Namco 163 audio: Difference between revisions

From NESdev Wiki
Jump to navigationJump to search
(Undo revision 16774 by Tepples (talk): It also affects 5B. The situation is more complex than I thought and must be clarified elsewhere.)
(→‎Channel Update: Updated update frequencies per NTSC CPU = 1.79 MHz and PAL CPU = 1.66 MHz.)
Line 175: Line 175:


{| class="tabular"
{| class="tabular"
! Channels || Update Rate
! Channels<br/>Enabled || Update Rate<br/>(NTSC) || Update Rate<br/>(PAL)
|-
|-
| 1 || 119318 Hz
| 1 || 119.333 kHz || 110.667 kHz
|-
|-
| 2 || 59659 Hz
| 2 || 59.667 kHz || 55.333 kHz
|-
|-
| 3 || 39773 Hz
| 3 || 39.778 kHz || 36.889 kHz
|-
|-
| 4 || 29830 Hz
| 4 || 29.833 kHz || 27.667 kHz
|-
|-
| 5 || 23864 Hz
| 5 || 23.867 kHz || 22.133 kHz
|-
|-
| 6 || 19886 Hz
| 6 || 19.889 kHz || 18.444 kHz
|-
|-
| 7 || 17045 Hz
| 7 || 17.048 kHz || 15.81 kHz
|-
|-
| 8 || 14915 Hz
| 8 || 14.917 kHz || 13.833 kHz
|}
|}



Revision as of 21:28, 9 February 2020

The Namco 163 offers up to 8 additional sound channels that play wavetable samples while the 175 and 340 do not. Each waveform can be of a configurable length, and each channel has linear volume control. It has $80 bytes of sound RAM shared by channel registers and wavetable samples; at least $40 bytes are dedicated to samples, with more available if not all channels are used.

The chip is unable to clock every channel at once, so it cycles though channels, updating one every 15 CPU cycles. Because of this, the chip allows the game to configure the number of enabled channels. When fewer channels are enabled, the channels are clocked more often, allowing for higher tones with longer, more detailed waveforms. When more channels are enabled, clocking slows down since each channel has to wait its turn, resulting in lower tones and a high-pitched whining noise at the switching frequency. Most games using this IC used only 4 channels.

Registers

Sound Enable ($E000-E7FF)

7  bit  0
---- ----
.SPP PPPP
 ||| ||||
 |++-++++- Select 8KB page of PRG-ROM at $8000-$9FFF
 +-------- Disables sound if set.

Sound is enabled on the 163 by writing a clear bit 6 to this register (0 is sufficient).

Address Port ($F800-$FFFF)

7  bit  0   (write only)
---- ----
IAAA AAAA
|||| ||||
|+++-++++- Address
+--------- Auto-increment

Writing to this register sets the internal address to write to or read from. If the 'I' bit is set, the address will increment on writes and reads to the Data Port ($4800), wrapping $7F->$00 (address is 7 bits wide)

Data Port ($4800-$4FFF)

7  bit  0   (read / write)
---- ----
DDDD DDDD
|||| ||||
++++-++++- Data

This Port accesses the 163's internal $80 bytes of sound RAM. Which of the $80 bytes is determined by the Address register ($F800). When read, the appropriate byte is returned. When written, the appropriate byte is set.

This RAM is used primarily for wavetables. The sound channel control registers are also set by writing to certain addresses in sound RAM:

Sound RAM $78 - Low Frequency

7  bit  0
---------
FFFF FFFF
|||| ||||
++++-++++- Low 8 bits of Frequency

Sound RAM $79 - Low Phase

7  bit  0
---------
PPPP PPPP
|||| ||||
++++-++++- Low 8 bits of Phase

Sound RAM $7A - Mid Frequency

7  bit  0
---------
FFFF FFFF
|||| ||||
++++-++++- Middle 8 bits of Frequency

Sound RAM $7B - Mid Phase

7  bit  0
---------
PPPP PPPP
|||| ||||
++++-++++- Middle 8 bits of Phase

Sound RAM $7C - High Frequency and Wave Length

7  bit  0
---------
LLLL LLFF
|||| ||||
|||| ||++- High 2 bits of Frequency
++++-++--- Length of waveform ((64-L)*4 4-bit samples)

Equivalently, the waveform length = 256 - %LLLLLL00 samples.

Sound RAM $7D - High Phase

7  bit  0
---------
PPPP PPPP
|||| ||||
++++-++++- High 8 bits of Phase

The high byte of the 24-bit phase value directly determines the current sample position of the channel. Every time a channel is updated, the 18-bit frequency value is added to the 24-bit phase accumulator, which is stored in these three registers.

The phase registers may be written to immediately set the phase of the wave. It is good practice to set the frequency to 0 before doing so.

Sound RAM $7E - Wave Address

7  bit  0
---------
AAAA AAAA
|||| ||||
++++-++++- Address of waveform (in 4-bit samples)

Sound RAM $7F - Volume

7  bit  0
---------
.CCC VVVV
 ||| ||||
 ||| ++++- Linear Volume
 +++------ Enabled Channels (1+C)

Note 'C' is available on register $7F ONLY. Those bits have no effect in other registers.

  • When C=0, only channel 8 enabled
  • When C=1, channels 8 and 7 enabled
  • When C=2, channels 8, 7, 6 enabled
  • etc...

Other Channels

Above Sound RAM register descriptions ($78-$7F) are for the 8th channel. The other 7 channels are accessed via the same pattern, but each 8 bytes before the last:

Channel 8:  $78-$7F
Channel 7:  $70-$77
Channel 6:  $68-$6F
Channel 5:  $60-$67
Channel 4:  $58-$5F
Channel 3:  $50-$57
Channel 2:  $48-$4F
Channel 1:  $40-$47

Again note that the 'C' bits in the final register is only available at address $7F.

When channels are disabled, their registers are unused, and can be used for waveform data instead.

Waveform

Each enabled channel cycles through its waveform at a rate determined by the 18-bit frequency value 'F'. Each step in the waveform is 4-bits wide, and the number of steps is determined by the 'L' bits ((64-L)*4). Two samples are stored to a byte, which is little-endian (unlike the Game Boy's wavetable channel).

The 'A' bits dictate where in the internal sound RAM the waveform starts. 'A' is the address in 4-bit samples, therefore a value of $02 would be the low 4 bits of address $01. A value of $03 would be the high 4 bits of address $01.

For a visual example, assume you have the following sound RAM:

$00:    00 00 00 A8 DC EE FF FF EF DE AC 58 23 11 00 00
$10:    10 21 53 00 00 00 00 00 00 00 00 00 00 00 00 00

And assume a channel has an 'A' value of $06, and an 'L' value of $38. That channel's waveform would be a sine wave, looking like the following:

F -       *****
E -     **     **
D -    *         *
C -   *           *
B -
A -  *             *
9 - 
8 - *               *
7 - 
6 -
5 -                  *             *
4 -
3 -                   *           *
2 -                    *         *
1 -                     **     **
0 -                       *****

Channel Update

Namco's 163 does not internally mix its channels. Instead, each channel is output one at a time. It takes exactly 15 CPU cycles to update and output one channel. When multiple channels are used it will cycle between them. With 6 or fewer channels, the time to update all channels is a rate faster than any audible frequency, and the difference between this serial output and mixing cannot be heard, but for 8 channels it creates a very loud and apparent noise at the update rate. For a Famicom through RF output, this noise is attenuated during demodulation (which performs a lowpass filter), but through A/V output to a TV that does not filter high frequencies, it can be very unpleasant. Only two games used all 8 channels: King of Kings and Erika to Satoru no Yume Bouken.

Channels
Enabled
Update Rate
(NTSC)
Update Rate
(PAL)
1 119.333 kHz 110.667 kHz
2 59.667 kHz 55.333 kHz
3 39.778 kHz 36.889 kHz
4 29.833 kHz 27.667 kHz
5 23.867 kHz 22.133 kHz
6 19.889 kHz 18.444 kHz
7 17.048 kHz 15.81 kHz
8 14.917 kHz 13.833 kHz

The following is a speculative version of a single channel update, occurring every 15 CPU cycles:

* w[$80] = the 163's internal memory
* sample(x) = (w[x/2] >> ((x&1)*4)) & $0F
* phase = (w[$7D] << 16) + (w[$7B] << 8) + w[$79]
* freq = ((w[$7C] & $03) << 16) + (w[$7A] << 8) + w[$78]
* length = 256 - (w[$7C] & $FC)
* offset = w[$7E]
* volume = w[$7F] & $0F
phase = (phase + freq) % (length << 16)
output = (sample(((phase >> 16) + offset) & $FF) - 8) * volume

The output will be held until the next channel update. The 24-bit phase value will be stored back into w[$7D/$7B/$79].

The sample value is biased by -8, meaning that a waveform value of 8 represents the centre voltage. This means that volume changes have no effect on a sample of 8, will tend negative if <8 and positive if >8.

Frequency

The wave position is driven by the high 8 bits of a 24-bit accumulator. Every 15 CPU clocks, one channel will add its 18-bit frequency to the accumulator. Because only one channel is updated per tick, the output frequency is thus divided by the number of channels enabled.

f = wave frequency
l = wave length
c = number of channels
p = 18-bit frequency value
n = CPU clock rate (≈1789773 Hz)

f = (n * p) / (15 * 65536 * l * c)

Mixing

The relative volume of the IC varies from game to game, unfortunately. The following samples have been recorded using various test programs.

Difference between loudest APU square and loudest N163 square in 1-channel mode (dB)
Game NewRisingSun jrlepage Rainwarrior
Final Lap 12.7 11.2
Sangokushi II: Haou no Tairiku 12.9
Megami Tensei II 13.0 11.9
Rolling Thunder 16.9 16.0 16.9
King of Kings 18.0 17.3
Mappy Kids 18.6
Erika to Satoru no Yume Bouken 18.8 18.9
Youkai Douchuu-ki 18.9
Sangokushi: Chuugen no Hasha 19.5

Based on these measurements, the following submappers were allocated:

INES Mapper 019 submapper table
Submapper # Meaning Note
0 Default Expansion sound volume unspecified
1 Deprecated Internal 128b RAM is battery backed, no external PRG-RAM is present.

No expansion sound. (Equivalent to submapper 2 with 0 in PRG-NVRAM field.)

2 No expansion sound
3 N163 expansion sound: 11.0-13.0 dB louder than NES APU
4 N163 expansion sound: 16.0-17.0 dB louder than NES APU
5 N163 expansion sound: 18.0-19.5 dB louder than NES APU


Because the high frequency generated by the channel cycling can be unpleasant, and emulation of high frequency audio can be difficult, it is often preferred to simply sum the channel outputs, and divide the output volume by the number of active channels. For 6 channels or more, where the switching frequency crosses the threshold of audibility, this approximation will become slightly too loud as it fails to compensate for the transferred energy.

References