Author Topic: A ranking of the Cave console ports based on arcade accuracy [REMOVED]  (Read 4910 times)

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
UPDATE: 7 November 2018:

Ranking removed due to some inaccuracies and a general lack of detailed info. A new, more accurate and detailed ranking will follow soon in a new thread.

EDIT: New thread up: http://cave-stg.com/forum/index.php?topic=1767.0
« Last Edit: January 05, 2019, 11:02:47 PM by EOJ »
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
Several updates are needed for this old post, which I'll add in the near future. Particularly the addition of Akai Katana, SDOJ and Mushihimesama Xbox 360 ports.
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline ookitarepanda

  • Trade Count: (+1)
  • Maniac Member
  • *
  • Posts: 341
I second the comments about Deathsmiles. It explains why my highest score is on my 360 version and not on my PCB...  :facepalm:

Offline xygax

  • Ballet Hell
  • Trade Count: (0)
  • Maniac Member
  • *
  • Posts: 15
  • It Hurts
Yup that's the list I was looking for, thanks EOJ  :)

PS: would be nice to have the 32bit ports in the list too even if just brief mentions, the crappy DDP ones are well documented, but DP not really.

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
Hopefully someone else can chime in about the 32bit ports as I am not qualified to give an opinion worthy of serious consideration.
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline xygax

  • Ballet Hell
  • Trade Count: (0)
  • Maniac Member
  • *
  • Posts: 15
  • It Hurts
BTW since there's talk about emerging lag reduction features in emulators right now, I was thinking "well that will help making CV1000 games a bit more playable in MAME", then I remembered those play at incorrect speeds anyway.

I don't know if people remember but the main issue was the missing slowdowns, and a single feature called blitter delay (which has a slider in the MAME ingame menu and a basic setting in DEmul) could help emulating those at least partially in some games.
For instance it was said that for Futari and DeathSmiles 63% is the best/closest value, producing results similar to the 360 ports.

But that was a long time ago and the MAME driver has been revised a few times, the topic then fell into oblivion.

It would be interesting for reference to have the opinion of very educated players like you EOJ about that, what blitter delay value seems best (if any!), and add a word mentioning emulation accuracy in that list or a separate/addendum.
IIRC someone said there haven't been many revisions of the pcb's fpga so there should be only as many 'correct' values to find to cover all titles.
Of course there's still a ton of games to check so that'd be baby steps updates, I've tried to find out about some myself but I'm no good.
« Last Edit: April 20, 2018, 09:46:06 AM by xygax »

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
I don't know much about the MAME stuff. I have the roms and have tried them out, but they all played so horribly (I'm talking about the SH3 games) I never touched them again. If someone else knows about this stuff maybe they could chime in. For me no matter what I did the slowdown was all screwed up and input lag was horrendous. I guess they could be fun to play around with (service menu, etc). And it's one of the few ways most of us will get to try out Ibara Kuro until some kind company (come on, M2!) ports it.

For Cave games I just play X360, PS4 and PCB. I might buy their Steam releases too just to check them out.  :righton:
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline cstle

  • Trade Count: (0)
  • Maniac Member
  • *
  • Posts: 10
the Steam releases are on sale until the 23rd if you're interested at the moment  :)

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
I finally updated the main post, adding in DFK BL, MMP, PS, SDOJ, Mushi X360 and Akai Katana. I also made some minor adjustments to other games in the ranking.
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline xygax

  • Ballet Hell
  • Trade Count: (0)
  • Maniac Member
  • *
  • Posts: 15
  • It Hurts
Yeah in short there are things you can do to attend to the emulation issues a bit, but to judge the worth of it it requires close examination from some who know the games well but also got the right hardwware for the job: a pc with a fast cpu, and display+controls as lagless as possible.
AFAIK GroovyMame can shave off nearly two of mame's extra frames if not more with a recent experimental feature, making the SH3 games playable in that field.
Then finding the right blitter delay values. Maybe someone will look for that one of those days yup. ^^

note: not long ago iirc mamedev were trying to gather as many cave pcbs as possible again to document the actual behaviour (slowdowns, glitches etc) in an attempt to rework the drivers for better accuracy.
but that was when most pcb prices already skyrocketed, talk about impossible

Anyway thanks a lot for the list update!  :righton:

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
I would like to add Guwange and Dangun Feveron to the list. Is there anyone here who is knowledgeable about these games and has played the PCBs and the ports?

Also, I made more revisions to the main post.
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
I have been playing several PCBs a lot, and then playing the ports soon after. Lots of testing back and forth. I think the Futari 1.5 port is still pretty accurate, but not quite as accurate as I previously stated. Black Label is even less accurate than I originally stated (and noticeably less accurate than the 1.5 port). So, I will have to adjust the list accordingly to reflect this (i.e. split 1.5 and BL up, instead of lumping them together), among other things.

I would also now classify the MMP port as "poor" instead of "decent". It really is not very good, in several respects, and I am convinced it is impossible to score as well on it as the PCB. Of course, the same is true of the Pink Sweets port, but for different reasons.

I think it will be more useful to simply have different tiers, instead of a number ranking. After all, it is difficult to quantify these things, and when I say a port is 90% accurate in regard to slowdown, that is just an estimation, not some precise measurement.

Also, I welcome any other opinions on these matters from other players who have spent time with the PCBs and ports, particularly those who own PCBs I do not have access to ATM (which is everything except Futari BL, Futari 1.5, Espgaluda 2 and MMP).
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline Kaneda

  • Trade Count: (+3)
  • Maniac Member
  • *
  • Posts: 300
I would also now classify the MMP port as "poor" instead of "decent". It really is not very good, in several respects, and I am convinced it is impossible to score as well on it as the PCB.

I don't think it's so bad, but I agree that it's impossible to get as high of a score on the port as on the PCB. I'm also someone who doesn't lose their mind over input lag like some others seem to, so that says something about how picky I am about port accuracy. To me, if I don't see a marked difference in scores between the PCB and the port or MAME, then it's all academic.

The major difference between the two (besides the baffling moved enemy pattern towards the end of stage 5) is in milking the stage 1 boss. I'm starting to think that the PCB, maybe just due to heavier slowdown, produces more waves of patterns to milk when the boss is firing its strongest blue pattern. The 360 port is just not as generous.

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
Thanks for your input!

Re: MMP port vs PCB: There are several other important scoring spots that have little to no slowdown in the X360 version. The ships at the end of Stage 1 are a big one, you just can't score with them like you can on the PCB because the slowdown isn't there. Same deal with the same ships in the second half of Stage 5. The second half of Stage 4 is another. These spots aren't big point-makers by themselves, but they add up over the course of the game.

The MAME emulation of slowdown is more accurate (but then you have to deal with more input lag, so it's a tradeoff). But this depends on what version of MAME you are using. I was using something in the 180s before (I think 187?), and the slowdown was pretty good (not perfect, but better than the X360 port). But on the newest 199 build there was pretty much no slowdown anywhere. I don't know what happened there.  :displeased:
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
Re: A ranking of the PS2 and X360 Cave console ports based on arcade accuracy
« Reply #14 on: September 05, 2018, 01:10:17 AM »
I've decided to redo this whole list, after playing a lot of the PCBs and ports back to back over the past few months. I think instead of a ranking I'm just going to list each game, and rate them (A-F) on three criteria:

-Slowdown accuracy
-Core port accuracy (enemy placement, bullet patterns, input lag, etc)
-Visual options (vertical screen (tate) options, scanline options, smoothing options, etc)
My score archive
twitter: cavexstg
XBL gamertag: eojx9999

Offline EOJ

  • Administrator
  • Trade Count: (+35)
  • Ultra Member
  • *
  • Posts: 7230
Re: A ranking of the PS2 and X360 Cave console ports based on arcade accuracy
« Reply #15 on: November 06, 2018, 09:29:11 PM »
I am working on a new thread with more updated and accurate analyses of the slowdown accuracy. I have been buying most of the PCBs over the past several months, testing them against the ports, and recording detailed notes. I will lock this thread now and delete the initial post, as it is largely outdated and several aspects are inaccurate. The new thread will follow soon.
My score archive
twitter: cavexstg
XBL gamertag: eojx9999