Crash Bandicoot: Emulator & Console - Load Comparison

It's the beginning of the end.
Time for a purge.

First the Consoles.
The 75k & 90k PS2 comes from CaneofPacci you can find him here: https://youtube.com/CaneofPacci/
The 5501 PS1 comes from CavemanDCJ you can find him here: https://youtube.com/CavemanDCJ/
CavemanDCJ also provided his 90k as Cane's showed some questionable FDS loading into The Lab. The results ended up the same.
In a full run of the game the 90k's issue in loading The Lab would still make up for earlier levels.

We're showcasing the NTSC-U version of Crash.
Why not the NTSC-J version you ask.

In speedrunning both U and J are ran, and people have preference on one over the other and the fact that some top runners have been able to trade positions on which system they use.
(Except for 100% cause J has some levels cut down giving it the advantage.)

The PS2 is there to show what runners use and the PS1 is to showoff the who's the closest to a real console.
All emulators must either be close to or match the PS2 on SDS or the PS1, if they're within a large margin they should be banned.
If they're closer to PS2 FDS they should be taken out back.

Now for the bios recommendation and emulator used.
SCPH5501 - mednafen, BizHawk (Octoshock), RetroArch (Beetle)
This trio is essentially the same emulator.

mednafen is known as one of the three most accurate emulators on the field but compared to another competitor it isn't that demanding of hardware.

BizHawk (Octoshock) is the current TAS emulator it is also the most out of date. At the time of writing BizHawk's Octoshock core is almost 4 years behind the upstream on mednafen.
But is that a bad thing? We'll find out soon.

RetroArch (Beetle) has a feature for increasing disk load speed. You may notice I went to x8.
x10 matches the x8 load times when compared and both x12 and x14 fail to pass the Naughty Dog logo. This is why we start from that logo.
These faster disk loads are used since RetroArch can nag for you not having the best load speed.

SCPH7001 - XEBRA
XEBRA is known as the king of most accurate, and CPU intensive emulator to that of a PS1 and should able to match the load times and lag accurately enough.

SCPH1001 - pSX (psxfin), PCSX-RR, PSXjin, no$psx, ePSXe 1.6 - 2.0.5
The ugly bios that emulators usually recommended.

pSX is a dead emulator with no use what so ever and should not be taken seriously by anyone.

PCSX-RR is the first PlayStation TAS emulator. It has the worst audio plugin in existence, unless paired with EternalSPU which is showcased here.

PSXjin is the second PlayStation TAS emulator. It's a major improvement to PCSX-RR supporting SOUND. It's a revolution I know.

no$psx is more or less treated as a debug emulator and that it's not meant to be taken for accuracy at all.

ePSXe is a clsoed source emulator full of trashy ass hacks and trashy ass plugins. What is this, N64 Emulation?
The emulator amonst all 8 are using their defaults, but for video they're all using the latest Pete's OpenGL2 Plugin (2.9).
Interpreter is a mode in newer ePSXe versions to be closer to that of a console. Does it do the job though?

HLE - RetroArch (PCSX ReARMed)
For some unknown reason the recommended bios which is the same as the Beetle core did not work, so this core is forced to use the cruddy HLE.
ReARMed is a great emulator for low end and ARM devices.

The PC running the emulators is an i7-9700k with an NVIDIA RTX 2070 SUPER and 8GBx2 3200MHz of DDR4 RAM.
This is one of the best case scenarios possible. Emulator footage is recorded with OBS. All emulators are clean installs with some exceptions already mentioned.

The Bottom Right Side is the accuracy corner. It's PS1, mednafen, RetroArch (Beetle), XEBRA and BizHawk (Octoshock) with ePSXe 2.0.5 Interpreter on the side.

Results:
RetroArch (Beetle x8) - 10:27.1
RetroArch (Beetle x6) - 10:31.52
RetroArch (Beetle x4) - 10:41.28
PS2 75001 (FDS) - 10:43.19
PS2 90001 (FDS) - 10:43.42
pSX - 10:57.15
no$psx - 10:58.59
ePSXe 1.7 - 11:01.58
RetroArch (PCSX ReARMed) - 11:02.18
PSXjin - 11:03.18
PSCX-RR EternalSPU - 11:04.51
PCSX-RR - 11:05.07
ePSXe 2.0 Interpreter - 11:05.57
ePSXe 1.9.25 - 11:06.35
ePSXe 2.0 - 11:08.17
ePSXe 1.8 - 11:09.10
ePSXe 1.6 - 11:11.39
- - - - -
RetroArch (Beetle x2) - 11:13.14
ePSXe 2.0.5 - 11:13.49
mednafen 1.24.0 - 11:14.19
ePSXe 2.0.2 - 11:14.25
ePSXe 1.9 - 11:15.38
ePSXe 2.0.2/5 Interpreter - 11:15.57
PS2 90001 (SDS) - 11:18.37
PS2 75001 (SDS) - 11:20.25
BizHawk (Octoshock) - 11:25.39
PS1 5501 - 11:27.22
XEBRA - 11:30.12

You may have noticed a line to the results.
To keep it as simple as possible.
An emulator that above that line should be banned in the Crash scene for this instance.
Crash 1 doesn't have crazy/any amount of lag, and serious runners would just get a PS2.
Those that want to try out the game should use the more accurate emulators.
The line is basically for those that are within 5 seconds of the 90k PS2 on SDS.