This is a really useful and comprehensive review, much appreciated. I don't have too much to add in addition to what you have already stated, but for the sake of completeness I will complement or add a little chunk of additional information.
Quote:
Originally Posted by MOV_Games
* Fireworks Factory has been notably pitch and speed altered for some reason, and is lacking the fadeout entirely. This almost seems like an intentional edit, possibly even to do with the sample content in that particular track (it's primarily made up of samples from Groovemaker that none of the other YOTD tracks here used so could be issues clearing them for streaming/sale and editing them gets around that). Might also just be a bad export, either way it's a shame.
|
I found something that points to a bad export with a high likelihood. I have compared the audio sourced from PSX (after resampling it from 37.8 kHz to 44.1 kHz) and compared the difference of pitch with respect to the version released by Stewart Copeland. It turned out that the latter is about 1.0884x the speed of the PSX version. Despite looking like an arbitrary number, it is not the case when we are talking about digital audio.
If we have an audio file formerly rendered at 44.1 kHz and we change the sample rate to 48 kHz (without actually resampling the song), the speed will increase in a ratio of (48/44.1)x, which is... voilà, ≈ 1.0884x. Taking into account this consideration, what I think that happened is the following: Stewart Copeland formerly rendered this song as a 44.1 kHz file, like the remaining ones. At some point, for whatever reason the header of the wav file was changed to 48 kHz without resampling, namely, exactly the same audio information, but instructed to be played at a higher speed. Afterwards, that audio file was converted back to 44.1 kHz, but this time through resampling, that is, interpolating between the samples so that the pitch is being kept the same than the prior 48 kHz file.
In short, if this theory is correct, it means that this song has suffered a process of quality degradation with respect to the original master, because besides the resampling process (which already means a loss of quality), there's less information stored due to the song being up-pitched. The last part implies that if one tries to recover the original pitch, a new resampling procedure using interpolation will be needed. We could say that this song has about a 91.88 % of quality with respect to the original master.
Finally, and continuing with the topic related to sample rate, this release is still a thing. Not only includes stereo versions of Spyro songs that were formerly encoded as mono files in the games, but we also got some of the songs with CD quality (uncompressed 16-bit PCM audio at 44.1 kHz), as opposed to the tracks that were featured on the PSX games (compressed 4-bit ADPCM audio at 37.8 kHz). Furthermore, there seems to be also a subtle but noticeable remastering process behind all the available tracks, improving their quality overall and respecting the dynamic range (namely, without distortion and saturation of the audio).
Quote:
Originally Posted by MOV_Games
* Finally, multiple Year Of The Dragon tracks have errors and gliitches in them, some examples being 0:19 in Country Speedway (weird scrape noise), 3:45 on the fadeout of Charmed Ridge (dropout/stutter) and 1:22 in Dino Mines (another dropout/stutter). The YOTD tracks in particular seem like they may have been taken from a worse source than the others?
|
Certainly. I think they all can be fixed with enough skill and patience in an audio editor, but I'm not fully certain because I haven't gotten to work on them yet.
If only the soundtrack from the whole trilogy was available without all the aforementioned defects and all the lacking songs... but it's safe to assume that the reason why these missing songs haven't been released is probably because the original masters / projects have been lost, sadly. Needless to say, I really wish to be proven wrong in the near future.