Even with the latest version, as I said before, the bass channel is significantly lower in volume than it should be in this mapper.
Firstly, thanks for hunting down those recordings.
I did an A/B comparison using my headphones.
While I agree that there is a slight difference in this particular patch (I used the same track as a reference you posted in the youtube video), my personal impression is that the difference is not nearly big enough to call it "significantly".
What I can also confirm again (which is what Asbrandt has been claiming) is the lack of low-pass filtering on my FPGA implementation. For example, the bell-like instrument sounds way sharper on my core.
What I do also hear in the original HW recording is a significantly "dryer" bass reproduction, compared to the youtube video you posted.
I therefore stand by my claim that the youtube video has been "enhanced" a lot by applying EQ. Using headphones, I can easily spot an additional chorus effect.
Just to show you what a bad idea it is to reference non-HW recordings:
https://www.youtube.com/watch?v=h_2wm6IzavA&index=8&list=PL2917FA48A56DB74CThe perfect scenario would be to raise the bass channel and lower the PCM drums, although I'm sure it's not that easy, since Necronomfive did mention getting clipping with the mix as it currently is.
You seem to completely ignore 2 important statements from my previous post, so let me reiterate them:
1. I never ever did claim that my mapper sounds 100% identical to the original HW. I even acknowledged Asbrandt's observation that the high frequency filtering is absent on the Everdrive N8 (crispy sound).
2. The difference in the bass reproduction CAN actually be attributed to the difference in HW mixing between the FPGA<>VRC7, and how the signal is further processed in the analogue domain on the Everdrive N8.
I already made a clear statement regarding this: I won't be adding any kind of EQ processing in the FPGA to countermeasure the difference in the analogue sound reproduction. All my concern is that the
digital model closely resembles the real VRC7 / YM2413.
In this regard, the only point which can still be improved are the parameters for the instrument rom (which have been, so far, reverse engineered by ear). But this issue also affects the implementation on the emulators. To get all parameters correctly, the VRC7 needs to be decapped.
Just one last point:
I work as a sound engineer in post-production for feature films myself, so I'd like to think my professional background would give my findings just a little bit of legitimacy
Frankly, I don't give a hoot about any attempts at argumentum ad verecundiam.
In my POV, regarding this topic, you'd gain more credibility if you would have created some music tracks on Yamaha FM HW before. Or did some VRC7 / Yamaha FM programming / reverse engineering.
Then, your point / argumentation would have looked quite differently right from the start.
I appreciate all kind of bug reports, but this here is clearly out of scope for this project, since it's not even a bug.