Why did Roland never fix the reversed panpot?

[Rhizome]

I only just found out about the panpot positioning being reversed on the MT-32/CM-32L/CM-64 until now.

However, why did Roland never fix this issue on the CM-32L which was released in 1989? The Roland D-110 which was released in 1988 doesn't have this reversed panpot (it appears Cloudschatze had his reversed). I've actually tested this as according to the D-110's manual, the first 2 bass drums should be at 2>, which as you can probably tell means slightly to the right. With the audio cables inserted correctly, the panpot was slightly to the right - I also tested it with a sustained channel by setting channel 2 to 7> and playing a few notes through it. Yes, it was completely to the right 

I've posted several samples from the 1.10 D-110 (the last revision) over at Alistair's forum. Not only is the panpot in their correct positions, but there is actually some tone differences compared to the recordings Cloudschatze did. The slap bass is much different for a start (listen to Chi Chi Bar 1 and the introduction) and the poly doesn't seem to drop as much - or hardly at all - on Space Quest 3's introduction. So if Cloudschatze did have his cables plugged in correctly  , it would mean Roland fixed this issue after ROM version 1.06.

Just to double check it wasn't the software I was using, I manually changed the panpot on the module it's self. Again, the positioning was correct either at <7, >< (center) or 7>.

I really don't understand why Roland never fixed this. It's obvious they could have since the CM-32L was released a year after the D-110 was 

[NewRisingSUn]

Compatibility with older MT-32 models, I suppose.

[Cloudschatze]

(it appears Cloudschatze had his reversed)

I could be wrong, but I'm pretty sure that the D-110 recordings I did are correct. Since Sierra (mostly) composed to the MIDI specification, which the D-10/20/110 adhere to, no channel reversal would have been necessary for anything but the Quest for Glory II recordings.

...but there is actually some tone differences compared to the recordings Cloudschatze did. The slap bass is much different for a start (listen to Chi Chi Bar 1 and the introduction)

The first recordings I did were with a D-10, which included those from Iceman. After it was discovered that the D-110 driver had not been optimized for use with the D-10, I pulled all of the D-10 recordings. I only recommend comparing output with what is presently available in that parent post (the D-110 recordings).

[Rhizome]

I could be wrong, but I'm pretty sure that the D-110 recordings I did are correct. Since Sierra (mostly) composed to the MIDI specification, which the D-10/20/110 adhere to, no channel reversal would have been necessary for anything but the Quest for Glory II recordings.

Well, what made me think about this was I have the audio cables in correctly and the percussion seems to pan to what the manuals state - here's a scan of the rhythm setup. Another thing I found odd was that the GM patch reverses the sustained channels (so then <7 is actually 7> etc). However, the rhythm channel isn't 

The first recordings I did were with a D-10, which included those from Iceman. After it was discovered that the D-110 driver had not been optimized for use with the D-10, I pulled all of the D-10 recordings. I only recommend comparing output with what is presently available in that parent post (the D-110 recordings).

Listen to Crazy Eights from Hoyle's Book of Games. The strings from your recording seem to fade out longer and you can also hear the difference in the slap bass. Unless that recording was from the D-10 too? 

[Alistair]

If you reverse the reversed MT-32, as in, have it panning like GM does, the drumkit/percussion will actually pan. Normally (as in, reversed), the drumkit will not shift from 64, no matter what you tell it to.

That in itself makes me think the MT-32 is supposed to be done like GM, although the 'centre' is a little to the left when you do that, I think.

- Alistair