krikic / nulldc

Automatically exported from code.google.com/p/nulldc
0 stars 0 forks source link

Shenmue - Incorrect Telephone tone #193

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
What steps will reproduce the problem?

1: Start a new game.
2: After waking up, try and leave the house.
3: Pick up the telephone on the table to the left.

The expected output is: A continuous low tone, identical to the Japanese dial 
tone.

But instead I see: Phone is picked up, a single "boop" is heard, but should be 
"boooooooooooooooo" until a number is dialed.

I use version: 1.0.4

I use nullDC_Win32_Release-NoTrace.exe

I use the following plugins:

PowerVR Plugin: 6/20/10
GDRom Plugin: 6/20/10
AICA Plugin: 6/20/10
ARM7 Plugin: 6/20/10
Maple Plugin(s): 6/20/10
Ext.Device Plugin: 6/20/10

On what kind of system?

My system specifications are as follows:

Operating System: Windows 7 x64
CPU: Intel Core i7, Q 720 @ 1.60ghz
Video Card: nVidia GeForce GTS 360M
Sound Card: Built-in "High Definition Audio Device"

Please provide any additional information below:

No special settings needed, download from the website, insert bios, load the 
.GDI.

(Apologies if this is sort of nitpicking, amazing job here everyone)

Original issue reported on code.google.com by jimgov...@gmail.com on 20 Aug 2010 at 2:24

GoogleCodeExporter commented 9 years ago
This is version r50, sorry for omitting that.

Original comment by jimgov...@gmail.com on 21 Aug 2010 at 3:22

GoogleCodeExporter commented 9 years ago
Upload a savegame close to the phone-booth pls, bored to wait all the intro and 
stuff :p. Its probably related to AEG handling, varius sounds are 'fading' 
because of bugs in the new code

Original comment by drkIIRaziel on 21 Aug 2010 at 4:21

GoogleCodeExporter commented 9 years ago
haha, agreed, ultra-long intro. Here's a Savegame from Ryo's room, just leave 
and pick up the phone on the way out.

Correct phone activity can be heard at approximately 4:48 of this video: 
http://www.youtube.com/watch?v=aoJ2ki0zNfc

Thanks!

Original comment by jimgov...@gmail.com on 21 Aug 2010 at 5:23

Attachments:

GoogleCodeExporter commented 9 years ago
Actually i don't hear a tone -- the volume is wayyy tooo low or smth. Nice bug 
spotting there :D

Original comment by drkIIRaziel on 23 Aug 2010 at 11:47

GoogleCodeExporter commented 9 years ago
whatever I can do to help you guys.

Original comment by jimgov...@gmail.com on 23 Aug 2010 at 12:17

GoogleCodeExporter commented 9 years ago
Okay, its a problem with the looping -- not related to AEG.

The game starts a channel with LSA=0x92, LEA=0xA4, in adpcm format. The current 
implementation of looping, adpcm decoding, or smth else leads into the adpcm to 
be decoded wrongly, leading to this problem.

Also happens on SA2 (the 'grinding' sfx on the first level) and possibly in 
more 'looping' sounds

Original comment by drkIIRaziel on 2 Sep 2010 at 10:58

GoogleCodeExporter commented 9 years ago
oh wow. that's pretty interesting, i figured it was another AEG handling thing. 
Glad you were able to discover it!

Original comment by jimgov...@gmail.com on 2 Sep 2010 at 6:18

GoogleCodeExporter commented 9 years ago
okay, it seems that the entire model of the aica decoding we have is quite 
wrong or smth. Managed to fix this bug, but breaks other SFX -BADLY-. Also 
found two more bugs on shenmue (clicky sounds) that seem kinda related .. so 
YAY?
I'l upload a test build for cross testing later on :)

For anyone that cares about the details:
It seems as if the adpcm state must not be reset on loop for this to work.

Original comment by drkIIRaziel on 3 Sep 2010 at 8:01

GoogleCodeExporter commented 9 years ago
guys is there any current solution to this problem or we'll just have to bear 
with it.
Thanks

Original comment by yau...@gmail.com on 7 Sep 2010 at 6:52