Deprecated: mysql_connect(): The mysql extension is deprecated and will be removed in the future: use mysqli or PDO instead in /storage/content/49/145849/famitracker.com/public_html/forum/classes/dbHandler.php on line 29
FamiTracker
Anyone who's spent some time with the actual hardware realizes that the volume of the expansion chips is not actually consistent - for example, one copy of akumajou densetsu will have a louder or quieter VRC6 compared to another one.
However, famitracker has a built in, unchangeable value set for all expansion chips. Which is fine, except the physical hardware varies, and all of the NSF players have either decided on a different standard, or have actually implemented their own mixer so you can decide for yourself what their relative volumes should be.
I am not asking for more advanced or finer volume control in the tracker itself - 0-F (and other values as appropriate per expansion) should be kept. I just want a basic mixer in the configuration so it's feasible to actually write a song for an NSF player, which I feel is an important feature for off-site compos. As it is right now, you have to guess, which (more often than not) requires you to go back and re-do all of the volume settings in the module.
A mixer is planned for that reason, to allow closer matching to certain hardware. That could be used for NSF players too; but best is to check the most commonly used players instead of targeting one specific too much, since anyone could use any player. Anyway there are a few things I need to figure out first, but it should appear in a future version.
The built in volume levels is based on my own hardware btw, so they are supposed to be accurate in at least that sense.
In my experience, the N163 is the most problematic expansion to deal with in terms of mixing. The issue with it is that the fewer channels you enable, the louder the chip gets compared to the 2A03. By itself this isn't a problem, but I found out through experimentation that different cartridges have different components in them that cause the mixing to differ. Generally speaking, cartridges designed for games which used all eight channels (King of Kings being the one I own, though I believe this is also the case for Erika to Satoru no Yume Bouken) have less aggressive damping, since the 8-channel configuration is the quietest one; and on the opposite end of the spectrum, games such as Final Lap and Megami Tensei II will make the N163 quieter because they use 4 channels, and their relative volume would drown out the 2A03.
If that were that, life would be easy. But then you have weirdos like Rolling Thunder and Mappy Kids (the latter of which I do not own, but I have deduced this by examining the NSF file), which use 4 channels but have the same mixing configuration as 8-channel games, meaning the software levels for N163 channels are lower to account for the fact that the chip itself will be louder.
So how do you deal with huge discrepancies like this between one cart and the next? This could be a tricky one to solve.
_______________________
Follow me on [url=https://twitter.com/jrlepage2a03]Twitter.
I record (some) NSFs on hardware. Feel free to [url=http://www.famitracker.com/forum/posts.php?id=3633]request a hardware render.
Well, there are exactly 2 games that use N163 8-channel, and one of them barely uses any 2A03 for the music. It's actually not that hard to find a default volume that's acceptable for all existing N163 rips. Also, the sample size for your tests is too small to really make assumptions about how N163 carts are mixed, and whether particular games are mixed lower or higher. You need multiple copies of the same game, and also multiple tests of the same hardware across multiple Famicoms. Until then, your speculations about consistent mixing variations are really very tenuous.
If you're playing back on hardware, you are already used to dealing with variations, since no two carts/machines are alike. In this situation you're just gonna have to mix it to your own taste anyway.
So... for these reasons I don't really think this is an issue. If you have an option to mix your devices in Famitracker that covers everyone's needs. You don't need two separate default settings for N163.
Well, I tested those things on two Famicoms and five N163 carts which is... more than most people have! Admittedly I don't have two copies of any game, but the NSF rips are enough to make conclusions about the intended mixing for that specific game (Mappy Kids and Rolling Thunder, for example, have the N163 channels really quiet compared to other N163 NSFs, whether commercial or homebrew).
_______________________
Follow me on [url=https://twitter.com/jrlepage2a03]Twitter.
I record (some) NSFs on hardware. Feel free to [url=http://www.famitracker.com/forum/posts.php?id=3633]request a hardware render.
[quote=jrlepage]NSF rips are enough to make conclusions about the intended mixing for that specific game[/quote]
I disagree on this point. You trying to pretend you have the same taste in mixing as the composer, and also presuming that the hardware he tested on wasn't significantly varied from the cart that was actually produced. These are two compound sources of error (actually 3, since the hardware variation is already a compound error of Famicom vs cart).
The worst problem for a default mix, I think, is the FDS, since the Sharp Twin variations are drastically different. At least cart to cart and famicom to famicom things are somewhat consistent. I still need to test my FDS and rewrite the emulation for it in NSFPlay, but across existing FDS games there's horrendous variation in mixing from game to game.
Anyhow, I don't want to sound like I don't appreciate the speculation-- it gives us something to go look at and shine light on. I'm mostly just saying that we should make too many assumptions based on tiny data sizes. There can't really be a good solution to the problem until it can be measured.
