Re: [maemo-developers] SDL vertical blank (vbl) sync.
Hi, > Yes. So to draw without tearing effect either X of framebuffer driver > should report when to draw (vblank period or which line is currently > drawn by hardware) Another possibility is that (SDL) application could ask X server to ask Framebuffer to flush the framebuffer contents to the LCD controller when it's done with a frame. It might even do that directly by calling the fb ioctls (if there would be such ioctl and it would have access rights to do it). AFAIK the framebuffer doesn't have constant update rate to the LCD and it doesn't immediately push every framebuffer update to the LCD controller. It does updates only if the framebuffer contents have been updated (there's an ioctl to tell that) and there's a certain amount of time since last update. So, if application could force the framebuffer flush, it would have time to update the screen before the next (automated or application forced) flushing happens. > or the controller has to have two buffers so single > or multiple SDL_UpdateRects could be pushed to visible screen in one > operation (SDL_Flip) in appropriate moment. Cannot/doesn't SDL do double buffering by itself when asked? (it would be much slower if it's done in the normal RAM where the framebuffer resides, but at least there wouldn't be tearing. :)) - Eero ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Tapani Pälli wrote: I think there's no guarantee that you will get "HW surface" from SDL, it will silently fallback to SW surface if it cannot have HW one. SDL draws to X, X draws to the framebuffer mapped by kernel which then updates that data to the lcd controller where screen gets finally updated. I would recommend to blit things in memory, then push only needed parts with SDL_UpdateRects trying to avoid fullscreen updates. Yes. So to draw without tearing effect either X of framebuffer driver should report when to draw (vblank period or which line is currently drawn by hardware) or the controller has to have two buffers so single or multiple SDL_UpdateRects could be pushed to visible screen in one operation (SDL_Flip) in appropriate moment. Frantisek ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
ext Frantisek Dufka wrote: > Visti Andresen wrote: > >> My "flashingTest" reports that: >> It got a hardware surface (I should be writing directly to "video >> memory" accordingly to the SDL docs) >> And the surface is double buffered. >> >> The size of the surface is 800x480 16 bit (R5G6B5) >> So there should be at least 1536000 bytes of video memory >> >> I suppose the kernel display driver is open-source? > > Yes, it is. And I _did_ look at it. And somehow I don't think SDL > hardware surface really means memory in Epson HWA742 controller :-) > That's why I asked. > I think there's no guarantee that you will get "HW surface" from SDL, it will silently fallback to SW surface if it cannot have HW one. SDL draws to X, X draws to the framebuffer mapped by kernel which then updates that data to the lcd controller where screen gets finally updated. I would recommend to blit things in memory, then push only needed parts with SDL_UpdateRects trying to avoid fullscreen updates. > Frantisek > // Tapani > ___ > maemo-developers mailing list > maemo-developers@maemo.org > https://maemo.org/mailman/listinfo/maemo-developers > ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Visti Andresen wrote: My "flashingTest" reports that: It got a hardware surface (I should be writing directly to "video memory" accordingly to the SDL docs) And the surface is double buffered. The size of the surface is 800x480 16 bit (R5G6B5) So there should be at least 1536000 bytes of video memory I suppose the kernel display driver is open-source? Yes, it is. And I _did_ look at it. And somehow I don't think SDL hardware surface really means memory in Epson HWA742 controller :-) That's why I asked. Frantisek ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
My "flashingTest" reports that: It got a hardware surface (I should be writing directly to "video memory" accordingly to the SDL docs) And the surface is double buffered. Don't trust SDL, SDL is a liar. In X backend the SDL trusts what X says so "HW surface" could mean actually "server sided pixmap". The size of the surface is 800x480 16 bit (R5G6B5) So there should be at least 1536000 bytes of video memory The video hardware isn't like in traditional desktop PCs, this in not quite right. The display hardware is actually a blackbox which contains it's own memory and outer memory space is fully unaware about this (AFAIK). I suppose the kernel display driver is open-source? AFAIK it is, so you could dig up the sources... -- Kuisma ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
On Mon, 29 May 2006 16:50:50 +0200 Frantisek Dufka <[EMAIL PROTECTED]> wrote: > Juha Yrjölä wrote: > > > Unfortunately we don't provide user-space with the vsync information (or > > "Tearing Effect" signal, as it's called in LCD land). > > > > If there's large enough need for this, we might consider implementing > > the support. The tearing effects should be minimal on the 770, however, > > because of the high (around 55 Hz) refresh rate of the display. > > > > Hello, > > well the effect is visible even when playing the Ice Age Trailer (I > suppose it is not already in the movie data). I think such feature is > quite useful for games and movie playback. BTW how big is the > framebuffer memory? Is there space for 2 frames so double-buffering is > possible? Also is there some datasheet for the Epson hwa742 chip? Can > you make it public? I tried to google for it (and for lph8923 too) but > found nothing useful. Thanks. My "flashingTest" reports that: It got a hardware surface (I should be writing directly to "video memory" accordingly to the SDL docs) And the surface is double buffered. The size of the surface is 800x480 16 bit (R5G6B5) So there should be at least 1536000 bytes of video memory I suppose the kernel display driver is open-source? > > Frantisek > ___ > maemo-developers mailing list > maemo-developers@maemo.org > https://maemo.org/mailman/listinfo/maemo-developers ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Juha Yrjölä wrote: Unfortunately we don't provide user-space with the vsync information (or "Tearing Effect" signal, as it's called in LCD land). If there's large enough need for this, we might consider implementing the support. The tearing effects should be minimal on the 770, however, because of the high (around 55 Hz) refresh rate of the display. Hello, well the effect is visible even when playing the Ice Age Trailer (I suppose it is not already in the movie data). I think such feature is quite useful for games and movie playback. BTW how big is the framebuffer memory? Is there space for 2 frames so double-buffering is possible? Also is there some datasheet for the Epson hwa742 chip? Can you make it public? I tried to google for it (and for lph8923 too) but found nothing useful. Thanks. Frantisek ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Kuisma Salonen wrote: Hi I was wondering if anyone knows how to sync. the SDL_Flip to the vertical blanking of the display. While it is nice to get 47+ FPS while memset'ing the display, I would rather be synced to the display and avoid the tearing at the cost of FPS. This is actually generic issue, not just SDL one. When you call SDL_Flip() (or SDL_UpdateRects() etc) it "draws" local data to the display HW, and this goes through X. Having synchronization in the display driver would solve this. Anyway I am not the person to say more about this as I am not enough aware of display HW internals to tell how hard implementing this would be and whether it would even be possible. Unfortunately we don't provide user-space with the vsync information (or "Tearing Effect" signal, as it's called in LCD land). If there's large enough need for this, we might consider implementing the support. The tearing effects should be minimal on the 770, however, because of the high (around 55 Hz) refresh rate of the display. Cheers, Juha ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
On Monday 29 May 2006 08:42, ext Kuisma Salonen wrote: > Visti Andresen wrote: > > Hi I was wondering if anyone knows how to sync. the SDL_Flip to the > > vertical blanking of the display. > > > > While it is nice to get 47+ FPS while memset'ing the display, I would > > rather be synced to the display and avoid the tearing at the cost of FPS. > > This is actually generic issue, not just SDL one. When you call > SDL_Flip() (or SDL_UpdateRects() etc) it "draws" local data to the > display HW, and this goes through X. Having synchronization in the > display driver would solve this. Anyway I am not the person to say more > about this as I am not enough aware of display HW internals to tell how > hard implementing this would be and whether it would even be possible. http://scanline.ca/smooth/ This is a good resource to look at. This guy (http://scanline.ca) is extremely good in computer graphics, you may read his code and if it still doesn't work, mail him. -- Gustavo Sverzut Barbieri INdT, Recife, Brazil Jabber: [EMAIL PROTECTED] MSN: [EMAIL PROTECTED] ICQ#: 17249123 Skype: gsbarbieri Mobile: +55 (81) 9927 0010 Phone: +1 (347) 624 6296; [EMAIL PROTECTED] GPG: 0xB640E1A2 @ wwwkeys.pgp.net ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Visti Andresen wrote: Hi I was wondering if anyone knows how to sync. the SDL_Flip to the vertical blanking of the display. While it is nice to get 47+ FPS while memset'ing the display, I would rather be synced to the display and avoid the tearing at the cost of FPS. This is actually generic issue, not just SDL one. When you call SDL_Flip() (or SDL_UpdateRects() etc) it "draws" local data to the display HW, and this goes through X. Having synchronization in the display driver would solve this. Anyway I am not the person to say more about this as I am not enough aware of display HW internals to tell how hard implementing this would be and whether it would even be possible. -- Kuisma ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
On Sun, 28 May 2006 14:47:29 +0200 [EMAIL PROTECTED] (Heike C. Zimmerer) wrote: > Visti Andresen <[EMAIL PROTECTED]> writes: > > > Hi I was wondering if anyone knows how to sync. the SDL_Flip to the > > vertical blanking of the display. > > You're sure there's anything like vertical blanking for the 770? My > understanding is that VBL is closely tied to the transmission of video > data through an analog line. Sort of, vertical blanking occurs when the electron beam retraces back to the top of the CRT screen. And yes it has a dedicated analog line (on a vga/monitor cable, it is shared with the luminance/colour data on a composite signal). > I doubt you will find anything like that > in digital-only LCD circuitry. There will, however, be some kind of > display refresh which uses some timing of its own. If you look that > direction, maybe there's some interval which can be used for similar > tasks. As you mention internal TFT/LCD displays use different and usually digital (until the LCD/TFT controller chip) signalling. But the controller chip still receives a "top of frame" or "vertical blank" signal (at least in those data sheets I have read for LCD displays). Anyway I would like to sync the SDL_Flip's to the "top of frame" or "vertical blank" signal. The reason for this wish is, that you really can't do anything animated if you refresh the display at a random point in time. Okay you can, but there will be artifacts, these are caused by the top part of the display containing one frame and the bottom one containing the next frame. For those not familiar with the problem I have attached two tiny programs (compiled for arm) flashingTest. Displays some "random" colours, I wrote it to "benchmark" the graphics performance of the 770 and to detect the native screen format (its 16bit R5G6B5 btw.). It do however show the artifacts. vblSync. Was written especially to show the artifacts. The high DPI of the 770 display makes them less noticeable but they are still there. Oh and for those of you who still have the IceAge2 trailer that came with the 770, the Nokia movie player also shows these artifacts quite clearly in the beginning where there are a lot of panning going on. I can just hope that there are a "hacker" out there who knows how get this kind of "scanline" information out of the 770. Otherwise I'll just have to live with sub Amiga 500 graphics capabilities :) > > - Heike > ___ > maemo-developers mailing list > maemo-developers@maemo.org > https://maemo.org/mailman/listinfo/maemo-developers flashingTest.tar.gz Description: Binary data vblSync.tar.gz Description: Binary data ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] SDL vertical blank (vbl) sync.
Visti Andresen <[EMAIL PROTECTED]> writes: > Hi I was wondering if anyone knows how to sync. the SDL_Flip to the > vertical blanking of the display. You're sure there's anything like vertical blanking for the 770? My understanding is that VBL is closely tied to the transmission of video data through an analog line. I doubt you will find anything like that in digital-only LCD circuitry. There will, however, be some kind of display refresh which uses some timing of its own. If you look that direction, maybe there's some interval which can be used for similar tasks. - Heike ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
[maemo-developers] SDL vertical blank (vbl) sync.
Hi I was wondering if anyone knows how to sync. the SDL_Flip to the vertical blanking of the display. While it is nice to get 47+ FPS while memset'ing the display, I would rather be synced to the display and avoid the tearing at the cost of FPS. ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
