Jump to content

EricBall

Members
  • Content Count

    2,361
  • Joined

  • Last visited

Posts posted by EricBall


  1. The 7800 has a 13 color palette from 256 shades. As stated, the palette registers can be updated mid-screen as well. (Note: the registers are not buffered, so it's recommended to update the registers after STA WSYNC to avoid changing them in the middle of the row.) How sprite data relates to which palette registers are used depends on the current read & write modes.

     

    You probably could do some kind of intelligent flicker routine to handle cases with either a large number of sprites (especially with a tile background). Just keep track of which sprites haven't been displayed, and add them to the display list before the ones which haven't been displayed.

    • Like 1

  2. Ars Technica has provided an overview of the various Steam Boxen shown at CES. While some of the boxes are small, the price tags certainly aren't. Ars also has a review of the Steam controller - it seems mapping the analog control pads to simple keypresses is less than perfect in practice.

     


  3. From Forbes.com (via digg.com): http://www.forbes.com/sites/insertcoin/2013/11/30/who-will-buy-a-500-steam-machine/

     

    In summary:

    1. A PC gamer doesn't need to spend $500 to play games on the TV, just a graphics card with HDMI output.
    2. And who says a PC gamer wants to play using a controller (no matter how leet) instead of mouse & keyboard - and only play games available for Linux on Steam.
    3. While $500 might be "cheap" for a PC*, it's the same price as a PS4 or XB1. Yet it doesn't have the AAA games available for those consoles.

     

     

    * I'd even argue it's not much cheaper than a custom built PC using similar components.

     


  4. http://www.theverge.com/2013/11/25/5146398/ibuypower-steam-machine-499-radeon-r9-270

     

    $500 for a third party SteamBox (includes controller). Dunno how it compares to the Intel+Nvidia Valve prototypes (or Ben Heck's "Steambox" http://benheck.com/07-11-2013/ben-heck-steam-box-concept-with-amd-apu) but I'm betting on the low end (the R9 270 is slightly better than the GTX 660).

     

    However, since it's basically a PC I wonder whether it's significantly cheaper than what you could build yourself (or buy from Dell...).


  5. It somewhat depends upon how you are measuring. A TV engineer measures horizontal timing from the start of the hsync pulse (Maria II Screen Layout), but from a programmatic perspective you want to know how many cycles from WSYNC to first pixel (7800 timings). M2 also gives some additional details on what is going on during that HBLANK period. MARIA takes care of all of the horizontal (and vertical, unlike the 2600) sync timings.


    In the end, the NTSC 7800 has a 7,159,241Hz (2x colorburst) pixel clock for 320 active pixels per line and 136 pixels of HBLANK. (160 modes are handled by displaying each pixel for 2 pixel clocks.) Note, this means a 7800 line is one pixel too long versus the NTSC spec (so the CPU clock is sync'd to the line). Most TVs should handle this, but it would have caused major problems with the "Laserdisc adapter".


    For the vertical, I'd use the 7800 timings counts. Adding an extra line to your DLL (243 vs 242) will prevent MARIA from running off the end. The 192 "active" lines is more a recommendation than anything enforced by hardware. Modern TVs are able to show 240 lines, but often will stretch the picture vertically to mask off some of the top & bottom anyway. Again, from a programmatic perspective, you can either use the CTRL register to detect VBLANK, or a DLI in your last DLL active entry to set a flag which your main application watches.
    • Like 1

  6. Hmm.. it seems Valve has made three different systems - high, medium and low. (Although even the low system is a decent PC - they aren't testing integrated graphics or Pentiums.) So part of this trial might be to determine how much difference Steam OS makes versus Windows & Linux.


  7. Kinda cool, but I wonder what kind of pricetag it's going to have on it. Also, will it come bundled with Steam Machines or will it be competing with other control options like keyboard & mouse and the Xbox 360 controller?

     

    The "Portal 2 Bindings" graphic seems to suggest the Steam Controller is emulating a keyboard + mouse rather than being a pure USB HID. While that might be great for games which don't support game controllers, it adds another layer of complexity - configure Steam Controller to K+M mapping, then configure the K+M to action mapping in the game. I wonder if it will be possible to map both touch pads to directional buttons.

     

    Hmm.. how will auto-centering work? With a normal joystick (other than the 5200 & CoCo) the springs return to center when I release. Will the touchpad read center when I'm not touching? Will I be able to rest my thumb on the edge of the touchpad and not send a buttonpress like I can with an actual button?

     

    Of all of the announcements, this is the most innovative. But only time will tell whether the Steam OS+Machine+Controller becomes a worthy competitor or a historical footnote.


  8. Well that is a whole lot of nothing. Apply to be one of 300 beta testers of a prototype Steam Box. But no actual specs, just that multiple partners will be offering Steam Machines.

     

    What do you bet that O+O is something regarding the already announced streaming or family options? Or maybe it's a Chromecast style adapter for streaming. I guess we find out on Friday.


  9. Those requirements are unique to Half-Life and it is likely due to the OpenGL API and multi-threaded code used for the Linux port. I think looking at the requirements for a modern game would put the platforms on more equal footing. In the case of Dota2 the requirements are the same for all three supported platforms.... OUYA is not a very powerful system. If Valve puts out a high performance system and gets AAA titles running on it natively it could do very well.

     

    The system requirements of Dota2 are similar to HL on Linux, so are probably indicative of what a Steam Box would require. Thus it's probably competing with the XboxOne, PS4 and PC instead of the $100 micro consoles. This means it won't be an impulse buy and will need a wide selection of titles, published and promised, and in particular exclusive titles to effectively compete with the incumbents.

     

     


  10. SteamOS is to game console/HTPC hardware what Android is to mobile hardware.

     

    While this may be what Gabe is trying to do, I just see the downsides to that plan.

     

    First, one of the major advantages of traditional consoles has always been uniformity: every one of the 78 million Xbox 360s are identical. So from a development perspective, unless the game requires a specific controller, you have 78 million potential customers and every one of them will have an identical experience. But for Android, Windows, Linux, and (to a lesser extent) IOS and OSX, this isn't the case. So does the developer go with lower requirements to increase the potential market or do the Crysis and only support the bleeding edge? Yes, suites like Unity help to smooth out the differences, but there's still a big advantage.

     

    Second, the current system requirements for Steam games on Linux is quite high which will push the system cost up. On Linux, Half Life requires a dual core 2.8 GHz CPU and a decent graphics card (while on Windows it will make due with a 500MHz processor and integrated graphics). A Steam Box won't be able to compete with the $100 OUYA and Vita TV. Plus, other companies won't be able to price a Steam Box as a loss-leader.

     

    This is assuming Valve hasn't found a way to run Steam games on the Raspberry Pi...


  11. I'm hoping the next two announcements answer some of the basic questions I have.

     

    As a potential customer I need to know the following:

    1. What (exclusive) games will it play?

    2. How much will the hardware cost?

    3. How much will the games cost?

     

    Quite frankly there's a serious glut in the future for game playing hardware and I don't have the same glut of time or money to invest in all of them. So if Gabe wants me to "buy into" his vision of a Steam OS future, he's got to do better than all of the other choices - including none of the above where I sit back and buy games for the hardware I do have.

     

    Note: I have a Steam account, from when I bought Half Life 1: Anthology, but I don't think I've bought anything using Steam. (Although I did link my PS3 Orange Box purchase.)

     

    Addendum: I read the SteamOS announcement and I think, "big whoop". So it's based on Linux - so's Android and my Kobo. It can stream from my PC/Mac to the TV - or I can plug my PC/Mac into my TV and not have any compression or lag. It handles online media services - so does my PS3, WiiU, and even my Smart TV.


  12. Thinking more about the Vita TV I think the biggest problem is going to be if Sony doesn't scale the PSP/Vita games to the HDTV panel resolution (720p, 1080i and 1080p). If it tops out at 1080i then a 1080p display is going to look worse than a native 720p display.

     

    And while Remote View may be cool for some, I don't know whether the PS4 is going to be sending a full 1080p signal or a quarter (960x540) which matches the Vita's resolution.

     

    There's also the naming. While Vita TV is good because it reflects the ability to play (some) Vita games, it may have been better to give it a different brand, like Playstation Mini, so games could be clearly marked as "Vita & PS Mini compatible".


  13. I agree that Nintendo seems to be the only company able to make a serious dent, but to be fair Sony is doing something here that Nintendo should have done first. The Wii u is more than capable of running 3ds games, and since Nintendo doesn't see 3d as terribly crucial to the games' success, it's not like it would hurt to play them in 2d at home.

     

    In fact Nintendo could honestly one up Sony in this case if they allowed the system to play downloadable games on the Wii u without any "3ds" attachment. Or heck, a 20 dollar 3ds cart reader that plugs in via USB.

     

    Hmm... use the Wii U GamePad for the bottom touch screen & controls and have the Wii U show the top screen on the HDTV. It would work for some games, but others wouldn't. Fer instance Dream Team has you turn the 3DS sideways for some of the boss fights. Other games use both screens to show double-height graphics.


  14. The Vita TV is a strange duck - almost like an escaped skunkworks project. Some Sony engineers took a Vita and stripped out the controller and display and hooked up a DS3 and a TV thenhad a blast playing PSP games on the bigscreen.

     

    It's biggest problem is compatibility with Vita games which use the touchscreen etc. features which don't exist with the DS3 controller. The DS4 controller might provide some additional compatibility, but requiring one might require waiting for the PS4 release - which has been pushed back to to February in Japan.

     

    So maybe the Vita TV is Sony's Japanese Christmas present instead of the PS4.

     

    IMHO it's biggest advantage is price. $100 (really 10K yen) makes it an impulse buy for many. (Although that price doesn't include the DS3 or memory cards.) However, I suspect a lot of reviews will be "The Vita TV is <negative aspect>, but it's only $100". However, it's possible Sony will listen to these reviews and release an improved version to address some of the complaints.

     

    And because others have, I'll mention the OUYA - another $100 console. The big advantage the Vita TV has over the OUYA is the existing software library. But the OUYA is more "open" so will probably appeal more to Indie developers and those already developing for the Android platform. The OUYA will also likely be more favorable to micropayment (aka free-to-play) and advertising supported games. But the Vita TV will appeal more to developers and content providers who want a locked-down platform.


  15. He would be using STA WSYNC within the DLI. I don't think it would be necessary but didn't know if there would be some slight jitter to when the vbi routine starts each time.

     

    Outside the DLI I know it could be problematic. IIRC, on the Atari 400/800s the keyclick routine used it which could cause DLI issues.

     

    Sorry, I meant using STA WSYNC outside the DLI. You can use STA WSYNC inside the DLI to sync the DLI to the next line (required to change the palette), but if you're just releasing the mainline then it isn't necessary.


  16. Any particular reason you're using 5 byte entries instead of 4? That would save you 9 cycles per zone (27 per sprite).

     

    I'd also recommend reviewing all of the code & calculations you're doing before the "Now Build the DL" code. Are you actually saving cycles/RAM/code by pre-calculating those values, or could they be done in the zone code? For instance ZONTEMP0/ZONTEMP1/ZONTEMP2 aren't required and can be calculated from VZLIST,X as required.

     

    Note: only try to eliminate CLC/SEC after you've finished optimizing everything. And always ask yourself "could C be different than I expect?" for all opcodes which change C.

     

    In http://atariage.com/forums/topic/104546-sample-7800-source-code-using-zpx/ I found a slight improvement using (ZP,X) instead of (ZP),Y to build the display lists. Although the store+increment is 12 cycles instead of 7 cycles, there's a savings on the start & end.

×
×
  • Create New...