Talk:Super NES Programming/SNES Specs

I never understood: Low-pass filter for improved quality of low-frequency (bass) tones

A low-pass filter lets everything UNDER a sepecific frequency through. A low-pass filter at 20Khz would let everything under 20Khz PASS through, that's why it's called a low-PASS. How a low-pass filter would improve bass is beyond me, as it would only let bass through!

I think what is really meant is that the SNES has a hi-pass filter to improve the quality of low-frequency sounds. But it's written the above way everywhere.

Also, address decoders are NOT required for ROMs over 32Mbits (they're just used because ROMS over 32Mbits were prohibitly expensive).

It doesn't have low pass or high pass, it has an 8-tap FIR filter that is applied to the ECHO output. It can improve bass tones because it can do a low-pass effect on certain channels and thus cut out most high frequencies generated from resampling/errors in the sample data.

As you stated, a low pass filter allows the lows to pass, hence the name. Applying a low pass filter to lower quality samples helps to allow more accurate reproduction, leaving out high frequency noise in something such as a bass guitar-effectively helping to reduce the bandwidth used by the sample. For example if the prevalent frequencies are under 5Khz and we apply a low-pass with cut-off at 4.8Khz, we are rolling off the high end of the spectrum of the sample, effectively reducing its bandwidth. For more information on filters and cuts, there are many examples on YouTube.
 * Response:

If you were to use a high-pass (only allowing high frequencies through) on a sample that was strong in the low end, you would effectively reduce the sound below the cutoff threshold, and end up with only high frequency noise. Remember that this is not to be confused with sampling rate, but actual frequency spectrum.

For example, a 30KHz Sampling rate has a range from 0-15KHz of recorded audio frequency range.

HOWEVER, this filtering is done digitally even though the methods I described were simplified, and you can find more information on this, (including FIR) in the Signal Processing Section @ http://en.wikipedia.org/wiki/Signal_processing.

I may have gotten some of this wrong, but just trying to help out :) If this information needs correcting, please feel free to!

Aetern142 (discuss • contribs) 10:23, 15 January 2012 (UTC)

Misinformation in "Most common display modes" that I tried to fix
I fixed the definition of Mode 1 to be "2 16 color per tile scrolling layers; 1 4 color per tile scrolling layer" instead of "16 colors per tile; 3 scrolling layers", as the latter implies that all layers in Mode 1 have 16 colors available per tile. However, only the first two layers have 16 colors per tile; the third layer only has 4 colors per tile. I'm unsure why my revision from July was rejected, as it was fixing misinformation. --180.150.49.44 (discuss) 12:08, 24 September 2019 (UTC)