Stunts 1.1 Mod 1.0 | A new original | an other STUNTS 30 years edition.

[Frieshansen]

As announced in the shoutbox, I made a few benchmarks between Stunts 1.1 and Mod 1.0. There is a clear result. Because a lot of different factors play in (codecs, recording frames ...), I rounded them to 0,1 seconds. An IBM PC AT (286, 6 MHz, 640 kB, EGA, DOS 6.22) was emulated with PCem (https://www.pcem-emulator.co.uk/) as a test system. Please see the detailed System Configuration an Performance in the attached images.
The setup was EGA with PC Speaker.

The first measurement was carried out from the start of stunts.com to the appearance of the first pixel of the Broderbund logo. The mod was started with skidega.exe.
Unmodified Stunts 1.1: 19,0 seconds.
Stunts 1.1 Mod 1.0: 10,6 seconds.

The second measurement was carried out from the first movement of the car after the trailer door was completely opened to the display of "Fasten your Seatbelt"
Unmodified Stunts 1.1: 4,8 seconds.
Stunts 1.1 Mod 1.0: 4,8 seconds.

EDIT: As llm noticed (thanks!), the times were not correct. I have now measured and checked everything again - now it should be pretty accurate.

[llm]

The first measurement was carried out from the start of stunts.com to the appearance of the first pixel of the Broderbund logo. The mod was started with skidega.exe.
Unmodified Stunts 1.1: 19,0 seconds.
Stunts 1.1 Mod 1.0: 4,7 seconds.

The second measurement was carried out from the first movement of the car after the trailer door was completely opened to the display of "Fasten your Seatbelt"
Unmodified Stunts 1.1: 4,7 seconds.
Stunts 1.1 Mod 1.0: 4,0 seconds.

The first measurement result show what impact replacing the whole laoding,unpacking process has

The second measurement result should less different because the game code isnt changed at all

[Duplode]

As announced in the shoutbox, I made a few benchmarks between Stunts 1.1 and Mod 1.0.

That's a nice result to have documented! It's always good to have extra ways to help people running the game on different systems, as Daniel3D mentioned early in this thread.

[Frieshansen]

The second measurement result should less different because the game code isnt changed at all

Thanks for noticing that! There were really some errors - these have now been corrected in the original post.

[llm]

still ~9sec faster

[dreadnaut]

In day-to-day DosBox use, it takes maybe a second to go from stunts.com to Broderbund logo. Is that because DosBox does not try to simulate a specific speed, but just goes as fast as it can?

[llm]

Yes, at least for memory, hard drive etc. access is goes full speed

But beware that the mod exes are the only ones that execute without problems on todays hardware under real dos, freedos COMBINED with some milliseconds of gained startup speed

[Daniel3D]

As announced in the shoutbox, I made a few benchmarks between Stunts 1.1 and Mod 1.0.  etc..

Wow, i noticed it being faster because i didn't see "Loading stunts..." any more. But on an old system it is quite a bit faster.
Thanks for the benchmark test.

[Daniel3D]

I have found a bug that should not be there.

Out of curiosity, I ran the mystery byte edited cars in EGA, CGA and Tandy graphic (emulation in dosbox) but didn't experience anything strange with the graphics. Although I did notice that the soundblaster music gets a bit funcky in CGA mode if you get back in the main menu after racing.
This has nothing to do with the cars (checked that) but is not present in the original game.

These files are made from source with the execombiner and not compiled from the restunts data. So they should be exactly the same as the original stunts...

It doesn't break the game, just sounds strange.

[Cas]

Daniël... you mean this sound problem only happens with cars that have their mystery bytes edited?

[Daniel3D]

Daniël... you mean this sound problem only happens with cars that have their mystery bytes edited?

No. Completely separated. Happens also in the original uploaded V11 M10 with original cars.

[llm]

its super easy to introduce bugs that are nearly not noticeable (by patching the binaries or adding asm code)
or firstly find old bugs that just get got not noticed before

the simplest code length change or move can produce unwanted and hard to notice bugs everywhere - that was the main reason for directly starting the C port

is the sound glitch also happening with the original floppy image version - not using the new exes but boring old load.exe etc.?
(please add a bugs.txt to the Mod to tell which bugs are new/known in Mod or original)

its very needed to test out every simplest change throught...

[Daniel3D]

is the sound glitch also happening with the original floppy image version - not using the new exes but boring old load.exe etc.?

No, in my clean original with crack it's not present.

[llm]

is the sound glitch also happening with the original floppy image version - not using the new exes but boring old load.exe etc.?

No, in my clean original with crack it's not present.

only for the CGA version is strange (its the same codebase as hercules - which is still missing)
maybe the combine process is not 100% correct here - but takes time to understand whats happening (i have not even an idea what can produce such problem)

[Daniel3D]

is the sound glitch also happening with the original floppy image version - not using the new exes but boring old load.exe etc.?

No, in my clean original with crack it's not present.

only for the CGA version is strange (its the same codebase as hercules - which is still missing)
maybe the combine process is not 100% correct here - but takes time to understand whats happening (i have not even an idea what can produce such problem)

Neither do I. The music plays fine on first entering the main menu. But after racing (and crashing because I am lazy to finish the race or hit ESC) it sounds really weird upon returning to the main menu.