't mislead in any way....but when ATI or nVidia say "this is our best software release to date" people expect more than what that actually means.

I was talking Hypotheticaly about the <Dissapointment> Poll at Rage3d

I am happy with the drivers as i said and of course the drivers are more important from the Control panel..

So even i was a little disapointmented

i am in general happy with these new drivers Stability and Image quality

oldsk00l
10-09-03, 07:38 PM
Honestly I find it ironic.

The community is reacting the same way it did back when nVidia first rolled out the "Detonators" for the old TNT's. nVidia's TNT2 Ultra opened up a can of whoop-ass on everyone with those sets. They utterly annihilated everyone, and don't anyone from ATI argue otherwise, Carmack, Sweeney, Drew Markham, everyone eventually settled nVidia took the crown....and the community said something to the effect to the Detonator 3's "I expected better, I'm dissappointed"

John Carmack at the time even stated that nVidia's OGL was considered "Gold standard"

So I dunno, "Good enough" would practically be a full game bundled in with the driver release :)

Mike89
10-09-03, 08:06 PM
I generally like the Rage3D forum but one thing I've found out. If you want to get an accurate assessment of any ATI driver, "DO NOT GO TO THE RAGE3D FORUM TO FIND IT". Jeez, those driver threads are comprised of a lot of whiners, trolls, and people would wouldn't know how to set up a computer if their life depended on it. Not to say there is not some good input there from people who do know what they're doing but there's way too much of the other. If I went strictly by those driver threads, I would never install any ATI driver and have my card here just to stare at.

Bottom line, these drivers are very nice (as all other ATI driver sets I have used) and I personally like the new Control Panel. ATI is all good as far as I'm concerned (and this is from a FORMER Nvidiot). ATI's built in refresh rate override works flawlessly (as it has since 3.1 and Nvidia still doesn't have it), flawless performance in all games I have played so far. My 9800 Pro likes em!

StealthHawk
10-09-03, 08:15 PM
Originally posted by CATALYST MAKER
Again be specific. I cant help unless I know exactly what you want. Telling me that you "expected better things" doesnt help me make it any better.

Please, please, please, make it so application profiles can be applied to certain .exe files! I want the ability to set video options for a game once and then forget about it, and have the options(FSAA, AF, vsync, etc) be automatically applied whenever I launch a game.

Keeping general profiles(I call the profiles that exist right now general profiles) would be good too. I can see some people have a use for that. I myself, along with others, would like application specific automatically applied per-launch profiles though.

Supposedly the tweaker Radlinker already offers such functionality, although I haven't used Radlinker myself.

ChrisRay
10-09-03, 08:25 PM
Originally posted by CATALYST MAKER
Be specific. What are these few things we need "fixed" (or is it things we just dont support that you want).
I hear Super Sampling (and we have said many times there is no real benefit for R300 and it would cost us a large amount to support it), and I hear saturation (but I feel our color control panel can do everything).

I want to help but please be specific as to what you want and why you want it.

Well. Since you want specifics I have few things that I might say. One Question is a possibility.


My One pet peeve with ATI's color controls. Being Saturated image. Since the only way to really deepen the values of these colors is to saturate them. For Example.

I tend to use a program called "CuteHTML". Which uses colors to discern HTML tags from each. Specifically red/green/orange being the most predominate. Saturating these colors makes a huge difference in there apearence when looking at them on the "fly". Now I admit this is a selfish reason to want saturated images. But surely you can see the need to increase color saturation of those level.

Now My Question. Deepening Colors Really isnt possible on ATIS Current configuration. This can only be done with indepentent software such as Coloriffic Or Monitor. ATI's current method of color changing. Tends to change the values of these colors. Adding shades of red to the gamma does not help increase the saturation of green for example.


Moving on. May I ask what you meant by "Costing you a large amount to support it?". My desire for Super Sampling is purely selfish. I understand that Super Sampling With Fog tends to create various issues. I "really" liked my ATi Radeon 8500's 4x/Quality Mode. IMO it looks better than the r300's 6x. And it's the best Looking Anti Aliasing I have ever seen. That Seen on ATis 4x/Quality mode.

So Am I understanding that would cost you alot in time? Or would cost you guys alot in benchmarks where people are too dumb to understand that its not a standard feature?

My Last Question. Is More of a question. How Come Anti Aliasing is not occuring on off rendered buffers? (pbuffers) Now I understand these buffers are not usually visible to the end user. This is an emulation question most importantly. (And note this is not just an ATI issue, so dont feel singled out on this) For example. The Pbuffer is used to store playstation primitives before they are finally rendered to the texture. This is done as a solution to the unfortunate slow framebuffer readback issue. Its essentially a render to texture trick. Currently ATI and NVidia cards dont kill this buffer. So my main question is. Is it even possible for AA to be applied to the offscreen pbuffer.

Mike89
10-09-03, 08:39 PM
Personally I can see why ATI chose not to try to implement supersampling. I just don't think there is wide wanted use of it for one reason and it comes down to being able to be used for Alpha textures vs the performance hit. I guess I could see having an option for it but I personally don't have any games where I would use it. I think it's the old, "the ends doesn't justify the means" thing. It's probably a very intensive thing for ATI to do and implement and probably would bring with it a whole new set of potential problems to have to deal with, and hardly anyone would really use it anyway to warrant the expense (and I'm not talking about just money).

ShoobyDooblyDoo
10-09-03, 08:56 PM
Originally posted by StealthHawk
...Supposedly the tweaker Radlinker already offers such functionality, although I haven't used Radlinker myself.

In brief, once Radlinker (http://www28.brinkster.com/chrisww1942//) is installed, it integrates itself into the Windows shell adding a "Create Radlink" right click option when a desktop shortcut is chosen. A Radlink of that shortcut is then created and it's inside this that you can select the very exact settings that each individual enjoys at the per-game level.

I have not found any other utility that can touch this for ease of use and for customisation. ATI really couldn't go wrong by implementing such features to future Catalyst revisions.

I have to add that the image quality is absolutely superb on the 3.8's :)

http://www.uploadit.org/files/101003-showperapp.jpg

If you haven't already, give it a whirl.

Kruno
10-09-03, 11:53 PM
Catalyst Maker, applying shaders in post process has loads of potential.

I'm sure not many people see it as a revolutionary feature but I am seriously looking forward to what you plan on doing with it.

Here is a B3D article on post process shaders:
http://www.beyond3d.com/articles/deflight/index.php?p=8

Interesting eh? ;)

(BTW: I'm not insinuating that you will implement SSAA, I know your stance on the issue and I will not push it nor will I talk to you about it)

StealthHawk
10-10-03, 12:10 AM
Something missing from that OGL Radlinker screenshot is the ability to force tripple buffer! (Long live tripple buffering)

ShoobyDooblyDoo
10-10-03, 05:15 AM
This is true. Then again development of Radlinker has ceased and the last build was prior to the triplebuffer option that ATI recently added to the driverset.

Kombatant
10-10-03, 06:33 AM
Originally posted by StealthHawk
Something missing from that OGL Radlinker screenshot is the ability to force tripple buffer! (Long live tripple buffering)

That's why you can enable it from within the control panel since the 3.7s :p

silence
10-10-03, 07:11 AM
Originally posted by saturnotaku
Some of us liked the old control panel better. Very clean and very effective. The new one seems...I dunno, not well thought out. I don't like it as much as the old panel.


ya, changing settings for dx and openGL is little bit confusing now....maybe not to ppl that already had ATi, but i see that those that will switch soon from NV will have more trouble setting up what they want....

i like new features....but IMO, u should have kept diff panels or merge them in some other way (joint AA&AF slider?) with only sfecific featurea that show when u click on link.....

something like.....sliders are there, but if u want to force tripple buffering in openGL, you click on openGL link/mark/tab/whatever....and openGL specific features appear under sliders......that would make more sense then this:)

i set both to 4xAA & 8xAF...so same slider might be good idea......currently its little bit complicated...

dunno if this is possible.....

sebazve
10-10-03, 03:15 PM
does this release fix the white textures on HALO?

i have a radeon 9100 and cats 3.7?
no probs with any other game

Red Dog
10-11-03, 12:38 AM
How does the 2xAA profile bug make it past the beta testers? I dont want to be rude, but it baffles me.

Smart Shader could be cool for making sig's.

Im glad the 16bit ogl bug is gone.

I wish ATI would incorporate something similar to coolbits.

I too thought this game profile feature would at least have some kind of application detection.

I have 2 monitors showing up as being connected, when there is only one.

Hype. Hype sux. None of the new features are earth shattering yet there was one month of hype and secrecy. ATI could have revealed the upcoming features weeks ago but it was more profitable to hype for a month. I hate when companies string us along with hype.

Nice, visible image quality increase with 3.8.

I still prefer 3.5's because I want 2xAA / 16x AF.


quote:
--------------------------------------------------------------------------------
Originally posted by oldsk00l
Another thing, ATI participates more in the nV oriented forums than, NV!!! Come on!!!
--------------------------------------------------------------------------------

That is a very good point oldsk00l.

john19055
10-11-03, 02:16 AM
Looking at Radge3d I thought there was something wrong with them,but after installing them they run everthing fine, the I.Q. is great and the 16b ogl bug is fixed,no problems.

Kruno
10-11-03, 02:54 AM
How does the 2xAA profile bug make it past the beta testers? I dont want to be rude, but it baffles me.


Funny as I don't have that bug.
The beta testers probably never got the bug either hence it was never reported.

Red Dog
10-11-03, 11:12 AM
What O.S. are you running?

Im on xp and I deff have the 2x bug, as do many others.

weird

Red Dog
10-11-03, 12:27 PM
I was thinking about it and if I had to characterize the catalyst drivers I would say, "Fix something, break something, fix something, break something."

last perfect set for me was 3.5

I skipped 3.6 and 3.7 due to the 16bit ogl bug.

I would give the cat team and its beta testers a 5 out of 10, but that is just my personal opinion. I find their work sloppy.

I expect minor bugs, but no 32bit textures in opengl and no 2xAA saved in the cp I consider to be show stoppers. I will prob revert back to 3.5.

I do appreciate ATI's image quality so I dont want to sound too critical. I am not comparing ATI's drivers to Nvidia's so dont go there.

I am thrilled with ATI's image quality, but otherwise their drivers leave alot to be desired.

Sotos
10-11-03, 08:06 PM
Originally posted by Red Dog
I was thinking about it and if I had to characterize the catalyst drivers I would say, "Fix something, break something, fix something, break something."

last perfect set for me was 3.5

I skipped 3.6 and 3.7 due to the 16bit ogl bug.

I would give the cat team and its beta testers a 5 out of 10, but that is just my personal opinion. I find their work sloppy.

I expect minor bugs, but no 32bit textures in opengl and no 2xAA saved in the cp I consider to be show stoppers. I will prob revert back to 3.5.

I do appreciate ATI's image quality so I dont want to sound too critical. I am not comparing ATI's drivers to Nvidia's so dont go there.

I am thrilled with ATI's image quality, but otherwise their drivers leave alot to be desired.

I do not have any problem

And the 2x bug is not serious to say that these drivers SUCKS

I have tested all my games and they run great with EXCELLENT image quality :)
So they are great drivers for me :)

Now the only thing that i do not like so much is the 3d settings of the new Control panel but the DRIVERS ROX

Red Dog
10-11-03, 08:41 PM
I never said they suck.

Any way, Ive had time this morning to mess around abit and I discovered I can enable 2xAA in conjunction with 16x using Rtool.

I absolutely love the IQ and performance of 3.8 but didnt want to give up 2x/16x. Now that Ive found a way around teh cp bug Im happy as a pig in crap.

Im happy to say I now think they are a solid 8.5 - 9 out of 10 and I will deffinately stay with them. A few more bugs squashed, a few cp layout tweaks and they will be a 10.

America's Army: Operations looks sooooo good @ 1280x960 2x/16x with everything cranked in game.

I you want some free eye candy and like shooters give AA a shot. Its a free after all.

Kruno
10-11-03, 11:20 PM
Red Dog I'm also using XP.
Like I said before, if the beta team didn't pick up the bug and report it the reason would most likely be because they didn't experience the bug.