Please humor me.  I'm hardly the erudite video game soundtrack listener as most of you seem to be.  Couple basic questions:

1) When people use the term "streaming" does that always mean music created with real instruments/orchestra/etc or licensed music (ie, GTA)?  And when you use the terms "MIDI/sequencer/sample library/etc" does that always mean the simulating of real live instrument sounds (ie, most Japanese RPGs)?

2) Does a gamerip mean it may be of lower sound quality than an official release?

Thank you.

Depends on the software and hardware you use to rip or record it. Me, I record it directly by plugging the audio cables from the console into my PC, and I use Cakewalk Pyro (software) to record it, and depending on the settings I choose, I can get an exact recording without any interference, just as good as the official thing.

It takes a hell of a lot of time though, and how long the song is depends on how long I feel like recording it. I have to choose the time to start recording so it sounds like it's fading in, and I have to edit in the fade-outs myself.

So there's no way to, say with PS1 and on systems, go into the CD/DVD and directly extract the sound file?--it's a purely manual, remarkably laborious procedure?

"Streaming" means that on the actual game disc, the music is stored as a streamed format.  Stream is equivalent to WAV, MP3, APE, FLAC, AAC, ATRAC3, that kind of thing.  It means that it is a recording.  Whether or not the music is orchestrated, synthesized, licensed, whispered, live, makes absolutely no difference.

"Sequenced" means that on the actual game disc, the songs are stored in two (ok sometimes more but two for the sake of this argument) parts -  the samples (which is typically recordings of single notes played by an instrument) and the sequence (which is the digital sheet music that tells the console how to play the samples.) - Equivalents of sequenced music include MIDI, IT, S3M, MOD - long ago, when both memory space of the console and media storage space of the console were limited, it was much more advantageous to have sequenced music, because it takes up much less space. (For comparison, ALL of the music data for Final Fantasy VII for PSX can be stored in a MEGABYTE or two, but the OST takes up four CDs - that's about 2GB uncompressed.   Tons of space saved there using sequenced music.)

This is a very tough question, but the answer is almost always "yes" - the only advantage one gains by RECORDING (not ripping, but that's a whole 'nother ball of wax for another time) tracks as played from the console is that the number of loops can be infinite.  However, the majority of released OSTs are mastered from the original soundboard or whatever using extremely high-quality equipment, so the quality at least of the recording is going to be higher.  The classic example of this seems to be Final Fantasy XII -- apparently the OST is not merely a recording of the game's sequences as played from the console, but actual recordings of a real orchestra playing the tracks.  I don't own the OST and have only listened to the PSF2s, so please someone correct me if I'm wrong.

http://groups.yahoo.com/group/psf_rippers

A wealth of information there regarding ripping (digitally extracting via the means of popping the disc in your computer and digitally converting the files found within to something a PC can read and play back)

PS1 is pretty easy to extract streamed tracks... they're usually compressed ADPCM or XA.  Do some googling and see what you come up with. I personally recommend the programs PSMPlay and MFAudio.  If the words "offset" and "interleave" scare you, you may not be cut out for ripping.

Leaving work now, be home later, hopefully this topic is still hoppin'.  It's one of my favorite things to harp on about.

It's just a matter of whether music is pre-recorded by composers' full equipment or played by console's embedded samplers, not live instruments or MIDI.

Even if streamed, a majority of game soundtracks are still based on MIDI, because game composers often don't have a capacity to arrange/orchestrate thier music and don't have an enough budget to hire studio musicians or a real orchestra.

The opposite can be true. Even if sequenced, game composers (and sound programmer/sythesizer manipulator) still can manage to use live instruments (e.g.  FFVII, FFVIII, FFIX, FFX, FFXI, FFCC, KH2).

Yes. When it's streamed music, since the bit depth and sampling rate of compressed format are always lower than 16bit 44kHz, it could sound unclean and somewhat noisy. When sequenced, it depends on what it is. It ranges from "identical" to "clearly different".

Nah, FFXII is a MIDI-based soundtrack, except for "Opening Movie", "Ending Movie", "Kiss me Good bye" and "Hope".

A big thank you to you both.  Really exceptional and easily understood explanations!  For the first time I finally feel like I understand the gist of the streaming vs sequenced issue.

I have all but the KH2 soundtracks but I don't remember there being much of any live instruments in those other OSTs besides the opening theme, vocal ballads, and such--and the FF9 Plus OST.  Or do you mean there was actually minimal use of live instruments throughout the tracks?

With today's consoles (360 and PS3), when they are using pre-recorded/live music are they streaming using WAV, MP3, APE, FLAC, AAC, or ATRAC3 files?  With the previous generation, what file type did they use to stream?  And a final curiosity: take a game like GTA3, when you're listening to the many licensed tunes in the game is it at CD quality or lowered quality, like the case of FF12?

Thank you again.

Been thinking bout couple more issues...

What would the differences be between streamed MIDI music vs sequenced MIDI music?  Simply disc space usage?

I thought over that statement and realized I don't quite understand it.  If sequenced music, as explained by Amber, is music that's created by having a sequencer play the samples it's been instructed to play, how can there be live instruments?  Or do you mean the samples themselves are the live instruments?

Thank you!

Sound quality. Sequenced MIDI music is liable to be uncleaner, more noisy, less colorful, less subtle and less loud. It sometimes has unbearable hiss noise or aliasing noise. That doesn't mean sequenced music is inferior to streamed music in all aspects, as the former tends to have a old school flavor. But that's another story. The best example is actually FFXII. Please compare the OST with the psf2 set which you can hear by following what Amber posted. I prefer the sequenced version of "Battle For The Freedom", since Sakimoto's trombones are sometimes awful. But that's another story.

Yeah,  I didn't mean to say live performances were featured throughout those soundtracks, except for FFCC. IIRC, there are only less than 10 tracks that have live instruments as short passages - e.g. "One winged Angel" (chorus),  "The place to return someday"(flauto traverso), "Selbina" (fiddle), "Lazy Afternoons"(sax) and blah. FFCC however may be an exception, as it sounds as if most tracks had early instruments or whatnot. My point is, just because music is sequenced doesn't mean it doesn't have any live performance.

Every post-SNES console has a sampler (not a sequencer or a sythesizer) as its music/sound function. Thus, it's possible to sample a short passage of live performance. However, I think "Otherworld" from FFX is an exception, as the entire vocal section of this track seems sequenced, not sampled as a passage.

Not sure, but I think PSX often used CD-XA PCM, while SS, PCE and PC-FC use CD-DA PCM. Today, some game companies use a standard PS2 ADPCM format (called VAG?), some have a license of ADX(AFS) format and SQUARE ENIX recently uses even an ATRAC-based format in FFXI and Dirge of Cerberus.

Lastly, I'm not the best person to talk about this matter. As there are more informed persons in this forum, I'd better leave before making hilarious statements.

Thank you for that.  There's still some stuff that's perplexing to me but it's more cause I don't really understand the functions of a sampler, sequencer well enough to process all the explanations.

Maybe I'll inquire little further if you don't mind.  A sequencer plays sounds it's been instructed from a bank of samples; what does a sampler do?  My understanding is that it holds a bunch of samples but as you say consoles since the SNES now all have a sampler (and not a sequencer), than how does sequenced music get played in the absence of a sequencer?  Does a sampler also handle the duties of a sequencer?

I suspect I'm way off.

[Actually, I think I need to do further research. One of the most important details I'm trying to clarify seems to have changed since the last time I read up on it]

(expanding a bit on what Cedille posted in Post #6: )

"Streamed MIDI music" means that the music was recorded beforehand using the composers' own sound equipment. Their equipment will likely have higher-quality instrument samples, special effects, and give them more control over the mixing. The game will contain these recordings, and it just has to play them back as they are.

"Sequenced", on the other hand, means that the game contains music data in the form of sequences and samples, and the game uses them to generate the music by itself. The console can't do the same stuff that the composers' dedicated sound setups can, so the original sequences may have to be scaled back, and the samples are of lower quality.

Summary:
Using pre-recorded streaming music gives the composer/sound team more control over how the music will eventually sound in-game. Sequenced music is more limited, and its sound is dependent upon the power of the console and the skill of the sound programmers.

I'm not an expert on audio hardware, but this is how I understand it:

A sampler is a piece of hardware that is given audio samples and information on how to play them (pitch, volume, length, etc.). This is the role of a console's sound chip (from the SNES on): it takes audio samples and plays them back based on the information they're given.
A sequencer is just software that records or reads musical sequences. It doesn't technically have the ability to play back music, it just interprets the data and sends it to something that can.

A game's sound program is what passes the samples and audio data off to the console's sound chip to be processed and played. How a program actually works likely varies from game to game.
The game's music program may take the form of a sequencer, reading back the music data and determining what samples need to be played and how to play them, then sends it all off to the sound chip (sampler) for actual playback.

And so, yes, a live performance would be recorded and included in the sample bank as an instrument, and when the sequencer comes across the instructions to play that "instrument", it sends the clip off to the sound chip to played.

...But then again, I could be totally wrong.

Sorry for getting off the topic a little, but what about some of the old "Redbook audio" stuff that used to be popular?  If we're using DVD formats now, I dont' think it would be that big of a deal to stream the audio from that.

I remember being lucky enough that my Lemmings 3D used Redbook audio to make it very simple to rip.

That's like streaming but using the Audio CD standard (which is called "Redbook" and makes one able to play it in normal audio CD players).

The problem is twofold: firstly the DVD Audio standard never became any popular so most people (including developers) just don't know about it, secondly in many cases developers would still opt for more heavily compressed sound data as that reduces the amount of disc accesses (different data is usually stored in different locations on the disc, so the laser has to seek for it back and forth, which can add significantly to loading times if not optimised). Now those more compressed streamed sound data isn't standardised, so you can't just play it back in some common player. But like mentioned before the popular ones of those (Sony ADPCM aka VAG, XA, ADX and the likes) can be found, ripped and/or played using some dedicated software.

With regard to one of my statements that confused loveydovey a lot, that meant nothing.

Since he/she wondered how come a console's sequencer could play live instruments as well as sampled notes, I tried to say all of post-SNES consoles had a sampling function, hence it was possible to store a live performance (even though in practice it's hard for either PSX's 512 kilobyte or PS2's 2 megabyte RAM to have a whole live performance passage along with sequence files (.seq) and sample files (.bd) for other instruments. That's why in few cases live performances are featured in sequenced VGM). Thus, when I wrote "Every post-SNES console has a sampler (not a sequencer or a sythesizer) as its music/sound function", I did emphasize that consoles had thier own samplers that could store sampled notes and/or passages and bend pitches of them. Also, maybe I thought sampler softwares could sequence music as sequencers (aka digital audio workstations). 

But it's not an apt remark. Every post-SNES game console has a sequencer as well as a sampler (though honestly I'm not familiar with it). Considering my poor language skill and little knowlege, yeah, I shouldn't have stated my smattering in the first place. It's sort of like an arcana of VGM.

That's exactly what I was hoping to clear up, and you managed to do it in far fewer paragraphs. I'll introduce my thoughts later, if another post provides a reasonable hook for me to salvage them.

Thank you again everyone for your replies.  This has been a really great read.

So if the sound quality of sequenced MIDI music is lower than streamed MIDI music (as Cedille stated earlier), what reason would a developer have for choosing sequenced MIDI over streamed MIDI?  Lower cost and lower consumption of console resources?

Here's a question that should go a long way in helping me understanding some basics: does a console's sampler already have a library of samples in it and the video game composer has to create music based on the samples available in that console's sampler or is a sampler completely empty, so that a composer writes music using his own samples which are then transferred into the console's sampler when you start playing the game (and removed from the console's sampler when you take the game disc out)?

Would that something usually be a sampler?

When you say "live performance would be recorded and included in the sample bank" are you saying an entire song can be fit into one single sample (meaning 1 sample equals 1 song) or do you mean all the instruments that make up a song are sampled individually (so if there's 5 instruments, there's 5 samples) and accompanied by sequencing instructions, or do you mean that every single note in a song by all the instruments are individual samples/sampled individually (resulting in hundreds of samples for one tune) which are accompanied by sequencing instructions?  Obviously I don't know the first thing about samplers.

Basic question:  the sequencer and sampler are considered components of a console's sound chip or separate from it?

So a sequencer without a sampler is nothing but a set of instructions and cannot play back music and a sampler without a sequencer is nothing but a bank of sounds and also cannot play back music?  Would that be a correct statement?

Thanks all!

Pretty much.

The instrument samples are originally part of the game data, and are loaded into the sound chip.

In this case, yes.

A sample is generally a very short audio clip, representing a particular instrument playing one particular pitch. (Sometimes, an instrument will have multiple samples for varying pitch ranges) The sampler can manipulate the sample so that it sounds at a different pitch.
The sequencer's job is to tell the sampler what samples to play, and what pitches to play them at.

A short section of a live performance can be used as a sample too. Longer performances would likely be broken up into smaller clips. The sequence would contain instructions to play the live-performed clips at the appropriate time. (With no pitch modifications, of course)

The sampler is part of the console's sound chip.
The sequencer is a program that communicates with the sound processor. I'm not familiar enough with console architecture to know if it's included on the sound processor, or if the games all provide their own.

Almost.
The sequencer isn't the set of instructions, it's what reads the set of instructions.
The sampler isn't a bank of sounds itself, it has to be given the bank of sounds before it can do anything.

Thank you, thank you, and thank you Warpster.  Great answers.

So that means if a song requires a saxophone to play notes from the C scale, all you would need to do is make a recording file/sample of any note in the C scale and the sampler would then be able to emulate any sound in the C scale?  If yes, is that what game composers usually do for sequenced MIDI and streaming MIDI music--sample one note from an instrument, then let the sampler create the rest of the needed notes via a MIDI keyboard connected to it?

Well, almost. The sampler can manipulate a sample into any note. They'd record the sax playing a C, and the sampler can use the clip to produce a C#, D, D#, E, etc.  Though, like I said before, one instrument may have other samples for higher/lower octaves.
The composer writes their music, and then the samples are used to play the notes in it.

In most samplers (even the professional ones) playing samples in different pitches is achieved by playing the sample slower or faster (think of the effect of playing an analogue tape or an LP slower or faster respectively). So if it was recorded in 44.1kHz, 22.05kHz is one octave below and 88.2kHz is one octave above the original recording's pitch. Since the sampler's output sample rate is fixed there ideally is some kind of interpolation done while mixing all samples together, e.g. SNES, PS1 and PS2 all use Gaussian interpolation, while N64 and GBA games' uses no interpolation at all to save CPU cycles as the mixing has to be done in software. Also the space available for all samples (and sfx) is limited (SNES: <64kb, PS1: 512kb, PS2: 2MB) so using looped samples and only one or very few recorded pitches as a base for instruments is very widespread (resulting in the typical unrealistic sounds).

"Streamed" MIDI also uses (hardware or software) samplers, but it could be called "streamed" MIDI (I personally heard it here first) since in many highest quality sample libraries every single pitch and plenty velocity levels are recorded separately and streamed off the HD (made popular primarily by GigaSampler and the likes as well as the abundance of RAM and HD space on PCs nowadays). This makes looping and interpolation unnecessary in most cases (but it is, next to video rendering, the most space sucking work one can do on a PC). So if the composer has access to good sample libraries and know how to combine and mix the instruments to sound realistic it can get pretty close to perfect live studio recordings.

If the sampler goes farther and farther up or down from an originally sampled note (say a C), does it sound less realistic than the C# a half step up from the originally sampled C or does the entire range of sounds created based on one sampled note sound uniformally realistic/unrealistic?  It seems in your last sentence, you seemed to be inferring the former, but I wasn't sure.

It will quickly sound unrealistic to trained ears, especially when the instrument's native pitch range (which is often rather narrow anyway) is crossed. The result may or may not sound good, but the instrument is mostly no longer clearly identifyable. Additionally lower pitches become unrealistic by losing the harmonics (every octave lower halves the sample rate which again halves the audible frequency range). Higher pitches become unrealistic by making looping obvious (it is hard to make a smoothly sounding loop point, the higher/faster a sample is played the more noticeable a clicking sound from the loop point can become hearable) or having a non-looping sample being played only for a very short time (the sample playback time is reduced by a half every further octave up).

If the sample's pitch is only changed a few couple notes up or down the result can still be acceptable depending on the recording and the instrument.

Btw. it's interesting that you mentioned a saxophone as example before. Since that instrument is highly expressive it's hard to get it sound good in canned sampler recordings.

Thanks Datschge.  Just the info I needed.