› Foros › Retro y descatalogado › Consolas clásicas
The way detection work was very simple but fundamentally changed the set-up. Whenever you fired a gun, it had a radius test and alerted the non-player characters within that radius.
If you fired the same gun again within a certain amount of time, it did a larger radius test and I think there was a third even larger radius after that. It meant if you found one guy and shot him in the head and then didn’t fire again, the timer would reset.
It wasn’t realistic but it meant the less you shot, the quieter you were, the less enemies came after you. If an NPC that hadn’t been drawn and was just standing in a room waiting was alerted by gunfire, it would duplicate itself and one went to investigate. You can see it happening sometimes – if you go to the right place and make a noise, you see more enemies spawning.
David: I was given a book, and in it the authors were talking about the original Doom on the PC, and that each level was composed of about 3,000 polygons, and that each level in Quake was composed of about 10,000. An average level in Turok runs between 250,000 and 300,000 polygons. That's a lot of mass, a lot of polygons.
C-Left, C-Right, C-Up, C-Down, R, L, L, R, C-Up, R, C-Down, R, L, C-Right
Objects out of ram(1) !!
JacintoCinete escribió:Buen hilo, me apunto!
WAVE RACE 64
- Al diseñar el juego se planteó la posibilidad de usar lanchas futuristas en vez de motos acuáticas, e imprimirle mayor velocidad a las carreras. Se descartó para evitar parecidos con el posterior "F-Zero X".
- Fué versionado para Gameboy y tuvo en “Wave Race: Blue Storm“ de Gamecube su secuela espiritual.
Sogun escribió:JacintoCinete escribió:Buen hilo, me apunto!
WAVE RACE 64
- Al diseñar el juego se planteó la posibilidad de usar lanchas futuristas en vez de motos acuáticas, e imprimirle mayor velocidad a las carreras. Se descartó para evitar parecidos con el posterior "F-Zero X".
- Fué versionado para Gameboy y tuvo en “Wave Race: Blue Storm“ de Gamecube su secuela espiritual.
En realidad el Wave Race de GameBoy es de 1992 y fue lanzado en Europa años más tarde tras ver el tremendo éxito de la secuela en 64 bits. enlace a la wikipedia
Aquí un vídeo de ese primerizo Wave Race 64 con lanchas futuristas y unos circuitos que nada tienen que ver con la versión final. La verdad es que uno se pregunta si no se trata de desarrollos distintos, porque es que ni el agua es parecida. Es como si hubieran rehecho el juego de cero.
Del Wave Race 64 siempre me llamó la atención que no tiene secuencia de créditos. Ni siquiera en el manual viene nada. Leí cuando salió que Miyamoto estaba implicado en el desarrollo y sé que la música es de Kazumi Totaka porque es de los pocos juegos donde aún están buscando su famosa canción.
gcc-core
gcc-g++
libmpc-devel
libpng-devel
wget
texinfo
make
cd ..
cd ..
cd usr
cd local
cd libdragon
cd temp
./build
cd ..
export N64_INST=/usr/local
make
make install
make tools
make tools-install
cd temp
cd libmikmod
make
make install
cd ..
cd ..
cd examples
cd bmb
cd filesystem
$N64_INST/bin/mksprite 32 conker.png conker.sprite
cd ..
make
cd ..
cd ..
cd usr
cd local
cd libdragon
cd examples
BMBx64 escribió:TUTORIAL: COMPILAR Y USAR LIBDRAGON EN WINDOWS
Bueno pues vamos con algo nuevo, esto solo tiene sentido para el que tenga un copión o flashcart y quiera trastear con la consola, las roms generadas no funcionarán en la mayoría de emuladores (los que solo miran libultra), funcionará en emuladores a bajo nivel como MESS o CEN64.
Asset creation was more of an issue with the restriction than the limitation itself. I really hated how we had to make the cars look in Top Gear Rally, because we allowed the users to edit the textures they had to have a very simple mapping which precluded us from making them look much better than they did.
kassanmoor escribió:excelente hilo, lei gran parte de las curiosidades, por aca existia la leyenda urbana de que si se te arruinaba la ranura de cartuchos de la consola, podias conectar cartuchos a la ranura inferior {la del disk drive}, aparte recuerdo que en las cajas de N64 las imagenes de mariokart tenian items inexistentes como una pluma y la bombas eran de colores, y en una revista salia Kamek en lugar de Donkey Kong, a ver si alguien sabe mas sobre eso
Waldo64 escribió:Me han sorprendido mucho los datos sobre el coste de la consola, y de los cartuchos, visto así no me extraña lo "barata" que era la consola, y lo carísimos que eran los juegos comparándolos con los juegos en formato CD de la competencia.
Sigue así @BMBx64 me lo estoy pasando muy bien con tus posts!!!
Anyone know the actual fillrate of the N64? I couldn't find it.
- Wouldn't be of much use, the theoretical numbers were much higher than what you'd see in practice since it was bandwidth limited for the most part. Plus you have to take into account things like stalls while loading the texture into the cache which the PS1 never had to deal with.
Turning off Z made a considerable difference, World Driver runs without a Z buffer for this reason, but it was a significant risk when we made that decision, because it was unclear if Nintendo would bounce a title for excessive Z fighting.
And the only Nintendo supplied uCode that made it practical to omit the ZBuffer didn't come until very late and was feature incomplete.
At a hardware level you build command lists in memory and DMA them to the local memory on the RSP, the RSP runs arbitrary code on the command list and eventually you output rendering primitives to the RDP. Most uCode writes this back out to memory into a circular buffer for the RDP to read (increasing the load on main memory), although it was technically possible for the RDP to read directly from RSP memory RAM was so restricted, the RDP tended to starve.
Reordering is certainly not as simple as on PS1, but you could run sets of static display lists in arbitrary orders pretty easilly.
I think most people used the Zbuffer because it was there.
It's actually an Amiga style tracker, channels are left or right, I can't remember how many channels we supported off the top of my head. N64 really didn't have good audio.
Audio like our later titles was a tracker, because that's what our sound guy (Barry Leech at the time) wanted given the very significant constraints. At the time the Nintendo libraries used the RSP for the mixing, but we measured a significant performance penalty for this, so I wrote a mixer on the main CPU, which balanced load a bit better. For WDC we moved the mixing back to the RSP because we had access to the uCode and it wasn't really the bottleneck.
TGR happened because we were located just up the road from Kemco's US office. They were looking for a dev to do a Top Gear racing game on Nintendos upcoming platform, and several of us wanted to do a racing title.
The prerendered sequence we did back then was largely to convince Nintendo to get devkits, it wasn't a common sales tactic in those day, but since it's become common practice.
The physics engine went in very late in TGR, There were actually two developed and the first prooved to be unusable. I wrote the second one to dig us out of a dev hole. In terms of what's modelled it's similar in principle to the WDC one, the tire model is simpler and it has some bonus bugs in it. The WDC one is also about 10x faster (which gives you how an idea of how much implementation can change performance) mostly due to improvements in the collsion representation.
The Snow tracks came about because an artist volunteered to do it in a meeting with Kemco. Everyone loves the Jungle in the Snow...
The paintshop was an idea that had been thrown around internally, and I always fought against (I lost) because it hurt the graphics quality of the cars significantly.
Q: Considering the fact that every voice costs processor time, how do you go about creating a song? Do you start with as many voices as possible and scale back, or do you limit the number of voices from the get-go?
Voss:
I prefer to shelve the processing resources. Voice-wise I usually make things work within 16 physical channels to eliminate the unknowns of voice stealing. ROM/RAM resources-wise I will start with the samples at a higher fidelity and then work them down from there.
Q: What's a good, average number of voices for music and effects?
Voss:
16 for music, 8-16 for sound effects, mixed at 44.1 Khz. RAM-wise somewhere around 512K-1MB for music/sound. I would say soundtracks are in a range of at least 10-25% of the experience of a game, psychologically speaking. So for resources in general, developers should mirror those numbers in how much space, processing, and budget they look at for the soundtrack. In games where the soundtrack is a more key element, there is the added benefit that if the soundtrack is strong enough it could stand on its own as an album, which could reciprocate some of its development cost.
Q: H2O basically created brand-new sound drivers for Tetrisphere. How powerful are they compared to the original SGI drivers?
Voss:
In retrospect they were close to the same, but allowed me to use a much better and more applicable editing environment. I sequenced the audio/animations/music using a tool called 'Fast Tracker 2' on the PC. This is an event-based music sequencer with everything internalized. It is very cryptic, but if you can handle that, it is a very viable alternative to a lofty MIDI sequencer. We converted that to a proprietary format for playback on the N64, which was calibrated to the N64's low-level audio functions. The replayer allowed me to avoid using AdPCM and use 8-bit samples (about 90% of Tetrisphere's samples are 8-bit). Then we added a simple compression which took about 40% out of all the samples' sizes. The whole sample bank was around 2.75 megs -- all music and sound samples -- with about 1/2 meg of audio pattern data. As an interesting sidenote I did not use any effects-bus -- built in reverb, chorus, delay, effects -- because we were unable to get it going in time along with the custom player. All the effects are hand-programmed delays and choruses within the music.
Since then I have worked with a programmer (Ian Luck) to develop a custom converter which translates a number of "tracker" formats perfectly to Standard MIDI files, along with variable options for calibration of playback to different environments, and "re-wiring" of MIDI controllers to control specific nonstandard effects or events.
In the end, I had 700k for music data and instrument samples, and about the same for sound effects… basically, all the data could have fit onto a 3.5inch floppy disk….so it was a, let’s say, a creative challenge cramming everything in.
BMBx64 escribió:3) Donkey Kong 64, este ya es un caso conocido, tuvieron que usar la expansión y incluirla con el cartucho porque el juego tenía un bug y se colgaba aleatoriamente (estarían pisando memoria reservada?), es un fallo que incluso puede estar presente en algunos módulos de expansión no oficiales, por lo que se recomienda comprar el oficial.
BMBx64 escribió:3) Donkey Kong 64, este ya es un caso conocido, tuvieron que usar la expansión y incluirla con el cartucho porque el juego tenía un bug y se colgaba aleatoriamente (estarían pisando memoria reservada?), es un fallo que incluso puede estar presente en algunos módulos de expansión no oficiales, por lo que se recomienda comprar el oficial.
DiGiCharatFan escribió:@BMBx64 esto lo sabras, que hace esactamente el PIF? a parte de ser el CIC seleccionando la región.
Lo digo poruqe en el grafico donde se meciona pone PIF CONTROLLER, que mas controla?
Aracnoid escribió:@BMBx64 por casualidad, tu no sabrás como rescatar las partidas guardadas de los cartuchos para poderlas jugar en el emulador (no quiero volver a pasarme todo el Zelda OOT)
Saludos y a seguir con este fantástico hilo!
Aracnoid escribió:@BMBx64 por casualidad, tu no sabrás como rescatar las partidas guardadas de los cartuchos para poderlas jugar en el emulador (no quiero volver a pasarme todo el Zelda OOT)
Saludos y a seguir con este fantástico hilo!
BMBx64 escribió:Cartucho original o flash? Aún así el Zelda OOT tiene bastantes cheats para situarte casi donde quieras.
Aracnoid escribió:Cartucho original, el tema seria copiar la partida del cartucho original y meterla para jugarla en emulador y/o en flashcard.
¿Como se pondrian esos cheats tanto en emulador como en flashcard?
3D clipping is expensive and should be avoided. Methods employed by the host application which can reduce the amount of geometry that gets clipped are a good idea. Crude visibility determination algorithms, geometric level-of-detail, and careful scene construction can help improve clipping performance dramatically.
The clipping algorithm is sensitive to large numbers and overflows.
Because microcode loading increases overhead, discretion in loading is recommended to obtain good performance. For practical purposes, this means intermittently switching between the microcode types used for rendering. As an example, a clip-capable microcode such as F3DLX would be used for rendering landscapes and fast microcode such as F3DLX.Rej used for rendering characters. As with previous releases, switching between F3DEX and L3DEX when rendering lines can be done without CPU involvement.
Triangle Rejecting Process
- Triangles which are entirely inside the rejecting rectangle are drawn.
- Triangles which stretch from inside the rejecting rectangle to outside, or those that are entirely outside the rectangle, are rejected.
- The rejecting box is 2-6 times the size of the screen rectangle. By default, the rejecting box is 2 times the size of the screen rectangle. The range can be changed using g*SPClipRatio. Values from FRUSTRATIO_2 to FRUSTRATIO_6 can be specified. In the Z direction (the depth direction of the screen), rejection is performed according to the Far plane. No rejection is performed according to the Near plane.
Please see "Rejection Processing" below, for additional information.
- Because of reject processing, a large triangle may not be rendered if one of its vertices falls outside the screen, even though it should appear in the screen.
- Reject processing with F3DLX.Rej and F3DLP.Rej can allow faster processing of “2 Triangles” commands. When making DL, it is advantageous to use g*SP2Triangles, if possible.
- The gspF3DLX.Rej.o version provides texture perspective correction, while gspF3DLP.Rej.o version does not. F3DLP.Rej is slightly faster than F3DLX.Rej.
- Both F3DLX.Rej and F3DLP.Rej do not support G_CULL_BOTH.
IGN: Forsaken 64, set for an April 98 release, looks great. Running at 60fps (frames per second) in single-player mode and a smooth 30fps with up to four people playing, it's also fast.
The Nintendo 64 version of the game will feature exclusive new levels and weapons as well as more sophisticated AI (artificial intelligence) than its PC predecessor.
Unlike Resident Evil 2 for Nintendo 64, though, Eternal Darkness runs on a fully 3D engine, enabling complete freedom of movement and, because backgrounds aren't pre-rendered, a totally scalable camera system. What this means for the player is 3D environments that are totally interactive and look beautiful.
The game utilizes a rock-solid 3D engine that outputs pre-rendered-quality visuals -- and by this we mean extraordinarily detailed textures and polygon models -- but with full freedom of movement and interactivity with the environments.
BMBx64 escribió:3) Donkey Kong 64, este ya es un caso conocido, tuvieron que usar la expansión y incluirla con el cartucho porque el juego tenía un bug y se colgaba aleatoriamente (estarían pisando memoria reservada?), es un fallo que incluso puede estar presente en algunos módulos de expansión no oficiales, por lo que se recomienda comprar el oficial.
1985a escribió:@BMBx64
Me gustaria ver un analisis de The Legend of Zelda: Majora's Mask.
Porque a nivel de poligonos se nota tanta diferencia con su antecesor. En especial, la nariz.
AxelStone escribió:Curioso, se supone que las direcciones reservadas deben venir especificadas en la documentación. ¿Documentación incompleta o desistieron de revisar miles de líneas de código?
In a recent Director's Commentary video for Conker's Bad Fur Day (embedded below, at 2:40), Chris Marlow, one of the game's programmers, let slip some behind-the-scenes details that reveal the true reason for the Expansion Pak's inclusion with Donkey Kong 64. According to Chris, there was a glitch in the game that caused it to "randomly crash...in the 4 meg only version," which apparently didn't occur with the Expansion Pak installed. Unable to track down the cause, Rare was forced to include the memory expansion for free "which cost [Rare] a fortune." Though as fellow Rare colleague Chris Seaver pointed out in that same video, it actually worked out in Perfect Dark's favor (released the following year), as that game "definitely needed it" in order to run most of its modes, which helped avoid most customers from having to purchase the accessory separately.