Because of the ridiculous amount of space you'd need even when compressed? (remember the Z80 isn't all that good at mixing)Quote:
Originally Posted by kool kitty89
Printable View
Because of the ridiculous amount of space you'd need even when compressed? (remember the Z80 isn't all that good at mixing)Quote:
Originally Posted by kool kitty89
Exactly what Sik said, we're not after streaming here, its completely worthless in majority of cases.
There's ADSR etc. but there's no space to hold your samples, only a tiny shared pool of memory into which you can fit some shitty quality SFX and music samples, and a simplistic playback program and music+SFX data... realtime updates are not worthwhile in an action game unless you want a ton of slowdown...
The echo buffer abuse is lot less attractive, you got to push in 16bit uncompressed samples in there, and there is no space to hold anything... you could stream in something but streaming in a game is not going to happen unless nothign is going on in that game, and that you got a shitton of ROM space, which you don't have...
Technically it can be useful if the SPC700 is generating samples on-the-fly (essentially synthesis). Still more useful than the YM2612 if you consider you aren't restricted to sine waves and can do more crazy stuff, but I don't know how fast is the SPC700 to be able to do all that (for all we know it could be so slow it's useless).Quote:
Originally Posted by TmEE
You got 32 cycle timeframe form what I recall, there's not a whole lot you can do during that time...
He means just having whatever collection of sounds you need in the same sample. Like a string ensemble, or chords. That was one of the boons of sampling in the early days, which saved both channel space and memory. Genesis can do simple fifths or chorus effects using 2x 2op signals, though it won't sound as "big" ass a full ensemble.Quote:
Originally Posted by kool kitty89
Of course, but I'm not sure it's useful to conflate those issues. BTW I think the music for the SNES launch titles is uncommonly good, whereas the Journey to Silius music doesn't stick in my mind at all (but maybe that's because I didn't like the game), so to me it's a strange head-to-head. Most of the NES music I like best, I appreciate mainly for its tunefulness or compositional depth, not so much for any overt technical wizardry.Quote:
Originally Posted by kool kitty89
Huh? The Actraiser VGM isn't weak at all, it's a masterpiece! Inasmuch as I talked about "canned" sounds, I meant that no one will mistake the instruments used for the real thing. Some people find that SNES VGM in general falls into "uncanny valley" territory; others find that the technical constraints add aesthetic value, like a Mellotron or the crunchiness of early samplers.Quote:
Originally Posted by kool kitty89
Well, up to a point, but I used some of that hardware BITD (like the Yamaha DX7/TX81Z which use FM synthesis), and I also used early-ish sample playback modules like the Proteus and Kurzweil K2000. When you're trying to write orchestral music the differences between FM synthesis and sample playback are pretty striking.Quote:
Originally Posted by kool kitty89
FM synths are valuable because they sound like themselves -- and it's a great sound! -- not because they're good at imitation. You can get by with them in commercial music that use "synth strings" et al., and in a VGM context they make sense, but stack up a FM piano to a sampled piano -- even an early cheesy one like the K2000 piano - and there's just no comparison.
And again, the difference between what's possible, and what's practical, is crucial here. One of the core assets of the SNES hardware is that it allows composers to pretty quickly set up a sound palette which is familiar to them and to listeners as well. Whether that's a good thing or not in aesthetic terms -- I personally think the SNES sound hardware is a great confluence of new possibilities and interesting limitations -- the point is that people respond to it, and it posits a soundworld which immediately feels different from anything based on synthesis at least in the way that synthesis was used at the time (especially on consoles).
The Tim Follin X-Men/Gambit track I posted earlier -- that's just impossible on the Genesis, and trivial on the Sega CD...or is it?
Money and aesthetics were definitely in play, too. And of course, those two things are interrelated: project studio dudes tend to write project studio dude music, most of which has aged very poorly indeed, but they're cheap and usually get the work done on time. The effect of CD-DA on consoles is extremely complex, but at heart it's the same perennial issue: as you remove impediments from a composer, you often find out that the thing he really wants to do is much less interesting than what he does when confronted with severe technical limitations.Quote:
Originally Posted by kool kitty89
Also paradoxically, it becomes damned expensive to "properly" record things like e.g. the Follin example, since there's no longer an excuse, inherent in the medium, for obviously-canned sounds. Of course nowadays you'd just use samples for everything and bring in a guitar player, but back then you couldn't really fake a full band without it sounding like, well, a fake full band.
OK, that makes a lot more sense than what I interpreted the comments as . . . but wouldn't the genesis be able to do that too, at least for 1 or 2 channels? (via Z80 PCM playback)Quote:
Originally Posted by GeckoYamori
If the SPC700 was too slow to do that, there's still some more limited advantages of using added 16-bit PCM channels. (including having the SPC decompress other formats like ulaw, 2-bit ADPCM, CVSD, etc)Quote:
Originally Posted by Sik
For the sort of sound the SNES pushed in general, I wonder if the best alternative for the time wouldn't have been a pure sampler-based system at all or Yamaha FM, but a wavetable synth unit like Ensoniq's chips. (which could also be used to play PCM, but the primary function was wavetable synthesis with FM abilities as well -iirc, including sample-based FM using fully programmable waveforms somewhat like the Saturn's SCSP allowed: a bunch of osccilators that can be used for direct PCM playback, additive synth, or FM synth -using waveforms in RAM- in various configurations)
That sort of set-up would be far more flexible and far more efficient with limited memory (both limited sample RAM and limited ROM), though since only PCM is supported, sampled SFX would either need to be uncompressed or decompressed using the CPU.
Even better would obviously be a set-up like the Jaguar with a powerful, highly programmable DSP, but that wasn't realistic for 1990. (well, a programmable DSP based system might have been practical -like a TMS320/M56k set-up, or a lower-end/custom alternative embedded DSP- but definitely not something on the level of the Jaguar's DSP in terms of flexibility/programmability -which is closer to Hitachi's SHDSP line in some respects)
I wasn't just talking FM either, but addressed a variety of synthesizers in general. (including wavetable -ie Ensoniq stuff- and Roland's LA Synth, not to mention older analog synths)Quote:
Originally Posted by goldenband
They're also great for producing sharp and bright sounds, unlike preceding analog synthesizers and samplers with limited memory and heavy filtering. (wavetable synth would also have that advantage in many cases though)Quote:
FM synths are valuable because they sound like themselves -- and it's a great sound! -- not because they're good at imitation. You can get by with them in commercial music that use "synth strings" et al., and in a VGM context they make sense, but stack up a FM piano to a sampled piano -- even an early cheesy one like the K2000 piano - and there's just no comparison.
Depending on the composer (and programmer) in question, setting up said palette would be quite reasonable on the MD too, it just depends on what synth methods the composers/programmers are experienced with.Quote:
And again, the difference between what's possible, and what's practical, is crucial here. One of the core assets of the SNES hardware is that it allows composers to pretty quickly set up a sound palette which is familiar to them and to listeners as well. Whether that's a good thing or not in aesthetic terms -- I personally think the SNES sound hardware is a great confluence of new possibilities and interesting limitations -- the point is that people respond to it, and it posits a soundworld which immediately feels different from anything based on synthesis at least in the way that synthesis was used at the time (especially on consoles).
In any case, a huge number of developers stuck with the generic samples available in the dev kits . . . or if they DID use their own samples, they tended to use the same generic sounds for almost all of their games. (like Capcom)
Actually, one advantage over the Sega CD would be memory . . . CD-DA has limited duration, and a number of games were constrained by that limit. (albeit there's workarounds there too, like streaming lower bitrate 8-bit PCM -Pier Solar uses that to great effect, and realtime synth in the MCD has the ricoh chip to work with as well, so you get the full 6 FM channels, 8 PCM channels, and PSG)Quote:
The Tim Follin X-Men/Gambit track I posted earlier -- that's just impossible on the Genesis, and trivial on the Sega CD...or is it?
I think this is a cop-out argument that's overused.Quote:
Money and aesthetics were definitely in play, too. And of course, those two things are interrelated: project studio dudes tend to write project studio dude music, most of which has aged very poorly indeed, but they're cheap and usually get the work done on time. The effect of CD-DA on consoles is extremely complex, but at heart it's the same perennial issue: as you remove impediments from a composer, you often find out that the thing he really wants to do is much less interesting than what he does when confronted with severe technical limitations.
Sure there are some cases where luck may have struck and forced programmers/composers to take different (arguably better) routes with chip synth than they might have with a fully open ended recording suite, but those cases are limited.
The vast majority of cases would tend towards better overall quality of music when such recording is allowed . . . and if CD-DA DOESN'T do better, it definitely still could. (I can think of several examples of games that got arranged versions that were arguably worse, but some cases where you even got different releases with different arrangements and some were poorer than arcade/console chip synth while some were just amazing -Raiden comes to mind, the PSX and PCE-CD tracks aren't nearly as good as the FM Towns one).
Hell, on PC games, you already had the semi-pro quality MT-32 being used for many games, so CD-DA could have just used recordings of those awesome compositions.
As far as economics, there's a wide range of professional synthesizers and samplers that could have been used to produce high-quality soundtracks for CD-DA well beyond anything possible on any of the cart systems (well, aside from using tons of ROM -ie Neo Geo games using streaming ADPCM tracks). So plenty of options well short of hiring a band/orchistra to perform.
There are many, many more cases where chip compositions/arrangements were done poorly and could have been far better with CD-DA. This is especially true for US developers who seemed to rarely get good arrangements on console/PC hardware (especially for FM -PC and MD- and to lesser extent SNES). In many cases, you even had some nice compositions, but poor quality arrangements due to composers not being good at working with the hardware . . . and a lack of good software tools to make up for that. (ie feature-rich trackers or such)
Beyond CD-DA, there are examples where quality tools with high quality synth hardware also fared consistently well with the majoirty of composers in the US. (the MT-32 is a prime example, many good -and some truly great- arrangements done using that MIDI module, but very few good -let alone great- arrangements on Adlib/SB in commercial games -a very few games and many more hobbyist chiptunes show the real potential of the OPL2 and OPL3 chips)
Just imagine if PC games in the early 90s had music of this quality: https://8bc.org/members/mad/
https://8bc.org/music/mad/Oskari+goe...ndblasterland/
https://8bc.org/music/mad/Oskari+Wins+the+Game/
https://8bc.org/music/mad/International+Crisis/
https://8bc.org/music/mad/FUCK!+Oska...a+New+Galaxy!/
(and that's without any use of PCM -though another odd thing with PC games was lack of using the SB's PCM channel for music, it was far simpler to use than the MD's Z80 playback, but the MD got far, far, far more use of PCM in music)
Those are all OPL3 tracks, though that chip is still technically weaker in some areas than the YM2612. (4-op synth modes aren't nearly as capable as the 2612, but it's got 18 channels to work with in 2-op mode -and lot of potential for pairing channels for additive synth and/or echo, etc)
The violins are almost absent from the Genesis version. It's pretty clear that the SNES version has a better balance between the lead and background music.Quote:
Originally Posted by Kamahl
One of my favorites. I honestly don't think a snes could generate that kind of power in a song. It's brutal.
http://www.youtube.com/watch?v=qFMsrHCFXj0
I really like the Genesis renditions of the snes songs. They really fit in FM.
1:58, just compare to this:
http://www.youtube.com/watch?v=hcL70l8MkV8
It's using the same sound engine as Mega Turrican without any new instrument patches it seems. I disagree in general though about the balance . . . Amiga may balance it a bit better, but SNES has the violin way, way too prominent. (at least towards the beginning) Regardless of balance, I still prefer the sound of the MD version by far over the other 2. The percussion in the SNES version is above average for SNES stuff and certainly better than those short/weak samples on the Amiga, but not up to par with the great drums from Mega Turrican. (a bit like comparing Mega Turrican and Turrican 3)Quote:
Originally Posted by gamevet
I'd rank it MD, SNES, then Amiga, though that might not be the case for all the tracks.
I'd thought there were some nice recordings of the ST music too, but the only decent quality ones are for game footage (albeit there's no SFX, so it doesn't hurt much), and none of those videos go up to level 3. (the gamplay and music videos that do have crappy sound -recorded too loud and distorted)
However, I think that the BETA SFII version of ALL tracks was always better. (they re-done it due to the PCM playback that would be interrupted by the SFX, but the Beta tracks are awesome.).Quote:
Originally Posted by Bastardcat
http://www.youtube.com/watch?v=TYO8jVmeCoA
I mean, compare this version:
http://www.youtube.com/watch?v=4feZtw4pRVkc
VS BETA version:
http://www.youtube.com/watch?v=QvQAAoYTjoU
:love:
Amiga one sounds pretty bad IMO. The SNES one is pretty good except the lead is a little weak sounding. Not enough 'bending' of the timbre and not pronounced enough. On the other hand, the Genesis one has a the perfect and best lead - but the rest is a little flat and rigid. But since the best part about the song IS the lead, the Genesis one all the way, baby! :cool: One of my favorite tracks for Genesis tunes.Quote:
Originally Posted by gamevet
The amiga one isn't very bad IMO, one of the big problems was the amount of ram wasted by the samples (and jim power uses rather good ones compared to turrican 2 for example). Title screen music on the amiga is usually really good.Quote:
Originally Posted by tomaitheous
Genesis version is the best for sure. The lead and drums sound perfect.
Btw for the discussion above, the violins comment was only referring to golden axe, jim power was a separate discussion.
I find those beta tracks pretty terrible in comparison...Quote:
Originally Posted by Delek