Μάλλον δεν θυμάσαι καλά ακριβώς. Το SNES έχει έναν ψιλοχάλια επεξεργαστή, απλά το video hardware υποστηρίζει πολλά διαφορετικά εφέ και "modes" που δεν τον επιβαρύνουν και ισορροπει κάπως την κατάσταση. Πολλά shmups στην αρχή είχαν πολλά slowdowns για αυτόν τον λόγο, μέχρι να μάθουν να εκμεταλλευονται καλύτερα το υπόλοιοπο hardware.
Η βασικές διαφορές του Neo-Geo από τις "απλές" 16bit κονσόλες είναι, πρώτον, το πολύ μεγαλύτερο address space και όχι τόσο τα μεγάλα roms. Το θέμα είναι να μπορείς να επεξεργαστείς όλα αυτά τα δεδομένα ταυτόχρονα και όχι να έχεις εκατοντάδες MB στην rom να κάθονται. Το Neo-Geo μπορούσε να εκμεταλλευτεί τις τεράστιες roms που είχε.
Δευτερον, το πολύ δυνατότερο sprite engine και ο ταχύτερος επεξεργαστής που είναι αναγκαίος για να χειρίζεται όλο αυτόν τον χαμό στην οθόνη, που ακόμα και αυτός αποδείχτηκε κάπως αργός για τα Metal Slug.
Τέλος τα χρώματα. Μπορεί να δείξει αρκετές χιλιάδες χρώματα ταυτόχρονα στην οθόνη, χωρίς να θυσιαζει sprites, εφέ, scrolling, κλπ, και όχι λίγες εκατοντάδες ή δεκάδες οπως οι κονσόλες και τα homes.
Ολα αυτά μαζί δίνουν το αποτέλεσμα που όλοι ξέρουμε. Μέχρι και οι 32 bit κονσόλες επρεπε να κάνουν υποχωρίσεις για να δουν Neo-Geo ports. Ούτε τα 3MB RAM του Saturn (μαζί με το 1ΜΒ expansion cart) δεν έφταναν για να χωρέσει το πρώτο Metal Slug με όλα τα sprites και τα frames. Το Neo-Geo CD είχε 7 ολόκληρα MB RAM για να καλύψει την αλλαγή από roms στο αργό CD.
Bρήκα τη συζήτηση. ΤΟ Reddit το έχω παρατήσει εδώ και καιρό αλλά αυτό το θυμάμαι.
Sorry, but Vas is correct, the Neo was nothing special.
I feel like I'm being generous to it when I say it's only real strengths were fill rate and the address bus width for the graphics as while the fill rate was decent, even there it struggled quite badly with many games hitting the limits.
Put on a table I'd pick a SNES or MD over the MVS/AES every single time. The Neo was good at a select few things, but really anything beyond Ninja Combat was a case of throwing money at assets, rather than the system being a glowing example of power.
While expensive for a home system (due to amount of ROM required for a single game and a degree of marketing hype) it was a budget arcade system with basic capabilities that was easy to develop for, which is also why it lasted so long in the arcades.
Many similar concepts can be found in earlier Alpha developed hardware, such as that used by Gang Wars and POW (the main thing the Neo added over those was basic zooming) I didn't realise just HOW close they were until I emulated Black Touch '96, a Korean game which basically clones that earlier hardware.
Giving those old hardware designs a slight bump in capability, then separating the games from the base hardware by using a cartridge format so that games could be sold to operators more cheaply, for a platform they already had installed, is what allowed the system to succeed. The MVS design was a stroke of genius too, as it allowed operators to try out games without having to remove one that was still relatively popular with players. Requiring every single game to adhere to those standards and work in such a set-up was maybe more important to the success of the system than the hardware.
From a pure hardware perspective, a truly powerful system is also well balanced, I don't feel the Neo is balanced. It's success was mostly down to other factors.
και συνεχίζει σε άλλο
There's no way an AES could replicate what a lot of the 16-bit systems (and even some 8-bit systems) were doing either.
It's all about what impresses you the most I guess; cheap effects that rely on huge amounts of ROM are less impressive to me.
Try making a Gals Panic / Qix game on the AES/MVS without having to resort to modifying the cart design to incorporate a bunch of RAM. The system, as sold / made available to developers by SNK offers no way to do that natively whereas you can do it completely for free on a MD. (I can envision maybe some weird hardware abuse where you encode every possibility for a set of pixels being visible / invisible in a 16 wide strip then use raster effects to switch those per screen line, but that's an absurdly inefficient way of doing it in terms of ROM space and CPU time)
Anything that isn't just "mass of tile based sprites" it simply can't do. Even a lot of traditional sprite effects such as one sprite cutting through another it can't do either because the list rendering is strictly in-order.
The Neo might have been able to wow people with big chunky graphics and (very wasteful) use of high (and expensive) ROM capacity but please stop confusing those for power. Were it a genuinely more powerful system it could likely have supported much more efficient storage of the sprite data for example (non-tile based, RLE compressed etc.) but the design simply expected developers to just spend a fortune on ROM capacity instead.
If we're talking versus other arcade platforms, even the NeoGeo zooming functionality which SNK really wanted to show off early on is quite limited; you can only shrink sprites, never enlarge them, which makes it ill-suited for 'Super Scaler' type games too as you'd be having to store every asset at the biggest size it'll ever be visible at. Now obviously the MD doesn't zoom either, but it's why the only ever game you saw trying to do that on the Neo was the dreadful Riding Hero because in that field it would have been completely unable to impress compared to other arcade titles, even ones from over half a decade earlier (and the sprite-based hardware was ill-suited to the road drawing part too, making that much more expensive in terms of CPU processing cost)
και το τρίτο
If you were to put 40+MB of data onto a cart for the equivalent systems you could certainly come close in many cases, a lot of the compromises in the home ports came down to keeping the games affordable which instead meant making more intelligent use of the power on offer rather than just throwing capacity at every problem.
As I said, you wouldn't be able to get a NeoGeo to pull off a Qix clone; my ZX Spectrum could do that. It's not some wonder piece of hardware, it never was, it's a budget arcade platform with very specific capabilities that the library ended up being built around.
The Neo can't even draw shadows, it has to resort to cheap and tacky looking 'on-off sprite flicker' every single time. Even for the genre of choice, that one is quite a noteworthy limitation.
You use Sidekicks 3 as an example, but the pitch rendering on that is god-awful looking as the Neo can't do native linescroll, so they're having to abuse the raster effects to shift sprites about midscreen. It's also a very poor football game compared to the offerings on the consoles yet it's 10 times the size of any game on the home systems, and likewise the same amount more expensive.
Even if the price points had been identical, I think the MD and SNES would have outsold the AES on the strengths of the games available alone. AES owners might have looked down on others back then, but in private everybody was laughing at them while enjoying the better games we had. Take note that when the Dreamcast came along, and 'finally' gave everybody 'arcade perfect' ports, it absolutely bombed.
