VOGONS


Reply 60 of 137, by jwt27

User metadata
Rank Oldbie
Rank
Oldbie
philscomputerlab wrote:
This is an issue with 3dbench. There are too versions. […]
Show full quote
jwt27 wrote:

I just tried slowing down my slot 1 machine. 3dbench runs at about 1 frame per 3 seconds, took 15 minutes to finish, and yet still says 62.1 fps... uh, okay.

This is an issue with 3dbench. There are too versions.

1.0c is part of the 486+ benchmark suite, this is the one used in the VGA Benchmark database project. It works great for fast machines, and ok for fast 386 machines, but slow machines it glitches out.

Version 1.0 glitches out with fast machines and it counts over after 99.9 frames, but it's much more accurate with slow computers: http://www.philscomputerlab.com/386-benchmark-suite.html

15 minutes is a very long time. What processor did you use?

Tried the 3DBench from your link, but it says "Packed file is corrupt". I've seen this before (on perfectly good executables) but don't recall offhand what caused it.
I used a Katmai 550MHz, stock speed, with Throttle at max and cpucache off. And then I still have the option to underclock the whole thing to 133MHz...

elianda wrote:

More interesting for benchmarks would be Topbench as it benches the CPU, memory and graphics subsystem independently. Usually this becomes quite unbalanced when slowing down a faster PCI based system compared to a 386 with ISA.

Nnd0dnV.jpg

Seems pretty consistent, total speed varies between about 4900-5300µs. Using acpi throttle without disabling cache is much less stable, total time keeps jumping up and down by about two orders of magnitude.

Reply 61 of 137, by PARUS

User metadata
Rank Oldbie
Rank
Oldbie
PARUS wrote:
Signal Lines Connection. ----------- ------- --------- GUS Ace MT-32 64Gold ----------- ------- --spdif out | […]
Show full quote

Signal Lines Connection.
----------- ------- ---------
GUS Ace MT-32 64Gold
----------- ------- --spdif out
| |
V V
analog in
--------------------spdif in
EWS64XL
--------------------spdif out
| s/pdif
V
--------spdif in
AV310
--------spdif out
| s/pdif
V
Digital Receiver DAC Amplifier.

This is my best DOS sound system.

This text is uneasy for looking and reading. It's better to see this picture:

059920975d2163c8d0a48bf763c6c902.png
That's right.

Reply 62 of 137, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
brostenen wrote:

Games released from 1987 to 1989 was mostly Amiga's first. Only a handfull of games, were PC only. I think Settlers was the first that the other way around. Being that the Amiga-port was a conversion of the PC edition.

One of the 1st IBM PC games ported to Amiga was King's Quest (1984), appeared on Amiga in 1986-87.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 63 of 137, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
dr.zeissler wrote:
please test this for me: […]
Show full quote

please test this for me:

- Does "Bodyblows" work with sound and no issues on the K6-III+ ?
- Does "Desert-Strike" run without problems?
- What about "pinball illusions"
- Has JazzJack2 smooth Hardwarescrolling with 640x480 or with 640x400 ?

That email didn't come through, but I managed to download and test some more games.

Desert Strike works fine. There is an error in Pentium mode, but in 486 mode the game runs fine. Nice graphics and didn't know the PC had such a game.

Also tried Bodyblows. There is music and digital sound effects (punches, screams, voices). Also works fine. Pinball Illusions I confirmed working earlier. I couldn't find this Jazz game, but I can already see that most games will work just fine with the cache method. I found the same on my Pentium. Ultima 7 is an exception as it enables caches when you start it 😀

I think that the issues you might have had are not so much to do with the processor, but not trying all the speed options / memory options or some other miss-configuration of the system.

YouTube, Facebook, Website

Reply 64 of 137, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
philscomputerlab wrote:

I couldn't find this Jazz game

Jazz Jackrabbit 2

most games will work just fine with the cache method

The aim was all games of peroid to work fine, with possible very low percent of exceptions. To understand this confidently it needs thorough testing. Or you may get say 20% of games wich will work incorrectly, but by testing only 10 popular titles it was not seen.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 65 of 137, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
Tertz wrote:

The aim was all games of peroid to work fine, with possible very low percent of exceptions. To understand this confidently it needs thorough testing. Or you may get say 20% of games wich will work incorrectly, but by testing only 10 popular titles it was not seen.

When I started this project, the goal was to play as many games possible, from as large time range as possible. I have found the cache method to be extremely compatible, apart from Ultima 7, I haven't run into a game that wouldn't work. These games I recently tried was because another forum member mentioned them, I assume he had issues with them on his computer but often these issues can be related to memory configuration or other aspects of a system.

Obviously popular titles will be played by most. Some of these games like desert strike or body blows, I haven't even heard of 😀 But anyway, they work fine. I'm happy to test other games, but please zip them up and provide them as I have limited time and work on other projects as well.

YouTube, Facebook, Website

Reply 66 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t

Desert Strike and Jungle Strike run fine on a PII-333 without changing anything.
With a K6-III+ the Game does not start, even with 2x66 Multi, disabling the CPU-Cache
makes it playable. The PII-333 has so much more Power, but it runs just fine.

Lotus III is the same, run just fine on PII, on K6-III+ not without disabling the caches.

Have you looked at the AmigaVersion of BodyBlows ? It's by FAR better then the PC-Port.
The GFX and the Sound are extremly good on the Amiga, much better then SB/VGA on the PC.

Retro-Gamer 😀 ...on different machines

Reply 67 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t

Jazz Jack Rabbit 1 has some issues. It does not start on a PII-333, using Throttle and other slowdown tools cost the "smoothness" of the scrolling, so one point for the K6-III+

Retro-Gamer 😀 ...on different machines

Reply 68 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t

Jazz Jack Rabbit 2 is a Win9x Game. It's really great. It has a "Hardware-mode" for the Scrolling.
This can provide a really smooth scrolling like old platformers "e.g Giana on the C64/Amiga", but
it does not work with all resolutions and only with two cards I own.

The Scrolling in JazzJack2 is smooth under the following circumstances:

- "Hardware-Mode" chosen
- GFX-Card must be "Nvidia RIVA128 or 3dfx Voodoo3"
- Resolution has to be changed to 640x400

Please test this.
Doc

Retro-Gamer 😀 ...on different machines

Reply 70 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t

Yes, one of a lot of good reasons for using a voodoo3.

Yes, I did. With the use of Throttle 1-7 the Benchmarks run alway smooth and I did not see any weird things like "slow gfx" with "high results",
3d-bench goes down from round 200 to 20. Your benches lack of the "shiny-mark" for dos and win9x. a good indicator for the high-end dos-games
with pentium-machines.

Retro-Gamer 😀 ...on different machines

Reply 71 of 137, by brostenen

User metadata
Rank l33t++
Rank
l33t++
PARUS wrote:

This is my best DOS sound system.

Hmmm.... Does it run DynaBlaster with full sound?
The only way I can get that working, is to run a P133 on a SS7-MVP3 board, disable motherboard cache, disable all serial and parallel and have IRQ5 free and use IRQ7 for sound. When I load the GUS-ACE drivers or enable serial/parallel. The game will not detect the soundcard.

And the machine is running MS-Dos6.22. I had real quirks running that on a Celeron 233 on a 440ex board.
Swapping between different cards, using all from Aztech-ISA/Creative-SB16-ISA to Ensonic-PCI/Creative-PCI cards, eighter Dynablaster, Lotus-III or Syndicate would not play, or even worse. The games would crash. And that was in Win98SE and Win95-Osr2/3.
Anyway... That was Windows and not dos. MS-Dos is actually even worse, trying to get Dynablaster working, when a "Fast" cpu is chosen.

So... I still stick to Phil's 4 in 1 sollution, regarding the motherboard-platform of choice, and just using a different CPU. (Pentium 133 - NON-MMX)
Using a different soundcard (Creative SB16) and using a CL-5446-PCI GFX card.

Last edited by brostenen on 2015-07-25, 10:38. Edited 2 times in total.

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk
My YouTube: https://www.youtube.com/user/brostenen

001100 010010 011110 100001 101101 110011

Reply 72 of 137, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
dr.zeissler wrote:
Yes, one of a lot of good reasons for using a voodoo3. […]
Show full quote

Yes, one of a lot of good reasons for using a voodoo3.

Yes, I did. With the use of Throttle 1-7 the Benchmarks run alway smooth and I did not see any weird things like "slow gfx" with "high results",
3d-bench goes down from round 200 to 20. Your benches lack of the "shiny-mark" for dos and win9x. a good indicator for the high-end dos-games
with pentium-machines.

Without Throttle please. Just at full speed, and then with caches disabled. It shouldn't take long with the benchmark menu system.

YouTube, Facebook, Website

Reply 73 of 137, by carlostex

User metadata
Rank l33t
Rank
l33t

FIFA 94 is an example of a game that seems to have trouble with K6 CPU's. With a Pentium MMX runs fine. On a K6 it gives a Memory segmentation fault.

Reply 74 of 137, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
philscomputerlab wrote:

When I started this project, the goal was to play as many games possible, from as large time range as possible.

If there will appear a significant % of non-working games then people will need to use other systems simultaneously anyway, while the practical sense (besides fun) was in replacing many computers by a single. CPU and other parts may affect. In times of low-level programming could be used exotic algorithms, so more wide testing is desirably. I suppose after time the situation of practical use of such systems become more clear.

Upper end of your build I'd say 1998. Later 3D games may to work slow in SVGA, especially by today 30 fps standards. Would be good to have 2nd better 3D card near a hand, what's on your recomendations too. This may expand time limits with a little of stiring. Instead of 128, - 256 Mb is better for Win9x.
And you haven't said about LCD monitors you may prefer for your build.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 75 of 137, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
Tertz wrote:
If there will appear a significant % of non-working games then people will need to use other systems simultaneously anyway, whil […]
Show full quote
philscomputerlab wrote:

When I started this project, the goal was to play as many games possible, from as large time range as possible.

If there will appear a significant % of non-working games then people will need to use other systems simultaneously anyway, while the practical sense (besides fun) was in replacing many computers by a single. CPU and other parts may affect. In times of low-level programming could be used exotic algorithms, so more wide testing is desirably. I suppose after time the situation of practical use of such systems become more clear.

Upper end of your build I'd say 1998. Later 3D games may to work slow in SVGA, especially by today 30 fps standards. Would be good to have 2nd better 3D card near a hand, what's on your recomendations too. This may expand time limits with a little of stiring. Instead of 128, - 256 Mb is better for Win9x.
And you haven't said about LCD monitors you may prefer for your build.

A lot of users are using Socket 7 machines with cache disabled. It's a very popular method and many have stopped playing on 386/486 machines and use this method exclusively. Maybe you can contribute with testing non-popular games, it seems to interest you.

The graphics card is fine, the CPU power is what is holding back the performance in some of the later games. You could use another card to run at 1600 x 1200, but you won't be getting any higher frames.

Monitors I use are standard 17" or 19" LCD monitors. But I also have two 18.5" widescreen with 4:3 aspect ratio support. I've tried more RAM and it does nothing really to boost performance for these games.

You see a lot of these other topics I've covered through other projects and videos. Might be worth checking out my channel because there is just too much information to pack into one project.

YouTube, Facebook, Website

Reply 76 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t
philscomputerlab wrote:
dr.zeissler wrote:
Yes, one of a lot of good reasons for using a voodoo3. […]
Show full quote

Yes, one of a lot of good reasons for using a voodoo3.

Yes, I did. With the use of Throttle 1-7 the Benchmarks run alway smooth and I did not see any weird things like "slow gfx" with "high results",
3d-bench goes down from round 200 to 20. Your benches lack of the "shiny-mark" for dos and win9x. a good indicator for the high-end dos-games
with pentium-machines.

Without Throttle please. Just at full speed, and then with caches disabled. It shouldn't take long with the benchmark menu system.

Throttle supports my chipset, every step is 10% slower, that's much better the turning the cpu-cache on/off.
btw. throttle also has the option "-c" so that the slowdown is without disabling the cache.

what software should I use to disable the cache? "cpucache.com" does not work with emm386, it's only functional with qemm.
I think throttle is by far the better one, are there any option beside that that work with emm386?

Doc

Retro-Gamer 😀 ...on different machines

Reply 77 of 137, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++

If you need to use Throttle, that's not a good thing. IMO it's the very last resort.

Just disable cache in BIOS for this benchmark. I want to see what speeds you're getting for comparison.

YouTube, Facebook, Website

Reply 78 of 137, by PARUS

User metadata
Rank Oldbie
Rank
Oldbie

brostenen, I'm not sure that DynaBlaster or any other game before 1993 will run on my Core 2 600MHz without issues. I said my system is mostly for 1993-2009 and I think that it has 100% or very close to 100% compatibility with this "era".
But I can to try. Firstly it should be without sound and if it will run without any issues then I'll try to play sound.

When I talk "it's my best DOS sound system" I mean other things at all. Most ISA cards have a terrible analog lines. When you listen by good acoustics the S/PDIF quality difference... Well, must to hear 😀

Reply 79 of 137, by dr.zeissler

User metadata
Rank l33t
Rank
l33t

Dynablaster has serious sound slowdowns on my PII, have to test it with disabled caches.

Retro-Gamer 😀 ...on different machines