25 March 2023, 06:12 | #41 | |
Registered User
Join Date: Mar 2018
Location: Hastings, New Zealand
Posts: 2,828
|
Quote:
On my A1200 with 50MHz 030 25 fps feels very smooth, but of course to get that I have to run it in a small window. Normally I run it 2 sizes down with low detail and get 11-20 fps (~15 fps on average), which is plenty fast enough and a satisfactory window size for me. A 386DX-40 should get similar results to my A1200, and a 486 would be faster so with those settings 20-25 fps seems believable. |
|
25 March 2023, 19:53 | #42 |
Registered User
Join Date: Sep 2004
Location: Tacoma, WA USA
Age: 52
Posts: 1,935
|
It is capped at 35 fps. If you do run it in higher resolutions you can benefit on faster processors. I get ~11 at 640x400 on Vampire A1200. Of course 35 fps at 320x200 & on the 060/50+ the same. Not quite as high at 640x400 but close.
|
26 March 2023, 04:12 | #43 | |
Registered User
Join Date: Feb 2009
Location: london/england
Posts: 1,347
|
Quote:
My 486 wasn't bad considering it was the cheapest money could buy at the time, it did about 22FPS for F1GP and Actua Soccer was fine on it too. It just wasn't quite enough for smooth 2D scrolling games the A1000 didn't even break a sweat doing. At some point I bought a Gravis Ultrasound for it and a CD-ROM. Still have the Gravis actually. I had quite a few PCs over the years, my fav' was a Packard Bell iXtreme which had an Athlon XP3000 or better CPU and it looked like something out of the movie 2001 with that cool curved white plastic case and touch sensitive DVD ROM tray eject buttons. Of all the PCs I have ever owned that was the nicest, the fan was near silent so great for watching movies on the TV via the ATI Radeon 9000 card with composite video output. Also had 5.1 audio output. I miss that PC. The only bad side was I had to use Windows on it. |
|
26 March 2023, 06:35 | #44 | |||
Registered User
Join Date: Mar 2018
Location: Hastings, New Zealand
Posts: 2,828
|
Quote:
Quote:
VL bus 386 boards were available in 1992, but if you say he only had ISA then we have to believe you. All I can say is that every benchmark I have seen puts a 386DX-40 with ISA card (even the ET4000 in my post above) at less than 10fps. Of course this is with standard settings (full window with hud, high detail). Quote:
The other thing most people don't talk about is how jerky and inaccurate a PC mouse was/is. Most VGA cards didn't have a hardware pointer so it had to be rendered in software. It would often flicker while the screen was updating, and sometimes leave 'mouse droppings' on the screen. The ET4000 was infamous for this if you used the standard VGA driver. Even on my 'modern' PCs I have difficulty getting the mouse movement right. On this Linux box the 'acceleration' was way too fast and I had to type a command in to reduce it. Sometimes the mouse freezes up when the machine is heavily loaded with virtual memory - which can take 20 minutes or more to clear! In comparison, even a slow old A500's mouse pointer is rock solid and accurately controllable. This is very important when eg. trying to draw stuff in a paint program. |
|||
23 June 2024, 12:49 | #45 | |
Registered User
Join Date: Aug 2020
Location: Sydney/Australia
Posts: 1,135
|
Quote:
Doom benchmark score for 486DX2-66 with SiS6326 PCI card is 31.37 fps. |
|
23 June 2024, 12:52 | #46 |
Registered User
Join Date: Aug 2020
Location: Sydney/Australia
Posts: 1,135
|
The statement "The graphics card does not matter for DOS Doom" is FALSE.
|
23 June 2024, 17:38 | #47 | |
HOL/FTP busy bee
Join Date: Sep 2006
Location: Germany
Age: 46
Posts: 32,286
|
Quote:
|
|
23 June 2024, 18:15 | #48 |
Registered User
Join Date: Dec 2015
Location: USA
Posts: 3,002
|
Adoom on my A4000 with Ragnarok and Voodoo 3 I get ~70FPS @ 640x480x8bpp.
Prboom with fancy smoothing on the same system gets ~57FPS @640x480x8bpp. |
23 June 2024, 20:42 | #49 |
Alien Bleed
Join Date: Aug 2022
Location: UK
Posts: 4,713
|
|
23 June 2024, 22:50 | #50 |
Ex nihilo nihil
Join Date: Oct 2017
Location: CH
Posts: 5,098
|
|
24 June 2024, 10:50 | #51 |
Registered User
Join Date: Sep 2005
Location: France
Posts: 520
|
I don't have the FPS i once measured before but one sure thing is that on my A4000, any Doom port flies with 68060/50MHz + AGA or RTG compared to my bridgeboard side with 486SLC2/50MHz + STB Nitro VGA with 2 MB which is an excellent SVGA card. With this bridgeboard setup, Doom is ok to play but the Amiga experience is way faster and smoother.
|
24 June 2024, 11:52 | #52 |
cheeky scoundrel
Join Date: Nov 2004
Location: Spijkenisse/Netherlands
Age: 43
Posts: 7,024
|
As per usual the truth is in the middle, making sweeping statements is mostly a bit of fun to pull people out of the woodworks Unfortunately just saying "you're wrong" doesn't really... matter.
|
24 June 2024, 12:02 | #53 |
Alien Bleed
Join Date: Aug 2022
Location: UK
Posts: 4,713
|
@gimbal
You know how it is. We all love to argue about the strengths and weaknesses of our platform by comparing software that was never intended to run on it and assumes a completely different graphics architecture. We're sensible like that. |
24 June 2024, 19:12 | #54 |
HOL/FTP busy bee
Join Date: Sep 2006
Location: Germany
Age: 46
Posts: 32,286
|
If you guys want to check out what playing classic Doom in 2024* feels like, I recommend having a look at Simulacrum which was released this month: https://www.doomworld.com/forum/topi...i-and-myolden/
*Might not work on either a 486 or an A4000 |
24 June 2024, 21:56 | #55 | ||
Registered User
Join Date: Mar 2018
Location: Hastings, New Zealand
Posts: 2,828
|
Quote:
OP wants fps counts for Doom 1 on:- Amiga 68040 Amiga 68060 Mac 68030 Mac 68040 I am also interested in those, especially stock A4000 with 25MHz 040 and 25MHz 030, and Mac 68040. Please don't bother us with results from souped up Amiga systems with 80MHz overclocked 060s, Vampires or PiStorms. This isn't a pissing contest. Estimating frame rates off YouTube videos is difficult and error-prone. To be a fair comparison people should benchmark their current machines (using doomattack on the Amiga, or adoom if that's all you have) on standard settings, noting any details that may be relevant such as the turbo card installed, CPU clock speed, doom version etc. I agree that we don't need more benchmarks for PCs, though if you have any the numbers could still be interesting (I will benchmark my 386SX-16 once I get Doom onto it). Settings to use for Amiga Doom benchmark: 320x200 screen mode, NTSC (if not possible then specify screen mode used) Window: 2 steps down from max Detail: high CLI command: doomattack -forcedemo -timedemo demo3 fps calc: 2134*35/realticks (valid only if gameticks = 2134) Quote:
Nextstation 68040 33MHz 9.8 fps IBM ThinkPad 486SLC-50 12.3 fps Amiga1200 68040-40 adoom 0.5 13.4 fps I486-50 OTI077 ISA 15.3 fps AMD486-66 Cirrus Logic 5428 VLB 20.8 fps Pentium-75 S3 PCI 23.2 fps Amiga 1200 68060-50 adoom 0.9 24.6 fps Vogons 'ultimate 486' Doom benchmark: 486SX-25 12.2 fps 486DX-33 16.3 fps Thandor.net: i386 SX25 (WDC ISA) 3.4 fps Am386 SX33 (ALi M1217) 3.86 fps Am386 SX40 (ALi M1217) 4.68 fps Am386 DX33 (ETEQ) 7.47 fps Cx486DLC 33GP (ETEQ) 7.86 fps Am386 DX40 (MX83) 8.24 fps i486 SX25 (UMC PCI) 14.06 fps i486 DX33 (UMC PCI) 18.3 fps |
||
25 June 2024, 12:07 | #56 |
Longplayer
|
Not fired up the Amiga in a longtime and had some problems to overcome to get it booting right and then displaying aga modes as its normally rtg only. But i managed to get a result out of it in the end.
Blizzard PPC 040/25 256mb ram (50ns no wait states) Doomattack C2P_040+ MMU Routines 1+2 NTSC 320x200, 2 steps down from fullwindow, high detail, emulated midi music. Timed 3863 gametics in 11453 realticks. (11.80fps) Note: The timedemo demo3 moves quite sluggish and nothing like normal play which i guess has plenty of frame skipping. AGA capture while running timedemo (Forgive quality, scandex scandoubler connected to aga output. [ Show youtube player ] AGA Capture playing demos with fps meter. [ Show youtube player ] EDIT: Timedemo demo3 With -NoSound -NoMusic Timed 3863 gametics in 10043 realticks. (13.46fps) ------------------------------------------------------------------------------------------------------------------------------------------ A1200 Blizzard PPC 040/25 Mediator Voodoo3 P96 RTG Voodoo3 320x200x8 Timedemo Demo3 No Music Nosound Timed 3863 gametics in 8763 realticks. (15.42fps) Timedemo Demo3 (with emulated midi music) Timed 3863 gametics in 9313 realticks. (14.51fps) Voodoo3 RTG Capture Showds timedemos and normal demo runs a few times with different settings. [ Show youtube player ] ------------------------- Additional note: A good copymemquick patch helps a little for an extra fps or two. Its a bit of a minefield finding a good one though for my card. if you have issue with startup items randomly failing, WarpOs exceptions or Ramlib errors, a copymemquick patch is a potential culprit if you have one installed. In the worst case you can even have file corruption when writing. On my system with BlizzardPPC 040 The effect on Doomattack is that the demos wont play properly. desyncing and even crashing. CMQ060, NewCMQ060 seem ok initially but crash Doomattack at some point. CopymemAIO was the worst offender even in safe mode causing what appeared to be fpu errors with triangles stretching during demo2 but usually crashed doomattack when the barrel is supposed to explode in demo1. For the benchmark runs I settled on Copymem (Matt Hey CopyMem040.safe) in safe mode with no notable issues with doomattack or AmigaOS itself. Last edited by Mad-Matt; 26 June 2024 at 13:46. |
25 June 2024, 17:01 | #57 |
Registered User
Join Date: May 2023
Location: essex
Posts: 580
|
Hmmm Doom 1 ran OK on my cobbled together 486SX25, 4mb RAM, 11mhz 16bit ISA Oak Technologies cheap SVGA card. Not slick 20-25 fps but somewhere between 15-19 FPS in proper full screen mode. Not very scientific but some of those benchmarks look really low. On 2 steps below full screen it was smooth.
|
29 June 2024, 11:27 | #58 | ||||
Registered User
Join Date: Mar 2018
Location: Hastings, New Zealand
Posts: 2,828
|
Quote:
I haven't got music installed on my system, but sound effects have very little effect on the frame rate (maybe 0.1-0.2 fps) so I test with sound effects on because it's less boring. Also note that if gameticks doesn't equal 2134 then you are either using a different demo or it's not playing back correctly. Different demos will have different average frame rates due to the amount of stuff going on, which makes the results less comparable. Quote:
Quote:
What's interesting is your 25MHz 040 seems to be about the same speed as my 50MHz 030. Theoretically the 040 does 1.1 MIPS/MHz while the 030 only does 0.36 MIPS/MHz or 1/3 the speed, so a 25MHz 040 should be 50% faster than a 50MHz 030. However the system bus timing may be slowing it down. Quote:
However here is an IBM PC-330 with 486DX2-66 and onboard 32 bit Cirrus Logic Video chip. This was also tested with a different demo that got 7001 gameticks, getting a frame rate of 17.04 fps running the game from Windows and 21.7 fps running from DOS. [ Show youtube player ] Assuming a 486 is the same MIPS/MHz as an 040 a 66MHz machine should get 2.6 times higher frame rate, however this is clearly not the case. This test was done with window size only 1 step down from max which would slow it down a bit, but you got 15.42 fps with an equivalent video card so this PC is only about 40% faster, not 160%. Another point is that the Amiga runs Doom from Workbench with full multitasking, so to be fair we really should be comparing it to the PC running Doom from Windows - and then it's only 10% faster! Here's another 486 getting 30.1 fps from DOS. This machine has a 486DX4-100 and 32 bit Trident video card. Unfortunately the game was run full screen so the fps number is a bit lower than it should be. However it is running the correct demo. [ Show youtube player ] Based on CPU clock speed you would expect the DX4-100 to be ~50% faster than the DX2-66 (100/66 = 1.52), and it pretty much is (30.1/21.7 = 1.39). Finally here's another video comparing a 386DX-40 to a 486SX-25. The 386 got 8.7 fps and the 486 got 12.9 fps. [ Show youtube player ] So here we have your 25MHz 040 and AGA beating a PC with 25MHz 486! Last edited by Bruce Abbott; 29 June 2024 at 11:33. |
||||
29 June 2024, 11:38 | #59 | |
Registered User
Join Date: May 2020
Location: Figueira da Foz
Posts: 447
|
Quote:
|
|
29 June 2024, 14:16 | #60 |
Longplayer
|
Is the demo difference a Doom vs Ultimate doom thing so not comparable ? Or Maybe doomattack not playing back as expected.
I have to admit though that I upgraded from an 030/33 which played doomattack ok around 5-7 average fps if I recall right and expected a fair bit more from the 040 upgrade. That said I also had AdoomPPC to play which plays well enough but misses all the cool doom attack features like transparent, spinning map, jump and crosshairs. I always looked out for a DoomAttack PPC back then but it wasnt to be. Has anything even happened after the source was released? Id have expected a newer c2p plugin at the least. Just to check, does 2-steps down window size mean from full window or fullscreen (no hud) as id forgot about nohud before pressing down size twice. |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Doom wad for weaker systems? | utri007 | support.Games | 36 | 04 December 2021 00:51 |
Pause emu and move frame by frame? | hexaae | support.WinUAE | 3 | 27 October 2019 01:18 |
Amiga & Consoles - frame rates, game quality and hardware specs | roondar | Retrogaming General Discussion | 65 | 28 February 2019 11:50 |
Frame rates in Adoom/DoomAttack | Joel_w | support.Games | 1 | 11 December 2018 18:45 |
Amiga native frame rates | mark_k | support.WinUAE | 6 | 25 January 2013 14:56 |
|
|