Video description, copied:

"This is a comparison of the time with and without loads between Xemu and Cxbx-r playing Barbie Horse Adventures: Wild Horse Rescue up to the gate opening cutscene in Forest Trail 1.

The runs compared are my (Soulcloset) fastest runs of Any% No Foals (played on Cxbx-r here: https://www.speedrun.com/barbie_horse_adventure_wild_horse_rescue/run/mk59vqvm) and Any% Collect Foals (played on Xemu here: https://www.speedrun.com/barbie_horse_adventure_wild_horse_rescue/run/mk53lj5m). I cut these videos starting at the frame the game responded to the A button press on the main menu (when the buttons start to fade), and ending on the first dark frame post-cutscene where barbie becomes visible.

For removing loads, I removed all frames where a loading screen was visible, without removing any frames before or after that may have been frozen due to lag. This could be changed, however, it would be difficult (in my experience) to tell exactly which frames are lag and which are just the game's framerate not matching with the recording framerate, or the player not doing an action frame-perfectly.

I propose that all BHA:WHR runs be timed/retimed with loads removed in the above detailed way, to keep competition mostly fair between emulators, and between emulator and console for the same platform. Due to its lag and the untested effects of Fast Disc Speed, PS2 is currently slower across the board than Xbox, emulator or not. However, many games have a fastest platform, and in my opinion, evening out the field between emulator and console creates a fair environment on par with other speedgames."

Since I am new to this, we could try using load removers and see if that is useful as I know some runs do that where the timer automatically stops during loading screens, but I am unsure if that is something that can be done via emulator. I will say that xemu s unless help is provided will still not be allowed due to accessibility of getting it to work. You and sneezeball both did it, maybe it is just me but I tried with everything and could not even get it to load. I can put in rules to from now on habe runs only ve times while In the actual game but it is difficult as you said. How would you stop the timer everytime a loading screen occurs, and if ya cant, then removing the loading screens every time sounds like a huge hassle.

That's fair, I see the accessibility angle there for sure. I'll look into load removers and if we can't get something working then I don't know how you'd want to handle Xemu runs already on the leaderboards. I'm fine with my all foals run on Xemu being taken down, but ultimately Cxbx would still be the fastest platform out of any console or emulator with Xemu removed if we still count loads.

Due to it being an oversight on my end, and unless others want to say anything about it, I think we can keep your runs. However any future runs on xemu will not be accepted. From what I have seen, the majority of Speedrunners here use emulators, minus one person. Unless they want to say anything to that effect, I think the rules are fair the way they are since cxbx, while buggy as all hell, still functions and gets the job done and overall is the fastes way to run this game. If you can look into load removers more, and start a new thread on how to do it for future runs, I have no problem with adding it to the rules.

That's a fair decision, makes sense to me. I found a program that uses screenshots of the game to function as an autosplitter, and it has a "pause" function so we could use it remove loads. It might need to be calibrated per-person for use with real console capture, however, with emulators it should be consistent since the image is the same every time. I'm testing it with cxbx-r and it works great! I'm gonna try to creat a full set of images that can make it work as a load remover/autosplitter and then do a test run.

So load removal will be for all consoles? I have the best run done on an actual PS2 and seeing it buried beneath a bunch of emulator runs with much faster loads is quite insulting

Hi there, sorry for the confusion. We're intending to remove loads for all runs done from this point forwards, however, there is a manual way to do load removal without playing on emulator. If you time all the loads in your run by counting frames either on youtube or a video editing software, you can subtract that time from the total and that's your "time without loads". I would be happy to do this for your best time on the leaderboard so it's fair!