ReClock DirectShow filter : try it ! [Archive] - Page 7

temporance

27th January 2003, 17:34

yes for now i made the choice not to distribute all the source code. I give only the GPL portions of the code (to respect the GPL licence). ReClock is freeware, but not open source. Maybe it will change in the future.@ogo: mate, I'm afraid you're not allowed to do this. If you use any GPL code, you have to release all the source code that you integrate with it. Read the GPL. So you should release all of your filter's source code (please :) ).

Sorry to be so confrontational, but it's really not fair to use other people's GPL code and then to keep your derived work secret.

EDIT: This is why Microsoft is $hit-scared that one of its programmers will include some GPL somewhere in Windows. MS would then have to release all the Windows source code under GPL!!!

Jorgosch

27th January 2003, 17:38

Ok, new "problem". The new core player is out (www.corecoded.com) and it has a problem with ReClock. With ReClock adjusting the speed on auto-setting, audio gets a hickup every 6 seconds. With speed set to "original" everything plays ok.
Could you guys work on a solution together, because the Core Player is REALLY fast and smooth.

EDIT: Why do I even bother *sigh* You guys code faster than I can write, the problem was with 99e, fixed in 99f...

Chibi Jasmin

27th January 2003, 18:27

Originally posted by ogo
@Chibi Jasmin: maybe you could do the "MPC + VMR7" test (with an old version of MPC) and if you get the tearing, just check if the tearing stays in sync with vertical sync ?

I tried it, but unfortunately I not only get quite bad playback on monitor with vmr 7 (choppy etc.), but only green junk on dvdmax output...tried this with older and newer version of mpc. It seems, DVDMax doesn't like VMR 7 renderer (neither does it like VMR 9)...so this test doesn't work, I'm afraid...

ogo

27th January 2003, 18:30

@temporance : I've already had a *LONG* discussion by email with someone knowing well the GPL. At first he stated i had to release all the code, but i explained him that :
- the parts in GPL are not at all the heart of my filter (or said diffrently my filter is not an extension of those gpl parts). I you read the GPL licence stuff, you will see that such a use is indeed authorized (i don't remember where but i need to re-read my emails to tell you)
- i can technically completely externalize those gpl parts, and use them as "plugins" (DLL), so i use 0 lines of GPL code i my filter. But i must be too lazy to do that :)

The reasons for which i don't want to give the source code now are quite long to explain, but to be short :
- it's not finished
- i don't want this to be stolen and used by someone to make profit
- people that could be involved in modifying this would be very few (but you are one of them :)), and i don't think that seeing many versions floating around there would give good good results, especially if they are buggy, etc... I'm not saying there that i'm the best out there to do this stuff, just that too many people working on this would not give good results. Remember DVD2AVI ? I released the first "mod" which was 1.82. It was working quite well. I stopped dues to lack of time. After that, there have been many mods starting in all directions (and many of them were buggy), and now this tool is nearly dead because nobody could lead the mods ...

Anyway i'm open to this discussion, maybe i'll change my mind ...

And of course, if i stop working on reclock, there are good chances that i'll give the source. I just started this as a personnal project, but i thought it would be useful to many people because i think it's really useful for people that knows what it does (experienced HTPC users are the more demanding) :)

@Chibi Jasmin: too bad ... Maybe DVDMax is working with an overlay

Chibi Jasmin

27th January 2003, 18:44

Originally posted by ogo
@Chibi Jasmin: too bad ... Maybe DVDMax is working with an overlay

Yes, it is...that's exactly what it does...it outputs the overlay surface fullscreen to the tv :-) I thought, you know that... :D

It actually gets triggered by the creation of an overlay...if there's no overlay present, tv is black, if an overlay window is created, dvdmax outputs its content fullscreen on the tv...

((( atom )))

27th January 2003, 18:53

@jwexqm:

so radeons have comp-sync out?

i have one! :)

do you have any mor info on this one?`i might go and solder a cable, if i'd be assured my card supports al the necessary. i am using a radeon 7000 btw..

i am looking forward to this.

ogo

27th January 2003, 20:21

@jwexqm: me too i'm interested :)

@Gabest: no luck with latest MPC version :
- with "default renderer": according to reclock, VMR7 is used and playback is FINE (no tearing)
- with "VMR7 renderer": tearing :(

Sumster

27th January 2003, 20:34

OGO... I fully support your decision regarding GPL. You have an amazing utility that will revolutionize home theatre PC's, and I would hate to see your hard work stolen and used in all video playback products without your work being recognized or compensated.

If you stay within the rules (and even externalize all the GPL code into .dll's as you mentioned), it would be well worth it!

Amazing work as usual.

temporance

27th January 2003, 21:33

@ogo:
I'd love to respect your decision but what you've done is not legal under the GPL. And putting the GPL parts in a DLL isn't enough: See http://www.gnu.org/copyleft/gpl-faq.html#MereAggregation . You should ask the author of the original code to ask if what you're doing is ok. They might allow you to use the code in ReClock without having to GPL your new code.
i don't think that seeing many versions floating around there would give good good results, especially if they are buggy, etc... The way to avoid this is to use a central CVS repositary. This lets you keep control of the source. Also, you can claim the name "ReClock" so that no-one else is allowed to use it. Look at xvid - if a project of that size can work by GPL then so can ReClock which, i presume, has a much smaller codebase. Also, don't worry, no company would steal your code. They could just take your compiled DLL and you could do nothing (because you use GPL code). Anyway, it would take a big company very little time and money to write their own reclock.

Anyway, keep up the good work, hope to see the source soon! ;)

ogo

28th January 2003, 00:34

I'd love to respect your decision but what you've done is not legal under the GPL. And putting the GPL parts in a DLL isn't enough: See http://www.gnu.org/copyleft/gpl-faq...MereAggregation . You should ask the author of the original code to ask if what you're doing is ok. They might allow you to use the code in ReClock without having to GPL your new code.

Sorry it's not what i understand when i read paragraph 2 of the licence :
"These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works."
reclock have 2 modules under GPL :
- the DRC : reclock doesn't need one to function
- the resampler : i can write a bad resampler in 10 lines of C++, and provide a DLL loader to use another one (a good one) for people that need one. It will solve the problem
But is this useful ? Did i fool someone ? Am i against the GPL spirit ? I don't think so :)
99% of the work is mine and has nothing to do with DRC and resampler, and even the GPL code was quite strongly modified by me :)

I'm too honest. I should never have said there were GPL portions in my code :D

temporance

28th January 2003, 01:04

Originally posted by ogo
"These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works."You are not distributing them as separate works - they are both integrated into ReClock.

Putting the GPL code in a DLL is not allowed either, as even dynamic linking is considered to be incorporating the code into a larger work. The GPL people thought of everything :)

Am i against the GPL spirit ?GPL and open source is all about sharing. I can write 1000 lines of code, someone else takes 100 of then and adds 800 of his own. I can then take back 200 lines into my original program. As soon as someone decides that they only want to take and not give back, then the system stops working.

Anyway, on the scale of GPL violations, this is pretty tiny so I'll try to resist posting on this thread again. Please think about respecting GPL :)

gabest

28th January 2003, 01:23

GPL is evil from a kind of aspect :devil:, same as standards and patents. After a critical time every solution of a problem will become opensource and will belong to someone who has a GPLed code of it, and then other programmers can't prove anymore that their code was made from sratch or using others work, after accusing them too much can even cause them to have bad reputation for the public.

ogo

28th January 2003, 01:46

Putting the GPL code in a DLL is not allowed either, as even dynamic linking is considered to be incorporating the code into a larger work. The GPL people thought of everything
I was more thinking about dynamic loading :)

GPL and open source is all about sharing. I can write 1000 lines of code, someone else takes 100 of then and adds 800 of his own. I can then take back 200 lines into my original program. As soon as someone decides that they only want to take and not give back, then the system stops working.
I totally agree with that (please have a look at my previous submissions to doom9). That's a way of sharing, but not the one i choosed for reclock for reasons given before :)

I give my work for free anyway. You cannot imagine what amount of time i spent to write this thing, respond emails. And it's very very tiny compared to people who wrote ZoomPlayer or Media Player Classic. The whole idea of my work has nothing to do with a resampler or a drc (no aggregation, extension or whatever you want), it's the application of a new idea never implemented before (to my knowledge). Think about this when you think GPL. Technically one could still argue that i somehow violate the licence, but i think i respect the spirit of GPL anyway by telling i use GPL source code and releasing modified versions of it :)

I'm sure DOOM9 would think like you anyway :) I remember some discussions in the past about this ...

temporance

28th January 2003, 11:46

Originally posted by gabest
GPL is evil from a kind of aspect :devil:, same as standards and patents. After a critical time every solution of a problem will become opensource and will belong to someone who has a GPLed code of it, and then other programmers can't prove anymore that their code was made from sratch or using others work, after accusing them too much can even cause them to have bad reputation for the public. No, this isn't a danger. GPL uses the copyright of the source code and protects the written lines of code only. It doesn't protect the idea within the code - anyone can reuse this in any way they like. So, for example, ogo could quite legally rewrite his own resampler that does exactly the same job as the GPL code.

gabest

28th January 2003, 12:56

Yes, he could. But after already seeing how it is done, would he do it differently? Or for example could anybody release a new divx/xvid/... decoder AND prove (!) that he hasn't used a single line from ffdshow or mplayer? If so, how?

oddball

29th January 2003, 13:42

BIGASS BUG REPORT

When you play an avi/ogm with dual audio it loads Reclock twice and plays back both audio streams at the same time. In Zoomplayer you also are unable to select the audio track you wish to take priority while Reclock is doing it (Greyed out option). You have to disable Reclock entirely by unregistering it to be able to play such files correctly.

((( atom )))

29th January 2003, 20:49

i have to actually confirm oddballs report.

this also happens with wmp but not with the core player, where i can still switch streams (built in feature like in zoom player).

it would be very nice have this issue fixed.

ogo

29th January 2003, 23:32

Yep i saw this problem with BIVX :( Didn't looked at it yet. Another one on the todo list ... Gabest, how are handled files with multiple tracks in MPC ? Why are 2 reclocks created in the graph ?

For now i'm worried with somes problems reported by some users :
- sound getting out of sync, or slowly going high pitched ... Do any of you have this problem ? If only i could have a sure way to reproduce this i could understand and solve that
- graphedit load/save still not fixed

gabest

30th January 2003, 08:22

Originally posted by ogo
Yep i saw this problem with BIVX :( Didn't looked at it yet. Another one on the todo list ... Gabest, how are handled files with multiple tracks in MPC ? Why are 2 reclocks created in the graph ?This is normal, every audio path is rendered if the audio switcher is not enabled or no morgan stream switcher is installed.

((( atom )))

30th January 2003, 08:32

This is normal, every audio path is rendered if the audio switcher is not enabled or no morgan stream switcher is installed. the core player is still able to switch, after installation and activation of reclock.

morgan stream switcher caused nothing but trouble for me, whatever i tried. on a former installation it slowed down the speed of the audio, while the video ran on and on my actual installation it completely garbles mp3-tracks while leaving ac3 intact..

can anyone recommend any other switcher?

gabest

30th January 2003, 08:46

Originally posted by ((( atom )))
the core player is still able to switch, after installation and activation of reclock.By rebuilding everything from the avi splitter, one path each time. It is also a way, but I think with the audio switcher it is better for two reasons: 1. switching is faster, 2. I could put the 44.1khz resampler, channel mapper and audio time offset features in the filter.can anyone recommend any other switcher? options/audio/enable audio switcher, and this should be on by default :P

Sumster

30th January 2003, 15:48

Ogo,

I haven't had a problem with audio/video slowly getting out of sync, or with audio becoming high pitched. I have watched several movies straight through. Perhaps it is because I am using audio-slave mode...

Haven't yet tried AVI's with multiple audio tracks...

Everything still working fine here. Are you planning dynamic reconnect in the next version?

-Sumster

ogo

30th January 2003, 20:18

Thanks. Yes it's fine with the audio switcher. Any news with the VMR7 tearing problem Gabest ?
Also i would like to ask you some advice about DirectShow. In a future version of ReClock i would like to find a way to be notified by the video renderer each time a frame is about to be drawn. I found no way to do that with the provided API in DX9. Could it be possible to insert a sort of null filter in front of the video renderer that would notify me each time it receive/sends a frame ? If so, how can i implement such a filter (i don't know how to do that) ?
My need is to monitor at what "distance" of the vertical sync each frame is drawn by the video renderer. It would allow me to validate that the clock is properly corrected, and also would give me the possibility to draw each frame just after vsync by tweaking the clock.
Thanks for your help :)

gabest

30th January 2003, 20:34

I don't think you could access any directdraw functions of the video renderer and putting a filter before the renderer is a very bad idea, you couldn't find out anything from the rendering time anyway.

But since there is only one display (usually :)) I think creating your own directdraw object would give you the same values for the scanline.

Here is my code to wait for the vsync (no idea why it wasn't working for you). m_pDD is a pointer to IDirectDraw7 and rDstPri is the target rectangle for the video (the inner client area of mpc's window).
DWORD dwScanLine;
for(;;)
{
hr = m_pDD->GetScanLine(&dwScanLine);
if(FAILED(hr) || hr == DDERR_VERTICALBLANKINPROGRESS)
break;

if((LONG)dwScanLine < rDstPri.top || (LONG)dwScanLine > rDstPri.bottom)
break;
}

// blt video pic on primary
Ah, and I have no idea how to do it with dx9, it has one singe call (Present) to show the backbuffer, no blting needed. I think MS wanted to make sure vsync can't be controlled from the app, but from outside. (Or I am totally wrong now :))

edit: rereading your post, I have a feeling that you want to know when a picture is blted and not just the scanline's position... Well, this looks impossible to me.

ogo

31st January 2003, 01:11

edit: rereading your post, I have a feeling that you want to know when a picture is blted and not just the scanline's position... Well, this looks impossible to me.
Yes that's more what i need. Knowing exactly when is drawn each frame by the video renderer. I know very well the scanline stuff, but it's useless in this case :)

I don't think you could access any directdraw functions of the video renderer and putting a filter before the renderer is a very bad idea, you couldn't find out anything from the rendering time anyway.
Why do u think it is a bad idea ? it would give me a close estimation of when each frame will draw ... But it's difficult (renderers can have multiple input pins, etc...). I would like to make the test, but i'm not sure i can force the graph to always put my "null" filter in front of the video renderer. Anyway i can make a test by manually inserting a filter in graphedit !
Another very very dirty way would involve some real hacking. If the MS video renderers are based on some CBaseRenderer classes of the SDK (i'm not sure they really are), i could manage to extract some information from them knowing their address ...

ogo

31st January 2003, 01:18

BTW, scanlines do not exactly match Y positions on screen. You can not really compare a position of a window with a scanline.

I'm not sure you saw my response about tearing with MPC with VMR7, but it's not fixed by your last MPC version.

oddball

31st January 2003, 04:02

I have two audio streams enabled when Reclock is not used on divx/xvid dual audio files. When Reclock is used the ability to switch to one or other tracks becomes disabled (greyed out). The 2 instances of Reclock show up but there does not appear to be a way of turning either off so both audio streams are played at the same time (Whereas when Reclock is not loaded as a filter you can select either audio stream).

This is in Zoomplayer 3 b5.5 BTW. I also tried WMP and BSplayer with similar results.

gabest

31st January 2003, 13:10

Why do u think it is a bad idea ?Format changes. They can arrive from both directions, from the decoder or the renderer, not easy to handle. Or there are mpeg2 decoders which are better to be connected directly to the renderer to allow hw acceleration.
BTW, scanlines do not exactly match Y positions on screen. You can not really compare a position of a window with a scanline.This is how vsync was implemented in dx7 sdk, any better idea?

Chibi Jasmin

3rd February 2003, 20:51

Originally posted by ogo
@Chibi Jasmin: maybe you could do the "MPC + VMR7" test (with an old version of MPC) and if you get the tearing, just check if the tearing stays in sync with vertical sync ?

I reported that DVDMax didn't seem to like VMR7 renderer...by accident I today discovered it is something else...if I set MPC to use VMR7 my tv output is screwed...if I set MPC to use System default it actually uses VMR7 (also reported by ReClock) and TV picture is fine. There must be something screwed with forcing VMR7 in MPC.

I will do the VSync tearing test, as soon as I find the time...please let me know any version number of MPC that has this tearing problem with VMR7 for sure!

jwexqm

4th February 2003, 02:41

ogo: I've collected a ReClock log generated at 25Hz progressive here (http://www.jwexqm.vispa.com/reclock_log_25hz_non-interlaced.html).

I have a friend who owns a Dell Inspiron laptop that is fitted with an ATI Mobility Radeon 7500c. I asked him if he'd mind installing PowerStrip to discover if his chipset supports composite sync and interlaced modes and it turns out that it does. I suspect yours will support it too.

This makes sense really as I am aware of a few people on AVSforum who are using composite sync on desktop versions of the 7500. I can vouch that my old 8500 and my new 9700 both offer these features too.

I have put up a basic page which details the VGA to RGB SCART direct drive cable here (http://www.jwexqm.vispa.com/vga_rgbscart.html).

Please do be careful though as I'm sure that 75Hz progressive VGA straight into a TV's SCART socket could be damaging!

I really do think it would be fantastic though if ReClock could support interlaced display modes, particularly as this SCART arrangement looks promising.

James

((( atom )))

4th February 2003, 03:28

james -nice one, thx a lot!

also you get my personal 5-star to be the first guy who thinks of writing [solder side] to an image where it belongs.

* * * * *

[EDIT] i would actually suggest to open a new thread on this one, since i can image there might be quite some people interrested in this goodie and they might not even find it hidden in this thread here..

ogo

4th February 2003, 11:46

jwexqm> Very interesting stuff. My laptop use a M6 mobility chipset.
M6 drivers are never up to date, but i can use the latest catalyst drivers by tweaking the INI files of the setup and telling them my card is a radeon 7500. So my chipset must be very close to a Radeon 7500.
I now understand why 50hz interlaced is useful :)
I will make a cable at the end of the week and test that setup with my laptop and an old TV set. If all goes well i will be able to see if i can modify reclock to detect and use interlaced mode. Looking at your logfile, i already see it detects horizontal frequency correctly. That's promising
My only concern with this cable is impedance. Is the VGA connector designed to be loaded at 75ohms ?

BlackSun

4th February 2003, 12:19

Originally posted by gabest
By rebuilding everything from the avi splitter, one path each time. It is also a way, but I think with the audio switcher it is better for two reasons: 1. switching is faster, 2. I could put the 44.1khz resampler, channel mapper and audio time offset features in the filter.options/audio/enable audio switcher, and this should be on by default :P

Yeah I'm reconnecting the Pins manually from the AVI splitter, and re-rendering the pin. It's fast but actually work only with audio streams having the same decoder.

Btw there is no buffer in Reclock ? When I launch Mozilla for example the renderer stop for a few seconds while mozilla is loading.

Chibi Jasmin

4th February 2003, 18:24

@ogo: Now I got MPC to render with VMR7 I still fail to reproduce the tearing bug...which mpc version has the bug for sure?

gabest

4th February 2003, 18:37

All except this last posted here. But I'm not sure that fix worked since ogg said he could still see it.

Chibi Jasmin

4th February 2003, 19:40

Originally posted by gabest
All except this last posted here. But I'm not sure that fix worked since ogg said he could still see it.

Strange, I failed to reproduce the bug...well, I'll try again later...btw: please have a look into the option forcing vmr7 renderer...it gives green garbage on dvdmax output, while system default also uses vmr 7 and works on tv...

BTW: How can I set, which renderer (audio and/or video) is my system default?

gabest

4th February 2003, 19:58

When vmr7 is used through "system default" then it runs in overlay (vmr9 won't run in overlay anyhow), but that mode can't be used for subtitling and so full screen overlay mirroring on tv-out won't work either.

Chibi Jasmin

5th February 2003, 11:14

Originally posted by gabest
When vmr7 is used through "system default" then it runs in overlay (vmr9 won't run in overlay anyhow), but that mode can't be used for subtitling and so full screen overlay mirroring on tv-out won't work either.

I see...that's why it works with DVDMax output then...does the tearing bug also occur, when vmr7 is run through system default using overlay?

ogo

5th February 2003, 11:57

no tearing with VMR7 through "default renderer" on my system (using a geforce 3)

Chibi Jasmin

6th February 2003, 10:49

Originally posted by ogo
no tearing with VMR7 through "default renderer" on my system (using a geforce 3)

So I guess, I just can't use this as a test, if TV clock on Matrox is totally correct or not...any other ideas?

BigT

6th February 2003, 22:19

@BigT : why is it the case ? I don't understand why putting a display in primary mode would give bad quality. Matrox is doing bad work here

Sorry, i have not had a look at this thread forsome time i fear :(
My guess is that DVDMAX simply uses a lower custom resolution. Another poster (jwexqm) suggested to use powerstrip in order to enable such modes without the need for DVDMAX. Personnally i prefer to keep a 1024*768 windows desktop on my TV and switch to DVDMAX when needed (with video). I have just coded a very simple application that allows to enable/disable DVDMAX manually today by the way rather than Matrox's "automatic DVDMAX" when overlay is detected (in case someone else is interested). It requires Borland C++ runtime libraries though.

((( atom )))

7th February 2003, 01:48

i have a strange issue here: reclock doesn't apply to avis anymore. when playing mpeg-files, it loads, though..

i tried reinstalling/reregistring it, but that doesn't help it.

has anybody had this happen before?

Chibi Jasmin

7th February 2003, 11:38

Originally posted by BigT
Sorry, i have not had a look at this thread forsome time i fear :(
My guess is that DVDMAX simply uses a lower custom resolution. Another poster (jwexqm) suggested to use powerstrip in order to enable such modes without the need for DVDMAX. Personnally i prefer to keep a 1024*768 windows desktop on my TV and switch to DVDMAX when needed (with video). I have just coded a very simple application that allows to enable/disable DVDMAX manually today by the way rather than Matrox's "automatic DVDMAX" when overlay is detected (in case someone else is interested). It requires Borland C++ runtime libraries though.

DVDMax uses 720x576/480 (PAL/NTSC) resolution...

What exactly does your little app do? You can enable/disable DVDMax in the Display Properties...I don't need an app for that...or did I get something wrong...

Maybe it can display something else than overlay content on the TV?

BigT

7th February 2003, 13:11

I have the HTPC connected to a TV only without monitor which is not what you do if i understood correctly. In your case DVDMAX is probably not nearly as annoying as here.
In clear: In order to use DVDMAX one has to check it in the Dualhead properties. The only problem is that it activates itself automatically when overlay is detected. Let's take the example of a TVTuner application for example. You start it and Matrox's software switches to DVDMAX automatically preventing you to switch channels until you switch back to your desktop (as i said i have no monitor; just TV).

What i did was getting rid of the automatic behaviour. Basically you enable DVDMAX "when you want" or disable it respectively. Since you have a monitor connected to the G400 you probably have no use for this (for other people like me the original behaviour of DVDMAX is disturbing as i said). The way this works is not exactly rocket science either. Source code is not even 2kB.

Chibi Jasmin

7th February 2003, 14:43

Originally posted by BigT
I have the HTPC connected to a TV only without monitor which is not what you do if i understood correctly. In your case DVDMAX is probably not nearly as annoying as here.
In clear: In order to use DVDMAX one has to check it in the Dualhead properties. The only problem is that it activates itself automatically when overlay is detected. Let's take the example of a TVTuner application for example. You start it and Matrox's software switches to DVDMAX automatically preventing you to switch channels until you switch back to your desktop (as i said i have no monitor; just TV).

What i did was getting rid of the automatic behaviour. Basically you enable DVDMAX "when you want" or disable it respectively. Since you have a monitor connected to the G400 you probably have no use for this (for other people like me the original behaviour of DVDMAX is disturbing as i said). The way this works is not exactly rocket science either. Source code is not even 2kB.

Ah, I see...so I don't need your app then, but thanx for the explanation...

Chibi Jasmin

8th February 2003, 17:50

Originally posted by oddball
BIGASS BUG REPORT

When you play an avi/ogm with dual audio it loads Reclock twice and plays back both audio streams at the same time. In Zoomplayer you also are unable to select the audio track you wish to take priority while Reclock is doing it (Greyed out option). You have to disable Reclock entirely by unregistering it to be able to play such files correctly.

I encountered the same problem here...didn't play a dual audio avi for a long time, I guess... :)

ZoomPlayer can't switch and The Core Media Player and MPC seems to crahs sometimes on these files, but can switch...

Grimmi Meloni

10th February 2003, 00:02

Hi Ogo,

I´m having trouble using reclock. The sound does not work. First I thought it´s because I´m nailed down on waveout (my SPDIF does not work via Direct Sound), but I saw somewhere in the changelog that it´s ought to work. Reclock Properties show correct values during playback, but it´s all just silence I get. I´ve checked the Log, and found something.

1.73s 0000bc CreateAudioBuffer()
1.73s 0000bc Getting reg key WaveDevice = 1
1.74s 0000bc waveOutOpen result=32 tries=1 Dieses Format wird nicht unterstützt oder kann nicht übersetzt werden. Benutzen Sie die Funktion 'Capabilities', um die unterstützten Formate zu ermitteln.
1.74s 0000bc waveOutOpen result=32 tries=2 Dieses Format wird nicht unterstützt oder kann nicht übersetzt werden. Benutzen Sie die Funktion 'Capabilities', um die unterstützten Formate zu ermitteln.
1.75s 0000bc waveOutOpen result=32 tries=3 Dieses Format wird nicht unterstützt oder kann nicht übersetzt werden. Benutzen Sie die Funktion 'Capabilities', um die unterstützten Formate zu ermitteln.
1.75s 0000bc waveOutOpen result=32 tries=4 Dieses Format wird nicht unterstützt oder kann nicht übersetzt werden. Benutzen Sie die Funktion 'Capabilities', um die unterstützten Formate zu ermitteln.

The german message means: "This format is not supported or cannot be translated. Use function "Capabilities" to determine supported formats."

Any idea what´s this all about?

Thanks for your support.

downloada

11th February 2003, 17:06

hi,

i really like reclock, thanks ogo for creating it!

but i have a problem with multi-audio-files, too. the sound of both audio-streams is played back at the same time. oh, and i do have morgan stream switcher, but it isn't in the graph anymore when i enable reclock. the graph with reclock looks completely different (viewed with graphedit).

reclock disabled: audio goes from avi stream01 and stream02 to two mp3-decoders (haven't tried with dual-ac3 yet) then to morgan and finally to the audio renderer.

but with reclock enabled: audio goes to the mp3-decoders and gets to reclock directly without passing through morgan, so reclock is in the graph twice.

i think you could implement an easy, quick'n'dirty workaround: just let the user mute the audio in the filter-properties of the decoding-filters. this isn't too comfortable for the user, but it'll surely work (at least for mp3, don't know if it would work for ac3-passthrough) until you found a better solution.

btw: why is morgan being excluded from the graph if reclock is enabled? anyone got information on that?

edit: i just tried to insert morgan into the graph with reclock manually, but it doesn't work. zoomplayer says: "unable to create filters for: ... this may be due to a corrupt file or missing decode filters."

cu