Difference between revisions of "Super Game Boy timing"
From SDA Knowledge Base
(removed a dead link, added excel file zip) |
(moved picture, highlighted multiplier) |
||
| Line 4: | Line 4: | ||
I spent the better part of two nights researching what to settle on for the ratio. My method was based on recording music. The games I used were the two in question, Link's Awakening and Metroid II. For LA I made a new file, picked up the shield, exited Marin's house and then saved & quit. Then I got ready to load that file, and started recording as soon as I pressed A. Link stood outside, unmoving, for 10 minutes while I recorded the music. I did it again for 20 minutes for better results. In Metroid II, I used a new file and recorded the main theme while standing at the ship. I also recorded 10 minutes of sound from my actual game boy using a microphone. | I spent the better part of two nights researching what to settle on for the ratio. My method was based on recording music. The games I used were the two in question, Link's Awakening and Metroid II. For LA I made a new file, picked up the shield, exited Marin's house and then saved & quit. Then I got ready to load that file, and started recording as soon as I pressed A. Link stood outside, unmoving, for 10 minutes while I recorded the music. I did it again for 20 minutes for better results. In Metroid II, I used a new file and recorded the main theme while standing at the ship. I also recorded 10 minutes of sound from my actual game boy using a microphone. | ||
| − | |||
| − | |||
A quick look in Cool Edit showed that SGB was clearly faster than GBP and GBP seemed to be the right speed compared to GB. I made this picture to show a comparison of the three waveforms. Now I just needed to settle on the ratio. I took my 20 minute recordings and began to take sample points by zooming in on the waveforms and finding the place where the music repeats. Because of the patterns in the music, they were easy to see in the waveform and hear too. I made an excel file with all the sample points and then did a bunch of calculations. | A quick look in Cool Edit showed that SGB was clearly faster than GBP and GBP seemed to be the right speed compared to GB. I made this picture to show a comparison of the three waveforms. Now I just needed to settle on the ratio. I took my 20 minute recordings and began to take sample points by zooming in on the waveforms and finding the place where the music repeats. Because of the patterns in the music, they were easy to see in the waveform and hear too. I made an excel file with all the sample points and then did a bunch of calculations. | ||
| − | The resulting ratio is 1.024115. If you record a run on SGB you can multiply your time by this number to get an idea for what your run will be listed as. You can download my [[Media:gbComparison.zip|excel file]] and look at my numbers and a few graphs if you want. | + | [[Image:gbComparison.png]] |
| + | |||
| + | <font color=red>'''The resulting ratio is 1.024115.'''</font> If you record a run on SGB you can multiply your time by this number to get an idea for what your run will be listed as. You can download my [[Media:gbComparison.zip|excel file]] and look at my numbers and a few graphs if you want. | ||
This number is supported by the MHz clock speed of Game Boy and Super Game Boy listed at [http://www.overclocked.org/emugb.htm overclocked.org]. | This number is supported by the MHz clock speed of Game Boy and Super Game Boy listed at [http://www.overclocked.org/emugb.htm overclocked.org]. | ||
Revision as of 13:44, 5 August 2008
(originally written by Radix)
The purpose of this page is to explain what's wrong with the "Super Game Boy", an addon for the Super Nintendo that lets you play Game Boy games on TV. It's one of the ways that can be used to record speed runs of Game Boy games, but there's a serious problem with the hardware - it runs fast. What should take 1 minute, it does in about 58.6 seconds. The issue was first pointed out by carlmmii concerning my original any% run on Metroid II. I meant to do a test but I was lazy/busy and never did one until 4 months later after Lucid Faia's Link's Awakening run on SGB brought up the issue again.
I spent the better part of two nights researching what to settle on for the ratio. My method was based on recording music. The games I used were the two in question, Link's Awakening and Metroid II. For LA I made a new file, picked up the shield, exited Marin's house and then saved & quit. Then I got ready to load that file, and started recording as soon as I pressed A. Link stood outside, unmoving, for 10 minutes while I recorded the music. I did it again for 20 minutes for better results. In Metroid II, I used a new file and recorded the main theme while standing at the ship. I also recorded 10 minutes of sound from my actual game boy using a microphone.
A quick look in Cool Edit showed that SGB was clearly faster than GBP and GBP seemed to be the right speed compared to GB. I made this picture to show a comparison of the three waveforms. Now I just needed to settle on the ratio. I took my 20 minute recordings and began to take sample points by zooming in on the waveforms and finding the place where the music repeats. Because of the patterns in the music, they were easy to see in the waveform and hear too. I made an excel file with all the sample points and then did a bunch of calculations.
The resulting ratio is 1.024115. If you record a run on SGB you can multiply your time by this number to get an idea for what your run will be listed as. You can download my excel file and look at my numbers and a few graphs if you want.
This number is supported by the MHz clock speed of Game Boy and Super Game Boy listed at overclocked.org.
