Media Player .NET (MPDN) - D3D HQ GPU Video Renderer [v2.49.0/v1.31.0 27 Dec 2018]

[burfadel]

EDIT: Fixed in v2.29.1.

Thanks, works well now!

[Zachs]

Hi guys,

Just a quick heads up - there's a new MPDN Extensions release (v1.2.0).

Changes:
- Removed conditional usage option for ImageProcessor (use the Conditional script instead)
- Added script editor for ImageProcessor
- Conditional script revamped (note: you may need to delete your config folder if you encounter an error)
- Conditional no longer loads from Script Group's preset - it now allows you to select from a list of render scripts

[foxyshadis]

I get this error with the new extension pack:
===================================

c:\av\MPDN\Extensions\RenderScripts\Mpdn.ScriptedRenderChain.MpdnScriptEngine.cs(33,21): error CS0104: 'ScriptEngine' is an ambiguous reference between 'Microsoft.ClearScript.ScriptEngine' and 'Mpdn.Extensions.Framework.ScriptEngine'
c:\av\MPDN\Extensions\RenderScripts\Mpdn.ScriptedRenderChain.MpdnScriptEngine.cs(122,46): error CS0246: The type or namespace name 'Clip' could not be found (are you missing a using directive or an assembly reference?)
(CSScriptLibrary)

------------------------------
Program Location:

at csscript.CSExecutor.ProcessCompilingResult(CompilerResults results, CompilerParameters compilerParams, ScriptParser parser, String scriptFileName, String assemblyFileName, String[] additionalDependencies)
at csscript.CSExecutor.Compile(String scriptFileName)
at CSScriptLibrary.CSScript.LoadWithConfig(String scriptFile, String assemblyFile, Boolean debugBuild, Settings scriptSettings, String compilerOptions, String[] refAssemblies)
at CSScriptLibrary.CSScript.Load(String scriptFile, String assemblyFile, Boolean debugBuild, String[] refAssemblies)
at MediaPlayerDotNet.CsScript.†††
†††ˆ�–(Object , Object , Boolean , Object )
at MediaPlayerDotNet.CsScript.†††
††††‡�(Object , Object , Boolean )
at MediaPlayerDotNet.CsScript.†††
††††‡˜(Object , Object , Object , Object , Object )
at MediaPlayerDotNet.CsScript.LoadScripts(String scriptDir)
at MediaPlayerDotNet.MainForm.†††
†††‡‰™()

Sorry about the garbage, must be the lack of symbols.

[Zachs]

You didn't remove the old extension files, did you? :)

[foxyshadis]

Ah, should have noticed the datestamp. Looks like things have changed a LOT in the last few days, much more broken out. At least my little extensions still work!

[Zachs]

Oh what extension is that if you don't mind me asking?

[ryrynz]

Ah, should have noticed the datestamp. Looks like things have changed a LOT in the last few days, much more broken out. At least my little extensions still work!

I sometimes wish there was an installer that took care of everything for me lazy style.. I think I might just write a batch file to hose the MPDN config and install directory as well as extract out the ZIP.
Seems like more often than not I have to do one or the other or both.

Hey Zach any chance of an option for storing the application config in the registry?

[Zachs]

I sometimes wish there was an installer that took care of everything for me lazy style.. I think I might just write a batch file to hose the MPDN config and install directory as well as extract out the ZIP.
Seems like more often than not I have to do one or the other or both.

Hey Zach any chance of an option for storing the application config in the registry?

Nope. Registry should never be used to store app settings. That's what the app data folder is for.

[stax76]

Nope. Registry should never be used to store app settings. That's what the app data folder is for.

I agree but where I don't agree is startup time is the single most important feature for a player. Currently it feels like 2 seconds, I believe 150 ms is even for a .NET player possible. Obviously it's not doable loading twenty assemblies before playback starts or deserializing or parsing large files.

[Shiandow]

I agree but where I don't agree is startup time is the single most important feature for a player. Currently it feels like 2 seconds, I believe 150 ms is even for a .NET player possible. Obviously it's not doable loading twenty assemblies before playback starts or deserializing or parsing large files.

Does loading time improve if you delete the settings folder?

[stax76]

Does loading time improve if you delete the settings folder?

I tried it a few days ago with a fresh setup.

[Belphemur]

I sometimes wish there was an installer that took care of everything for me lazy style.. I think I might just write a batch file to hose the MPDN config and install directory as well as extract out the ZIP.
Seems like more often than not I have to do one or the other or both.

Hey Zach any chance of an option for storing the application config in the registry?

Could you describe what you would want in the installer of MPDN ?

For now the installer is taking care of removing the content of the Extensions to avoid any problem. And when uninstalling you can ask it to remove all the configuration.

[Magik Mark]

Zachs

Is there a way to make switching to FSE faster? It takes me around 3 seconds. Comparing it with MPC BE with madvr it only takes 1 second to switch back and forth between windows mode & FSE

Thanks

[ryrynz]

Zachs

Is there a way to make switching to FSE faster?

Thanks
What d3d version you using and what card / driver? I don't have any issues with performance here personally.

[Magik Mark]

What d3d version you using and what card / driver? I don't have any issues with performance here personally.

D3D 11 GTX 550Ti Driver 353.06

[Zachs]

Takes less than half a second for my 560gtx. 1 second is painfully slow.

[Magik Mark]

Takes less than half a second for my 560gtx. 1 second is painfully slow.

Is there something I could Tweak? I was just wondering why its a lot faster in mpcbe with madvr

[ryrynz]

Is there something I could Tweak? I was just wondering why its a lot faster in mpcbe with madvr
Perform a clean install of that driver and test d3d 10 and 9 and report if it's the same speed as 11.

[Zachs]

I agree but where I don't agree is startup time is the single most important feature for a player. Currently it feels like 2 seconds, I believe 150 ms is even for a .NET player possible. Obviously it's not doable loading twenty assemblies before playback starts or deserializing or parsing large files.

Not sure how saving and loading from the registry is going to make things any faster though?

[ryrynz]

I agree but where I don't agree is startup time is the single most important feature for a player. Currently it feels like 2 seconds, I believe 150 ms is even for a .NET player possible. Obviously it's not doable loading twenty assemblies before playback starts or deserializing or parsing large files.

Only reason I said it was so if I nuked the install directory the settings would still be remembered, but I'll just rely on the installer from now on.
It certainly wouldn't make any difference with regards to loading times. Yeah about two seconds on my machine too (currently using a HDD on this one) It's not too bad really.

Could you describe what you would want in the installer of MPDN ?


It's fine. I was using the zip, I didn't realize the installer deleted the extensions first.

[stax76]

I just take a look at StaxRip on what feels slow, take a look what could be optimized and use a class called stopwatch to find bottlenecks from time to time, binary deserialization takes most time in my case, I have room for optimization but really don't have a need since the startup is still very fast. What takes most time in your case can easily be measured. I've extended many applications with mostly personal scripts and add-ons, Firefox, MediaMonkey, DVBViewer, Windows Explorer, VisualStudio. Some of this apps are on the slow side but it's mostly acceptable considering the amount of powerful features. For a media player however the fastest possible startup time is crucial for me, for others it's probably not so important. I use MPC and VLC sometimes but most of the time when I don't have special needs like subtitles I use a personal player built with VB.NET with special behavior hard coded, the code for a old version is still available here (https://sourceforge.net/p/staxmedia/code/HEAD/tree/trunk/StaxPlayer/). It's using winforms and a WPF window for OSD but that was crashing on Win10 so I'm currently implementing OSD with EVR. It's somehow unfortunate that there isn't a player yet I like, I looked at all in detail, unfortunately I have to say I'm not impressed by most, they can all still be improved and I could contribute but I'm mostly fine with my personal player and occasionally MPC and VLC. Sorry about the rant, just that it's not for me doesn't mean that it's bad. :)

[Keiyakusha]

huh, startup time? I never noticed before you mentioned. And now that I did and measured for myself, I still fail to see how it matters. You probably don't know what is slow, MPDN is like 3 times faster than Photoshop, haha ^_^ But seriously, I can understand why for example slow FSE switching can be annoying (I'm not saying it is slow, I don't use it myself) but startup time is just a unnecessary nitpick.

acceptable considering the amount of powerful features
Isn't MPDN falls into this category too? I hope by "powerful" you don't mean stuff like additional internal filters, 1001 old renderer to select for no reason, encoding capabilities and other bloat, some other players may have.

[stax76]

Startup time is not a concern for everybody, too me it's very important. With powerful I mean applications you need minimum 10 years to build.

[Keiyakusha]

I understand your point, however I want to contest the following statement too. I hope its not too much offtopic.
With powerful I mean applications you need minimum 10 years to build.

No application or even OS needs this much time to build. For some of them it takes this much time, or even more but only because they are changing all the time. With this respect, some of the software will never be completed.
For some windows players it took around 12 years to get them where they are now. Are they more powerful than MPDN? Highly debatable I should say. Shouldn't we be saying quite the opposite thing, that it is OK for MPDN to have slower startup because it got this powerful with this short amount of time?

[foxyshadis]

Quick question: With the new Conditionals, can I use the other Renderer properties as well, like InputFormat? Turning off debanding for P010 and Y410 would be useful, I rarely if ever need it for 10-bit vids.

[Zachs]

Quick question: With the new Conditionals, can I use the other Renderer properties as well, like InputFormat? Turning off debanding for P010 and Y410 would be useful, I rarely if ever need it for 10-bit vids.

InputFormat isn't accessible yet but it's very easy to add to the Clip class. It's a one liner. I'll add that in there when I get the chance but if you can figure out where and how to add it then fell free to propose a git merge.

EDIT: Rather than exporting InputFormat, I'd prefer it to be a simple int property called 'SourceBitDepth'.

[Zachs]

v2.29.2 without extensions now startup (fully usable) in less than 1s.
Takes an additional 1s if you have extensions installed. It should be a non-issue now even for the pedantic users.

I prefer human readable text based XML serialization and will never go for binary serialization where app settings are concerned. I'd rather sacrifice a bit of startup time.

MPC-HC starts up much faster than MPDN but opening a media file when madVR loads the first time is hell of a lot slower. When MPDN starts up, its renderer is already fully loaded and initialized.

[madshi]

MPC-HC starts up much faster than MPDN but opening a media file when madVR loads the first time is hell of a lot slower.
Depends on the selected algorithms, though. E.g. when using OpenCL stuff, madVR doesn't start rendering until OpenCL is initialized, the kernels are compiled etc. That's on my to do list to improve. If you don't use any OpenCL or DirectCompute stuff, MPC-HC/madVR should start rendering without much delay. But that can probably still be improved a bit, too. There's always something that can be further optimized... :)

[huhn]

madVR/mpc-hc can present a frame in less than a couple 100 ms.

it is nearly instant when the file is on a sdd the bottle neck is clearly the HDD these days.

i tested mpdn 2.29.2.3117 without extensions it is a clearly slower than madVR/mpc-hc. but the difference is meaningless to me.

[stax76]

v2.29.2 without extensions now startup (fully usable) in less than 1s.
Takes an additional 1s if you have extensions installed. It should be a non-issue now even for the pedantic users.

I prefer human readable text based XML serialization and will never go for binary serialization where app settings are concerned. I'd rather sacrifice a bit of startup time.

MPC-HC starts up much faster than MPDN but opening a media file when madVR loads the first time is hell of a lot slower. When MPDN starts up, its renderer is already fully loaded and initialized.

Did you measure how long the completely deserialization takes? StaxRip was VBScript -> VB6 -> .NET 1.0 so I'm using binary serialization since .NET 1.0, the reason was if I remember right ease of use and flexibility, since .NET 1.0 binary serialization was significant improved giving even more flexibility, XML serialization was surely improved too, I don't know how both compare today, binary serialization allows you to serialize any object graph no matter how complex it is with little effort, you can serialize practically everything, all you need to do is adding SerializableAttribute to classes and take care of versioning for which various customizations are available.

What you can do to speed up serialization is serialize only values that are different from the default value and you can split things up to several files and use lazy loading.

I'm impressed how fast SharpDevelop 4.4 starts considering it's a huge WPF application with addin support, it's start practically instantly, I'm using it to translate C# to VB.NET which works incredibly well.

You can also soon use the brand new x64 JIT compiler RyuJIT:

Microsoft’s next generation just-in-time compiler for .NET, RyuJIT, is now the default compiler for 64-bit applications on .NET 4.6., if you have installed VS2015 Preview or newer, RyuJIT will be enabled.

I don't know when the final ships, probably with Win10 so pretty soon.

[Zachs]

The bulk of the time isn't in serialisation. That said, why are we still having this conversation? Is it still too slow for you when it starts in less than 1s without the extensions?

[Zachs]

Depends on the selected algorithms, though. E.g. when using OpenCL stuff, madVR doesn't start rendering until OpenCL is initialized, the kernels are compiled etc. That's on my to do list to improve. If you don't use any OpenCL or DirectCompute stuff, MPC-HC/madVR should start rendering without much delay. But that can probably still be improved a bit, too. There's always something that can be further optimized... :)
When I last tested it, first time I open a media file it feels like madvr DLL only gets loaded then. It took around 2-3 seconds before playback starts and that's with just hardware bilinear. Subsequent playback was a lot faster though.

Anyway it's a little out of topic for this thread.

[ryrynz]

Could you describe what you would want in the installer of MPDN ?


For the uninstaller to remember the directory I installed to, uninstall from that directory and for it to reference it in it's dialogue box. The Installer would also offer to reinstall to that directory by default.

[Anime Viewer]

Lately I seem to be encountering quite a bit of shuddering/jerky playback (with Fluid Motion on) on certain types of videos. Especially on horizontally panning scenes. Anyone one else have similar experiences on videos with the properties listed below?


From MPDN Media Info:

Video
ID : 1
Format : AVC
Format profile : Main@L3.1
Format settings : CABAC / 6 Ref Frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 1417251
Bit rate : 0
Width : 848
Height : 480
Display aspect ratio : 1.006
Frame rate mode : CFR
Frame rate : 23.976
Chroma subsampling : 4:2:0
Bit depth : 8
Scan type : Progressive
Bits/(Pixel*Frame) : 0
Stream size :
Writing library : x264 - core 120 r2120 0c7dab9
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x1:0x111 / me=umh
/ subme=8 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 /
trellis=1 / 8x8dct=0 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 /
threads=4 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 /
constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 /
weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=23 / scenecut=40 /
intra_refresh=0 / rc_lookahead=50 / rc=2pass / mbtree=1 / bitrate=768 / ratetol=1.0 /
qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 /
vbv_maxrate=1536 / vbv_bufsize=3840 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

Language :

Audio
ID : 2
Format : AAC
Codec ID : A_AAC
Duration : 1417252
Bit rate mode :
Bit rate : 0
Channel(s) : 2
Sampling rate : 44100
Bit depth : 0
Compression mode : Lossy
Stream size : 0
Language :

Subtitle
ID : 3
Format : ASS
Muxing Mode :
Codec ID : S_TEXT/ASS
Language :


Edit: It seems Fluid Motion was actually causing the problem (or was related in some way to it) (and was constantly flashing on and off in OSD). Disabling it resolved the issue (it appears).

CTRL+J shows: (When Fluid motion is turned off)
Dispay 59.824999Hz
Video frame rate
23.976Hz (says source filter)
23.972Hz (avg last 2s) (seems to fluxuate between 23.9 and 23.7)
23.97625Hz (ref clock)
Render Queue 12/12
Render speeds ~ 19.04ms

CTRL+J shows: (When Fluid Motion is on and fluctuating)
Display 59.82501 Hz
Video frame rate
23.976Hz (source filter)
16.012Hz (avg last 2s) (also goes down to 14)
23.97618Hz (ref clock)
Render Queue 0/12
tons of dropped and delayed frames
Render speeds ~ 61.73ms

[ryrynz]

Lately I seem to be encountering quite a bit of shuddering/jerky playback (with Fluid Motion on) on certain types of videos. Especially on horizontally panning scenes. Anyone one else have similar experiences

I've encountered this too, I wasn't sure what it was related to. I did some testing last night but didn't encounter it though. I have been getting the odd delayed and dropped frame but it seems somewhat random, just when I think I got it what's causing it a retest proves me wrong.

[Zachs]

Hint: Render duration 61.73ms. Your source is 23.976fps.

[Anime Viewer]

Hint: Render duration 61.73ms. Your source is 23.976fps.

I know the render duration being so high is why the dropped frames occur, but it would seem that is the effect of something else (my current best guess is the Fluid Motion flashing on and off constantly instead of remaining in one state or the other). Fluid Motion isn't so taxing that it should more than triple render times compared to when it is switched off. I get the impression that Fluid Motion isn't correctly detecting what state it should be in, and so it is continually triggered on and off.

ryrynz,
Have you noticed it with particular types of files (for example the 23.976 source files)? If you go to the MPDN Media Info area does it list CFR for the frame rate mode of the video(s) you see issue with?

I'm not sure how to interpret the report of the frame rates for the (last 2s) part. That being under the source and screen Hz is obviously an issue, but is it causing the other issue(s) or is it a result of the other issue(s). Could a low frame rate trigger Fluid Motion to switch on, and then when the frame rate increases Fluid Motion feels its no longer needed (and thus switches off), then when its detected as dropping again it switches on again, and thus a continual loop of switching on and off?


Chicken or the Egg which came first?

[Zachs]

The only time it will get switched off in your case would be render queue dropping under 2. For some reason it's causing such high render time when it's on.

FWIW I ran fluid motion on an old laptop with AMD mobility x2300 and the difference was around 2ms between on and off. That GPU doesn't even have a certified Windows 8.1 drivers. Does it only happen with one particular file?

[Anime Viewer]

The only time it will get switched off in your case would be render queue dropping under 2. For some reason it's causing such high render time when it's on.

FWIW I ran fluid motion on an old laptop with AMD mobility x2300 and the difference was around 2ms between on and off. That GPU doesn't even have a certified Windows 8.1 drivers. Does it only happen with one particular file?

No I've had it happen with more than one particular file. Like ryrynz mentioned it doesn't happen every time one of the files that has the problem is launched. Today when I launched that same file (as yesterday) it didn't have the problem occur. Other currently unknown variables are involved. When it happened yesterday it had just switched from playing one file in fullscreen mode to one of the problematic files that was on the playlist. I think the problematic file was third on the playlist. The second file on the playlist is always skipped after the first is played, so I thought that (or it switching from one file in fullscreen mode to another in fullscreen mode) might be related to the problem, but trying to repeat the same situation has failed to trigger the issue (in other words it wasn't switching fluid motion on and off continually). I'll continue experimenting, and report back if I can find a common variable when it is triggering.

Edit: I got it to occur again this time with other files. It started occuring after I went in and changed a setting in Fluid Motion. Specifically I went in and unchecked Only when video frame is less than half of display rate. The properties of the other two files I'm seeing it occur on are:


Video
ID : 1
Format : AVC
Format profile : Main@L3.1
Format settings : CABAC / 6 Ref Frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 1435185
Bit rate : 0
Width : 848
Height : 480
Display aspect ratio : 1.006
Frame rate mode : CFR
Frame rate : 23.976
Chroma subsampling : 4:2:0
Bit depth : 8
Scan type : Progressive
Bits/(Pixel*Frame) : 0
Stream size :
Writing library : x264 - core 120 r2120 0c7dab9
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x1:0x111 / me=umh
/ subme=8 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 /
trellis=1 / 8x8dct=0 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 /
threads=4 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 /
constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 /
weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=23 / scenecut=40 /
intra_refresh=0 / rc_lookahead=50 / rc=2pass / mbtree=1 / bitrate=768 / ratetol=1.0 /
qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 /
vbv_maxrate=1536 / vbv_bufsize=3840 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

Language :

Audio
ID : 2
Format : AAC
Codec ID : A_AAC
Duration : 1435179
Bit rate mode :
Bit rate : 0
Channel(s) : 2
Sampling rate : 44100
Bit depth : 0
Compression mode : Lossy
Stream size : 0
Language :

Subtitle
ID : 3
Format : ASS
Muxing Mode :
Codec ID : S_TEXT/ASS
Language :


and


Video
ID : 1
Format : AVC
Format profile : Main@L3.1
Format settings : CABAC / 6 Ref Frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 1452244
Bit rate : 0
Width : 848
Height : 480
Display aspect ratio : 1.006
Frame rate mode : CFR
Frame rate : 23.976
Chroma subsampling : 4:2:0
Bit depth : 8
Scan type : Progressive
Bits/(Pixel*Frame) : 0
Stream size :
Writing library : x264 - core 120 r2120 0c7dab9
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x1:0x111 / me=umh /
subme=8 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 /
trellis=1 / 8x8dct=0 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 /
threads=4 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 /
constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 /
weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=23 / scenecut=40 /
intra_refresh=0 / rc_lookahead=50 / rc=2pass / mbtree=1 / bitrate=768 / ratetol=1.0 /
qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 /
vbv_maxrate=1536 / vbv_bufsize=3840 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

Language :

Audio
ID : 2
Format : AAC
Codec ID : A_AAC
Duration : 1452244
Bit rate mode :
Bit rate : 0
Channel(s) : 2
Sampling rate : 44100
Bit depth : 0
Compression mode : Lossy
Stream size : 0
Language :

Subtitle
ID : 3
Format : ASS
Muxing Mode :
Codec ID : S_TEXT/ASS
Language :


Both of those files seem to have similar properties.

It is not occurring when running in the video's native window size. Only during fullscreen (I'm testing to see if it is FSE or FSE D3D 11 related).

Edit #2:
I didn't encounter it in D3D 11 Fullscreen Windowed mode, but then I also didn't encounter it after switching back to D3D 11 FSE mode either...
When the issue has occurred I'd also been running in "Multiple displays: Extend these displays mode", but I'm not sure if that has any barring. Next time it occurs I'll try switching to single display mode to see if that may be a factor.

Edit #3: When it reoccurred again I switched to single display, and D3D 11 Windowed Fullscreen, and Fluid Motion continued to switch on and off.

[ryrynz]

No I've had it happen with more than one particular file.

Could be wrong but I think it has something to do with switching between fullscreen and windowed modes.

[Garteal]

Anime Viewer: Are you using the latest playlist? That should've been fixed a long time ago unless something broke again.
Also what's your Render Script queue?

[Anime Viewer]

Anime Viewer: Are you using the latest playlist? That should've been fixed a long time ago unless something broke again.
Also what's your Render Script queue?

I'm using the playlist that is bundled with the latest installer. In other words: http://mpdn.zachsaw.com/Latest/Installers/MediaPlayerDotNet_x64_2_29_2_3117_Installer.exe
I'll run an uninstall, and then delete anything that the uninstaller is leaving behind to make sure I get a fresh playlist installed.
I'll also go into C:\Users\MyName\AppData\Local\MediaPlayerDotNet and delete the PlaerExtensions.64 folder to make sure that is reset properly. If I still have problems with it skipping the second file in the playlist after installing after that I'll let you know.

By render script queue are you refering to the Render Queue value I have set in Options -> video Render -> Queues -> Render Queue it is 12 slots (default?). (If you have a different recommended setting let me know). As for the render scripts I currently have set they are: Script Chain containing ChromaScaler -> Deband -> SuperRes

Edit: Yep, its still skipping the playlist items. I launch a video, go to view, go to playlist, and then drag and drop the videos I want to play after the current video finishes. In this case the first thing I added to the playlist (the second video if you count the one I originally opened, or the first if you're going off what is listed at the top of the playlist) is skipped. The third video (or the second one listed in the playlist) is played when the first video ends.

[ryrynz]

Yep, its still skipping the playlist items.

Garteal's latest dev release (https://github.com/Garteal/MPDN_Extensions/tree/dev) has this fixed (I can't reproduce using it)
This new release will make it into MPDN very soon.

[Zachs]

No I've had it happen with more than one particular file. Like ryrynz mentioned it doesn't happen every time one of the files that has the problem is launched. Today when I launched that same file (as yesterday) it didn't have the problem occur. Other currently unknown variables are involved. When it happened yesterday it had just switched from playing one file in fullscreen mode to one of the problematic files that was on the playlist. I think the problematic file was third on the playlist. The second file on the playlist is always skipped after the first is played, so I thought that (or it switching from one file in fullscreen mode to another in fullscreen mode) might be related to the problem, but trying to repeat the same situation has failed to trigger the issue (in other words it wasn't switching fluid motion on and off continually). I'll continue experimenting, and report back if I can find a common variable when it is triggering.

Edit: I got it to occur again this time with other files. It started occuring after I went in and changed a setting in Fluid Motion. Specifically I went in and unchecked Only when video frame is less than half of display rate. The properties of the other two files I'm seeing it occur on are:


Video
ID : 1
Format : AVC
Format profile : Main@L3.1
Format settings : CABAC / 6 Ref Frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 1435185
Bit rate : 0
Width : 848
Height : 480
Display aspect ratio : 1.006
Frame rate mode : CFR
Frame rate : 23.976
Chroma subsampling : 4:2:0
Bit depth : 8
Scan type : Progressive
Bits/(Pixel*Frame) : 0
Stream size :
Writing library : x264 - core 120 r2120 0c7dab9
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x1:0x111 / me=umh
/ subme=8 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 /
trellis=1 / 8x8dct=0 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 /
threads=4 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 /
constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 /
weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=23 / scenecut=40 /
intra_refresh=0 / rc_lookahead=50 / rc=2pass / mbtree=1 / bitrate=768 / ratetol=1.0 /
qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 /
vbv_maxrate=1536 / vbv_bufsize=3840 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

Language :

Audio
ID : 2
Format : AAC
Codec ID : A_AAC
Duration : 1435179
Bit rate mode :
Bit rate : 0
Channel(s) : 2
Sampling rate : 44100
Bit depth : 0
Compression mode : Lossy
Stream size : 0
Language :

Subtitle
ID : 3
Format : ASS
Muxing Mode :
Codec ID : S_TEXT/ASS
Language :


and


Video
ID : 1
Format : AVC
Format profile : Main@L3.1
Format settings : CABAC / 6 Ref Frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 1452244
Bit rate : 0
Width : 848
Height : 480
Display aspect ratio : 1.006
Frame rate mode : CFR
Frame rate : 23.976
Chroma subsampling : 4:2:0
Bit depth : 8
Scan type : Progressive
Bits/(Pixel*Frame) : 0
Stream size :
Writing library : x264 - core 120 r2120 0c7dab9
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x1:0x111 / me=umh /
subme=8 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 /
trellis=1 / 8x8dct=0 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 /
threads=4 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 /
constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 /
weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=23 / scenecut=40 /
intra_refresh=0 / rc_lookahead=50 / rc=2pass / mbtree=1 / bitrate=768 / ratetol=1.0 /
qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 /
vbv_maxrate=1536 / vbv_bufsize=3840 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

Language :

Audio
ID : 2
Format : AAC
Codec ID : A_AAC
Duration : 1452244
Bit rate mode :
Bit rate : 0
Channel(s) : 2
Sampling rate : 44100
Bit depth : 0
Compression mode : Lossy
Stream size : 0
Language :

Subtitle
ID : 3
Format : ASS
Muxing Mode :
Codec ID : S_TEXT/ASS
Language :


Both of those files seem to have similar properties.

It is not occurring when running in the video's native window size. Only during fullscreen (I'm testing to see if it is FSE or FSE D3D 11 related).

Edit #2:
I didn't encounter it in D3D 11 Fullscreen Windowed mode, but then I also didn't encounter it after switching back to D3D 11 FSE mode either...
When the issue has occurred I'd also been running in "Multiple displays: Extend these displays mode", but I'm not sure if that has any barring. Next time it occurs I'll try switching to single display mode to see if that may be a factor.

Edit #3: When it reoccurred again I switched to single display, and D3D 11 Windowed Fullscreen, and Fluid Motion continued to switch on and off.


So you can't find a consistent way to make it happen?
What happens if you try it without render scripts?

[Anime Viewer]

So you can't find a consistent way to make it happen?
What happens if you try it without render scripts?

I have been running 6-7 files in the playlist over and over again with render scripts set to none, and have yet to have the issue occur. I'll keep testing, and report if anything changes. If ryrynz's theory is correct and the issue is in some way related to it switching between Fullscreen and Windowed modes, but its not occurring with scripts off its probably not the debanding script. That would leave the ChromaScaler and SuperRes scalers. I would guess I'm a rarity running the two combined. I have a feeling other people that have encountered the issue probably weren't running ChromaScaler, and were more likely to be using SuperChromaRes with SuperRes. That would likely mean SuperRes is the common denominator. Something else is would likely be tied to the issue otherwise more people running SuperRes would have encountered the issue.

ryrynz,
What if any scripts did you have running when you noticed the issue on your system?

[ryrynz]

What if any scripts did you have running when you noticed the issue on your system?

Wasn't running any.

[Zachs]

Quick question: With the new Conditionals, can I use the other Renderer properties as well, like InputFormat? Turning off debanding for P010 and Y410 would be useful, I rarely if ever need it for 10-bit vids.

Done.


if SourceBitDepth == 8 ... // turn on debanding

[Zachs]

Hi guys,

Starting from v2.30, MPDN's installer no longer installs MPDN extensions as there's now a separate installer. This makes more sense as they have different release cycles.

For end-users, you only need to download the MPDN installer (pick either 32-bit or 64-bit edition, or both if you like) and the MPDN extensions installer. Install MPDN first before you install MPDN extensions. If you have both 32-bit and 64-bit editions installed, the MPDN Extensions installer will install to both editions (you can change this in the installer).

Cheers.

[Anime Viewer]

Download Notes:

Download the software certificate and install it (extract all files and run InstallCA.bat - you need to do this just once) on your system to let Windows trust software from zachsaw.com (i.e. MPDN). This also prevents you from getting desensitised from the unknown publisher prompt each time you run MPDN and stops other apps from masquerading as software from zachsaw.com. It also stops people with malicious intents from altering MPDN and making it available online as an alternate download. Make sure you only download the certificate from www.zachsaw.com/downloads/ZachSawDotComCertInstaller.zip.

Doesn't look like it exempts the extension installer (https://github.com/zachsaw/MPDN_Extensions/releases/download/v1.4.1/MPDN-Extensions_v1.4.1_Installer.exe) from the warnings.

[Zachs]

Oh the extensions Installer isn't a software by zachsaw.com. I'll have to get an open source cert.