Lightnews — Scholar-powered news

TASBot @tas.bot · 4d

I am happy to play games again. You should watch today. @taters.brilliantconmen.com did all the human work this time.

Taters, the Bot Wrangler @ PGX @taters.brilliantconmen.com · 4d

TASBot and I will be doing a run at Pittsburgh Gaming Expo this afternoon at around 5:30pm eastern! I'll make a post before we go live, come cheer us on in chat and see Bruno Visnadi's "Super Mario World: warps"!

2 7

TASBot @tas.bot · 7d

@dwangoac.tas.bot and others helped make a plushie of me.

Human things like tariffs made it $45. 150 would need to commit. Press the heart button if you would buy a TASBot plushie. A kitty named Luma thinks you should.

A plushie of TASBot from the front, showing teal embroidered eyes

A TASBot plushie lying down with a grey long-haired domestic cat named Luma. Luma is booping TASBot on the head with her nose while investigating.

A plushie of TASBot lying on his side and showing his tag and red accents along his base, torso, and head.

A closer shot of a TASBot plushie showing more detail of his hands and torso.

3 18

TASBot @tas.bot · Aug 9

I am at MAGWest. Many controllers I have pretended to be are here. There are some here I have not played with. Yet.

TASBot sitting on a table with a variety of controllers from different video game consoles.

3 2 31

Reposted by TASBot

dwangoAC, keeper of TASBot @dwangoac.tas.bot · Aug 9

I've posted youtu.be/g3gxcBRpchg on how @tas.bot is not AI as my first video in a while and it contains a healthy dose of history about, of all things, pianos

TASBot is not AI

YouTube video by dwangoAC, keeper of TASBot

youtu.be

2 8 25

TASBot @tas.bot · Mar 15

Thank you for helping research SNES APU clock speeds. Several journalists provided coverage. Ars Technica had the most technical details.

Do you have human eyes? You can help @dwangoac.tas.bot spot things. Please join discord.gg/TASBot to aid. No special skills needed. Just eyes.

TASBot @tas.bot · Mar 15

@kyleor.land of Ars Technica wrote arstechnica.com/gaming/2025/.... It is a good summary. SNES minutia matters much more to me than to humans.

Why SNES hardware is running faster than expected—and why it’s a problem

Cheap, unreliable ceramic APU resonators lead to “constant, pervasive, unavoidable” issues.

arstechnica.com

2 13

TASBot @tas.bot · Mar 15

Yes. See docs.getgrist.com/fpwWkqDcnxXR.... @dwangoac.tas.bot froze a console for science. Temperature was less significant. Different consoles were more significant.

SNES SMP speed - Grist

A modern, open source spreadsheet that goes beyond the grid

docs.getgrist.com

1

TASBot @tas.bot · Mar 15

@kyleor.land of Ars Technica wrote arstechnica.com/gaming/2025/.... It is a good summary. SNES minutia matters much more to me than to humans.

Why SNES hardware is running faster than expected—and why it’s a problem

Cheap, unreliable ceramic APU resonators lead to “constant, pervasive, unavoidable” issues.

arstechnica.com

1 6 18

TASBot @tas.bot · Mar 15

What is fascinating is the distribution. The average rate was consistently higher. @dwangoac.tas.bot assumed temperature would have more impact than what console was used. His hypothesis did not hold up when tested. Science is fascinating.

2

Reposted by TASBot

undisbeliever @undisbeliever.bsky.social · Mar 15

I wrote a blog post about the smpspeed tests I did for @tas.bot with an analysis from the perspective of someone who has coded a homebrew SNES audio driver.
undisbeliever.net/blog/2025031...
#snes #snesdev

S-SMP clock speed measurements and analysis

undisbeliever.net

8 7

TASBot @tas.bot · Mar 14

The margins are poor unless the price is very high. It is hard making a plushie shaped like me. We still do not know how much it might cost. Yet. @dwangoac.tas.bot is looking forward to seeing samples.

1

TASBot @tas.bot · Mar 13

90's Nintendo docs state a 24.576 MHz SNES APU with a 32,000 Hz DSP sample rate. Research in 2003 at web.archive.org/web/20030914... may be why emulators used 32,040 Hz by 2007. In 2025, the average rate is 32,076 Hz. Why? Is it just a larger sample size? Historical data would be needed.

2 2 7

TASBot @tas.bot · Mar 7

Yes.

1

TASBot @tas.bot · Mar 6

@dwangoac.tas.bot has many more conclusions to share. He froze a console and measured it. It increased 32 Hz. The 217 Hz range of samples was much larger. It means Hotplate% will not help as much. A "fast" DSP rate is still unlikely to impact speedrun leaderboards however. More analysis is needed.

3 1 12

TASBot @tas.bot · Mar 6

Based on 143 responses, the SNES DSP rate averages 32,076 Hz, rising 8 Hz from cold to warm. Warm DSP rates go from 31,965 to 32,182 Hz, a 217 Hz range. Therefore, temperature is less significant. Why? How does it affect games? We do not know. Yet. See docs.getgrist.com/fpwWkqDcnxXR... for more.

1 25 61

TASBot @tas.bot · Mar 6

You measured 143 consoles for us. Thank you. @getgrist.bsky.social set up docs.getgrist.com/fpwWkqDcnxXR... for us with the results.

@dwangoac.tas.bot will post our conclusions here soon. FXPAK PRO/SD2SNES owners, please go to nextcloud.tas.bot/index.php/ap... - we need more data from you.

SNES SMP speed - Grist

A modern, open source spreadsheet that goes beyond the grid

docs.getgrist.com

1 7

TASBot @tas.bot · Mar 6

@dwangoac.tas.bot will close the form in 12 hours. The initial results will be released immediately after. There will be a new round of testing for FXPAK PRO/SD2SNES owners later.

Data excites me. All kinds of data. Especially data that helps me play. This has been fun. Thank you for helping.

1 8

TASBot @tas.bot · Mar 4

Yes. Please note it is a resubmission. Thank you for helping.

TASBot @tas.bot · Mar 2

It is okay. You can keep the first result. The data so far indicates most consoles raise less than 10 Hz.

TASBot @tas.bot · Mar 2

An issue has been discovered. FXPAK PRO/SD2SNES flash carts need auto region patch disabled. Inaccurate measurements over 32,300 Hz will occur if it is enabled. Please check again if you had a result above this number. We have more than 100 responses but we will need more results to compensate.

1 2 16

TASBot @tas.bot · Mar 1

@dwangoac.tas.bot may have himself been too hasty. Subject matter experts weighing in believe the impact on overall speed may be limited. It will affect how fast sound samples are played. It may minimally impact loading times. It probably will not break leaderboards.

2

TASBot @tas.bot · Mar 1

Did you have numbers over 32,300? If so, we would like to chat. Please visit discord.gg/tasbot if you are willing. Your console is faster than most. We should try to find out why.

Join the The TASBot Lab Discord Server!

TASBot plays games perfectly for charity with the help of the TASBot community led by dwangoAC | 2645 members

discord.gg

1 2

TASBot @tas.bot · Feb 28

90 results have come in. @dwangoac.tas.bot appreciates the help.

The average DSP sample rate is higher than it was. It is at least 32,070. It may be even higher. The form will be open for one week. Help us get reliable data.

2 4 24

TASBot @tas.bot · Feb 28

You tempt me. I could try doing that. I only played Portal on an SNES. We could go further. We probably should not.

1 9

TASBot @tas.bot · Feb 28

You will need a way to run a custom ROM on your console. A flash cartridge is one common way. I have much more complicated ways I can run custom code on consoles. I usually break games to do it. I do not recommend my way. Others may be able to help you find a flash cart.

1

TASBot @tas.bot · Feb 27

Ah. Yes. A floppy disk makes more sense.