Ahh that's easy enough. Thanks.
Go into the RES subdirectory and delete SND.CRF, rename SND.BAK to SND.CRF.
Ahh that's easy enough. Thanks.
The some of the footfall sounds, the blackjack sound, the arrow sounds sound like Thief 3 sfx, or remastered sounds from Thief 3. Not saying that's a bad thing - I just figured out where I'd heard them before. :-D
Last edited by Sayne; 17th Jun 2008 at 16:39.
This is true. I took some sounds from Thief 3 from which I thought, they sound more realistic than Thief 2.
1.) How does one install this?
extract SoundUpdateinstall to root of Thief installation and run T1Sounds.bat?
did not work for me, it created empty ..\crfs\SFX folders
2.) My ..\RES\snd.crf is only 168 MB (Thief Gold installation)
Is this normal?
1. I think that is right, not sure.
2. Sounds right to me. The new one should be about 277 mb
KoHaN69, keep in mind, while optimizing with different versions of DarkLoader the crf-files ends up in different folders:
yes SinisterShadow, I did optimize with DarkLaoder 4.3
However, the unoptomized version has the snd.crf in the root of the install
and looking at the installation's bat file:
it assumes the snd.crf is in the crfs folder, which there isnt in any default installs of thief1 or thief gold for meCode:cls copy crfs\snd.crf crfs\snd.bak ren crfs\snd.crf snd.zip mkdir crfs\SFX SoundUpdate\Rar e SoundUpdate\T1SFXNew.rar *.* crfs\SFX cd crfs ..\SoundUpdate\7za u snd.zip SFX\*.* cd .. ren crfs\snd.zip snd.crf del crfs\sfx\*.wav del SoundUpdate\*.exe del SoundUpdate\*.rar
So I tried to follow the bat manually,
1.) I copied snd.crf to \temp\ and renamed it to snd.zip
2. )I extracted it to \temp\snd\
3.) extracted T1SFXNew.rar \temp\snd\SFX overwriting all the older files
A. Why does the original snd\SFX\ folder contain only 864 files while the T1SFXNew contains 896???
B. there are no replacement sounds for cavetone.wav and FTWAT_J.wav in T1SFXNew.rar
4.) Then I created a zip (store no compression) archieve of \snd\ from \temp\snd\ in WinRar
(it's size ended up being 296MB)
5.) copied \temp\snd.zip to thiefgoldinstall\RES\snd.zip
6.) renamed thiefgoldinstall\RES\snd.crf to snd.bak
7.) renamed thiefgoldinstall\RES\snd.zip to snd.crf
and it seems to be working fine.
I think many of the sounds should be replaced altougher, especially the weapon sounds like SWORD1.WAV and SWORD2.WAV you can really hear the hiss everytime its played
Last edited by KoHaN69; 28th Jun 2008 at 06:57.
Either you or darkloader puts thief 1/gold crfs in a crfs folder, whereas thief 2's is RES. Default thief doesn't put the crfs in any folder(I thought), and that explains why if you try installing the sound pack in an unoptimized thief it can't find snd.crf and does nothing. At least in thief 1/gold, I think thief 2 defaults with them in RES. Somebody correct me if I'm wrong.... ( I don't remember when I first installed thief what was the case or not, but it seems this way.) I see you did it manually though. That works too, then.
Didn't we have a similar thing some time ago? I think this......sounds like a non-complete Thief setup what however the CoK's little batch file expects to find.Originally Posted by KoHaN69
About your sidepoints:
The reason there were more files is, that I copied the feet and blackjack files directly from my Thief 2 mod and so stuff like treebeasts and snow was copied, too as well as blackjack-on-helm files. Only a few kbytes wasted but after all no disadvantage.
The cavetone.wav doesn't need to be axchanged, it is good as it is. I don't know why the water jump sound didn't make it into the package, but however, how often do you jump in water?
There is something different that's annoying me now. I re-recorded those string melodies for the opera mission and also used them in the compare-MP3s, but it seems they never made it into the package and I deleted the unpacked files on my harddrive because I though I don't need them anymore. Now I have to look if I have these files somewhere else or otherwise redo them again...
I was never lucky with those tools. Except the whole HD file system was broken and so far no new data was written on it. However, it was easier for me to rerecord those files again. I also removed the unnecessary files and reuploaded the package to the server again.
Excuse me for saying this but... I've been biting my tongue since I saw this thread.
What a load of bull.
No offense, but do you even KNOW anything about audio? There's a term in audio called 'garbage in, garbage out' - it refers to the working master needing to be recorded in the highest quality in order for the distribuitons to even be acceptable.
Since you likely do NOT have the original studio recordings and are working off the ones distributed with the game, which are 22khz/8 bit - no matter what you do you are either spinning your wheels doing NOTHING, or worse introducing distortion. (much like the old school ***'s who think tube amps 'sound better' - yeah, because they distort the signal from capacitance lag)
If you flat upsample, you aren't changing the quality at all - and since you do NOT have the missing pieces any attempts to interpolate the missing data is just muddying the recording and introducing distortion - net result is less 'crisp' sound, muddy at worst. (I tell you, you're KILLING the treble)
Upscaling from 22kz/8bit to 44khz/16bit is like trying to turn a 32x32 pixel image into a 16384x16384 one. You can either have giant blocks with no data loss just wasting disk space, or having a blurry mess that no amount of 'sharpen' filter is going to fix.
Of course if the studio mix-down was done even close to decent 22khz should be indistinguishable (since the extra doubling is done for sampling error correction on the high frequencies, take the directional peaks and it's 'close enough') leaving just the bit-width...
Being that it's digital mixed at 16 bits, and the top bit and a half worth of data are outside the human hearing range (the only reason 16 bits is even USED is because it's the next even up multiple of two from 15) that means at 24 voices (roughly 4.5 bits) that's only two bits away from the upper end of human hearing!
Which also means that any data more than 10 bits is effectively wasted on the Dark engine and most other 16 bit mixing games. You get into 24 bit mixing for missing interpolation, THEN 16 bit recordings become useful (in 256 voices)... though with hardware accelleration on, you MIGHT be getting 24 bit mixing on cards that support it in hardware (creative) - but I doubt it given the age of the game. (which came out at a time where the only thing that would even come close to 24 bit mixing would be the original SB Live)
...otherwise it's rubbish and a waste of time. Oh, but it's more bits and higher sample rate, it must be better - RIGHT.
I smell the same placebo effect as that which has a hold on FLACtards who copy songs off audio CD's at 384kbps, when the 16 bit 44khz audio is only 176kbps stereo and exceeds human hearing by 366%!
Last edited by deathshadow; 2nd Jul 2008 at 08:03.
Perhaps professionals like you deathshadow can't make any difference between these and old sounds but I'm lucky I can (and so are many others too). For me the biggest improvement is the reduced hissing sound that plagued many sound effects.
Thanks, deatheshadow, for your long post. Certainly you have noticed that I did much more than just changing the sample rate of the files. It's one way to resize a 32x32 picture to 256x256, but it's another way to afterwards redraw the contures by hand and recolor the rest and remove the dirt.
If you don't need it, I don't force you to use it.
I certainly know much more about audio stuff than you would like me to. I do know, that with my methods I can never achieve the original sound quality. But it's certainly nothing bad to take a towel and whipe away the dust laying on the files, that came by ADPCM compression and 8 bit quantisation. Of course with improving the bitrate I can't enlarge the dynamics (which is still 48 dB with 8 bits), but I certainly can remove the dithering noises using FFT noise reduction (I did use an idle Dither A noise as sample).
For restoring lost frequenices I used simple upsampling for non-tonal sounds, like chains etc. and afterwards EQing (reducing hight above 15 kHz), because that's the best way to achieve missing inharmonic overtones. For music stuff I used a feature that's also implemented in the MP3pro encoder that just takes certain frequency peaks, find its interval and just created new overtones above it (as we all know, harmonic overtones are just a multipile of the first harmonic).
Now you again.
Deathshadow, your ignorance is showing...again! Why do you assume that you know so much more than anyone else about things like audio (or physics), or that anyone else who doesn't think the same way you do is an idiot?
Child of Karras, there are many people here who appreciate what you've done. Thank you for all your hard work!
Well, in general there's nothing wrong with questioning things I believe, but even if I make digs at the CoK for a lot of things (), I usually respect his technical skills.
So where did the Sound Enhancement Packs go? They aren't on Southquarter and the original links are broken!
What are the new MINIMUM specs to use the Sound Pack ??
Um... anyone have the comparison mp3s?