• Hello MLAers! We've re-enabled auto-approval for accounts. If you are still waiting on account approval, please check this thread for more information.

Macworld Nubus video card tests, 1994

Written in haste. I've been looking for a good analysis like this for years, and stumbled across it just last night. To continue:

The interesting thing about this piece is that Macworld basically showed that the speed improvement gained by use of high end nubus cards over the stock video on a Centris 650 (same as Quadra 650 and so Quadra 800, and more or less all the high end 1994-era 68k hardware) was not spectacular in real usage scenarios, mainly because many of the functions users wanted accelerated were not dependent on QuickDraw, but rather, on routines in the proprietary software used. QuickDraw-specific functions were much improved, however, so scrolling was faster and panning smoother, but not much else.

The real benefit was seen in the user's ability to use large screens in 24-bit colour, at the same speeds as stock machines of the period provided using the typical smaller monitors of the period.

The piece deals only with the high end 68040 machines of the day. Presumably the situation would be different on a 68030, which would be where a good Nubus card would surely make much more of a difference. Also, no mention of gaming is made in the piece, which is significant. At that time, people (ordinary people, at any rate) didn't really lay out a fortune on a nubus video card for the sake of gaming, but for the purpose of time and money efficiencies in business and especially publishing.

Useful fodder, I think.

Trash, wherefore art thou?

 
Last edited by a moderator:
edit: :lol: apparently typing at exactly the same time as you were! You beat me to the send button. I hope my review of the article is helpful to you. BTW Centris 650 was 68LC040, the others you mentioned were full blown 68040s, so their FPU would have helped performance significantly.

Dunno, I don't think the persons responsible for the flashy title of the article and page layout ever read the article.

It was fatally flawed from the get go, saying up front that all testing was done at 16" resolution. Results were skewed to let average users make an informed decision in making a jump to that set of parameters. The reader was asked to extrapolate gains to be had by going from 14"(?) to 16" resolution for making the jump to 21" resolution, which was assumed to be out of their budget.

They never tested for the needs of the headlined user.

Did get something out of it though! If I click on the scroll bar (and I'm assuming the PageNav keys) instead of using the hand tool when moving around in a big graphic I'll be better using QuickDraw acceleration to its full potential. [:)] ]'>

 
Last edited by a moderator:
Also, no mention of gaming is made in the piece, which is significant. At that time, people (ordinary people, at any rate) didn't really lay out a fortune on a nubus video card for the sake of gaming, but for the purpose of time and money efficiencies in business and especially publishing.
For gaming conclusions, just head over to the flawed article over on lowendmac comparing the speed of NuBus QuickDraw acceleration to the Mac's built in video performance.

Built in video will always beat NuBus (IMO based on my overview of the situation) for gaming and average use. Again, it's a question of the user's needs. Games of the period were intentionally designed for high performance on lowendmacs. RoadApples to Apples comparisons. Duh! ::)

 
Those cards were made for high resolution 24 bit video work and some offered stuff like Gworld RAM and DSP acceleration for audio and graphics work.

Its kind of like comparing todays built in video to the latests Geforce cards today, outside of gaming or CAD 99% of what we do does not need the extra horsepower.

 
As we are on the subject, and as I've not seen it addressed, there were zoom features on some of these cards, generally invoked by a keystroke combination. Were they particularly prized, and for what reason? Cleaning up pixels manually rather than by filter? Or were there other uses?

 
One thing I read of late (perhaps on df) was that the Mac could pick which particular 256 or thousands of colors it was displaying. Onboard video was probably fine for almost anybody except for professional graphic artists, photographers hopping on the digital train early, etc. 256 colors renders most games fine, and the 2-meg-VRAM Quadras (900, 950, 840, IIRC also the 800 and 650) can do thousands all the way up to the 21-inch resolution.

I don't have time to go try to assemble shopping lists, but I do wonder based on certain targets if there were situations where it was more efficient to budget for an 840 or a 950 and a 16-inch display, compared with, say, a 610 or 650 and a video card and a 16-inch display to achieve millions at that resolution.

Of course, MacWorld's assertion here is essentially that if you're doing anything with graphics at all, 24-bit/millions is a must, and that's kind of interesting considering the financial impact that would have.

Very neat read-through though.

 
Back
Top