Why my Parhelia review isn't up

B

ben6

Regular
Matrox sent the board in time , by normal means (UPS Next Day) for the board to arrive on Thursday/Friday. Unfortunately, I received the Parhelia on Monday in the afternoon. Not wanting to run a few benchmarks , shoot a screenshot or two and call it a review , I decided to spend some time with the board instead of making the NDA lift date. It's unfortunate that US Customs held the board, as I had planned to spend the weekend with it, but them's the breaks...

I hate missing a NDA lift date, but there was no help to it...
 
Ben

you would do far more service to do a quality review on this than rushing it out to hit a date. I am glad you waited as your reviews have been good.
 
I agree, I'd much rather wait a week for a great review than have a quick one ready when the NDA is lifted.
 
Ben 6
Please use Il2-Sturmovik for a good comparison of the Anistropic and AA qualities of the card. If there ever was an underdog for system chugger of the year, Il2 takes it. And i would love to see what the scenery from way up high might look like. Esp over the cities.. :)
cough *berlin* cough..
:D
-rip
 
Hrm well I'm having a similar issue as Anand did with their card . Here's the list of games I intend to show screenshots with in the review

IL-2 Sturmovik (heh yes it's on the list ripvanwinkle)
Morrowind
Dungeon Siege
Jane's Attack Squadron
NFS High Stakes
NeverWinter Nights
SOF2
JK2 JO
FS2002
GTA3
Madden 2002
Final Fantasy VIII

whew I think that about covers most genres . I may remove one or two games as that's a lot of work :)
 
Nice to see NFS 4: High Stakes on that list... :) Some time ago I had replay file for benchmarking (with every 3rd party eyecandy available
set up for the 4 laps race.) but I am not sure if I still have it..

Though, those who are still playing it are using all kind of tweaks and no one really uses original EA cars, because those are so unrealistic. (EA Cars: 700-1000 polygons, usual nowadays 3rd party cars: 2500-5000 polygons, but even over 10 Kilopoly cars have been seen.)

for example, using EA Rash driver from NFS PU, gets about 35-60 % boost to framerates.
 
I think Morrowind, Dungeon Siege, and either JK 2 OR SOF 2 are the most important. Also, I'm not sure which of IL-2 Sturmovik, FS 2002, or Jane's differs most from standard "FPS" games in demands to be more informative, so I don't have an opinion.

You didn't ASK for an opinion, but I'm offering one as to what I think is most important for analysis anyways...you can just skip the post if it isn't welcome. ;)

Morrowind because of my uninformed impression of shader usage. There is the water, of course...and I haven't seen an Ash storm yet, or any type of sand storm or extreme weather (just rain) but I don't think anything else out right now stresses cards in quite the same way.

Dungeon Siege because it atleast seems to be DX 8+ oriented, though I haven't noticed anything drastic difference from my 7200 in my very brief revisit to it.

JK2 or SOF2 because they are both advance Q3 engines, though I think JK2 has been established to have some weak texturing for IQ analysis, based on a vague impression on discussions I've observed.

A flight simulator of some type because of the different demands it places on the card, and the opportunity to stress the new AA implementation on the Parhelia in interesting ways.

I think the rest of the games are low priority (but I May Be Wrong), and think you should include benchmarks for the above if possible.

It doesn't hurt to ask someone to provide the analysis you want when you don't have to do the work yourself (and don't have access to the card anyways), so that's my excuse. :LOL:
 
I should state that my benchmarks will be different than my gameplay though there will be some overlap , benchmarks:

3DMark2001SE
Fablemark (got tired of Villagemark too :D)
Comanche4
Dungeon Siege
Jedi Knight II Jedi Outcast
Aquanox
Serious Sam 2
ChameonMark
Shark Mark
 
Comanche4, besides the entire question of detecting pixel shaders on non-nVidia cards, just doesn't seem to me to be a very useful benchmark of a graphics card, even though it is all the rage. If you have time, and inclination, and a differing opinion, I would appreciate an education as to why it is. :-?

You are going to give a rough indication of fps ranges in the "non benchmark" games you run, correct? Well, my opinion list still stands and for the same reasons.

I'll also add the motogp demo that someone mentioned somewhere today as an interesting image quality analysis tool, even if used briefly...I find the portions of the course going nearly straight north or south a possibly interesting stress test of texture filtering, especially with the texture sharpness turned up, due to the lighting used on the track. Also, the pixel and vertex shader options, and the mysterious (to me) "Full screen special effects" Anti-aliasing option (which seems to be related to a speed blur effect, at a guess, and does seem to do good anti-aliasing as well) have to be good for investigating SOMETHING about the Parhelia's image quality.
 
My choices on games are mainly based upon games that I have played recnently and are on my hard drive right now. It wouldn't make much sense for me to use games I don't play. If you have an issue with the games I chose, that's fine, but I tried to get a mix in the order that's important to ME:

1. RPGs . Man I love RPGs Morrowind , Nvverwinter Nights and Dungeon Siege are among the best of this genre in years imho. Finsl zg
2. Flight Simultors combat flight sims- AA is VERY important FS2002 is Microsoft's latest entry
3. Racing games (perhaps I'll add Nascar 2002) - AA is important
4. FPS- AA is less important, however JK2 JO is one of the best of the recent bunch and SOF2 was just recently released
5. Sports Games I dunno I just like Madden 2002.
6. GTA3 why ? Because I loved this game on the PS2 and love it equally on the PC though I've had numerous problems on my 8500 because of a fog issue that's now fixed with a patch


As to Comanche4 as a benchmark, it's honestly mostly CPU limited. However, I don't base my opinion on a single benchmark or two as a couple of the reviews today have done..

I'm trying to do something a little different. Will I get negative responses/positive responses for the choices I make? Sure, but you get that whenever you write a review or state a opinion.

I'm sure my opinions will be equally criticized and I welcome it.
 
So so long as ATI is included I'll be a happy camper. . .

I'd politely suggest you post ATIs aniso' performance/quality comparisons to ensure a fair overall picture. Of course I'm sure you where going to but with the likes of Toms not posting them I thought I'd better drop the hint.

:LOL:
 
Wasn't trying to dump on you, and I hope you understand that, and the reference to "you can just skip the post if it isn't welcome" is made to indicate that I won't throw a fit if you don't follow my gospel, ;). I'm just providing my input, and, yes, it is "criticism", but it isn't meant to be as negative as many people interpret the word. Basically, anything that distances itself from the brain dead, distorted, and incomplete reviews that seem to be the curse of new card launches will make me a happier person.
 
It seems to have been benchmarked almost to death now. The conclusion must be - for gamers or game developers - dissappointing.

This might just be a perception issue. But I feel this card can't really claim to be a gamers card nor a FSAA card in its present form given its demonstrated issues.

Why not analyse what went wrong and why the new kid on the block isn't a demon? In 2-3 months this card will have to compete against R300 and NV30 - is it going to look like a joke next to those cards?
 
Final Fantasy VIII
Click to expand...

Marvellous!
Fantastic!
I realise that I might be approaching uniqueness with my weird obsession with Final Fantasy (which obviously is strong enough to lure me out of the lurking shadows), but nevertheless I find it most pleasant when someone even considers using it in a review. I regard such games - slow-moving ones where you cannot choose arbitrary resolutions - very AA-friendly, whereas raw speed often is of lesser importance. Moreover, I often find myself playing them...

So, even if Final Fantasy is omitted in the review, it is encouraging that it was even thought of.
 
No matter which game you decide to benchmark I think that it is key to log/plot FPS so that everyone can see how low the frames goes when the card is really pushed.

Check the tech-report for the use of Serious Sam SE to plot the FPS:

http://www.tech-report.com/reviews/2002q2/parhelia/index.x?pg=5

ssam-1024.gif


BTW: It's kind of interesting to see that the GF4 ATI 8500 and Parhelia shows the same pattern. Are we somewhat CPU limited here?
 
IL-2 Sturmovik rocks. Sort of off-topic question: How do you go about benchmarking it? I looked all over, and couldn't find any definate way to benchmark it. I got the demos that you can bench, but how do I get it to start and stuff? Any help would be most appreciated. :)
 
ben, is there some way to change mem/core clocks (powerstrip). I would definitly like to see how much parhelia really profits from its wide mem bus, so it would be great if you could include a graph that shows relative performance degradations in regards to (don`t know the correct english term) lowered mem clock. While you're at it, some comments about its overclocking caps would be appreciated.
 
You must log in or register to reply here.

Similar threads

S
Why Does My Heart Feel So Bad
2 3
Replies
42
Views
7K
Sage
S
Back
Top