ScoreHero
Home | Forum | Wiki
Inbox [ Login ]Inbox [ Login ]
SearchSearch MemberlistMemberlist
ProfileProfile Log inLog in
SlowHero Charts Project
Goto page Previous  1, 2, 3 ... 17, 18, 19, 20  Next
 
Post new topic   Reply to topic    ScoreHero Forum Index -> Technique, Style, and Gameplay -> Guitar/Bass
View previous topic :: View next topic  
Author Message
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Sun May 20, 2012 11:12 pm    Post subject: Reply with quote

LocalH wrote:
tma wrote:
I can certainly give it a go. Is that a custom? I don't have a PS2 I'm afraid, in fact I've never even tried customs, however PS2s are cheap... I could be coerced into it. ;)

Trogdor, man, Trogdor. I dunno the differences between the 360 and the PS2 in terms of GH2 and the strum limit, but nothing custom required, just a PS2 and an original GH2 disc. I do recommend you find a fat PS2 and a network adapter, especially if you have a decently sized IDE HDD laying around, so you can save the laser. If all you can find is a slim, you can still save the laser with slower loading methods but they work fine with GH2.


Heh I guessed it was Trogdor shortly after I wrote my post.

I have the PS2 mids for GH2 (along with the X360 ones) so I'll use them as the source for the bot, once it's up and working. I'm watching a GH2+guitar on ebay now, and I think I can borrow a PS2 from a mate locally. So yes, I will be able to test it.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Sun May 20, 2012 11:19 pm    Post subject: Reply with quote

Getting back on topic:

While I've only given the apathtoofar code a brief look over, it looks like some of the obvious shortcuts I'd mentioned have already been taken into account.

I think the next step would be to analyse the paths that are considered and look at how we can cull these down before sending off a bunch of paths for final analysis. Obviously for songs that don't have that many overdrive/star power sections this isn't a big problem, but once it gets up there, or if you're looking at multi-instrument pathing, the processing time required increases radically.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
Cabanon  
 




Joined: 29 Oct 2007
Posts: 5778
Location: Quebec, Canada

PostPosted: Mon May 21, 2012 3:16 am    Post subject: Reply with quote

what's the time taken for calculation of those chart ? does it take more time because there's let's say 10 activation compared to another song with only 2 ? does it take 5 sec ? 2 min ? 15 min ?
_________________
Back to top
View user's profile Send private message Send e-mail Visit poster's website MSN Messenger PSN Name: Cabanon
eddaket  
 




Joined: 19 Jul 2007
Posts: 2012
Location: Burillville, Rhode Island

PostPosted: Mon May 21, 2012 3:26 pm    Post subject: Reply with quote

These numbers are from my old program:

Billion Dollar Babies (Live):
7 Phrases
3 Activations

Time:
real 0m6.033s
user 0m5.980s

---

Sulfur:
12 Phrases
7 Activations

Time:
real 2m32.303s
user 2m32.290s

Of course, I wrote this in python which is going to make it slower anyways. Most songs take no time at all since they are short, but there are definitely a few stupidly long songs we would have to worry about. I tried to time RTL for comparison but I had to stop it at 15m so I could do something else.
_________________
Back to top
View user's profile Wiki User Page Send private message AIM Address PSN Name: eddaket
Naruto42  
 




Joined: 13 Dec 2008
Posts: 1284

PostPosted: Mon May 21, 2012 4:23 pm    Post subject: Reply with quote

I am personally very interested in such codes, may I have a look at them please ?

Also in GHTPC there is a bot, it may help you create yours.
_________________
632/660 - Best: One GH3 (PC) - My acc. thread - My Youtube


Back to top
View user's profile Send private message Send e-mail MSN Messenger XBL Gamertag: Naruto92250 Wii Friend Code: 7287116137472646
LocalH  
 




Joined: 30 Oct 2006
Posts: 1325
Location: Kingsport, TN

PostPosted: Mon May 21, 2012 5:29 pm    Post subject: Reply with quote

Naruto42 wrote:
Also in GHTPC there is a bot, it may help you create yours.

Nice idea, but I can say it probably won't help, as he's writing a bot that works external to the system, whereas GHTCP's bot is just enabling the one already built into the game code for testing purposes.
_________________
Read my blog! The Incoherent Ramblings of a Lowly Geek

SCPH-39001, Free HDBoot 1.93, OPL 0.9.2 via 250GB HDD (all retail GH/RB games plus customs) - no memory card needed

Alakaiser wrote:
POST BECAUSE YOU HAVE SOMETHING TO SAY, NOT BECAUSE YOU HAVE TO SAY SOMETHING.
Back to top
View user's profile Send private message
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Mon May 21, 2012 11:02 pm    Post subject: Reply with quote

LocalH wrote:
Naruto42 wrote:
Also in GHTPC there is a bot, it may help you create yours.

Nice idea, but I can say it probably won't help, as he's writing a bot that works external to the system, whereas GHTCP's bot is just enabling the one already built into the game code for testing purposes.


Correct. The bot I'm building is, effectively, a computer interfacing directly with a guitar hero controller and triggering the buttons and strum bar following a custom program that will be derived from the game files to get the exact timing.

I've already built a proof of concept - I have interfaced with an Xbox 360 wireless controller and can press buttons using code. I have also done most of the hard work to convert the mid files I'm familiar with (basically any Guitar Hero mid up to GH:M) into the timing program I need for the Arduino processor.

Currently I'm in the build process, soldering together a board to hold all the relays and interface with a modified GH controller and the Arduino board. If you're interested I can post some pics, but it's only a board with some IC sockets and LEDs at the moment.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Tue May 22, 2012 12:39 am    Post subject: Reply with quote

Just doing some basic benchmarks for interest's sake. The system I'm running these on is a Linux based Intel Core2 Duo CPU E4500 @ 2.20GHz, which is not under a lot of load.

Code:
tarragon@tyrell:~/git/apathtoofar/optimizer $ time ./rbopt.pl maps -v -g rb2 -p ~/cvs/spopt/assets/gamefiles/RockBand/x360
Gathering midi data...
Parsing mid...
Using ridiculously accurate whammy.
Calculating notes in OD phrases and in solos...
Checking for errors in rounding...
Running optimizer...
100% Complete - Estimated time remaining 00:00:00
Reading opt file...
Optimal score: 88195
Wrote text path to tmp/maps.expert.guitar.0.0.0.txt
Total processing time: 52 seconds

real    0m52.042s
user    0m53.599s
sys     0m0.228s


Code:
tarragon@tyrell:~/git/apathtoofar/optimizer $ time ./rbopt.pl foreplaylongtime -v -g rb2 -p ~/cvs/spopt/assets/gamefiles/RockBand/x360
Gathering midi data...
Parsing mid...
Using ridiculously accurate whammy.
Calculating notes in OD phrases and in solos...
Checking for errors in rounding...
Running optimizer...
100% Complete - Estimated time remaining 00:00:00
Reading opt file...
Optimal score: 293589
Wrote text path to tmp/foreplaylongtime.expert.guitar.0.0.0.txt
Total processing time: 271 seconds

real    4m30.732s
user    4m32.697s
sys     0m0.592s


I'm running Green Grass and High Tides at the moment but for some reason this has gone into a second pass and expects to be finished in about 2.5 hours...

Code:
tarragon@tyrell:~/git/apathtoofar/optimizer $ time ./rbopt.pl greengrass -v -g rb2 -p ~/cvs/spopt/assets/gamefiles/RockBand/x360
Gathering midi data...
Parsing mid...
Using ridiculously accurate whammy.
Calculating notes in OD phrases and in solos...
Checking for errors in rounding...
Running optimizer...
100% Complete - Estimated time remaining 00:00:00
Reading opt file...
Running optimizer...
5% Complete - Estimated time remaining 02:34:18


Note that the actual image generator code is broken, this is just generating the text path. Also note this is the "apathtoofar" code, I'll do some similar tests using the "spopt" code once this pass is finished.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
eddaket  
 




Joined: 19 Jul 2007
Posts: 2012
Location: Burillville, Rhode Island

PostPosted: Tue May 22, 2012 12:45 am    Post subject: Reply with quote

That's far more efficient than mine is, though some of that could be the system it was being run on.

I just got an idea that I am going to run with. Either tonight or tomorrow I'll get working on something.
_________________
Back to top
View user's profile Wiki User Page Send private message AIM Address PSN Name: eddaket
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Tue May 22, 2012 12:51 am    Post subject: Reply with quote

Further note on the second pass of Green Grass: on reading the code it appears to re-run the optimizer loop a bit different if it detects a "bad" path or whammy value during the first pass. So that partly explains that.

Also note that it's only running a single thread, so it's only using one of the CPUs in the dual-core. Some extra efficiency might be gained using some forking tricks.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Tue May 22, 2012 3:59 am    Post subject: Reply with quote

And the Green Grass run is done:

Code:
tarragon@tyrell:~/git/apathtoofar/optimizer $ time ./rbopt.pl greengrass -v -g rb2 -p ~/cvs/spopt/assets/gamefiles/RockBand/x360
Gathering midi data...
Parsing mid...
Using ridiculously accurate whammy.
Calculating notes in OD phrases and in solos...
Checking for errors in rounding...
Running optimizer...
100% Complete - Estimated time remaining 00:00:00
Reading opt file...
Running optimizer...
100% Complete - Estimated time remaining 00:00:00
Reading opt file...
Optimal score: 768558
Wrote text path to tmp/greengrass.expert.guitar.0.0.0.txt
Total processing time: 9767 seconds

real    162m47.285s
user    165m15.352s
sys     0m10.421s
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
LocalH  
 




Joined: 30 Oct 2006
Posts: 1325
Location: Kingsport, TN

PostPosted: Tue May 22, 2012 5:12 am    Post subject: Reply with quote

tma wrote:
Also note that it's only running a single thread, so it's only using one of the CPUs in the dual-core. Some extra efficiency might be gained using some forking tricks.

Alternately, you could process the charts in parallel, so that they'll get done faster overall, even if the extra cores don't help for single charts. Good for batch processing :P

A bit over two and a half hours doesn't seem all that bad for something that wouldn't need to be done again unless some modification to the algorithm is needed (such as if someone busts your upper bound and you can determine why :P) I'd imagine there are a few charts out there that would take even longer though, but GGHT is a good stress tester in general, I think it was one of the ones they used in RB1, either that or RttH.
_________________
Read my blog! The Incoherent Ramblings of a Lowly Geek

SCPH-39001, Free HDBoot 1.93, OPL 0.9.2 via 250GB HDD (all retail GH/RB games plus customs) - no memory card needed

Alakaiser wrote:
POST BECAUSE YOU HAVE SOMETHING TO SAY, NOT BECAUSE YOU HAVE TO SAY SOMETHING.
Back to top
View user's profile Send private message
THABEAST721  
 




Joined: 26 Dec 2007
Posts: 1999

PostPosted: Tue May 22, 2012 5:49 am    Post subject: Reply with quote

In case you didn't know, it's going to be hard to determine if your algorithm is working due to a very stupid glitch. People can adjust their lag to allow for higher amounts of whammy than would normally be possible. I don't know how you are applying the timing window or early whammy, but there are some unrealistic squeezes/activations they can get doing this.

Ehh I wasn't very clear... I am against the glitch. I don't think you should account for high lag values in this thing.
_________________
This does not leave my sig!!!

Back to top
View user's profile Send private message XBL Gamertag: THA BEAST 721
LocalH  
 




Joined: 30 Oct 2006
Posts: 1325
Location: Kingsport, TN

PostPosted: Tue May 22, 2012 6:27 am    Post subject: Reply with quote

Yeah, I did mean if someone manages to bust upper bound legitimately. Calibration abuse is bullshit.
_________________
Read my blog! The Incoherent Ramblings of a Lowly Geek

SCPH-39001, Free HDBoot 1.93, OPL 0.9.2 via 250GB HDD (all retail GH/RB games plus customs) - no memory card needed

Alakaiser wrote:
POST BECAUSE YOU HAVE SOMETHING TO SAY, NOT BECAUSE YOU HAVE TO SAY SOMETHING.
Back to top
View user's profile Send private message
tma  
 




Joined: 03 May 2007
Posts: 1414
Location: Australia

PostPosted: Tue May 22, 2012 6:59 am    Post subject: Reply with quote

LocalH wrote:
Yeah, I did mean if someone manages to bust upper bound legitimately. Calibration abuse is bullshit.


As it stands, neither spopt or apathtoofar allow for calibration abuse, and I've no intention of adding it.
Back to top
View user's profile Wiki User Page Send private message XBL Gamertag: zzUrbanSpaceman
Display posts from previous:   
Post new topic   Reply to topic    ScoreHero Forum Index -> Technique, Style, and Gameplay -> Guitar/Bass All times are GMT
Goto page Previous  1, 2, 3 ... 17, 18, 19, 20  Next
Page 18 of 20

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum




Copyright © 2006-2017 ScoreHero, LLC
Terms of Use | Privacy Policy


Powered by phpBB