Regarding 4:3 and 8:7. . . .

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
bocchi
Posts: 9
Joined: Mon May 16, 2022 7:35 pm

Re: Regarding 4:3 and 8:7. . . .

Post by bocchi »

I agree that pixel aspect ratio is a lot more useful and less misleading to talk about than screen aspect ratio, but I'm afraid the thread starter probably is correct about the latter being what you see discussed everywhere else by non-technical people. And even emulator developers often make the mistake of taking the 256x224 active display area and stretching that into a 4:3 rectangle for their 4:3 "corrected" mode, leading to a slightly too wide picture. Or even worse, doing the same with the 256x240 display in NES emulators. I think a lot of the confusion comes from not being aware the 256x224/240 area is not the whole video frame : that includes black borders on the sides as well as the top and bottom in 224 line mode. A picture says more than words so I have attached a captured frame from a real SNES:
mpv-shot0045.jpg
Note how the full frame is indeed 4:3 (by definition it has to be) but the active display area is actually narrower than that. Any comparision of whether square pixels or the actual SNES aspect ratio looks "better" would have to be made against an actually correct aspect ratio like this, not the naive 4:3 stretch often seen in emulators.

The "Perfect Pixel" naming of square pixel mode in recent Nintendo emulators is...bad.
User avatar
Dwedit
Posts: 4924
Joined: Fri Nov 19, 2004 7:35 pm
Contact:

Re: Regarding 4:3 and 8:7. . . .

Post by Dwedit »

I do like the idea of making a list of whether games were drawn with Square Pixels in mind, or Fat Pixels in mind.

A similar thing happened for MS DOS games. Many of them weren't drawn with the tall pixels of 320x200 in mind, they were assuming Square Pixels when the art was designed.
Here come the fortune cookies! Here come the fortune cookies! They're wearing paper hats!
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Regarding 4:3 and 8:7. . . .

Post by lidnariq »

Dwedit wrote: Sun May 29, 2022 12:58 pm I do like the idea of making a list of whether games were drawn with Square Pixels in mind, or Fat Pixels in mind.
The problem with this is that we already have a bunch of instances of games where some assets are drawn for 1:1 PAR and some are drawn for 8:7 PAR, in the same game, sometimes even at the same time.
Drag
Posts: 1615
Joined: Mon Sep 27, 2004 2:57 pm
Contact:

Re: Regarding 4:3 and 8:7. . . .

Post by Drag »

lidnariq wrote: Sun May 29, 2022 1:06 pm
Dwedit wrote: Sun May 29, 2022 12:58 pm I do like the idea of making a list of whether games were drawn with Square Pixels in mind, or Fat Pixels in mind.
The problem with this is that we already have a bunch of instances of games where some assets are drawn for 1:1 PAR and some are drawn for 8:7 PAR, in the same game, sometimes even at the same time.
Think about what the purpose of this kind of list would be: specifically artists who want to find examples of art assets drawn for different PARs. So therefore, if an artist is looking for examples of aspect-corrected art, this list should be able to point out examples from games where there is at least one example of assets being "obviously and deliberately" drawn aspect-corrected. Games where everything is aspect-corrected will be the exception rather than the norm, so it won't make sense to just list titles with no other information.
Last edited by Drag on Sun May 29, 2022 2:25 pm, edited 1 time in total.
iNCEPTIONAL

Re: Regarding 4:3 and 8:7. . . .

Post by iNCEPTIONAL »

lidnariq wrote: Sun May 29, 2022 1:06 pm
Dwedit wrote: Sun May 29, 2022 12:58 pm I do like the idea of making a list of whether games were drawn with Square Pixels in mind, or Fat Pixels in mind.
The problem with this is that we already have a bunch of instances of games where some assets are drawn for 1:1 PAR and some are drawn for 8:7 PAR, in the same game, sometimes even at the same time.
There's 1700+ games to research. The logic that you seem to be following (as far as I can tell), that it's basically not worth checking and listing a larger snapshot of the total library because potentially a few of them have a mix of both, is slightly misguided imo. My first-hand checking of the first thirty games on that Top 100 list that I posted, shows that the vast majority of SNES artists back in the day appear to have actually drawn their art to look proportionally correct with square pixels (as in, 1:1 PAR, "pixel perfect", and an 8:7 display aspect ratio), regardless of them eventually getting stretched once output from the SNES to some old 4:3 CRT TV. So testing an even bigger chunk of games is more than worth doing as far as I'm concerned, to definitively confirm this.

Now, personally speaking, there's a whole lot of people claiming we should all be creating our SNES art today with stretched pixels because the games were stretched on 4:3 TVs back then, but the research and list I've started in this thread already shows that is in fact not how the vast majority of artists back in the day making those 1700+ SNES games went about it. So, my conclusion here is that maybe those 1700+ artists who made all those SNES games back then are a good indicator of what approach might be the best-suited [and simplest] one for the vast majority of artists [not programmers] creating the art for their new SNES games in 2022 and beyond too. And I also think that is true whether those artists in 2022 and beyond are creating their new SNES games to run on an original SNES, a 3DS, a SNES Classic Mini, a Switch, a modern emulator or whatever. Or they can pre-stretch them to look proportionally correct at 4:3 IF they so choose--but no pressure. That's my informed conclusion anyway.

Or, just listen to what Drag said, as it sounds to me like he actually has a good use for such a list existing.
Last edited by iNCEPTIONAL on Mon May 30, 2022 12:06 am, edited 3 times in total.
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: Regarding 4:3 and 8:7. . . .

Post by Pokun »

I too think such a list would be great to have, but like others already said I'm not clear on how you got to that conclusion (besides that you haven't looked at nearly half of the 1700+ titles before getting to that conclusion). You need to have an objective method that others can follow exactly in order to repeat your research and come to the same conclusion. Looking at the majority of the game and make a subjective judgement which aspect ratio it works best in is not an objective method. Everyone would make their own list and come to their own conclusions, so the research would not be meaningful.

An example of an objective method would be to measure circles. Also you would have to document exactly what parts of each game where you identify your findings or it would be impossible to follow your research.
But I don't think that many games has natural circles like the moon in Chrono Trigger. Besides, the difference in aspect ratio is too subtle to matter that much in the most cases, but maybe that is your point?
bocchi
Posts: 9
Joined: Mon May 16, 2022 7:35 pm

Re: Regarding 4:3 and 8:7. . . .

Post by bocchi »

An example of blatant ignoring the PAR: in the intro sequence of Axelay there's this huge planet burning:
Axelay (Japan)_000.png
Axelay (Japan)_000.png (9.96 KiB) Viewed 777 times
An example of care being placed where it might not strictly have been necessary however is in Super Mario World:
Super Mario World - Super Mario Bros. 4 (Japan)_000.png
Super Mario World - Super Mario Bros. 4 (Japan)_000.png (7.87 KiB) Viewed 777 times
The sideways Mario is actually drawn a bit squashed to account for PAR!
iNCEPTIONAL

Re: Regarding 4:3 and 8:7. . . .

Post by iNCEPTIONAL »

Pokun wrote: Sun May 29, 2022 2:53 pm I too think such a list would be great to have, but like others already said I'm not clear on how you got to that conclusion (besides that you haven't looked at nearly half of the 1700+ titles before getting to that conclusion). You need to have an objective method that others can follow exactly in order to repeat your research and come to the same conclusion. Looking at the majority of the game and make a subjective judgement which aspect ratio it works best in is not an objective method. Everyone would make their own list and come to their own conclusions, so the research would not be meaningful.

An example of an objective method would be to measure circles. Also you would have to document exactly what parts of each game where you identify your findings or it would be impossible to follow your research.
But I don't think that many games has natural circles like the moon in Chrono Trigger. Besides, the difference in aspect ratio is too subtle to matter that much in the most cases, but maybe that is your point?
There honestly is no vastly better way or more "objective" method imo. The fact a lot of people in here seem to think there is a better way is, I would say, what's caused a lot of the conflict. You don't judge whether art looks proportionally correct at one display aspect ratio vs another using maths or physics or code or "measuring circles" with rulers or counting pixels or whatever (most of the art isn't going to be patent circles or squares anyway, and some of it won't even be flat-on at all); you judge it through the eyes of someone who understands what correct proportion looks like in these games (almost certainly an artist in this case, or just someone with a good eye for such things). And they do it by viewing a large chunk of the game, using their understanding of correct proportion [and indeed art in general] to judge what looks right across a whole bunch of the art assets, and, once they've come to their conclusion, they put it in the appropriate column. I hate to say it but this is not some kind of programmer-mentality exercise; it's deciding what display aspect ratio the art looks proportionally correct at and putting it in the appropriate column. Personally, as a trained artist, with a GSVQ 3 in art and design, an advanced certificate in design studies, and an honours degree in animation and electronic media, plus years of experience creating video game art (including pixel art for Rare on Donkey Kong Country Color and Banjo Kazooie for GBC) I'm sure I can look at a SNES game in both aspect ratios and know what ratio the original art was created in and what ratio it looks proportionally correct in. But that's just me speaking for myself.
Last edited by iNCEPTIONAL on Sun May 29, 2022 3:57 pm, edited 6 times in total.
iNCEPTIONAL

Re: Regarding 4:3 and 8:7. . . .

Post by iNCEPTIONAL »

bocchi wrote: Sun May 29, 2022 2:58 pm An example of blatant ignoring the PAR: in the intro sequence of Axelay there's this huge planet burning:
Axelay (Japan)_000.png

An example of care being placed where it might not strictly have been necessary however is in Super Mario World:
Super Mario World - Super Mario Bros. 4 (Japan)_000.png
The sideways Mario is actually drawn a bit squashed to account for PAR!
To be clear, those are examples of ignoring or taking into account the 8:7 PAR. There is the 1:1 PAR that can also be taken into account or ignored. So, technically, the first game absolutely takes the 1:1 PAR into account (which is almost certainly the PAR the original art was drawn at); it doesn't take the 8:7 PAR into account. Similarly, the Mario game takes the 1:1 PAR into account for the majority of the game (which is, again, almost certainly the PAR the original art was drawn at), but it adjusts one art asset to take the 8:7 PAR into account (probably because, unlike most of the rest of the art assets, the stretch on Mario sideways looked a bit too extreme on 4:3 CRT for the artist to comfortably ignore).

Just want to make sure people understand there is not only one single acceptable PAR for [making, judging, debating] SNES games and their visuals, which literally goes right back to my entire point of creating this thread (and the list, so far, very clearly demonstrates this).

Outside of that, yes, those are good examples of such things.
Last edited by iNCEPTIONAL on Sun May 29, 2022 4:02 pm, edited 1 time in total.
iNCEPTIONAL

Re: Regarding 4:3 and 8:7. . . .

Post by iNCEPTIONAL »

bocchi wrote: Sun May 29, 2022 12:53 pm I agree that pixel aspect ratio is a lot more useful and less misleading to talk about than screen aspect ratio, but I'm afraid the thread starter probably is correct about the latter being what you see discussed everywhere else by non-technical people. And even emulator developers often make the mistake of taking the 256x224 active display area and stretching that into a 4:3 rectangle for their 4:3 "corrected" mode, leading to a slightly too wide picture. Or even worse, doing the same with the 256x240 display in NES emulators. I think a lot of the confusion comes from not being aware the 256x224/240 area is not the whole video frame : that includes black borders on the sides as well as the top and bottom in 224 line mode. A picture says more than words so I have attached a captured frame from a real SNES:
mpv-shot0045.jpg
Note how the full frame is indeed 4:3 (by definition it has to be) but the active display area is actually narrower than that. Any comparision of whether square pixels or the actual SNES aspect ratio looks "better" would have to be made against an actually correct aspect ratio like this, not the naive 4:3 stretch often seen in emulators.

The "Perfect Pixel" naming of square pixel mode in recent Nintendo emulators is...bad.
If this in on an old 4:3 CRT TV, are you 100% sure that isn't just a result of you not adjusting the TV's display settings so the picture takes up the full screen (or as much as possible)?

I mean, I may not be recalling this perfectly, but I don't ever remember seeing black borders at the left and right sides of the images when playing SNES games back in the day. I think I remember black borders at the top and bottom--I'm from the UK, which meant I was playing on PAL TVs, so the extra scanlines were basically par for the course--but not to the sides.

Honestly, there's been a lot of times recently where I've wished I just had an original SNES and an old SD CRT TV to go check half the things I read in here directly. LOL
Last edited by iNCEPTIONAL on Mon May 30, 2022 12:12 am, edited 1 time in total.
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: Regarding 4:3 and 8:7. . . .

Post by Pokun »

iNCEPTIONAL wrote: Sun May 29, 2022 3:04 pm
Pokun wrote: Sun May 29, 2022 2:53 pm I too think such a list would be great to have, but like others already said I'm not clear on how you got to that conclusion (besides that you haven't looked at nearly half of the 1700+ titles before getting to that conclusion). You need to have an objective method that others can follow exactly in order to repeat your research and come to the same conclusion. Looking at the majority of the game and make a subjective judgement which aspect ratio it works best in is not an objective method. Everyone would make their own list and come to their own conclusions, so the research would not be meaningful.

An example of an objective method would be to measure circles. Also you would have to document exactly what parts of each game where you identify your findings or it would be impossible to follow your research.
But I don't think that many games has natural circles like the moon in Chrono Trigger. Besides, the difference in aspect ratio is too subtle to matter that much in the most cases, but maybe that is your point?
There honestly is no vastly better way or more "objective" method imo. The fact a lot of people in here seem to think there is a better way is, I would say, what's caused a lot of the conflict. You don't judge whether art looks proportionally correct at one display aspect ratio vs another using maths or physics or code or "measuring circles" with rulers or counting pixels or whatever (most of the art isn't going to be patent circles or squares anyway, and some of it won't even be flat-on at all); you judge it through the eyes of someone who understands what correct proportion looks like in these games (almost certainly an artist in this case, or just someone with a good eye for such things). And they do it by viewing a large chunk of the game, using their understanding of correct proportion [and indeed art in general] to judge what looks right across a whole bunch of the art assets, and, once they've come to their conclusion, they put it in the appropriate column. I hate to say it but this is not some kind of programmer-mentality exercise; it's deciding what display aspect ratio the art looks proportionally correct at and putting it in the appropriate column.
Well I don't fully disagree. Measuring things like circles is certainly math and it makes the most sense to use that in such cases as that burning planet, the Chrono Trigger moon or other things that most people agree are supposed to be round, like the clock face on Big Ben or the sun. But when it comes to things like characters it might just be the artist that want characters to be a little squashed, so in that case a fully objective method might not be possible. You would have to apply a qualitative judgment method which is basically judging subjectively. In that case you would have to describe the objects you find are better in one aspect ratio or the other, and why you think so, like Bochi did with sideways Mario in SMW. I'm an artist too and I agree with Bochi's conclusion in this case.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: Regarding 4:3 and 8:7. . . .

Post by rainwarrior »

bocchi wrote: Sun May 29, 2022 2:58 pmAn example of blatant ignoring the PAR: in the intro sequence of Axelay there's this huge planet burning
One might argue that a lot of plannets are wider than spherical due to their rotation and other factors, but I'd tend to agree it's just ignoring PAR.
bocchi wrote: Sun May 29, 2022 2:58 pmThe sideways Mario is actually drawn a bit squashed to account for PAR!
These are both really great examples. A list of specific examples like this with pictures and explanations I definitely think would be very useful to have.

A list of 100 games summed up only as whether iNCEPTIONAL thinks it's 4:3 or 8:7, with no other explanation, would not be useful. It would only start arguments, like the one that's ongoing.


Another relevant thing is that pixel artists for these games tended to use PCs of the time as tools. PC98 in particular usually had a 1:1 PAR. There are a billion ways to do it, monitors are often adjustable, and I think most artists don't just work at their computer in a vacuum and do actually look at the game on a TV once in a while... but I do think plenty of game art was just made at whatever PAR the artist's workstation had and never intentionally adjusted, whether or not they saw that it was "off". As others have been pointing out, even individual games are hardly consistent about this.

Probably to get a consistent game, you'd have to have an art director that cares about the PAR issue and whips people into shape about it.
iNCEPTIONAL

Re: Regarding 4:3 and 8:7. . . .

Post by iNCEPTIONAL »

Pokun wrote: Sun May 29, 2022 3:39 pm
iNCEPTIONAL wrote: Sun May 29, 2022 3:04 pm
Pokun wrote: Sun May 29, 2022 2:53 pm I too think such a list would be great to have, but like others already said I'm not clear on how you got to that conclusion (besides that you haven't looked at nearly half of the 1700+ titles before getting to that conclusion). You need to have an objective method that others can follow exactly in order to repeat your research and come to the same conclusion. Looking at the majority of the game and make a subjective judgement which aspect ratio it works best in is not an objective method. Everyone would make their own list and come to their own conclusions, so the research would not be meaningful.

An example of an objective method would be to measure circles. Also you would have to document exactly what parts of each game where you identify your findings or it would be impossible to follow your research.
But I don't think that many games has natural circles like the moon in Chrono Trigger. Besides, the difference in aspect ratio is too subtle to matter that much in the most cases, but maybe that is your point?
There honestly is no vastly better way or more "objective" method imo. The fact a lot of people in here seem to think there is a better way is, I would say, what's caused a lot of the conflict. You don't judge whether art looks proportionally correct at one display aspect ratio vs another using maths or physics or code or "measuring circles" with rulers or counting pixels or whatever (most of the art isn't going to be patent circles or squares anyway, and some of it won't even be flat-on at all); you judge it through the eyes of someone who understands what correct proportion looks like in these games (almost certainly an artist in this case, or just someone with a good eye for such things). And they do it by viewing a large chunk of the game, using their understanding of correct proportion [and indeed art in general] to judge what looks right across a whole bunch of the art assets, and, once they've come to their conclusion, they put it in the appropriate column. I hate to say it but this is not some kind of programmer-mentality exercise; it's deciding what display aspect ratio the art looks proportionally correct at and putting it in the appropriate column.
Well I don't fully disagree. Measuring things like circles is certainly math and it makes the most sense to use that in such cases as that burning planet, the Chrono Trigger moon or other things that most people agree are supposed to be round, like the clock face on Big Ben or the sun. But when it comes to things like characters it might just be the artist that want characters to be a little squashed, so in that case a fully objective method might not be possible. You would have to apply a qualitative judgment method which is basically judging subjectively. In that case you would have to describe the objects you find are better in one aspect ratio or the other, and why you think so, like Bochi did with sideways Mario in SMW. I'm an artist too and I agree with Bochi's conclusion in this case.
Sure, you can absolutely measure circles using math. I don't personally think anyone needs to do that in this case, and I won't be forcing anyone to do so either. If some people want to though, have at it.

Now, I won't be personally providing examples every time I put a game on one side of the list or the other, but, if other people want to, I would certainly be happy to see them do so.

I certainly agree that his two examples were drawn to fit a 1:1 PAR (and they do it very well) and not an 8:7 PAR (other than that one sideways Mario sprite). And, assuming this is true across all or the majority of the game art in both cases, they both go in the 8:7 [display aspect ratio] column.
bocchi
Posts: 9
Joined: Mon May 16, 2022 7:35 pm

Re: Regarding 4:3 and 8:7. . . .

Post by bocchi »

iNCEPTIONAL wrote: Sun May 29, 2022 3:34 pm If this in on an old 4:3 CRT TV, are you 100% sure that isn't just a result of you not adjusting the TV's display settings so the picture takes up the full screen (or as much as possible)?

I mean, I may not be recalling this perfectly, but I don't ever remember seeing black borders at the left and right sides of the images when playing SNES games back in the day. I think I remember black borders at the top and bottom--I'm from the UK, which meant I was playing on PAL TVs, so the extra scanlines were basically par for the course--but not to the sides.

Honestly, there's been a lot of times recently where I've wished I just had an original SNES and an old SD CRT TV to go check half things I read in here directly. LOL
It is a direct capture from the SNES's video output, there is no TV of any sort involved. That is the picture the (NTSC) SNES puts out.
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: Regarding 4:3 and 8:7. . . .

Post by Pokun »

I think it's enough to say that the moon looks wrong, show a screenshot of it and most people will agree. Your brain does the math for you.

Again if you don't describe your discoveries in any way I don't think your list would be very interesting or useful.
Post Reply