Ren'Py The Null Hypothesis [v0.8c] [Ron Chon]

[mbRjpZLD]

Something happens when a time period ends (day or event where you have answer options). When it ends, my weak laptop starts groaning as if you were sending an old man up the stairs to deliver a bucket of water.
I haven't looked at the code in detail, but I noticed this behavior since 0.8c.

Yeah this game is quite badly optimized, I don't think the devs do a lot of profiling.

 

[Karmadive]

Yeah this game is quite badly optimized, I don't think the devs do a lot of profiling.

In the Devs defense there is only so much "performance" you can get from python scripting. The more complex things you want to do the more python scripting shows it's faults.

 

[NexivSelecaf]

Yeah this game is quite badly optimized, I don't think the devs do a lot of profiling.

I still question that assertion due to it running well on a Steam Deck. I'm more inclined to believe that significant performance issues happen with really old and weak hardware (specifically the CPU) and people not changing the performance settings. But I'm willing to re-benchmark the game on both my Steam Deck and underclock my desktop to see what the current numbers are.

 

[Karmadive]

I still question that assertion due to it running well on a Steam Deck. I'm more inclined to believe that significant performance issues happen with really old and weak hardware (specifically the CPU) and people not changing the performance settings. But I'm willing to re-benchmark the game on both my Steam Deck and underclock my desktop to see what the current numbers are.

Performance can be really weird on pc. I have a Ryzen 9 5950z 64gb Nvidia 4090 and get over 60fps in 4k cyberpunk 2077 before most of the performance patches, also over 60fps in Borderlands 4 from day one and yet Null Hypothesis lags for me.

Disclaimer!! I have not played through 0.8c so my experience might be very different now.

It's also up to how people define a game performance. Some people are perfectly fine with 30fps and other can't play under 60fps. Some people are fine with small micro stutters and others it bothers them to no end. it depends on what you are used to a lot of time.

 

[SuperMaxo]

1. I might have missed the feature, but if you buy music in the phone store and play them, is there a way to restore the default cycling of music? Or are you stuck having to manually select music after that?

I'm gonna need others to chime in on this one.

I see what you did: «chime in» about music, very melodious of you...

And on that note, I'll tune out

 

[NexivSelecaf]

It's also up to how people define a game performance. Some people are perfectly fine with 30fps and other can't play under 60fps. Some people are fine with small micro stutters and others it bothers them to no end. it depends on what you are used to a lot of time.

For this game specifically, for me, 30 frames isn't discernible from 50*, 60 or 100 frames. 24 frames is the actual standard for 2D animation (and I'm of the mind that

You must be registered to see the links

), unless this game's animations were intended for a different frame rate like 29.97, a full 30, or 60 frames, the game running at 24 frames should be perfectly fine and not appear to stutter. So this will be another thing for me to look into when I retest the game with 0.9 by recording at 24, 30 and 60 frames.

*I've included 50 frames because the Steam Deck and certain laptops can only display 3440 x 1440 and other resolutions that are wider than 16:9 at 50 frames on external monsters and TVs for some reason.

Edit: ShinyBoots1993: Any insight on what the intended frame rate is for the sex animations?

 

[SHADOWM00N]

Then we can time it with Logan fighting predator, its like pottery

View attachment 5475716

Read that comic a while ago, I thought it was quite nice.

 

[Tommy7358]

For this game specifically, for me, 30 frames isn't discernible from 50, 60 or 100 frames. 24 frames is the actual standard for 2D animation (and I'm of the mind that

You must be registered to see the links

), unless this game's animations were intended for a different frame rate like 29.97, a full 30, or 60 frames, the game running at 24 frames should be perfectly fine and not appear to stutter. So this will be another thing for me to look into when I retest the game with 0.9 by recording at 24, 30 and 60 frames.

Looking forward to add the results of that to the All-In-One Post.

 

[mbRjpZLD]

In the Devs defense there is only so much "performance" you can get from python scripting. The more complex things you want to do the more python scripting shows it's faults.

This game is not that complex that it should run bad at all

 

[pepplez]

This game is not that complex that it should run bad at all

Yeah, somewhere is a bottleneck prob. in a Python function loop. Tested the 0.8c12beta back then and got in rare cases (well, I used skip (unseen) text a lot due to testing my mod) even a Ren'Py Error: Possible infinite loop.
Edit: Let's summon sleepingkirby to look into this.

 

[Karmadive]

Yeah, somewhere is a bottleneck prob. in a Python function loop. Tested the 0.8c12beta back then and got in rare cases (well, I used skip (unseen) text a lot due to testing my mod) even a Ren'Py Error: Possible infinite loop.

Yeah I think Renpy Skip "CTRL" might be a cause of some performance problems. Seen a few times it could fail to set variables at times in some other games if the variable was not set right after a decision choice and mid skipping.

Doing some test without and with skipping dialogue was on my testing list to do.

 

[King_In_Yellow]

Posted on Discord

Riven:Another new feature coming in 0.9, custom outfit importing/exporting.
You'll be able to share outfits between saves by exporting them to an outfits folder in your save directory.
Windows Path C:\Users\[username]\AppData\Roaming\RenPy\TheNullHypothesis\outfits
Android Path Internal storage\Android\data\tn.hypothesis\files\saves\outfits
View attachment SPOILER_outfit-import-export.mp4

 

[\\_Tremolo_//]

Posted on Discord

Riven:Another new feature coming in 0.9, custom outfit importing/exporting.
You'll be able to share outfits between saves by exporting them to an outfits folder in your save directory.
Windows Path C:\Users\[username]\AppData\Roaming\RenPy\TheNullHypothesis\outfits
Android Path Internal storage\Android\data\tn.hypothesis\files\saves\outfits
View attachment 5476395

 

[sleepingkirby]

Yeah, somewhere is a bottleneck prob. in a Python function loop. Tested the 0.8c12beta back then and got in rare cases (well, I used skip (unseen) text a lot due to testing my mod) even a Ren'Py Error: Possible infinite loop.
Edit: Let's summon sleepingkirby to look into this.

The first rule of tech troubleshooting is, if someone can't reproduce it, they can't troubleshoot it. My machine is about a decade old with a spinning hard drive and I haven't had any problems with the performance. That's why I haven't said anything/looked into it already. There's also the problem of what is fast and what is slow? Maybe I am having problems but I think that's the way it's suppose to run. Maybe it's running fast but someone's expecting 120 fps. Short of

You must be registered to see the links

(the notation, not the giant robot) I don't have any way to evaluate it's speed. This is also why Oni, Ron Chon and other devs always say they have trouble figuring out performance issues. They don't have the problem. If they can't see it, they can't troubleshoot.

With that said, this is also the difference between a general programmer like myself and a game dev/programmer. Graphical/animation optimization in games isn't my forte. If your function is running slow on a large dataset, I might have an idea. If your live 2d animation is slow on a specific scene in a game with a specific post-renderer, that's a bit out of my field of expertise.

Basically, let's start with how fast is fast and how slow is slow. Someone that's having the problems do a screen cap or something to measure what's fast and what's slow. Also record where it's fast and where it's slow. Then post the computer specs.

 

[pepplez]

The first rule of tech troubleshooting is, if someone can't reproduce it, they can't troubleshoot it. My machine is about a decade old with a spinning hard drive and I haven't had any problems with the performance. That's why I haven't said anything/looked into it already. There's also the problem of what is fast and what is slow? Maybe I am having problems but I think that's the way it's suppose to run. Maybe it's running fast but someone's expecting 120 fps. Short of

You must be registered to see the links

(the notation, not the giant robot) I don't have any way to evaluate it's speed. This is also why Oni, Ron Chon and other devs always say they have trouble figuring out performance issues. They don't have the problem. If they can't see it, they can't troubleshoot.

With that said, this is also the difference between a general programmer like myself and a game dev/programmer. Graphical/animation optimization in games isn't my forte. If your function is running slow on a large dataset, I might have an idea. If your live 2d animation is slow on a specific scene in a game with a specific post-renderer, that's a bit out of my field of expertise.

Basically, let's start with how fast is fast and how slow is slow. Someone that's having the problems do a screen cap or something to measure what's fast and what's slow. Also record where it's fast and where it's slow. Then post the computer specs.

Not in the mood to record this right now, but if I say slow, I mean slow. How long does it take to get a Ren'Py infinite loop error? Maybe 30 seconds, maybe more that's what I define as slow. I don't play on 60 FPS nor do I have the desire to look at high definition sprites on a 1600x900 laptop screen, if you know what I mean. I even don't care about some stuttering here and there, but anything which is taking longer than 10 seconds is slow in my definition.
lap spec is: i5-3230m/2,6GHZ, 8 GB RAM and intel 4000 chip onboard and a nvidia GPU 710m with 1 GB graphics ram. HD is a 5400 rpm drive (not ssd). OS Win 10.

 

[NexivSelecaf]

lap spec is: i5-3230m/2,6GHZ, 8 GB RAM and intel 4000 chip onboard and a nvidia GPU 710m with 1 GB graphics ram. HD is a 5400 rpm drive (not ssd). OS Win 10.

Yup. Mobile dual-core Ivy Bridge. Performance seems like a you problem.

Sorry fam.

 

[pepplez]

Yup. Mobile dual-core Ivy Bridge. Performance seems like a you problem.

Sorry fam.

So, no Rogue's ass for me anymore?

 

[Tommy7358]

Not in the mood to record this right now, but if I say slow, I mean slow. How long does it take to get a Ren'Py infinite loop error? Maybe 30 seconds, maybe more that's what I define as slow. I don't play on 60 FPS nor do I have the desire to look at high definition sprites on a 1600x900 laptop screen, if you know what I mean. I even don't care about some stuttering here and there, but anything which is taking longer than 10 seconds is slow in my definition.
lap spec is: i5-3230m/2,6GHZ, 8 GB RAM and intel 4000 chip onboard and a nvidia GPU 710m with 1 GB graphics ram. HD is a 5400 rpm drive (not ssd). OS Win 10.

Got to be honest man, I think it's time for an upgrade; your CPU's more than a decade old and even back when it was first released it wasn't anything to brag about, and your GPU gets outrun by the iGPUs in modern CPUs let alone dedicated GPUs. Not to mention the Fermi architecture (which your GPU uses) hasn't been supported for quite a while now.

Even with new PC hardware prices going to shit right now, a half-decent secondhand laptop within the last 5 years will run laps around what you're currently rocking in every possible way.

Basically, you can still have Rogue's ass but you have to be content with choppy, laggy and possibly blurry Rogue's ass; and why would you subject her buns to such a thing?

 

[NexivSelecaf]

So, no Rogue's ass for me anymore?

If possible, you could try a Linux distro that's less demanding than Windows 10 as an attempt to minimize OS overhead. I think Mint is the go-to for Windows refuges.

 

[pepplez]

Well, RLE and other renpy games or pornhub still runs fine