ReClock DirectShow filter : try it !

[ogo]

ffdshow gives me purple colors with clip2. clip1 seemed ok

no problems with divx 5.0.3 pro + reclock with both clips in whatever player i tried (MPC, ZP, WMP) ...

hope this helps !

[ogo]

Hi,

The next version of ReClock (0.99h, and still some room until Z ) is available on my website.

It has some interesting new features ...

@kallekill : and it corrects your problem

OGO

[kallekill]

I'm sorry to say it but I still got desyncted refreshrate, even if I enable the VSYNC monitor and run a clip that Reclock can sync the refreshrate to. I have a short clip that falls out of sync every time at a certain point. Ogo, I can send the clip to you so that you see what I mean. By the way, Media Player 6.4 doesn't like the VSYNC monitor at all. It crashes when opening a movie.

[ogo]

That's why i said it's experimental

Yes please send me your clip !

[Chibi Jasmin]

Originally posted by ogo
Hi,

The next version of ReClock (0.99h, and still some room until Z ) is available on my website.

It has some interesting new features ...

@kallekill : and it corrects your problem

OGO

Hi there!

Great work with the VSync monitor...it works fine here with my G400 DH card, no vsync problems! [See UPDATE below, might have a few still]

But what I still don't get is how I can use that to determine, if my video playback on the TV is jerky (well, kind of micro-jerky, to the naked eye it's fine)...in my video window on the monitor the vsync dots stay close together as they should and the bar doesn't move too much [UPDATE not always, sometimes I have the bar moving around a bit and then go steady again for some time]...so fine so good...this image is sent to the TV and of course the bar and dots is exactly the same on the TV, but it's a copy of the monitor image...couldn't it still be, that the TV VSync is another than the monitor refresh rate? Or does that thing in the readme apply, that the TV adjust itself automatically to the VSync in a certain limited range...and apart from that, we still don't know, if the G400 internally syncs it's TV-clock to the (perfectly adjusted (thanx to Reclock)) monitor-clock...or did I miss something? I can tell you now, that ReClock syncs fine on my monitor...can I conclude somehow it also syncs fine on the TV on my card? If yes, how?

[Dr_Colossus]

I have a clip that doesn't play very well at all with the new version but plays fine with the old version, I tried remuxing it with avimux but I still get pauses. Here's a link to the clip. This happens in any mode I choose.

[ogo]

@Dr_Colossus : you're right there is something wrong with this clip. I'll have a look into that and let you know. Thanks for your report. Please keep the file available a few days.

@Chibi Jasmin : i'm nearly sure the tv clock is the same as the clock used by the ramdac. To answer this question, i should add something like a "tearing test" into vsync monitor : in this test, i could display a fast continuous moving vertical bar across the video. If a different clock is used to display picture on tv, then you should see some tearing (the bar would be broken somewhere) because i believe parts of 2 frames would be displayed on the same time, or jerky movement of the bar.

@kallekill : I've not been able to crash WMP 6.4 with the vsync monitor. What OS/video card do you use ?

Anyway, as blight told me, new beta 6 of ZoomPlayer have an audio stream switcher that should work with reclock

[cult]

Ogo may I make a suggestion?Could you please place your site in your signature?

[Ewi]

But although the Vsync Bar is Jumping around with VMR9, the playback is much smoother for me under VMR 9 compared to VMR7. I tested it with Sumsters Godfather 3 clip (the one he posted a few weeks ago). Even if the VsyncBar is between the two limits, the height is rather small and the location of the bar is more or less stable, VMR7 is clearly disadvantaged.

MPC 6.4.3.1
WinXP
AthlonXP
Geforce2Ti

[kallekill]

I have a Geforce 4 Ti 4200 card running on Windows 2000 SP3.

Ogo, if you go online on icq I can send you the clip that always makes the picture fall out of sync at one point.

[kallekill]

MPEG1 files play just fine, but Divx-movies hangs Media Player 6.4 when enabling VSYNC MONITOR.

MPEG1 files makes MP 6.4 hang too. Sometimes it just takes a few seconds.

[gabest]

ogo: There is a new build of MPC posted in its topic (v6435). I changed the vsync code of vmr7 and the rm renderer, and now the jitter tests that karl lillevood has posted to somewhere are running flawlessly.

[mazzo]

1. ReClock doesn't work with my laptop, because the LCD seems to have a refresh rate fixed to 60 Hz.

2. When using a TV as monitor, the refresh rate is lowered to 50 Hz, which should be okay, but when I enable ReClock, the video will run like h..l, almost in double speed.

Do you have any suggestions?

[Chibi Jasmin]

Originally posted by ogo
@Chibi Jasmin : i'm nearly sure the tv clock is the same as the clock used by the ramdac. To answer this question, i should add something like a "tearing test" into vsync monitor : in this test, i could display a fast continuous moving vertical bar across the video. If a different clock is used to display picture on tv, then you should see some tearing (the bar would be broken somewhere) because i believe parts of 2 frames would be displayed on the same time, or jerky movement of the bar.

Adding this tearing thing would indeed be great, will you do it? PLEASE!

[Chibi Jasmin]

Originally posted by gabest
ogo: There is a new build of MPC posted in its topic (v6435). I changed the vsync code of vmr7 and the rm renderer, and now the jitter tests that karl lillevood has posted to somewhere are running flawlessly.

In its topic? Where would that be (if it's meant for the public), please?

[gabest]

Chibi Jasmin: It is under the players forum.

ogo: I've got another idea how to detect when the renderer receives a frame. Do you know delegating from .net? Why don't you hook onto IMemInputPin::Receive? Replacing function pointers is not nice, but it might just work.

[ogo]

@cult : done

@Chibi Jasmin: it's already done, i'll post it tonight on this URL (just for testing) : http://ogo.nerim.net/reclockfilter/setup_latest.exe

@gabest: i nearly gave up in my attempt to make the monitor work with DVD playback. Even worse some MPEG2 decoders simply refuse to connect to it because they don't know it (i suppose for anti piracy reasons). So it's not the way to go.
I thought replacing the allocator provided by the renderer by giving it mine (with NotifyAllocator), and calling back the original one, so i would be notified each time a new sample would appear. But you're idea seems much better (i could be notified AND overwrite the sample to add my little stuff). But how could i find and replace this pointer ? I can only obtain a COM interface to IMemInputPin. Overwriting the vtable ? I like that

[ogo]

@cult : done

@Chibi Jasmin: it's already done, i'll post it tonight on this URL (just for testing) : http://ogo.nerim.net/reclockfilter/setup_latest.exe

@gabest: i nearly gave up in my attempt to make the monitor work with DVD playback. Even worse some MPEG2 decoders simply refuse to connect to it because they don't know it (i suppose for anti piracy reasons). So it's not the way to go.
I thought replacing the allocator provided by the renderer by giving it mine (with NotifyAllocator), and calling back the original one, so i would be notified each time a new sample would appear. But you're idea seems much better (i could be notified AND overwrite the sample to add my little stuff). But how could i find and replace this pointer ? I can only obtain a COM interface to IMemInputPin. Overwriting the vtable ? I like that

[mazzo]

well, thanks, everybody ......

[gabest]

Originally posted by ogo
But how could i find and replace this pointer ? I can only obtain a COM interface to IMemInputPin. Overwriting the vtable ? I like that

Once you have the interface pointer it is only the matter of tricky castings. At codeguru you can find ideas how to do it:
http://codeguru.earthweb.com/cpp_mfc/Delegates.html
http://codeguru.earthweb.com/cpp_mfc...StandardC.html