ffdshow tryouts project: Discussion & Development

[foxyshadis]

Is there somewhere a page, which would collect known bugs and their status? Also it would be nice to have somewhere feature requests collected.

I think it would help both users and developers.

Souceforge bugtracker (http://sourceforge.net/tracker/?group_id=173941)

This should probably be linked in the first post, even though it technically falls under the project page.

[haruhiko_yamagata]

Just a bump to let you know I'm well on my way for writing up an ffdshow reference page: mewiki.project357.com/wiki/FFDShow_reference (http://mewiki.project357.com/wiki/FFDShow_reference)
I've written up info for the three most important pages, but haven't written any more yet. I won't be writing up any info on any of the transform filters, but I do hope to write info on the equivalent pages for the ffdshow audio component, and maybe a little more on the other 'meta' pages of ffdshow. I would of course welcome it if others chose to write them :) I've posted the link here for two reasons:
o to get people to point out my mistakes (and fix them)
o to find some suckers to write the rest for me
Any volunteers? :)Thank you for your documentation and place. I have posted "resize.htm" there.

Well, my goal is to create ffdshow.chm (microsoft htmlhelp document). So I need html. Writing two version (html and wiki) increases my work...

But if we get some volunteers by using MeWiki, it's worth while. Anybody can perticipate and contribute MeWiki, which I think a really ideal solution.

[haruhiko_yamagata]

Gradual denoise crush on my system on some movies(not any law, it's irregular). Win XP SP2 Pentium 4 631. All ffdshow setting are default, just only turn on denoise. Bug appears both in GCC and ICL builds.
I cannot reproduce.
What is your CPU? Which build do you use?
What is the size (width, height) of the video?

[haruhiko_yamagata]

@haruhiko_yamagata
Found another incompatible Winamp DSP plugin: Adapt-x v3.61c (http://www.winamp.com/plugins/details.php?id=138579). This one does not work with ffdshow and if it is in the list for usage it will make MPC not work properly, after opening a few files when you try to open a new one MPC says "Failed to query the needed interfaces for playback". I was wondering if you can add it to the DSP plugin ban list as well.

Is not possible to use Direct X audio plugins from within ffdshow right?OK, I'll look into it.

// EDIT
Excluded. I'm not sure why it does not work. Maybe ffdshow's bug, just for the time being...

I cannot find the way to know which plugins are compatible with mlltichannel audio. I'm going to add a check box so that user can manually choose.

[foxyshadis]

Thank you for your documentation and place. I have posted "resize.htm" there.

Well, my goal is to create ffdshow.chm (microsoft htmlhelp document). So I need html. Writing two version (html and wiki) increases my work...

But if we get some volunteers by using MeWiki, it's worth while. Anybody can perticipate and contribute MeWiki, which I think a really ideal solution.

Wilbert has a script that can convert from Wiki into plain html, for avisynth's offline documentation. ffdshow's could probably work with the same thing. You can send him a PM and see if he has it handy.

[chros]

I cannot find the way to know which plugins are compatible with mlltichannel audio. I'm going to add a check box so that user can manually choose.
I'm hardly waiting to test it !!!! Let me know when it's integrated! Thanks !!!

Another question: when DFX plugin is used it stoles the focus (even in hidden mode, which is my default) from the player (eg. MPC). Can you do something about this ?

Thanks...

[thuan]

What does Vista Optimized with Albain's build mean?

[iron2000]

For the HW deinterlacing at the output page, must the player by set to overlay mixer to use it?
No VMR?

HW deinterlacing does auto-deinterlacing, isn't it?

How I wish there would be some really thorough indepth guide for ffdshow something like the ATI Catalyst guide at TweakGuides. Maybe it can be a wiki for fast updates on the status of the features.

[Leak]

I will extract the first 2 minutes but I also put another clip out (see post #1041) that is 1 minute long that exhibits the behavior a few times:
approx 3sec in - oscillates from 41->33->41
approx 10sec in - oscillates from 41->33->41
approx 41sec in - oscillates from 41->33
approx 1:00min in - oscillates from 33->41
approx 1:03 min in - oscillates from 41->33
(first of all - sorry for taking my time with answering this, but I'm a bit stressed out at work currently so this had to wait until the weekend...)

Okay, I've tested the Video Essential clip several times in a row, logging the OSD values (esp. AviSynth info and Frame duration) to a csv file - there's an option on the OSD config page if you want to do the same.

It played at a constant 41.7xxx ms per frame for me, except for two or three scene changes in the beginning where the frame duration dipped down to 33ms for a single frame - which is surely caused by a bad edit (wonky interlace flags caused by cutting and/or pasting parts of the video after applying the flags in postproduction). If that's what you meant by "oscillate", that's perfectly normal and happens all the time; also, it should be almost unnoticeable.

It's caused by my filter calculating the new timestamps directly from the old ones so the timestamps stay in sync - I could assign all frames the same duration by making up timestamps myself, but that would make the audio/video sync drift apart over time...

But what I meant by "oscillate" was a repeating pattern of "41 41 41 33" for the whole length of the video causing massive stutter - and that hasn't happened for me with my last version, no matter how hard I tried.

np: Richie Hawtin - TZ Entry Point (DE9 - Transitions)

[clsid]

What does Vista Optimized with Albain's build mean?
The only difference is an additional audio feature that was added specifically for Vista.

From SVN:Added audio device selection for multichannel streams (for Vista only). This lets deal with the audio end points problem in Vista : only one entry point can be used at a time. This feature will let use another device than the default one for multichannel streams.

[haruhiko_yamagata]

For the HW deinterlacing at the output page, must the player by set to overlay mixer to use it?
No VMR?

HW deinterlacing does auto-deinterlacing, isn't it?

How I wish there would be some really thorough indepth guide for ffdshow something like the ATI Catalyst guide at TweakGuides. Maybe it can be a wiki for fast updates on the status of the features.
"Use overlay mixer" is historical naming. In this era of VMR, the name of the check box doesn't make sense in my opinion.

Technically speaking, "Use overlay mixer" force (or allow by intermediate state) ffdshow to use VIDEOINFOHEADER2.
If not checked, VIDEOINFOHEADER is used.
VIDEOINFOHEADER2 is superior in the point that it can specify pixel aspect ratio (SAR), interlacing flag (called HW deinterlacing there).

Intermediate state in VMR7/9 is same as checked.
Intermediate state except VMR, first try VIDEOINFOHEADER2 and if failed to connect, try VIDEOINFOHEADER.

If I remember correctly, on Windows 2000, when default video renderer of OS is specified by the application, "Use overlay mixer" works just like its name. If checed, overlaymixer, If not, old renderer.

Could anyone fix my English and post it to MeWiki?

I would like to rename the check box. What should it be called? "Send pixel aspect ratio to video renderer"?

[clsid]

Perhaps "Use advanced renderer capabilities" or "Send control flags to renderer".

[haruhiko_yamagata]

A lot of people are having trouble finding the 'presets' functionality in ffdshow because it is not in a very obvious place.

Some suggestions to improve the GUI:

* Rename 'Image settings' to 'Profiles/Presets'

* Which do you prefer 'Profiles/Presets' or 'Profiles/Preset settings'?

* Rename 'Show/hide filters' to 'Filters' and remove the indentation.

I tried, but it did not look better to me.

* Remove indentations for 'Decoder options', 'Queue & misc' and 'Output'.

The indentation means that they are in preset settings (The indentation begins 'Image settings') though this idea of milan is less likely to be understood. Cosmetically, removing indentation does not look better to me.

[haruhiko_yamagata]

typedef struct tagVIDEOINFOHEADER {
RECT rcSource,
RECT rcTarget;
DWORD dwBitRate;
DWORD dwBitErrorRate;
REFERENCE_TIME AvgTimePerFrame;
BITMAPINFOHEADER bmiHeader;
} VIDEOINFOHEADER

typedef struct tagVIDEOINFOHEADER2 {
RECT rcSource;
RECT rcTarget;
DWORD dwBitRate;
DWORD dwBitErrorRate;
REFERENCE_TIME AvgTimePerFrame;
DWORD dwInterlaceFlags;
DWORD dwCopyProtectFlags;
DWORD dwPictAspectRatioX;
DWORD dwPictAspectRatioY;
DWORD dwReserved1;
DWORD dwReserved2;
BITMAPINFOHEADER bmiHeader;
} VIDEOINFOHEADER2;

The difference is aspect ratio, interlace and copy protect.
ffdshow does not set copy protect flag, AFAIK.
dwInterlaceFlags is set only if "HW deinterlacing" is checked.

[fastplayer]

..."Send control flags to renderer".
Sounds good to me. The tooltip should contain a short explanation of all 3 states like
Checked: ffdshow uses advanced control flags like SAR and interlacing...
Greyed-out: ffdshow tries to use advanced control flags but falls back if unsupported...
Unchecked: ffdshow does not use any of the advanced control flags...

[haruhiko_yamagata]

Sounds good to me. The tooltip should contain a short explanation of all 3 states like

Checked: ffdshow uses advanced control flags like SAR and interlacing...


Well, if "HW deinterlacing" is not checked, only SAR is the advanced control flag.

[fastplayer]

It was just meant as a template. Until now I didn't know what this check box meant. The tooltip was as meaningful as last week's newspaper...
Maybe something like this:
"ffdshow uses advanced control flags like SAR and interlacing (if selected below) etc."

[clsid]

* Which do you prefer 'Profiles/Presets' or 'Profiles/Preset settings'?
Preset settings

[clsid]

Btw, is there any reason to keep the "use overlay mixer" checkbox? Could we not just make the intermediate state behavior the standard behavior?

[jidelite]

(first of all - sorry for taking my time with answering this, but I'm a bit stressed out at work currently so this had to wait until the weekend...)

No problem. I know how stressful work can be. I'm just glad you have taken them time to review this.

Okay, I've tested the Video Essential clip several times in a row, logging the OSD values (esp. AviSynth info and Frame duration) to a csv file - there's an option on the OSD config page if you want to do the same.

It played at a constant 41.7xxx ms per frame for me, except for two or three scene changes in the beginning where the frame duration dipped down to 33ms for a single frame - which is surely caused by a bad edit (wonky interlace flags caused by cutting and/or pasting parts of the video after applying the flags in postproduction). If that's what you meant by "oscillate", that's perfectly normal and happens all the time; also, it should be almost unnoticeable.

Yes, this is what I meant by "oscillating" but here is the frustrating part for me; when I use your ffdshow version and the frame duration dips to 33ms (like the clips I provided) my video stutters/freezes for a couple of secs. All other ffdshow versions I use handle the frame dips very smoothly without a problem. I will continue to test your version(s) but it will sure make testing easier when you add the switch to turn buffering on/off so I don't have to continue un-installing your version of ffdshow to go back to one that plays smoothly for me. I really wish someone else could confirm that this is happening to them but as it stands, if its only me, then forget about this as I'm sure there are other things you can do with your valuable time.

Thanks,
jidelite

[zambelli]

HW deinterlacing does auto-deinterlacing, isn't it
VIDEOINFOHEADER2 is superior in the point that it can specify pixel aspect ratio (SAR), interlacing flag (called HW deinterlacing there).
I think "Hardware deinterlacing" is a misnomer. Setting the dwInterlaceFlags in VIH2 enables the downstream DirectShow filter to treat the content as interlaced. In most cases the next filter after a decoder (such as Ffdshow) is the Video Renderer. VMR7, VMR9 and EVR know that if dwInterlaceFlags has AMINTERLACE_IsInterlaced set, they should try to perform deinterlacing. Whether the deinterlacing is truly performed in software (CPU only) or hardware (GPU assisted through DXVA) is irrelevant to the decoder. So I think calling this feature "hardware deinterlacing" is not most accurate. If all it does is enable VIH2 and sets the AMINTERLACE_IsInterlaced flag, then it would more apt to name it "Enable interlaced signaling" or something like that. After all, there's no guarantee that the next filter in the graph is a renderer - so there's still no guarantee deinterlacing will be performed if that checkbox is checked.

[foxyshadis]

Btw, is there any reason to keep the "use overlay mixer" checkbox? Could we not just make the intermediate state behavior the standard behavior?

Agreed unless anyone knows of a case where another filter rejects VIH2, even though it works when forced on.

[vlada]

Hi, I just read about a deinterlace filter in MPlayer called Yadif (http://forum.doom9.org/showthread.php?t=124284). It is a smart bob filter, which doubles framerate, has both spatial and temporal detection and an edge directed interpolation. I haven't tried it yet myself, but in a test I've seen it looks very good.

I'm wondering, if it would be possible to add this filter in ffdshow.

[Kurtnoise]

@devs : lavc doesn't compile with MSVC in the last svn rev (due to the latest atrac3 mods)...Seems to be ok with gcc though.

[HeadBangeR77]

Checked: ffdshow uses advanced control flags like SAR and interlacing...
Greyed-out: ffdshow tries to use advanced control flags but falls back if unsupported...
Unchecked: ffdshow does not use any of the advanced control flags...
This has made the option finally clear to me. Cheers!

Just one quite a noobish question: As long as I remember the option has been checked & greyed out in my case. What does it depend on? Graphic card's drivers? Graphic card's capabilities? OS as well? Or maybe all of them together to some extent?

Thanks in advance.
cheers,
HDBR77

PS. A quick look at video render pin revealed I've got "formattype: FORMAT_VideoInfo2" for all the formats I checked.

@ zambelli:
If all it does is enable VIH2 and sets the AMINTERLACE_IsInterlaced flag, then it would more apt to name it "Enable interlaced signaling" or something like that.
Do you assume then, if it's unchecked, that the video render (no matter which is being used) won't get the signalling as to whether the video is interlaced or not? If it's true, would the render always try to perform deinterlacing with this option checked? :confused:

[vlada]

One more question: It used to be possible to install ffdshow even with a user account. Now I can't install ffdshow even as a poweruser, the installer tells me I need to have an administrator account. Why?

[haruhiko_yamagata]

Btw, is there any reason to keep the "use overlay mixer" checkbox? Could we not just make the intermediate state behavior the standard behavior?
Hmm... seems to be a good idea. Only one negative point of intermediate state that I know is that it slows down connection process compared to "checked" state.
User-friendliness may prioritize over it though.

But I'm not brave enough to remove it, because ffdshow has a huge variety of users. There may be some users who need that feature for the reason that I cannot imagine.

[haruhiko_yamagata]

@ zambelli:

Do you assume then, if it's unchecked, that the video render (no matter which is being used) won't get the signalling as to whether the video is interlaced or not? If it's true, would the render always try to perform deinterlacing with this option checked? :confused:
If it's unchecked the video renderer won't get any information regarding interlacing. They should be treated as progressive (some video renderers may try IVTC for DVD, though I'm not sure).
If it's checked, the flag is set if it is interlaced or not. For progressive stream, the flag is set progressive. What matters a lot is DVD, they are always interlaced but in most cases they were originally film, i.e. progressive.

[wozio]

If it's unchecked the video renderer won't get any information regarding interlacing. They should be treated as progressive (some video renderers may try IVTC for DVD, though I'm not sure).
If it's checked, the flag is set if it is interlaced or not. For progressive stream, the flag is set progressive. What matters a lot is DVD, they are always interlaced but in most cases they were originally film, i.e. progressive.
Isn't it VIDEOINFOHEADER2 set for each media sample passed to renderer? So only thing to do for decoder is to pass flags from encoded stream. It is up to renderer (or in fact drivers if we are talking about VMR) to detect correct deinterlacing mode.

[Keepitsimple]

maybe time for next official beta?

[haruhiko_yamagata]

Isn't it VIDEOINFOHEADER2 set for each media sample passed to renderer?Yes.
So only thing to do for decoder is to pass flags from encoded stream.ffdshow pass interlacing flags only when "HW deinterlacing" is checked. Otherwise ffdshow sets the flag progressive. Most of DirectShow decoders have this kind of option.
It is up to renderer (or in fact drivers if we are talking about VMR) to detect correct deinterlacing mode.Decoders can choose weave and bob.

[zambelli]

Yes.
ffdshow pass interlacing flags only when "HW deinterlacing" is checked. Otherwise ffdshow sets the flag progressive. Most of DirectShow decoders have this kind of option.
Decoders can choose weave and bob.
Correct. In most cases "weave" is the same as progressive (or no VIH2) since it instructs the renderer to simply leave the frames weaved as is. "Weave" basically means "don't touch".

Most decoders will signal AMINTERLACE_DisplayModeBobOnly when they want the renderer to deinterlace the video, but this doesn't imply that bobbing must be used. Depending on the VMR and its mode (VMR7 acts differently in regular vs YUV Mixing mode), same-framerate deinterlacing or double-framerate bobbing might occur. That part's up to the renderer and the video HW.

Furthermore, I've noticed that most GPU deinterlacers do a good job of detecting true interlaced vs progressive signaled as interlaced and adapting the renderered framerate on the fly. I encoded a video once of a 1-minute source where the first half was truly interlaced, and then the second half was the same as the first half but deinterlaced. The first half was encoded in VC-1 as field interlaced, and the second half was encoded as progressive - but the whole ASF file was signaled as interlaced. When decoded back via DXVA and VMR9, the first half was played back at 59.94 fps, and then at the midpoint the framerate started dropping until it reached 29.97 fps. I think I tested this on an Nvidia 6600 GPU.

[wozio]

Depending on the VMR and its mode (VMR7 acts differently in regular vs YUV Mixing mode
I noticed the same for VMR9 but I think it is not regular mixing vs YUV mixing but more non mixing vs mixing mode (as far as I can tell yuv mixing is almost the same as rgb mixing but without last step of converting to RGB and passing it as a output, but I'm trying it only with one HW and only in my own app so I probably messed something).

Furthermore, I've noticed that most GPU deinterlacers do a good job of detecting true interlaced vs progressive signaled as interlaced and adapting the renderered framerate on the fly. I encoded a video once of a 1-minute source where the first half was truly interlaced, and then the second half was the same as the first half but deinterlaced. The first half was encoded in VC-1 as field interlaced, and the second half was encoded as progressive - but the whole ASF file was signaled as interlaced. When decoded back via DXVA and VMR9, the first half was played back at 59.94 fps, and then at the midpoint the framerate started dropping until it reached 29.97 fps. I think I tested this on an Nvidia 6600 GPU.

Modern GPUs can do even more, for example regular film telecined material broadcasted with scrolling text in regular ntsc frame rate can be properly deinterlaced to maintain both ivtc and smooth text movement :)

One more thing to add here. On my HW (ATI X1650) renderer (VMR9) input colorspace must be set to NV12 for advanced deinterlacing to work. With other colospaces driver does simple bob. So I have to disable all other colorspaces in ffdshow to force NV12 output. It would be good to add some kind of list with possibility to prioritize proposed colorspaces. This is really done well in coreavc, there is a list with colorspaces and they are proposed downstream in order (it is possible to move colorspaces on list as weel as disabling some of them).

Anyway for me this HW deinterlacing setting in ffdshow is not needed and confusing.

[alter4]

I cannot reproduce.
What is your CPU? Which build do you use?
What is the size (width, height) of the video?
My CPU is pentium 4 (model 631) video size 540x300
which build is not important, bug appears in all builds

[iron2000]

Anyway for me this HW deinterlacing setting in ffdshow is not needed and confusing.

Wow, sure did confuse me.

So when checked the next filter after ffdshow will know if the video is interlaced or not and when unchecked ffdshow will mark all videos as deinterlaced?

What if the driver has more method other than weave and bob, does it mean that ffdshow restricts it only to the 2 methods?

The guide at MeWiki is a pretty good start, hope that it will cover all the options in the future or maybe put them up on a ffdshow-tryouts wiki.

[haruhiko_yamagata]

Wow, sure did confuse me.

So when checked the next filter after ffdshow will know if the video is interlaced or not and when unchecked ffdshow will mark all videos as deinterlaced?Yes. As a hint for downstream. But downstream filter may detect the stream interlaced (or telecined DVD) and try to deinterlace (IVTC). It's downstream's choice.

What if the driver has more method other than weave and bob, does it mean that ffdshow restricts it only to the 2 methods?
It's just flag. Downstream filters may ignore the flag. You can force the downtream filter to ignore the upper stream's flag, if the filter supports the configuration.

[vlada]

What if the driver has more method other than weave and bob, does it mean that ffdshow restricts it only to the 2 methods?

As zambelli already said, weave is not a deinterlace method. It means, that the renderer treats the video as if it was progressive (no matter if it really is). So the only choice for an interlaced video is (smart) bob. There are usually no other deinterlace methods. But if you had for example a graphic card with more deinterlace algorithms, you'd have to choose them in driver properties. There's no way how ffdhsow could control it.

[cc979]

on ffdshow svn version 1122 is there a typo on
line 124 of makefile_c.inc

should it be this ?
--def $(GCCDEF) $(GCCDLLENTRY) -o $@ $(OBJS) -lpthread $(GCCLIBS)


as it did not compile before i changed it

[_xxl]

I would like to uncheck TomsMoComp,kernelDeint and x264 from msvc 2003(2005) solution files.
http://img17.imagevenue.com/loc992/th_77370_Capture_122_992lo.jpg (http://img17.imagevenue.com/img.php?image=77370_Capture_122_992lo.jpg)
TomsMoComp,kernelDeint should be compiled by ICL and x264 by GCC.

[zambelli]

I have a question regarding the Audio Decoder Configuration option under Output named "Don't use WAVEFORMATEXTENSIBLE header when not needed"...

Would it be possible to make this into a dropdown list with options:
* Never use WAVE_FORMAT_EXTENSIBLE
* Don't use WAVE_FORMAT_EXTENSIBLE when not needed
* Always use WAVE_FORMAT_EXTENSIBLE

As it is right now, I can prevent Ffdshow from outputting WAVE_FORMAT_EXTENSIBLE headers, but I can't force it to output WAVE_FORMAT_EXTENSIBLE headers.

[check]

The second option would be better worded "Only use when needed". Double negatives suck :P

[Leak]

TomsMoComp,kernelDeint should be compiled by ICL and x264 by GCC.
No argument about x264 from me, but what about us folks who only have MSVC (and of course GCC, since it's free)?

And isn't there a separate ICL9 project anyway?

[clsid]

ICL9 binaries for TomsMoComp and kernelDeint are in the SVN. So there is no need to compile them.

I agree that they should be removed (not just unchecked) from the main msvc solution files. There are separate project files present for them anyway.

[TheShadowRunner]

2 little suggestions:

- for the installer: if it could remember the language used during previous ffdshow installations.

- for ffdshow: for the "logoaway" feature, if we could have a feature for setting the position easily. Like a preview either in red or as negative image, to set the position accurately, and then we would disable it for the normal processing.
Hope you see what i mean ^^;
Later,

TSR

[clsid]

2 little suggestions:
- for the installer: if it could remember the language used during previous ffdshow installations.That is not possible with the install software that we use.

[haruhiko_yamagata]

My CPU is pentium 4 (model 631) video size 540x300
which build is not important, bug appears in all builds
Thank you. Now I can reproduce the bug.
The debug may be hard because it's random and rare.

[Seb.26]

Hi world ! :)

is there a way to disabling the TV->PC level remapping done when input is YV12/YU2 and output is RGB32 ?!
( = just output same level in RGB32 ... )

Thanks !
Seb.

[_xxl]

I have added SSE3 & SSSE3 detection code:
Patch
http://rapidshare.com/files/28159079/ssse3.rar.html
Bin
http://rapidshare.com/files/28158873/ffdshow_rev1124_20070427_xxl_test.exe.html
Yasm 0.6.0 is needed.
If anybody is interested in testing and helping me to fix it.
This will be used in xvidcore 1.2 and x264.

[Px]

I have added SSE3 & SSSE3 detection code:
Patch
http://rapidshare.com/files/28159079/ssse3.rar.html
Bin
http://rapidshare.com/files/28158873/ffdshow_rev1124_20070427_xxl_test.exe.html
Yasm 0.6.0 is needed.
If anybody is interested in testing and helping me to fix it.
This will be used in xvidcore 1.2 and x264.
Intel E6300
http://users.i.com.ua/~finish/ffdshow/ffd_c2d.png
Intel P3-450 :D
http://users.i.com.ua/~finish/ffdshow/ffd_p3.png

Also small speedup compared to 1090 icl9
1090
Null
User: 44s, kernel: 0s, total: 44s, real: 46s, fps: 26.4, dfps: 25.5
o/m
User: 45s, kernel: 0s, total: 45s, real: 46s, fps: 26.1, dfps: 25.6
1124
Null
User: 43s, kernel: 0s, total: 43s, real: 44s, fps: 26.9, dfps: 26.3
o/m
User: 44s, kernel: 0s, total: 44s, real: 45s, fps: 26.5, dfps: 25.9

[_xxl]

Intel E6300 doesn't have 3dnow! ext.
3dnow! ext can't be disabled...