Tonic Trouble Soundtrack

[kooz]

Exactly the type of input I was looking for.

Doc's Cave 3 loops just like that with an abrupt transition in the actual game (1-09 - Doc's Cave - Part 1 from original rip). I remember needing to pause the game to record that one. It only plays for about 5 seconds in the tunnel/slide that drops you into Doc's Cave. That may be the reason, actually. It was probably never intended to play all the way through to the point of looping.

The end of Pressure Cooker 1 doesn't have a "gap" per se... but I do hear what you mean - specifically, the reverse cymbal-to-cymbal crash. I'll try tightening that up.

Thanks for the feedback!

[kooz]

It's done. Enjoy!

Tonic Trouble SE (beta) Original Soundtrack v3 [FLAC]
Tonic Trouble SE (beta) Original Soundtrack v3 [V0 MP3]

Tonic Trouble SE (beta) Original Soundtrack v3.1 [FLAC]
Tonic Trouble SE (beta) Original Soundtrack v3.1 [V0 MP3]

[Droolie]

Fantastic! I already listened to some tracks and the quality upgrade is very noticeable. It also looks like you improved the parts I told you about, well done! I will be sure to put it up on RayTunes some time after my final thesis presentation next week.

[kooz]

Glad you're liking it! Take your time putting it up. I'm going to end up doing at least one additional revision tomorrow... I swear I triple quality-checked everything before posting, but here I am doing an end-to-end listen through while at work, and there is a nasty pop at the very end of North Plain. I'll investigate and fix it when I get home.

I have saved all project and temp files, so any further revisions (if necessary) should be a snap.

Fixes complete. What's new in v3.1:
- Corrected issue that caused the pop at the end of North Plain.
- Vegetable HQ 2 faded out too early. Now properly ends with ~15 second fade after loop repeats.

Download links:
Tonic Trouble SE (beta) Original Soundtrack v3.1 [FLAC]
Tonic Trouble SE (beta) Original Soundtrack v3.1 [V0 MP3]


I think that should be the last revision for a while, hopefully. Of course, if anyone finds any issues that may warrant further changes, please speak up.

[Harpic fraîcheur]

Once again, you did a great work kooz.

[Hunchman801]

The thing you have to keep in mind is, what you call "quality" is the fact that there are (or not) frequencies in the high-end of the spectrum. If you want to use the Crystalizer to strengthen the spectral folding under the assumption that it sounds better to you, then good for you. But that's not the "quality" we're looking for. We (as in, at least Droolie and I) are (or were, now that it's done) looking for a bit-perfect extraction of the soundtrack, which should show absolutely nothing in the spectrum above 11kHz (since they were sampled at 22kHz). The only frequencies we'd want above that threshold are the ones that got lost in the 44k > 22k resampling process, and which are irretrievable.
Anyway, I'm getting carried away. We've done awesome work, and I think there might be things left to dig.

Or, if I may use an image analogy, this is not a higher quality version of our beloved Mr Green smiley:

[kooz]

Wow, how did you get Mr Green looking so good!?

[PluMGMK]

I think there should be a "cutting through the bull" GIF for occasions such as this.

[deton24]

I'd compare it to an analogical situation when we had run R2 in 32 bit depth and game has a 16 bit textures, and in result, game looked slightly different.

I think that actually everything has been said already in this topic.

I regret only that we couldn't run R2 with different snd libraries on 44khz.

[deton24]

I found one interesting thing which may influence music and sound quality during ripping.
And maybe it will help with achieving 44kHz in R2:
viewtopic.php?t=26077
edit. There is no 44kHz, but still, it's better.

[deton24]

I found CTPAX-X group which was able to finally convert audio files in MPX files of Tonic Trouble beta.

It took them one week, and they want some donation for their servers.

You can write to support(at)ctpax-x.org in order to ask for some samples of the work, in order to verify whether e.g. it's not a scum, and just a copy of kooz rips, or just heavily processed rip with audio plugins.

Ask for unique sample for each user, and feel free to publish it here.

They use game DLL files to make their converter work, because it would be too time consuming to reverse the library. They said both, final and beta libraries work, and sources in "C" will be published for the tool.

As for donation methods, they say they have also a bitcoin, but generally there's a kiwi, yandex money, Webmoney, all on their main site in the left corner of
CTPAX-X.org, but not PayPal.

If I will receive some samples I asked, I'll let you know.

[Droolie]

Hi deton24

I just checked my old email since you wrote about it here yesterday (yes, I still read this forum and also got notified about this whole deal on XeNTaX) and saw this thing about CTPAX-X.

I'm not too happy about them asking you to "donate" in return for their work. The fact that they didn't tell you that payment was required before they started working on it seems fishy, and I don't think much good will come out of you paying them for this.

Even if it's true that they succeeded in using the game's DLLs to decode the files, I'm not too interested anymore.
Kooz managed to fully and reliably record all of the MPX files, and converting them with CTPAX-X's tool is probably not going to give us any higher quality files, considering they also use the same DLLs.

What's more, Tonic Trouble has an official soundtrack available with many tracks in 6-channel 48000Hz AC3 format! Much higher quality than what you'll find in the MPX files. Some parts are missing/abridged so it's a R2 PS1-like situation, but unlike that, this soundtrack also has entirely new parts! I discovered this last year and talked to kooz about it, but I have been extremely busy since then so I didn't prepare it for RayTunes yet. Here's what I sent kooz (also includes the link to the raw files):

Apparently, the Special Edition of Tonic Trouble was also released as a DVD, which comes with a video section that has many of the music tracks in 6-channel AC3 format, at 48000Hz.

If you're interested and have some time, I'd love to hear your opinion on these files.
I demuxed the tracks from the video files, you can find them here: https://mega.nz/#!ZKQ1HQgA!X_Gp2hxnYXQk ... vKDpyqBXH4
You can get the full DVD here (via the Multiregional DVD link) if you want to check the original video files (though I didn't modify the audio): https://www.myabandonware.com/game/toni ... l#download

The tracks include:

Code: Select all

01 - Intro, in many different languages
02 - "Here comes Ed", a song from a promotional video
03 - Glacier Cocktail
04 - Pyramid
05 - Grögh HQ
06 - Bonus/Crazy Town
07 - Pressure Cooker
08 - North Plain
09 - South Plain
10 - Credits
11 - A track that I think isn't used in any version (do you recognize it?) and the ending
12 - Doc's Cave (part 3?)
13 - Canyon
14 - Ski Slope

No Vegetable HQ or the rest of Doc's Cave, it seems. Interestingly, the other tracks also have some differences with the ones in the final version and the Special Edition rip that you did.

Do you have any idea if/how these could be reliably mixed back into stereo files? I've tried a few different approaches but nothing seems to sound good enough.

[deton24]

Hi Droolie.
It's cool to see that you're still with us.

Well, it's pity that you haven't written it before.
I engaged lot of people asking for help to decode these MPX files, and it would spare the whole time CTPAX-X group.
And it turns out it was worth of it anyway

Since Special Edition MPX files are also abridged in comparison to retail version (don't know in what extent to DVD) of course you may have right.

The tracks can be easily converted with Foobar and AC3 decoder plugin to FLAC stereo (and also played). There is also 5.1>2.0 downmix DSP (works for convert option and playback (place downmix on the top of DSPs if you have more of them in playback chain; during conversion you will use only this one for sure).
AC3 files sounds the best without conversion to FLAC 2.0 (with just downmix DSP activated during AC3 playback).

Really cool find. You rock.
Thank you

Greetings

PS. Track 11, no, I don't recognize it neither. On the beginning It reminds Pirate Patrol from Revolution... later it is like remix of one of the tracks. I think around three different themes are there. So coool...

The files are exactly 6ch AC3 ATSC A/52 16 bit/48kHz/448kbps.
Main tracks doesn't reach above 44kHz.

My favorite tracks are here. I can die happy xd

edit. I've added index of Tonic Trouble rips in post #4:
viewtopic.php?p=543758#p543758
There are... 13 versions in total.

2.0 DVD downmixed FLAC soon (sooner probably you will do it yourself due to my very slow connection).

Well. Kill me, but I accidentally hit on an information that during conversion from AC3, 32 bit output probably will be enough because there will be 20 bit precision or something. Whatever the author had in mind, I checked it, and there is less quality deterioration in comparison to original files than using 16 bit.
Don't know really why that works, since these files are identified as 16 bit, but maybe it has something to do with processed downmixing by plugin.

So finally it's gonna be 1GB for the whole FLAC 2.0 upload. Doubler than 5.1 AC3. On 20KB/s upload.
Man.

[deton24]

I had a dream.
Diving and breathing underwater.
Passing through mysterious sanctuaries of Glade of Dreams, and its outer worlds.
When I came to myself later tonight, I received a message.
They released the tool to public.
And it was all over. The final gunshot was an exclamation mark to everything that had led to this point. I released my finger from the trigger.
A bit closer to heaven.

Somehow it seems like it extracts some missing segments from PC version, sounding better than V3.1 rip. Judging by samples received two days ago. Exceptional.
Source codes are attached if someone is interested.
Interestingly, someone committed this thing 4 August, didn't notice:
https://github.com/losnoco/vgmstream/co ... d3cbe4fd0f

[kooz]

Somehow it seems like it extracts some missing segments from PC version, sounding better than V3.1 rip. Judging by samples received two days ago. Exceptional.

While I am extremely impressed that a proper tool has finally come along, I have to disagree regarding any quality difference. The end result is the same as my 44.1k rip, just achieved in a much more elegant fashion. Direct comparison:




MARMTH0_test1.wav: CTPAX-X tool output
MARMTH0_test2.wav: MARMTH0_00000000.mpx.flac from my 44.1k rip pack (viewtopic.php?p=1183961#p1183961)

There is no difference.

[deton24]

Nah. They reacted different with various DSP when I checked.
Just without them you will not notice any difference.

Making a gif from these images will show the difference between them more clear.
And preferably from Spek.

But if there is no difference in normal listening for such person like you, there won't be any also for most people. But technically the files are not the same though they achieve similar frequencies.

And it's still not strictly proper decoding tool, since it uses game library, which we don't know whether it doesn't cause downsampling.

[kooz]

Making a GIF would downsample the color depth of the images to 256 colors... The PNGs are lossless. You can just open each image in a new tab and view them at a 1:1 pixel ratio. Alternate between tabs by pressing Ctrl+PgUp/Ctrl+PgDn.

I'd love for you to provide some examples of where you can see/hear a difference, as well as any newly revealed samples you believe were missing before.

Here's a fun experiment you can do at home. Open both images up in a photo editor, like GIMP, for example. Merge the two using the "difference" option:


I've also attached the resulting merged image:


Feel free to play around with the brightness, contrast, etc. just to be sure it's not just something normal people can't see.

[deton24]

Allright. I did a comparison in Webm and Spek:
https://imgur.com/y6pocBK

Furthermore, the files are different in case of loudness:
https://imgur.com/a/csSmBuw

It is like with ReplayGain. It's not recommended for bit-perfect playback. Files will sound different. You can't skip this.

[kooz]

Somewhere in your testing you are altering one (or both) files. Perhaps it was in the sample/bit rate conversion (to 96/24) clearly seen in your REAPER screenshot.

A hex comparison shows these files are, in fact, bit-for-bit identical to one another, with the exception of the last few bytes (which was caused by Audition creating a fade to 0 where I cut off the CTPAX-X tool output sample down to match the duration of my original 44k rip sample). I could cut it down again with smoothing turned off, but I don't need to do that to prove my point. Any possible difference in decoding would have resulted in two completely different files from beginning to end, while these are a perfect binary match. Visual comparison by waveform or spectrum analysis is moot at this point.

The last 682 bytes:


Here are my test samples/results. Run them through your DSPs and tell me you still think there's a difference.
https://mega.nz/#F!EM82zQoZ!df-YigK2MROlXdZYaJac-Q

[deton24]

Visual comparison by waveform or spectrum analysis is moot at this point.

Yeah. Especially seing that noise above 16kHz.

In your 2-second files, volume is probably the same. At least automatic algorithm of ReplayGain detection shows the same values.

But on the file gathered from your V3.1 rip, volume is higher than on your provided sample.

Comparison from different DAW with 16 bit/44 kHz.
https://imgur.com/a/yukFloq

And it's not an effect of any ReplayGain information (if it was originally added at all) because all DAWs omit it.

On my comparison samples (CTPAX vs 3.1), readings from ReplayGain scan differ (+4.37 vs +1.39 dB).

When I was listening to your files, they sounded the same, but after more complicated post processing it turned out that test1 is a bit louder. But they sounded really, very similar.

The result sound on this video will be more or less distorted, but you should hear or see on your spectograms what I mean:
https://drive.google.com/file/d/1G6kWds ... sp=sharing

It's just an effect of various plugins with the same settings.

Look like files you sent me are very close to bit perfect. But your V3.1 rip gathered from your previously posted link:
https://mega.nz/#!dZ1hXDiD!8rBQhyPtLPjI ... QQOqs0VEiQ
Is different from both files you provided me just now.

Did you newly record it?

PS. Long time ago you asked about missing R2 Top of The World track. It's done:
viewtopic.php?p=1361667#p1361667
_________________________________________________________


Tonic Trouble DVD OST FLAC 2.0
https://yadi.sk/d/YK-SFtXLE4IfTA

Source file was 16 bit, but FLAC was encoded at float 32 bit. It resulted somehow in better quality in comparison to 16 bit, presumably due to appearance of downmix DSP and/or AC3>FLAC conversion (in normal cases it's shouldn't be necesssary). It gave 24 bit files with 32 bit setting in encoder anyway (probably encoder or F2K bug) but even in this case, 32 bit instead of 24 set in encoder (though result files are 24 bit) resulted in better quality. No dithering used. Furthermore, I used 1.33 FLAC x64 encoder, which officially is not available for Windows. I wrote to devs about it, but no sign of life. I gathered it compiled from external source. I think it's better than on 1.3.2 for some reason. Probably another isolated incident.
The most important is - after all of it, it seems to sound better than original AC3 files with just downmix DSP enabled. Personaly, quality just satisfies me.
Thank you Droolie. It would be nice to at least link to this and/or to your original:

AC3 5.1 version
https://mega.nz/#!ZKQ1HQgA!X_Gp2hxnYXQk ... vKDpyqBXH4