marshall@wind55.seri.gov (Marshall L. Buhl) (11/01/90)
root@shawn.uucp (0000-Admin(0000)) writes: >What's the REAL difference between using a fixed or a multisync? You can't run at higer resolution with programs like Windows or CAD programs if you go fixed frequency. Many folks have SVGA cards they run at 640x480 for normal DOS applications but can be boosted to 800x600 or 1024x768 when running Windows or other specialty programs. If all you run are character based applications like 1-2-3 and WordImperfect, save your money. If you run Windows, the high resolution really helps you get more information onto the screen. -- Marshall L. Buhl, Jr. EMAIL: marshall@seri.gov Senior Computer Missionary VOICE: (303)231-1014 Wind Research Branch 1617 Cole Blvd., Golden, CO 80401-3393 Solar Energy Research Institute Solar - safe energy for a healthy future
medici@dorm.rutgers.edu (Mark Medici) (11/21/90)
Fixed frequency monitors are seldom used anymore. Fixed "standard" monitors, however, are still useful. The difference is a fixed frequency monitor will handle only a single horizontal and vertical frequency. EGA, VGA and above require variable frequency monitors, though these are not generally referred to as "multisync" monitors. EGA requires 15.75kHz (or thereabouts) horizontal and 60Hz vertical sync frequencies. EGA uses 22kHz horizontal, but also needs 15.75kHz to display CGA compatible graphics (it steps down to the CGA spec). VGA monitors have a fixed horizontal frequency of about 31kHz, but a variable vertical frequency from 60 to 90Hz (if memory serves correctly). Most SuperVGA systems require 31kHz and 35kHz horizontal and 50 to 90Hz vertical frequencies. These are not normal ranges for single standard VGA monitors, but are for so called "multisync" monitors. Whether you need a multisync type monitor depends on what video card you are using now, or plan to use in the near future. Experience has proven (at least to me) that purchasing a multisync monitor as insurance for future compatibility is not a good investment. Monitor manufacturers do not drive the video standards -- IBM does with few exceptions (dispute this if you will, but first look at the history). Multisync monitors purchased in the EGA era did not do a good job of displaying VGA. The ones purchased early in the VGA era did not do a good job of displaying SVGA. I expect that the ones available now may do a good job at XGA, but probably not at whatever comes next (Super-XGA?). So my advice is to determine what graphics system you will be buying in the next 12 months (assuming you haven't already bought one) and select the lowest cost monitor that can clearly display that standard. Before you select your monitor, you should DEFINATELY go to a dealer or collegue that has one, and use it for a while. Do not rely on specifications, adds, or popularity when choosing your monitor. Your own eyes will tell you which one is best for your work. -- ---------------------------------------------------------------------------- Mark Medici/SysProg3 * Rutgers University/CCIS * medici@elbereth.rutgers.edu ----------------------------------------------------------------------------
medici@dorm.rutgers.edu (Mark Medici) (11/22/90)
I believe this information may help clarify my previous artical. In mail message of 20-Nov-90, Michael Kersenbrock <michaelk@copper.wr.tek.com> writes: > In article <Nov.20.17.41.51.1990.7352@dorm.rutgers.edu> you write: > >Fixed frequency monitors are seldom used anymore. Fixed "standard" > >monitors, however, are still useful. The difference is a fixed > >frequency monitor will handle only a single horizontal and vertical > >frequency. EGA, VGA and above require variable frequency monitors, > >though these are not generally referred to as "multisync" monitors. > > > >EGA requires 15.75kHz (or thereabouts) horizontal and 60Hz vertical > >sync frequencies. EGA uses 22kHz horizontal, but also needs 15.75kHz > >to display CGA compatible graphics (it steps down to the CGA spec). > >VGA monitors have a fixed horizontal frequency of about 31kHz, but a > >variable vertical frequency from 60 to 90Hz (if memory serves > >correctly). > > I think you are technically correct, but miss one important thing. Most > all (actually, 100% all, but I'm sure there are ones that don't) of the > graphic boards that I've looked at (a good number, I'm in the market for > one) "do" EGA, CGA, VGA (and some, hercules) modes ALL on a fixed > frequency VGA (31Khz) monitor. The boards "simulate" the modes as such. > This is why having a fixed freqency monitor is still very useful -- if > it's a VGA monitor. > > Your other comments are quite good. I personally am looking at > dual-frequency monitors 31/35Khz ones, where the 31Khz covers applications > that expect CGA through regular-VGA, and 35Khz handles 800x600 and a > flickering 1K x 768 interlace. I'll leave XGA types for next time. > I really don't want to get into a debate over semantics. However, your assertion that 31kHz VGA monitors are fixed frequency is not correct. Please refer to the appropriate IBM technical reference or specifications for VGA Only monitors. You will find that, while the horizontal frequency may be fixed at 31kHz, the vertical sweep frequency must be variable from 90Hz (for 350 line resolution) to 60Hz (for 480 line resolution). Why can't a single frequency be used for all display resolutions? The number of pixels to be drawn, the rate at which the pixel information is transmitted from the computer to the monitor, the speed of the monitor's electronics to act on the signals, and the speed of the phosphers act in concert to determine horizontal and vertical sweep frequencies. There are limits (technical and financial) on how fast the pixel information signals (dot clock rate) can be transmitted to and acted upon by the monitor. And, of course, there are practical limits to how wide a range the display controller and monitor can accept. Generally speaking, displaying more horizontal pixels on the monitor requires a faster dot clock and horizontal sweep; otherwise the pixels would extend beyond the edges of the viewable CRT area. The vertical frequency is determined by multiplying the time to display each hori- zontal line (plus flyback and setup time) by the number of horizontal lines displayed. If the horizontal frequency is fixed, the vertical frequency MUST vary to display different vertical frequencies. Likewise, if the vertical frequency is fixed (as is in the case of straight EGA monitors), the horizontal frequency MUST be variable (EGA switches horz. freq. between 15.75 and 22kHz). In the case of SVGA modes, both horizontal and verical frequencies are variable. Therefore, when running at 350 lines of resolution, a fixed horizontal frequency VGA monitor will have its refresh (vertical sweep frequency) at a faster rate, thus filling the screen. At 480 lines of resolu- tion, the refresh will be slower to fit more lines onto the screen. Since the VGA standard permits CGA, EGA and VGA resolutions, the horizontal resolution is fixed at 640 dots (or a value that can be evenly multiplied to result in 640, such as 320 and 160) so the horizontal frequency can also stay fixed. So, how can these multi-standard video cards work on a fixed 31kHz monitor? They do this by mapping the lower resolution to a higher resolution. For example, when displaying CGA they double each verti- cal line, providing a display resolution of 640x400, though the actual descernable resolution is still 640x200. To get 720 horizontal pixels, the adapter speeds its dot clock to pack more information into each horizontal scan at the same fixed frequency. However, the vertical frequency must be variable to permit 350, 400 and 480 lines of resolution defined as acceptable by the VGA standard. If the vertical frequency was fixed, the image height would shrink and grow with each change in vertical resolution (as happened on early monochrome adapters that offered CGA emulation). How do I know all this? I worked for Princeton Graphic Systems for two years. But, as that was over 3 years ago, I have forgotten some of the exact specifications. Nonetheless, the theory is still the same, and I have not completely ignored this area of the industry (old habits die hard). -- ---------------------------------------------------------------------------- Mark Medici/SysProg3 * Rutgers University/CCIS * medici@elbereth.rutgers.edu ----------------------------------------------------------------------------