The "Is it possible to rip this?" Thread by VGSB at 8:44 PM EST on February 19, 2009
So before I assume that something can't be ripped right now and start recording it, I should ask a few questions first.
1) Has anybody been able to rip music from the Wii channel songs that can be saved to Flash card? 2) Has anybody been able to directly rip music from firmware channels such as the Mii Channel? 3) Does anybody have a .brsar of Wii Fit, and can anything be extracted from it? 4) If I have a .wad of a WiiWare channel, what's the general method for extracting the music? 5) I've only been able to rip the pong hockey music from Wii Play. If anybody has been able to rip any more, please tell me.
On a somewhat related note, for the Wii songs that can not be ripped, I'm wondering what the best recording settings would be. Even though first-party Nintendo game streams are generally 32 kHz, I wonder if the samples are at a higher rate.
6) Is there a way to tell the sample rate of songs that are sequenced? 7) Are there any games that use an audio sample size greater than 16-bit?
Re: sample rate, the sample rate of the sample doesn't matter. The GC and Wii sound hardware is a single DAC, so it only works at a single rate that everything else is resampled and mixed into. Likewise if they did happen to have a sample size > 16 bit (most are only 4 bit fyi though decoded to 16), it wouldn't matter as the DAC only does 16 bit.
hcs, what is that single rate that everything is remixed into? And if the sample was at a lower sample rate though and upconverted, wouldn't the quality sound basically the same if I record the audio at the sample rate of the highest quality sample? Or would I lose artificial quality (such as how there is a perceivable better quality when DVD-Video players up-sample to HD resolution)? Same question for sample size. Knowing this info could really help compression efficiency.
I assume that it is variable. If you sample at a lower rate you are not losing "artificial quality", you are just sampling at a different rate. If you have a digital channel to the console you're possibly missing samples, if you're making an analog recording you're likely getting somewhat interpolated values between the actual digital samples anyway. Also note the Nyquist sampling theorem. What I meant with regards to the instrument sample quality is that the fundamental feature of a synthesizer is resampling samples to different speeds. A 32khz sample will only be played at 32khz for middle C or something, and then it is still resampled to whatever rate the DAC is running at. The upshot of all this is don't worry about the frequency if you're doing line in recording; you won't get something precise anyway unless you're really grabbing the digital samples before the DAC at the same rate that it is running.
I think I just now understood what you were referring to when you said you assume it's variable (the sample rate used by the DAC). If that's the case, then is there a way to tell the maximum sample rate capability used by the DAC?
-------
Thank you hcs for taking the time to explain and link me to that article. One question I don't believe you've answered (and I don't know how to find the answer to) is this: Is there a way for me to tell what the DAC resamples at? Let's take Mario & Luigi: Partners in Time for example. The DS emulation is also emulating the DAC resampling you mention, correct? How do I find out the maximum sample rate? Is the user setting for sample rate modifying the DAC resampling, or is it a resampling of a resampling? If it's the former, is there a way to tell the max DAC sample rate used by the actual hardware when playing the song? Does this value vary from game to game or song to song?
I think I sort of understand the theory, but one thing I don't get is why it wouldn't be optimum to use the maximum resample rate of the DAC, if it is the last in the chain to modify the audio. I understand that the analog connection adds noise, but it still seems to me that using the DAC sample rate would still produce the most efficient results for later compression (when trying to squeeze as much accuracy out of a high variable bitrate mp3).
Is there a general "likely maximum sample rate" that a DAC resamples to for sample playback such as the Nintendo Wii?
Allow me to answer your question with a series of desperate flailing motions.
Is there a general "likely maximum sample rate" that a DAC resamples to for sample playback such as the Nintendo Wii?
The way these things are typically constructed is the DAC rate is some fraction of a master clock rate. If it's 100 MHz (I don't know what the GC/Wii uses but this is a reasonable ballpark), a DAC "rate" (properly "period") would be set somewhere in a hardware register to a value like 2267, 100 MHz / 2267 = approx 44111 Hz, a decent approximation of 44100 Hz. This works by waiting 2267 cycles of the master clock to update the DAC with a new sample. If you want to consider the maximum rate of the DAC, it would be the master clock rate (obtained by setting the period to 1). Of course nothing uses anywhere near that frequency.
The DS emulation is also emulating the DAC resampling you mention, correct? How do I find out the maximum sample rate? Is the user setting for sample rate modifying the DAC resampling, or is it a resampling of a resampling?
For the DS I don't know, it has hardware synthesis which might have a fixed output frequency for the final mix. If your emulator allows you to set a sample rate then it is doing its own resampling on top of that which is implicit in the synthesis. Note that the last step, the DAC, is typically not a "resampling" step in hardware, it just converts the digital value to an analog voltage level, fetching a new digital value updating the analog level at some constant rate.
I understand that the analog connection adds noise, but it still seems to me that using the DAC sample rate would still produce the most efficient results for later compression (when trying to squeeze as much accuracy out of a high variable bitrate mp3).
I linked you to the Nyquist-Shannon sampling theorem to specifically address this. If you actually know the rate at which the DAC is running (say 32 KHz), and you are recording by sampling at some rate (as you will with a line-in recording unless it happens to be digital and synchronized end-to-end), if you actually want to capture the signal at 32 KHz you must record at at least 64 KHz, else you risk losing high frequency components. Note that the source signal is 32 KHz itself so that it can contain recordings of 16 KHz components. If you only want to capture approximations of those then 32 KHz is fine, but if you want to reproduce the signal that the console is putting out you will need a higher rate. Considering that you're converting to MP3 anyway, reproducing the signal as accurately as possible is probably not a concern to you. So I reiterate that you should just stick with what sounds good. Of course in the case of the emulator, if you can find a way to have it not resample the output from whatever the internal frequency is, that would be preferable.
Thank you hcs for the flailing motions! The article went above my head in a few areas, and so it was difficult for me to digest the most important information. Your last paragraph helps a lot.
I am going for accuracy to some degree, as I'm offering lossless in addition to mp3. But recording audio at 88200 Hz to retain those high ends would be too much overkill in terms of bandwidth and project backup. I'll use 44100 Hz for everything, except the sample rate determined by your Winamp USF plugin, as well as 32000 Hz for SNES audio. Thank you once again!
Below is directed towards everybody, not just hcs:
I've sabotaged the thread I started with those last questions though. Here's the most important question that prompted this thread in the first place: is there any tool or guide for ripping music from Wii channels? If I don't know the answer to this question, I'm going to be making recordings with SFX and ambient sounds.
I would assume Wii Channels have sequenced music. Wii Fit has music in the .brsar and trying to rip it is still under thought. I don't have a clue on how to rip from a .wad but someone else might know. I haven't taken a look at Wii Play but I will soon. :)
A couple of tools can unpack the wads, the one I've used is just called wadunpacker. You'll end up with a bunch of numbered .app files, which can themselves be unpacked with something like u8tool. However some of those (0000002.app in my experience) are compressed executables, I don't recall the program I used to decompress those. The individual files in the other app are very much like those in Wii games, though there is some tendency for them to be compressed. I think u8tool handled those decompressions as well. Some wiiware games have streams, but I think you'll find that the channels are likely sequenced.
Sorry to butt in randomly like this, but seeing this thread compelled me to ask something.
Has anyone ever managed to rip the ambient music that plays in the Gamecube BIOS? You know, the background music you hear while you're dealing with options, memory cards etc. Is it at all possible to rip it? Not that I have the necessary equipment to do anything myself...