Converting simple audio samples to Namco 163 sound chip

Sorry, I took too long to get around making this video, hopefully I didn't made the person who requested this hopeless, lol.
Good luck on your album!
Download the tool that I made:
drive.google.com/open?id=1KWh...
Audacity N163/FDS wave analyzer script (added up to 240 sample size):
pastebin.com/S9ScdGYx

[@detective_l2413]

This is how VinylCheese's N163 'Seinfeld' slap bass is made

[@canaconn2388]

This tutorial saved me! There was little to no real information on how to easily do this, other than running programs with no idea on how they function. Thank you!

[@nathanlawrence2484]

Thank you for this! The famitracket wiki being down made me almost give up on the n163 chip!

[@tux1468]

Roddy making highly specific tutorials? Awesome.

[@nestronroger8203]

YES

[@nestronroger8203]

Oh lord, thank you so much!

[@Nicszerg]

This is very great, I've been trying to figure out how that audacity waveform converter works for a long time.

[@ssg-eggunner]

THE NAMCO N613 IS SO FUCKING POWERFUL

[@N3Selina]

thank you so much. highly appreciated

[@TheCRTman]

That's craaaaaaaaaazy!!!!!

[@TheMuzykant]

I was looking for a tutorial on this long ago. Thanks man!

[@RoddyDev]

Glad to help! Hope you could understand this properly since looking back at it, I explained this quite poorly.

[@TheMuzykant]

@@RoddyDev I understood it well. You could do another video if you want to improve the explanation!

[@4IN4ND]

Interesting

[@MillenniumHero]

Neat tutorial! Might try this out later.

[@DeMOSic465]

Me too

[@controlleronahanger]

Nyquist returns the value "1" when I try and run the Wave Analyzer script/
Do I need a specific version of Audacity for it to work?

[@RoddyDev]

Just tested on the most recent version of Audacity and the exact same script on the pastebin link, and it works perfectly. You probably will need to check "Use legacy (version 3) syntax." on the Nyquist prompt, but it should automatically check it when it runs. You probably made a typo somewhere in the code while pasting?

[@controlleronahanger]

@@RoddyDev Just had to run it via the Analyze... part of the interface.

[@Heyzeus18]

gloverD

[@BlazeAlphaz]

Negócio da peste é isso?

[@groovingood]

Chiptune

[@pr0t0gam3r]

how do you get the spc to it tool

[@isaNYyN]

I'm intrigued by the code behind your tool
I fully understand the sampling problem that will arise because of the frequency differences, but can't the problem simply be addressed by dividing the sample length by 15, without using the script?
In your case, 3229/15 is 215.3, which can be rounded up to 215.
If you consider your list of 15 entries as an arithmetic sequence in which we go from one term to the next by adding 215, we end up with your values. Is the idea behind your script simply to recover waves at points evenly distributed along the length of the sample?

[@RoddyDev]

You are completely right!
Here is the code for the whole program: pastebin.com/5YVdYXwm
I was a very bad programmer back then (I still am! I don't code as much anymore)
The idea was: Get a list of numbers evenly spaced out based on your inputs, while making sure they don't overlap with each other (initial value + sample size, so me saying it was unnecessary was completely wrong)
But then I looked at the code... It runs the logic for that but the program doesn't even use it when writing the text file... It only grabs the evenly spaced out numbers. (As seen on Line 24). What was wrong with me at that time?
The worst thing is that you will need to look at the waveforms itself and make a guess where the shape starts and ends... All done manually. That's why I said in the video that you should only do this with simple shaped waveforms without chords.
Now, it's very possible that if I rewrite it using more modern tools with Python and some audio libraries, I could even generate complete FTI files without even needing manual inputs and Audacity. I don't know why I haven't updated this... It's possibly the worst tooI I have ever made because it does almost nothing. It really surprises me that it still helped some people.

[@isaNYyN]

@@RoddyDev Well, you're way too harsh on yourself, this was actually a valid way to approach the problem. Actually I didn't try the method myself, but the video demonstrate that it is efficient. I do am interested in this topic, though. N163 is a very interesting chip with unsuspected capabilities, and this video was great in order to demonstrate that. Another question, why did you specifically pick 15 as the sampling number? Is it some sort of compromise related to RAM considerations? What if the frequency isn't 60Hz anymore (which happens whenever you tweak the tempo i guess?)?
Thanks for your last answer!

[@RoddyDev]

@@isaNYyN 15 samples was returned by the formula floor(audio sample length / (sample rate / refresh rate)). If you get a longer audio file you will get more samples. If the refresh rate is different, let's say 50Hz for PAL, it would return less samples, to account for the slower clock. If it was the same as 60Hz, it would sound too slow.
The only N163 limitation is in it's 256-byte RAM, which is shared with the wavetable and audio channel data at the same time (16 bytes per active channel). And that's the reason why we use the Nyquist script to compress the waveform to make it fit.

[@isaNYyN]

@@RoddyDev Thank you for your answer

[@tails_the_god]

hey is there a it one for sega genesis vgms?

[@RoddyDev]

that's quite impossible since the YM2612 uses FM not samples

[@brockpiano]

what version of windows is this?

[@RoddyDev]

Windows 7

[@izuki8233]

Is there a way to do this with wav files?

[@RoddyDev]

You can import WAV in both OpenMPT and Audacity. I recommend converting it to mono first.