Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
Patashu wrote:
I think if you want to stress a sound codec to its limits, you need to try it on a song that is very complex with lots of subtle instruments and effects. May I recommend the La Mulana (not the steam but the original version) soundtrack?
There's also the consideration that games with music from waveform generators probably need smaller quality setting than ones with full PCM music.
So it depends on platform and game too...
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
It seems that basic support for Opus audio codec in .mkv is near, so I thought to do some testing of quality settings....
http://ilari.tasvideos.org/chronointro.zip
The zip contains original capture of the Chrono Trigger intro song loop (29.561s), plus opus (32-128kbps) encoded (and then decoded to flac) versions.
The chronointro-0.flac is the original lossless.
File sizes:
119686 chronointro-32
150130 chronointro-40
184745 chronointro-48
212899 chronointro-56
235226 chronointro-64
253943 chronointro-72
275710 chronointro-80
294350 chronointro-88
322561 chronointro-96
351176 chronointro-104
368069 chronointro-112
397047 chronointro-120
425976 chronointro-128
All settings except bitrate at defaults (so it uses VBR mode). I used libopus 1.1 beta.
Comments?
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
flashthe wrote:
so it is not possible to follow the marble into the green portal at the bottom and come out on top?
Doesn't look to be possible:
1) All directions except upper left and right are solid and unphasable.
2) Upper left is not an option because one can't squeeze through diagonal of solid and unphasable blocks (one could squeeze between spikes, but that's not useful here).
3) Right is not an option because those spikes would kill you.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
JohnnyG wrote:
At least this is a known problem.
Thanks for you answer, maybe someone got a solution later.
What might work (but is kinda tedious):
1) AVI Capture from start to first desync.
2) Play the movie without AVI capture and make a savestate a bit before it would desync.
3) Play back from that savestate while capturing AVI. The desync point might move to later point (or the movie would finish).
4) Stitch the two AVI captures together.
Methods like this are used to deal with PCSX-RR (urgh) video capturing.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
JohnnyG wrote:
Hey there,
I recorded some gameplay of SVC Chaos. Everything went fine and replaying the movie file syncs perfectly. But the moment I try to dump an .avi the playback desynchs at the first fight.
Yeah, FBA is kinda bad that dumping affects sync...
Some encoders have had to deal with this, but I don't know the specifics...
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
What Windows version? IIRC, for most of the games that can even gotten to run, one wants Windows XP in virtual machine.
Also, AFAIK, Hourglass has various options that can affect if game runs or not.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
MrWint wrote:
All of this discussion is only valid for polled input, as interrupt-driven input can't be accurately recorded in any emulator that I'm aware of. It can happen anywhere and would require subframe input capabilites accurate to the CPU cycle.
Such capabilities would also require movie format that is pretty totally different from most formats.
Subframe capabilities accurate to CPU cycle wouldn't be hard to implement, apart from that interface design could turn nasty.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
GhostSonic wrote:
Ilari wrote:
Tested with modified bsnes core (more accurate emulation of autopolling). Works fine.
You think there's a chance of this being console verifiable then?
I mainly wanted to check against exploiting that known emulation error.
Also, verifying this means controlling 4 controllers independently.
GhostSonic wrote:
I don't currently own Yoshi's Island so I can't test it at the moment. I wonder if the bsnes core emulates the Super FX 2 accurately enough to work properly.
AFAIK, there are known emulation errors with SuperFX (such as the CPU-SuperFX memory access race).
I think usually that error just results some missed lag frames (because it emulates that operations occur faster than the equivalent in real world).
GhostSonic wrote:
I remember the old run not syncing at all on newer version of Bizhawk and lsnes.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
MrWint wrote:
creaothceann is right in that it would be most accurate to record buttons not once a frame, but every time it is read by the game. Since this is during VBlank in the games I know (which aren't many, a more elaborate evaluation might be needed to confirm the hypothesis), separating the frames by VBlanks seems to the a more accurate way to handle frame timing. Making every video frame the same length is less important in my opinion, since it isn't the case on the actual console, either.
The frames all need to be on average the same length or issues crop up, both for timing of movies and with video dumping.
Well, with video dumping, the framerate can technically be changed, but doing such thing leads to problems when trying to encode the thing.
Yeah, some sort of constrained VFR mode that kept the clock counter in bounded interval containing the nominal value might work...
I don't think GB can generate video frames faster than 35112 cycles per frame (but it can generate those slower).
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
alec kermit wrote:
I'm having trouble dumping with a/v sync. It dumps and syncs fine from the start of the game but when I load a savestate or try to dump audio to AVI after loading a savestate dolphin crashes.
Yeah, avhack doesn't like loadstates. Known problem.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
It is done that way to ensure that video output frames happen at consistent intervals (1 video frame every 35112 audio samples) so audio and video will be in sync.
Also, if input and video frames are nowhere near aligned and game polls on VBlank, the poll should go in the middle of video frame.
Also, even if there are multiple input frames per video frame, those can differ (subframe input).
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
Zeyn0 wrote:
Yes, there are just four that do so in 15 rooms, but I was thinking that if there are some really slow rooms/bosses, an extra room could still be better...
We have to do at least two more rooms (R and T or S and T) if we choose another path, that's why i put them aside
No, there are routes with just one extra room (unless some doors are mysteriously closed), but it is unlikely that those are faster than either of 15-room routes.
Zeyn0 wrote:
Also, spikes (esp. in room B) are instakill, right?
You want to die to save time ? How could it works here ?
I was thinking that in case those spikes just did damage, one could save time there by taking damage...
But if those spikes kill you, dying there is unlikely to save any time.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
Zeyn0 wrote:
Hi Ilari. I will try to correctly answers your questions :)
quote]3) Or are rooms G and F designed so that one can't make (in sane time) from lower exit to upper exit (or vice versa)?
I am not sure to understand your question. Sometimes you can choose the door you wanted to take. And sometimes a door is closed. The graph show you when you can choose and when you can't. [/quote]
Well, it doesn't matter anyway.
Zeyn0 wrote:
I show eight different paths but there are just four paths that leed to the last room (the U room) in only fifteen rooms, other paths need more of that,
Yes, there are just four that do so in 15 rooms, but I was thinking that if there are some really slow rooms/bosses, an extra room could still be better...
When the 15-room path going through E is probed, one could probably calculate an approximation if any 16-room path would be even close (probably not, especially as I am not seeing anything really horrible in any avoidable room, B and C yes, but those can't be avoided).
Also, spikes (esp. in room B) are instakill, right?
Also, doesn't all exits dumping you at start of the room cut down the number of 15-room paths to probe in half (the order one does bosses in shouldn't matter then)?
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
Okay, I drew a graph of the final base based on that tree in previous post...
1) You can't go from S to T or from R to T before at least two color bosses are defeated?
2) Is it possible to go back from boss room after defeating boss (e.g. G->Green Boss->G could be useful if so!)?
3) Or are rooms G and F designed so that one can't make (in sane time) from lower exit to upper exit (or vice versa)?
4) If you take two exits that take to the same room (say, D->F and P->F), do both dump you on the same position in room F or not?
Oh, and in case I understood correctly, there are 8 paths that are 16 rooms:
A->B->C->D->F->{Ye->M,Bl->L}->P->R->G->{Gr->O,Pi->N}->Q->S->T->U.
A->B->C->E->G->{Gr->O,Pi->N}->Q->S->F->{Ye->M,Bl->L}->P->R->T->U
I think it is possible to simplify the routefinding using Vitterbi algorithm:
The 15-room routes.
1) entrance of A -> entrance of F.
2) entrance of F -> entrance of P, defeating two bosses. Discard slower.
3) entrance of P -> entrance of T.
4) entrance of A -> entrance of G.
5) entrance of G -> entrance of Q, defeating two bosses, Discard slower.
6) entrance of Q -> entrance of T. Discard slower vs. #3.
The 16-room routes:
1) <#1 from 15-room>
2) entrance of F -> entrance of P, defeating one boss, discard slower.
3) entrance of P -> entrance of G
4) entrance of G-> entrance of Q, defeating one boss, discard slower.
5) entrance of Q -> entrance of T.
6) <#4 from 15-room>
7) entrance of G -> entrance of Q, using faster route from step 2)
8) entrance of Q-> entrance of F
9) entrance of F -> entrance of P, using faster route from step 4)
10) entrance of P -> entrance of T. Discard slower vs. #5.
Emulator Coder, Experienced Forum User, Published Author, Skilled player
(1142)
Joined: 5/1/2010
Posts: 1217
EEssentia wrote:
Yes, for Windows development, C# is a pretty good choice.
That is, if you don't think you need to EVER make the program run well on anything except Windows.
Oh, and regarding using C++... Stay away from Win32 API as much as possible (especially for GUI tasks), as it is just bad.
And regarding WIn8 Metro, my advice is: Stay FAR away.
