Matrox on the verge of collapse?

[rubank]

By no means, but I don´t know of any other sites that do this kind of test. And I do constantly point out "according to tecchannel".
Subjective comments aren´t really the same thing, is it.

BTW, I´d really like to see close-up photos of actual screen output, at various resolutions, but noone seems to do that.

 

[Mephisto]

Does the claims of ONE site with ONE card invalidate the rest of the available data?

Yes, because this is the only site that measures the raw data with expensive equipment, all the other sites only give biased subjective OPINIONS.

 

[T2k]

Does the claims of ONE site with ONE card invalidate the rest of the available data?

Yes, because this is the only site that measures the raw data with expensive equipment, all the other sites only give biased subjective OPINIONS.

Very strong arguments, indeed.

(Do you feel it necessary to proclaim this kind of sweeping statements?)

BTW, do you own this product?

 

[Joe DeFuria]

Yes, because this is the only site that measures the raw data with expensive equipment, all the other sites only give biased subjective OPINIONS.

I don't really want to get into the middle of this, but I do have a relevant comment:

When it comes to judging image quality, sometimes "all the expensive and objective equipment in the world" doesn't give as accurate a representation of the results as lots of subjective opinions.

 

[arjan de lumens]

"Subjective opinions" are mostly useful in blind-tests - if people know what card's output they are looking at, there may be bias present (which could favour e.g. the model they already own, the newer or more expensive card, or the Matrox card etc), which is hard to correct for.

 

[Ty]

Double-blind would definately be the way to go but can't the monitor (brand/model) make a difference as well where one monitor might make one card look better than another?

 

[arjan de lumens]

Monitors can indeed vary noticeably in quality even with the same model - to correct for this, you need to, when doing the double-blind test, also swap the monitors randomly between the cards under test as well.

Another issue as well: it appears that some of the lesser cards try to boost the signal levels supplied to the monitor, making the picture brighter than it should be - this could possibly affect "subjective opinion" the same way that "subjective opinion" tends to consider the louder of two stereo sets to be the better one ..?

 

[Tagrineth]

Monitors can indeed vary noticeably in quality even with the same model - to correct for this, you need to, when doing the double-blind test, also swap the monitors randomly between the cards under test as well.

Another issue as well: it appears that some of the lesser cards try to boost the signal levels supplied to the monitor, making the picture brighter than it should be - this could possibly affect "subjective opinion" the same way that "subjective opinion" tends to consider the louder of two stereo sets to be the better one ..?

Well, for one, 3dfx doesn't do it on purpose - they were the ones who were 'RIGHT' about gamma levels, everyone else got it 'wrong'... oddly enough

 

[Mephisto]

When it comes to judging image quality, sometimes "all the expensive and objective equipment in the world" doesn't give as accurate a representation of the results as lots of subjective opinions.

By reading reviews, I often get the impression the reviewer only echoes something from hearsay, so he heard that Matrox boards offer good 2D image quality, so what he sees on his 100 USD 17" discount monitor must be good image quality then. He heard T&L is good, and therefore a card is bad because there is no T&L, Multisampling is good and therefore ... and so on.

I definitly prefer raw data measurements therefore which is independent of the monitor (normaly I don't own the same monitor as the reviewer does, do you?) and unbiased. Or do you argue a sinus signal with different voltages for RGB can result in a better image than what a G550 outputs?