mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support
-
@Riverstorm said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
but I'm curious on the 292. Can you show me the formula you're using to get to that resolution?
By resizing "source"-image (256x240) with a 8:7 PAR to one of 1:1 -> 256*8/7 = 292.571428(repeating), rounded down to even (292x240). Just to mention it: by reproducing the steps mentioned in the nesdev-link in PSP12 (adding 2x12px border, then (pixel-)resize to DAR 4:3) I am getting 293px for the game-image-area, but imho even numbers are easier to handle.
-
@Ashpool - Thanks for sharing the link. That was useful. I was getting there but more of a round about way. I could only get there with the DAR and not the aspect ratio.
aspect ratio = ?
(width / height) / DAR = pixel squareness
width / pixel squareness = 292.55999...or as 8:7
(width / height) / aspect ratio = pixel squareness
width / pixel squareness = 274.28571...or as 4:3
(width / height) / aspect ratio = pixel squareness
width / pixel squareness = 320I guess I usually trust the core to do a good enough job on resolution and conversion adjustments or like you pointed out even certain PARs like 6:5 (1.2:1) are damn close. For me to raise a red flag of this doesn't like right it has to be pretty far off.
If it helps here's a script I use to generate shader cfg's. It wouldn't fit the bill of "exactness" you're looking for but it shows the basic parameters of creating a custom cfg. to hit that 292:240 exactly. Do you know what the core generates for a resolution by default? I take it it's not 292:240? Are you scaling up to 1080p? How are you scaling up (manually; integer scaling to keep the AR?) It's a big task to adjust every game without some type of script to generate custom cfgs.
https://github.com/Wilstorm/retropie_shader_configs/blob/main/retropie_shader_configs.py
For vertical games it fills the y axis and adjusts the x axis based on the inputted target resolution to attempt to make them divisible by whole numbers within a user set % tolerance and more or less vice versa for horizontal games. Basically it helps clean up shader artefacts at the slight expense of some stretching on the x or y axis.
The math is somewhat similar. I update the databases once in a while. Mainly m3plus and FBNeo change and add games but the rest are fairly static. The newer MAME DAT's like mame2015 and mame2016 have a lot of unneeded entries that I try to pull based on the driver such as xbox or xerox820 or zrt80, etc. but the extraneous cfgs generated don't hurt anything but I continue to tidy them up. The newer ones are just huge!
-
@Riverstorm said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
I guess I usually trust the core to do a good enough job on resolution and conversion adjustments or like you pointed out even certain PARs like 6:5 (1.2:1) are damn close.
1st: by 6:5 i've meant the DAR not PAR you get from resizing 256x240 by the PAR of 8:7, without adding the borders needed to get a 4:3 DAR covered - or maybe we may need another term instead of DAR here? PAR1-AR? [AFAIR setting a DAR in the core is in reference to the targeted/source DAR and will be blanked/bordered if the used display is of other dimensions/aspect-ratio |my assumption: -> so, if MAME is simply ommiting the step of adding the blanks from 256 to 280, 6:5 is the DAR to be set in core options to get the "292", instead of the "320" the core is stretching to by assuming a "direct" DAR calculation from SAR to 4:3|]
2nd: I too, in general, keep faith in the cores... But somehow they are (in regards to one emulation they share (nes in our current context)) in some form of disarray (For instance by just using the SAR size of 256x224 (ommiting the 2x8px invisible bars on a analogue crt screen (if you ever captured a video tape or analogue tv signal you will remember the areas in the captures you have not seen on tv's screen - more real visible picture left/right, but also some blank/bar above and garbage below the image (which in fact could have incorporated digital data transmissions))) you are not changing the PAR, so 292x240 (with the 2x8px top/bottom offscreen bars) should also be identical to the image in 292x224, and 272x224 is simply the +-~same Image-AR covered, but gained from a wrong pixel-stretch... And now you got me (eventually) hooked .:. It reminds me on old VHS captures and getting the captured signal right (guides and information) ... and well .. I am not sure who (in general; not me nor you meant) is right, or more close to be exact... and maybe this should go to some seperated threat in general or wherever if we continue it (on the example of, and solemnly for the nes?) ;)p.s.: Thanks for the link to your script! I have/had a project to calculate a custom viewport running (also in python)... but with a different goal/aim in mind (unfinished by now) - and such things/observations as this one are always telling me that there is another case to be captured (my scripts build upon the assumption Arcade (MAME) is 4/3 or 3/4 per screen if the xml information ain't providing contratictinging data)...
-
@Ashpool - ooops! Easy now, just a typo on PAR…I’ve got PAR/DAR/SAR on the brain. 292:240=73:60 should work to calculate squareness. I was just trying to help you get to that perfect res. ;)
No need to make a new thread. I have no opposition to your math or point of view. I simply don’t play enough NES to care about the finer details. I forgo the perfect dimensions to use shaders. All these questions should be easy enough to answer by firing up a core and see what it outputs with different settings. Worst case scenario screen capture should grab any “dark” pixels as well as the picture.
-
@arcadez2003 since the catver.ini wasn't updated in a while, and as I just brought it up2date for my personal use (in regards to the last datfile commit from december 22), I would like to ask whether forking the rep and opening a pull-request for it would be ok and if so, are there any conditions I should be aware of for the format of the catver.ini? For now I just inserted/removed in place, but by that there may be a chance that I eventually will break/have broken the clones under parent-name ordering as it is now (at least that is what I assume to be the reason of all the breaks in the pure alphanumerical ordering of catvers I am aware of).
-
@Ashpool said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
@arcadez2003 since the catver.ini wasn't updated in a while, and as I just brought it up2date for my personal use (in regards to the last datfile commit from december 22), I would like to ask whether forking the rep and opening a pull-request for it would be ok and if so, are there any conditions I should be aware of for the format of the catver.ini? For now I just inserted/removed in place, but by that there may be a chance that I eventually will break/have broken the clones under parent-name ordering as it is now (at least that is what I assume to be the reason of all the breaks in the pure alphanumerical ordering of catvers I am aware of).
First i've spotted this having been busy with code and things, you could always open an issue on the MAME2003+ git then the whole team
can have a chance to discuss this further.Regards.
-
@arcadez2003 Thanks & Done...
-
Sorry to bother again [!], and not shure whom i should @ here, @arcadez2003 / @administrators (?).
In regards to the doc-pages for retropie/mame2003-Plus...Currently cite for the catver.ini is:
The MAME 2003-Plus catver.ini also contains data on games definitively known not to work, as well as sorting data for mature games or other, less desirable, romsets.I am not a native speaker, but for me it implies that the catver.ini holds the metadata for the working state within the emulation (plainly wrong assumption), but maybe it could also be read as "it also includes genre informations for games known as not working" - the information working state is (AFAIK) part of the mame datfile, not of the catver and I would either like to change that phrasing to a plain definition of the catvers function (assigning genre/subgenre and mature state (and of no relevance in this case, but for younger mame/catver generations additional mame version information) ) or remove the mentioning of the catver from the docs.
Nevertheless, as said, I am not a native speaker and I am not sure whether i missed some meaning of that phrasingand, or if not how it could be rephrased into something simpler and short that just implies/states that as the m2k3+ core diviates from the pure mame branch and the catvers (from the progetto-SNAPS page) for a certain mame version can't be matched to it, the projects git page provides its own catver.ini to supplement this core. -
@Ashpool - It might be just misreading it. The word 'definitively' isn't really required to make the point but it does let you know what is written is absolutely correct, without doubts. Pretty much by definition. I suppose you could use 'definitely' vs 'definitively' but they are kind of different. Similarities with two different uses.
It's just saying some entries exist that aren't part of the working set or basically extraneous entries. I think you're saying the same thing but it sure can't hurt to make the verbiage more clear if the doc folks feel the same.
"it also includes genre informations for games known as not working"
I think you need to make information(s) singular and I would change it to something like this:
"It also includes genre information for games known to not work."You could also drop 'genre' implying the subject is the catver.ini if it's been mentioned previously or is the section heading, etc.
"It also includes information for games that do not work."but that might be to close to what was originally written or by inserting definitively again.
"It also includes information for games that definitively do not work."There's definitely more than one way to skin a cat on this one! ;) Regional dialect can dictate the verbiage. I think you can have a lot of fun with it honestly but I'm not sure who maintains the docs.
-
@Ashpool - Another option is to post on the Github web page where you been submitting the catver.ini updates as they will get it more quickly or at least will reply more quickly I think.
-
@Riverstorm said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
The word 'definitively' isn't really required to make the point
Yup, I think you nailed the point of my confusion and I may consider asking/talking about it on the git-rep as you suggested in your 2nd reply (Here in my region the schools holidays (for my/our kid) just started, so I have some time to sleep over it ;]).
P.S.: And be it just because it then boils simply down to the fact that the docs statement, as it is now, is true (canon/lore) for all catvers/catlists (why even mention it then [edit: for a particular core]?) and doesn't explain why there [may be]&|[is] a need for a separate catlist for mame2k3+ with its state of being based on mame .78, but incorporating fixes/additions from/for later versions [1] and that therefore none of the regular catvers for mame can truly match the roms listed in the xml/ini for this core.
[Edit:]
1: and even incorporating homebrews/one homebrew available for purchase ["hypernoid"], which (of purpose?) are/is not included in the official mamedev branch, and so far unknown to the latest catvers for mame [AFAIK for now]. -
We're still dabbling with this core now and again and as a result you can play three more games one of which cant be played in current MAME.....
Deroon DeroDero
Just a Puzzler but alotta folks like this one and the music is kinda funky
Toukidenshou - Angel Eyes
Decent fighter this one as you'd expect from TECMO
Top Driving
World Rally Clone there were many of those but this is a decent wee game in it's own right
-
@arcadez2003 said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
exclusive to MAME2003+ for now
Actually, iq132 wrote a FBNeo driver for it a few days ago.
-
@barbudreadmon said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
@arcadez2003 said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
exclusive to MAME2003+ for now
Actually, iq132 wrote a FBNeo driver for it a few days ago.
Yeah your right i forgot about that and as a result i've edited my original post to reflect that :)
-
Here's a handy addition for MAME2003+ a new version of Dogyuun which means for the first time the game will have full sound and music in this core
Dogyuun (8/25/1992 location test)
Enjoy
-
Just a headsup you can now play this game using MAME2003+
Clockwork Aquario
-
Don't know if this is the right place for a bug report of sorts, but it seems the P2 start button isn't working on my
tinklpitin M3+. Can anyone confirm / is it even something fixable? -
@WeirdH said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
Don't know if this is the right place for a bug report of sorts, but it seems the P2 start button isn't working on my
tinklpitin M3+. Can anyone confirm / is it even something fixable?Thanks for the bug report a fix is on the way
-
Just a headsup following on from the recent work to get Clockwork Aquario up and running both of these are now playable using MAME2003+
DD Crew
You'll need to build a romset which uses decrypted roms for ddcrew similar to some other Sega games eg Turbo Outrun ect ect
Hammer Away
Rare Sega / Santos System 18 prototype which turned up several years back only two pcb's were known to exist
-
@arcadez2003 said in mame2003-plus: hundreds of new games, improved input, features, new bugs - now with runahead support:
You'll need to build a romset which uses decrypted roms for ddcrew similar to some other Sega games eg Turbo Outrun ect ect
Dumb questions: Will the dat file take care of that, i.e. are the decrypted rom files recognizably different from the encrypted ones for a romset builder like clrmempro? If not, how can I know myself which romfiles are decrypted?
Rare Sega / Santos System 18 prototype which turned up several years back only two pcb's were known to exist
I just love the retro scene for things like this. ❤️
Contributions to the project are always appreciated, so if you would like to support us with a donation you can do so here.
Hosting provided by Mythic-Beasts. See the Hosting Information page for more information.