James-F wrote:

That is why I asked for a recording of the game GODS where the music consists of the WAV and OPL combined so we can tweak the balance by ear even though the subject sound card has improper lowpass filter.

You might also want to try some of the Renaissance demos then, eg this: http://www.pouet.net/prod.php?which=2902
Or the separate player: http://www.pouet.net/prod.php?which=4168
They developed a tracker which used both FM and multiple digital channels (basically a sort of MOD player with software mixing).
It's also interesting to test whether clones have the same latency on OPL vs WAV. I know the demos don't work well in most emulators because the OPL and WAV aren't in sync. On a real SB it works flawlessly though.

Jepael wrote:

Just in case anyone is wondering, the OPL440 sets the channel to additive synthesis mode, so the sine wave amplitude is actually sum of two operators running at full volume.

I wonder how the pan law works when playing the opl440 test tone...
The YMF3812 is mono and the YMF262 is dual mono, does the test tone play using both channel on the YMF262?
If so the resulting sound would be 3db louder on the 262 than 3812.

James-F wrote:

I wonder how the pan law works when playing the opl440 test tone... The YMF3812 is mono and the YMF262 is dual mono, does the te […]
Show full quote

Jepael wrote:

Just in case anyone is wondering, the OPL440 sets the channel to additive synthesis mode, so the sine wave amplitude is actually sum of two operators running at full volume.

I wonder how the pan law works when playing the opl440 test tone...
The YMF3812 is mono and the YMF262 is dual mono, does the test tone play using both channel on the YMF262?
If so the resulting sound would be 3db louder on the 262 than 3812.

On SB 1&2 there are one YM3812 with single mono DAC, so mono digital stream is converted to analog audio and is then sent to left and right outputs.
On SB Pro (1) there are two YM3812 and two mono DACs, one for left and one for right outputs, so (effectively) same analog audio on left and right outputs.
On SB Pro (2) and SB16 there is one YMF262 with one stereo DAC. The amplitude of FM channel is the same whether it is sent to left or right or both outputs. And in OPL2 mode it is sent to both outputs, so again, same analog audio on left and right outputs.

So yes, because you have the test tone on both left and right outputs, the YMF262 plays identical audio on left and right digital outputs.
So no, I don't see why YMF262 playing a tone on both channels would be any louder or different than one or two YM3812 chips playing a tone on both channels.

New results for comparison between classic SB cards and DOSBox using a more accurate test than MPXPlayer,
I discovered that on my system MPXPlayer will not behave properly with a SB1 CT1320C card so I found a better source for PCM sound that translates between cards much better than MPXPlayer.
The new PCM reference is a constant Rifle (weapon 3) shot in Wolf3D with music and fx disabled..

SB1.5 CT1320C PCM: -7.70 wolf3d rifle 3 OPL: -13.31 opl440 d: 5.61 […]
Show full quote

SB1.5 CT1320C
PCM: -7.70 wolf3d rifle 3
OPL: -13.31 opl440
d: 5.61

SB2.0 CT1350B
PCM: -10.07 wolf3d rifle 3
OPL: -15.80 opl440.
d: 5.73

SBPro2 CT1600
PCM: -16.74
OPL: -21.4
d: 4.66

SB16 CT2230
PCM: -6.57
OPL: -11.66
d: 5.09

Audician 32
PCM: -5.3
OPL3: -9.3
d: 4.0 (OPL is louder than SBPro2).

Aztech Sound Galaxy AZT2316 (Multimedia Pro 16 ABI).
PCM: -17.04
OPL3: -22.33
d: 5.29 (very nice balance).

DOSBox 200 (vanilla 0.74).
Shot: -9.6
OPL3: -12.06, -12.04, -12.66 (nuked, compat, fast)
d: 2.44

DOSBox 150
Shot: -9.6
OPL3: -15.55, -14.54, -15.15
d: 4.95

DOSBox 140
Shot: -9.6
OPL3: -15.15, -15.14, -15.75
d: 5.55

Since DOSBox has a filter on the PCM (SB) that sounds closer to the SB1/2 I will choose 140 as the standard FM volume value.

Info:
the d value is the peak difference between FM and PCM.
PCM: Constant Rifle (weapon 3) shot in Wolf3D with music and sound FX muted (none).
FM: opl440.com test software.
The 140/150 value after DOSBox is a volume level for FM only in a DOSBox SVN test build.

I just measure peaks on a recording like this:

Sound Card Lowpass Filtering Measurement

To measure lowpass filtering of any sound card you can use Mpxplay for DOS (link, also attached) with the attached Test WAV package to play white noise and capture with spectrum analyzer.
You can also use Wolfenstein 3D rifle PCM sound (weapons 3 and 4) to compare lowpass filtering between sound cards.
Same for OPL3 lowpass, use Adlib Tracker 2 player called adt2play.exe to play the noise.a2m file and capture with spectrum analyzer.

Results should be similar to this:
* White noise PCM should be flat without filtering, so it's easy to see what LPF the sound card uses.
** SB Pro 2 (CT1600) is a simple 2nd order Butterworth at 3.2kHz.