madVR - high quality video renderer (GPU assisted)

[TheBreen]

madVR noobie here. Could someone please tell me what I have done wrong to produce these artifacts -- a movie that looks fine with VMR-9 looks as follows with the latest MPC-HC and madVR versions (default renderer settings; ATI 4850, WinXP):

http://i54.tinypic.com/aeobpd.jpg

[robpdotcom]

Looks like you're not outputting YV12 from your decoder. What codec/decoder are you using?

[dansrfe]

madVR REQUIRES YV12. If you don't supply it, it will spit on the decoder's face and reject the video and eventually force Video Renderer to take over.

[robpdotcom]

Ooops. I stand corrected.

It still looks like what I've seen when I used a decoder that, for whatever reason, wasn't compatible (Cyberlink Video Decoer, I believe). So, you're right, it must have been outputting YV12, but it looked just like what TheBreen's seeing.

[TheBreen]

I have ffdshow_rev3326_20100319_clsid handling most decoding. Is this something I should update or tweak?

I am getting these artifacts with all Blu-ray movies using madVR, but not using other renderers.

[Razoola]

That artifact across the bottom is a bug in madVR that happens sometimes when swapping from exclusive mode to windowed mode when the slider is used.

[ajp_anton]

That artifact across the bottom is a bug in madVR that happens sometimes when swapping from exclusive mode to windowed mode when the slider is used.
And it happens if and only if MPC-HC's internal decoders are used AND if the video is not mod16. Seen it with h264 and wmv so far.

[oldseeker]

Tried madVR (0.34) for the first time a few days ago, with my preferred player, kmp (3.0.0.1439). Thought it was just another renderer, was thinking like "How different could it be?"

I was just blown away by the quality. Nothing even comes close, nvidia purevideo, Haali or HQ VMR, even with very carefully tuned postprocessing.. there are just ZERO artifacts, no matter what you resize to. Amazing.

MadVR is breathtaking to the point I can overlook not having OSD messages from kmp (would that ever be possible technically?), but the thing is - EVERY time I pause, it crashes.

I've tried different decoders (2 internal decoders, external ffdshow), different audio renderers, disabled audio completely, different media types, super/high speed mode, different source filters, different video renderer settings, older kmplayer versions. It just dies on me every time I pause, no matter what I do.

I've searched the forums and teh interwebs high and low, I don't see any references to crashes on pause.

Any insight whatsoever would be greatly appreciated.
Thanks.

[leeperry]

It just dies on me every time I pause, no matter what I do. [..] Any insight whatsoever would be greatly appreciated.
try 0.32 and see if there's any difference: http://www.videohelp.com/tools/madVR/old-versions#download

otherwise, there's a debug version of mVR in the ZIP file, rename it and post the logs here(they'll be on your windows desktop) :)

I did use mVR in KMP a while ago, and it worked fine.

[oldseeker]

That was quick :)
I've tried 0.32, crashes on pause, same crash as before.
Also tried various mVR settings again, fullscreen exclusive on/off, different flushing/buffer options, etc..

Here's a link (http://www13.zippyshare.com/v/67917883/file.html) to two debug logs, one for 0.32, one for 0.34.
Hope it's something silly and easily fixed.

Thanks!

[trainspotter]

Hi,

I wanted to post the following bug report quite a while ago (when I initially registered). But before I received posting rights, RL got busy and I had soon forgotten about it (and kept using madVR 0.23).

Here's my initial post I saved from mid August:

Hi folks,

I've been following this thread on a daily basis until a couple of weeks ago.
Now that I am back from vacation I was delighted to see that 4 new versions and the long hyped exclusive mode have been released. Thanks for doing such a wonderful job, madshi :)

Sadly no version containing the new exclusive mode feature (v0.24+) seems to work on my primary machine.

Some parameters for my environment:
Geforce 8800 Ultra
Windows Vista x64 (fully patched)
Latest Nvidia WHQL Driver (258.96)
Latest MPC-HC Nightly
1280x1024@100hz on primary display, 1920x1080@24hz on secondary display


Opening any video file using the current madVR build will cause MPC to hang indefinately at "Opening...". Trying to close MPC conventionally will only result in MPC lurking in the background.
Even trying to kill it via task manager won't work - at some point (probably after some kind of timeout is sent), the MPC task just disappears.
If I wait long enough (10+ minutes) without trying to close MPC, I get the following error message and am able to end the MPC task conventionally:
http://img714.imageshack.us/img714/509/screenshotvvm.png

Happens on all files and all codecs. And in single and multi monitor setups.

I've created a debug log (http://rapidshare.com/files/438313911/madVR_-_log.rar) and uploaded it to help track down the problem.
Let me know if there is anything else I could do.


Regards and keep up the good work,
trainspotter


Now I wanted to update madVR and I am experiencing the same behavior with the most recent version of madVR.
Little has changed in my environment since my above quoted first post.
Updated to the latest NVIDIA WHQL drivers (260.99), using the current mpc-hc nightly (1.4.2786.0) and updated to the latest directX.
madVR 0.34 shows the same behavior as I described above.
Here's (http://rapidshare.com/files/438314308/madVR_0.34__-_log.rar) a log using the most recent madVR debug build.


Any advice on how to fix this would be appreciated.
Und mach weiter so, madshi! :)


Regards,
trainspotter

[rmp459]

1) Do any of you actually get playback of a movie with the osd not showing a single dropped frame? Even with my monitor at like 23.976hz and my movie at like the same exact fps i still get the occasional dropped frame and just starting/stopping the video or opening the osd will report dropped frames.

My rendering times on the gt220 are over 40ms and as a result the video comes in behind the audio. Its so annoying.

Im running mpc hc with coreavc and madvr. Coreavc uses my gtx 470 for decoding and madvr renders on a gt220 outputting to my hdtv. Im running catmull-rom x2 and softcubic 70 for the downscaling. Seems to be no bottleneck but it does this strange way of doing things since i run 3 monitors but use ultramon to activate a display profile that deactivates all but my hdtv. The gtx470 is for gaming on the other two and dealing with nvidia control panel each time would be a nightmare. Also i run aero off with fullscreen exclusive mode.

2). Is there a way to get mpc hc full screen autochange to work with full screen exclusive mode? Or am i going to have to do this manually each time? (change my refresh rate per film/video)



Thanks.

[scoz]

1) Do any of you actually get playback of a movie with the osd not showing a single dropped frame? Even with my monitor at like 23.976hz and my movie at like the same exact fps i still get the occasional dropped frame and just starting/stopping the video or opening the osd will report dropped frames.


I'm pretty sure everybody gets those dropped frames, thats just life? I don't think it really matters?


My rendering times on the gt220 are over 40ms and as a result the video comes in behind the audio. Its so annoying.


I'm thinking theres something else causing the audio desync issues. The renderer will drop frames to catch up to audio, delay frames to let audio catch up. If neither of those are happening then the renderer thinks its sync'd with the audio, so its thinking wrong for some reason. Have you tried using reclock as an audio renderer? Are you outputting audio by HDMI, optical, analog?


Im running mpc hc with coreavc and madvr. Coreavc uses my gtx 470 for decoding and madvr renders on a gt220 outputting to my hdtv. Im running catmull-rom x2 and softcubic 70 for the downscaling. Seems to be no bottleneck but it does this strange way of doing things since i run 3 monitors but use ultramon to activate a display profile that deactivates all but my hdtv. The gtx470 is for gaming on the other two and dealing with nvidia control panel each time would be a nightmare. Also i run aero off with fullscreen exclusive mode.


Would it be possible to give ffdshow a try? In theory they are both reference decoders(I think core has been accused of not de-blocking if it's tight on time? maybe thats old?) so they will give the same quality. Many people have reported problems using core.


2). Is there a way to get mpc hc full screen autochange to lwork with full screen exclusive mode? Or am i going to have to do this manually each time? (change my refresh rate per film/video)


I think madshi is planning on adding his own automatic change stuff in the future and that madvr currently will not work with MPC's built in autochanger.

Note that I don't know if I'll be able to help you at all with specific debugging ideas, I'm just reiterating the common suggestions by people who do actually know what they are talking about, with the thought that if you go ahead and report back with those results they may be able to help you quicker :)

[dansrfe]

madshhhhiiiiiiii......come baccckkkkkkk :)

[noee]

rmp459, as scoz responds, those drops are normal, cost of doing business. However, fwiw, I have run a movie through, start to finish, with 3 total drops, all happening at the very beginning (load/start). I'm back to Reclock and my playback monitor at 24Hz, aero off, fsX mode off. Others have reported issues with b34, however, in my setup, it is rock solid.

One thing to try would be to use windowed-mode only and try tweaking both scaling and flush options, see what happens. Madshi has made it clear in previous posts that the current fsx mode code path is a placeholder, in addition, he will be adding features to madVR to support resolution/refresh change. Anyway, at least in my case, FSX mode with MPC-HC in a dual-mon environment has never worked reliably.

HD2600XT, CCC10.12, Win7 x64, Aero OFF, dual-mon (primary 60Hz, sec 24Hz, both 1920x1080)
HMS/FFDshow/ReClock/madVR (windowed)/MPC-HC
3Dlut OFF|lumaDOWN=SoftCubic50|lumaUP=SoftCubic50|chroma=Bicubic75
flush|flush and wait(sleep)|don't|don't

[rmp459]

rmp459, as scoz responds, those drops are normal, cost of doing business. However, fwiw, I have run a movie through, start to finish, with 3 total drops, all happening at the very beginning (load/start). I'm back to Reclock and my playback monitor at 24Hz, aero off, fsX mode off. Others have reported issues with b34, however, in my setup, it is rock solid.

One thing to try would be to use windowed-mode only and try tweaking both scaling and flush options, see what happens. Madshi has made it clear in previous posts that the current fsx mode code path is a placeholder, in addition, he will be adding features to madVR to support resolution/refresh change. Anyway, at least in my case, FSX mode with MPC-HC in a dual-mon environment has never worked reliably.

HD2600XT, CCC10.12, Win7 x64, Aero OFF, dual-mon (primary 60Hz, sec 24Hz, both 1920x1080)
HMS/FFDshow/ReClock/madVR (windowed)/MPC-HC
3Dlut OFF|lumaDOWN=SoftCubic50|lumaUP=SoftCubic50|chroma=Bicubic75
flush|flush and wait(sleep)|don't|don't


It has been a nightmare for me. Im getting it more stable...

When i watch movies on my HDTV i switch display profiles and just run my single HDTV... other monitors are off. It seems the lower i run my refresh rate (ie: 23.976), the slower the rendering is with fullscreen exclusive mode... almost 45ms.

GPU usage doesnt exceed 60%.

Another option I had was to turn aero on... and I wrote a AHK exe that sit between the video files and my MPC-HC.exe and it simply launches mpc hc w/ the %1% file name. then waits for the file to load. Pauses it, then stops and starts the windows service that handles Aero... then unpauses the movie.. This works flawlessly in terms of allowing MPC HC to change refresh rate and then getting aero's composition rate to actually sync to the current film/video...

Ill have to play around with it more tonight, but last night I was getting horrible video delay compared to the audio and then it was taking almost 30 seconds for some files to even start playing in mpc hc... time to wipe and start fresh with my codecs and decoders and see where I can end up...

I feel like im getting closer though... When fullscreen exclusive mode is a quick as windowed mode with its current performance and madVR can control refresh rates... were in for a real treat.

[tdoll]

I used to use madvr on mpc-hd with my hd4650 catalyst 10.11. It worked like a charm. I never had a better movie picture.

For some other reasons I switched to a hd5770 catalyst 10.11. It seems to be that something is wrong in this setting. I get some vertical blanked out lines once in a while in the movie. I use two screens. One for the pc and a tv (50 inch plama) to display the movie. I even have this problem when i play the movie on the pc monitor. What I mean is that even in that setup the wallpaper on the plasma shows the lines.

If I use a different renderer as e.g. evr-cp everything works fine.

Any ideas?:confused:

[pankov]

tdoll,
I had the same problem when I used the ATI Overdrive, which I personally recommended to other users in the past so they can force the ATI card in "high power/frequency (both clock and memory)" mode.
Sadly when I installed Catalyst 10.11 this lead to the same "blanking" lines showing randomly (or not so randomly - I found a way to reproduce the problem) just as the ones that you talk about. The only solution was to disable the ATI Overdrive.
Recently I installed Catalyst 10.12 but forgot to test if it fixed this issue. I'll do it tonight and report back.

[tdoll]

pankov,

thanks a lot. I just switched off overdrive and everything seems to work properly again. By the way I tested it with 10.12 and it did not solve the problem. Anyway I'll keep on trying.

Thanks again

[rmp459]

totally just wacked out my filters in MPC trying to play around with decoders to get MadVR running well... now all my audio filters when I try to play a movie say "Failed to render pin out"... Any idea how to fix that easily ? Ive tried to clear out the registry and such but cant seem to reset it... Even the internal MPC filters wont work.

[tdoll]

As far as I know there is an reset option during the installation process of mpc-hc. Just download the latest version from xvidvideo.to and install it. Make sure you click on reset (might be called differently). Afterwords the internal filters should work.

Externally I just use ffdshow and Ac3Filter. The less codecs you use the better it is. I suggest to deinstall everything and install. Afterwords just install ffdshow and maybe Ac3Filter.

By the way. In the latest versions of mpc-hc the internal codecs seem to be very good to. I tend to use the external filters not that often any more.

[rmp459]

As far as I know there is an reset option during the installation process of mpc-hc. Just download the latest version from xvidvideo.to and install it. Make sure you click on reset (might be called differently). Afterwords the internal filters should work.

Externally I just use ffdshow and Ac3Filter. The less codecs you use the better it is. I suggest to deinstall everything and install. Afterwords just install ffdshow and maybe Ac3Filter.

By the way. In the latest versions of mpc-hc the internal codecs seem to be very good to. I tend to use the external filters not that often any more.

Ended up just reformating. Now using internal codecs for everything except aac->ac3 transcoding for multi ch aac audio. So far so good.

[madshi]

see below my old post with the results testing v0.32 and the exact line where the application freezes.
I didn't notice this before. It seems to me that the DirectShow class wrappers offered by MS internally have some kind of critical section which is responsible for this freeze. Right now I've a hard time figuring out, though, which other thread is holding the critical section. I have something in mind for a future version, but it might take a while until I get there...

I have had to go back to version 032, there is something in 034 that seems to not play nice with reclock and for the life of me I cannot create a log because I cannot get the problem to happen in the debug build.
I guess that's probably the new clock anti jitter algorithm. I'll disable it again in the next build, cause I didn't notice any obvious improvements, anyway.

need subtitle and screenshot taking support. I know it's the same thing everyone else has been saying but how much longer.
I generally don't give out time estimates for features that are not planned for the very next build. I don't know how long it will take...

i installed the latest version and when i choose madvr in mpc hc the video doesn't start. it just keeps "loading".
tried using the debug version and it worked once then stopped. anyway i found the log file on the desktop - 30mb 0_O - and thought i'd share some of it

[...]

does it help?
I need the whole log. Please:

(1) delete the log
(2) recreate the problem
(3) close/terminate MPC HC
(4) zip the log (important!)
(5) upload it somewhere

Thanks.

sorry. i already deleted the file :rolleyes:
so instead i retried running with madvr again and here's the new log
Is that the full log? I thought you mentioned something like 30MB? I really need the full log, not just some part of it.

When I enable madvr I am getting a strange green bar at the bottom AS well as everything looks blurry,almost as if the video is trying to convert to 3d.
Which decoder are you using? The internal MPC HC decoders are known to be buggy. Try some other decoder.

Will madVR support multi-lang. in future?
I can help to translate into Japanese, Simp. Chinese, Trad. Chinese!
Maybe later, but too early for that right now.

I occasionally get corruption when using madVR. Switching to EVR causes the problem to disappear.

http://stfcc.org/pics/i/71aee459daab84c59dea406895ee7d6c.png

Log: http://stfcc.org/misc/madVR%20-%20log%204.rar
I'll upload a sample if you need one.
Does the corruption always occur at the same frame? Or is it somewhere else every time? Are you sure that your RAM is alright? :p

1. does hdmi features like deep color, xvycc etc.. make a noticeable difference?
DeepColor means that the dithering noise added by madVR could be made less noticeable. Other than that it doesn't help much. If you do not see the dithering noise without DeepColor, then you'll likely see no difference with DeepColor. Currently madVR does not support DeepColor.

xvycc makes sense only if we get a new content format with a bigger colorspace. For DVDs, TV and Blu-Ray, xvycc makes no sense.

2. Is hdmi 1.4 or newer (1.4a) have any important/significant feature improvement over 1.3/1.3a so that i should emphasize 1.4 while buying a graphics card; if my preference are better colors & quality?
1.4 makes no difference in terms of quality. The main advantage of 1.4 is improved 3D support.

3. If a graphics card natively supports both hdmi & dvi & i use a dvi to dvi cable, will i get the hdmi features like deep color, xvycc etc?
Yes, I think so.

Reinstalled Windows on my HTPC today. Using the latest nvidia drivers and exclusive mode fails every time at high refresh rates now.
The log file says that switching to 1080p72 failed with the Direct3D error information "invalid call". That's not a very helpful error message from Direct3D, I don't really know what went wrong. Are you sure that your display actually supports 1080p72? I'm not sure if single link DVI/HDMI can do that, maybe you need double link DVI for that?

Powerstrip only works up to 4xxx cards, which is why I have a 4850 instead of a 5750.
The latest PowerStrip version seems to support 5xxx now, too. Don't know about 6xxx, though.

Ok, there seems to be another issue:
When I pause a video and then seek to different scenes, upon seek the whole frame goes out of focus and then comes back into focus (kind of when you try to focus with your camera).
This also happens when there is an abrupt change of scenery in the video.
The frame just momentarily becomes hazy and then becomes normal. WTF?!
Never seen or heard of this problem before. I don't think it's madVR's fault.

I have not seen the list of "known problems", so I'll write here.
If the path to madVR are non-Latin characters, the renderer does not work. :(
Will fix that in the next build.

this is a definite bug which I have found. When a dvd is played the first part of it including the fbi warning and such reports to madvr as 25 fps and when it comes to the movie part it is still at 25 fps when I know for a fact ffdshow is outputting progressive ntsc film at 23.976 fps. This in turn leads madvr to start it's correction process of trying to sync to 25fps
The FPS information seen in the OSD has strictly informational purpose, only. The current madVR version does not use it. So whatever problems you're seeing, they can't be caused by the incorrect FPS info. Besides, the FPS is reported by the filters connected to madVR, so the incorrect FPS is not madVR's fault.

Add manual colorimetery selections.
What for?

Fix FPS detection on some DVDs.
Not madVR's fault.

When you move the mouse to the bottom, there's like a half a second delay before the search bar shows up.
The reason for that is that madVR is prerendering the frames to fill the render queue. This is one of the tricks to make sure that smooth motion can be achieved. Now if you move the mouse to the seek bar, the seek bar is drawn immediately by madVR - but only to newly rendered frames. If your rendering queue is full, there are still 8 prerendered frames to be displayed, which don't have the seekbar visible. Of course I could dispose of all prerendered frames when you try to activate the seek bar. But if I do that, you might get some stuttering instead.

I have a problem with my interlaced Display (CRT TV) connected as second Display (Custom Interlaced Resolution created with Powerstrip). Downscaling does work perfect on the progressive Display (no banding, sparkling) but on the interlaced TV i got banding, sparkling (dont know how to call it) but only in Motion, no matter what Resizer is used (clear visible on Simpsons HD Recording).
Is this a bug ?
Is there a way to avoid it ?
Does any filter in your processing chain actually deinterlace the content? The problem is that you cannot properly downscale interlaced content. If you want to downscale interlaced content, you have to deinterlace it first, then downscale it, then reinterlace it for output. The current madVR version itself does not support deinterlacing, so you need to deinterlace in a filter before madVR. E.g. the ffdshow raw processing filter should be able to do that.

CoreAVC + DScaler IVTC Mod
Is Simpsons HD MPEG2 or h264? The DScaler IVTC mod should do the deinterlacing, if the content is film based, but the situation is different with h264. Furthermore if the content is video based, DScaler IVTC will not properly deinterlace.

Maybe a screenshot would help?

Isn't the YUY2 output "more" thrusty on the colour than YV12 ?
I don't know what "thrusty" means. Anyway, YV12 is the native format of DVD, Blu-Ray and broadcasts. So basically using YV12 means that the decoders are only decoding and not doing any processing. If the decoders would output YUY2 to madVR, the decoders would have to perform chroma upsampling. That's a bad thing, because madVR can do that in better quality.

For me, when switching from windowed to fullscreen there is a delay. It's not always the same length, sometimes just half a second, sometimes 2-3 seconds. During this time only a blank screen is shown.
I have a delay with my nVidia 9400 mainboard (Windows 7 x64), too, but not with my ATI HD3850 (XPSP3). I rather think that the OS plays the most important role in how long the delay is. Maybe I can find a way to get it down reliably to half a second. 2-3 seconds is really quite long. But I don't really know if I can do anything.

I don't know if this is related, but when switching from fullscreen to windowed for a very short time I see the frame that was shown when I entered fullscreen mode. After that the video continues normally in windowed mode.
That's a known problem and there's probably nothing I can do about it.

[madshi]

Also when I send Windows to hibernation while a video is open in MPC (not neccessarily playing) on restart this error message is shown in the MPC window (red on black ground)
This will be fixed in a future build.

With the script active, CPU load is under 20% (and only on 4 to 8 cores. GPU is under 40% (had to switch from a 5450 to a 5750 since the 5450 could not handle the load), yet I have about 30+% (300 over 10 seconds as an example) of all frames being dropped.
The GPU consumption being at 40% does not mean that the GPU can't be the bottleneck. Due to how madVR currently works, the GPU is rarely ever used at 100%. I might be able to improve that in a future madVR version, when having implemented the final exclusive mode rendering path.

Which is your display resolution? Does madVR have to scale? If so, please try to use Bilinear as a test. Does the solve the dropped frames? Also, do you use exclusive mode? If so, try to get rid of any "flush & wait" options in madVR settings. I know they are in the "windowed mode" section, but currently they do also affect the exclusive mode rendering, and for exclusive mode I've found that not waiting is better. Replace the "flush & wait" with only a simple "flush". Maybe that solves the problem?

What are the best madvr settings to use?
Matter of taste, decide for yourself.

PS: a request for madshi. Could you make the seekbar size configurable? or at least an option for a larger seekbar. On a large display sitting 8' away, it can be hard to see/click on that thing.
That's on my to do list.

Will madVR ever by able to take screenshots somehow?
Yes, sooner or later.

When I use tvlevels in madVR I get grey blacks although I would not expect this to happen with my configuration: beamer set to HDMI range normal (not expanded), MPC-HC render settings set to use tv levels, decoder (if setting available) configured to use tvlevels.

When I configure madVR to use PC levels the grey's become black again... is this correct?

Configuration: Windows 7 x86, Nvidia GT220, beamer connected as monitor 2 on extended desktop.
Please make sure you have created "custom resolutions" in your NVidia control panel. NVidia tends to convert all data to video levels behind madVR's back, if you use one of the preset HD/SD modes. By creating custom resolutions you can fix that.

This could be due to a bug in PotP, I'll run more tests on my previously problematic files: http://www.sat-universe.com/showpost.php?p=711140&postcount=45
Results?

YUY2 is 16-bit per pixel, YV12 12-bit per pixel. mVR only accept YV12 input, does all the computing in 16bit and then dithers to 8bit RGB.
Careful, it's easy to confuse people. YUY2 is 16-bit per pixel, YV12 is 12-bit per pixel. That's all correct. However, dithering is done to 8bit *per component* RGB. Without saying that it's "per component" people might think that YV12 has a higher bitdepth than madVR's output. So:

YV12: 12bit per pixel (native DVD/Blu-Ray/broadcast format)
YUY2: 16bit per pixel
madVR math: 96bit+ (floating point) per pixel
madVR buffers: 48bit (integer) per pixel
madVR output: dithered 24bit (integer) per pixel

hm if madvr only accepts YV12 input, how come I could see on a reclock screenshot "23.976 fps, 1920x1080p, YUY2 16 bits" ?
I've no idea, ask the Reclock programmer... :)

It's hard to believe that an original->very_big->target can be any better than an original->target resize.
Doing "original->big->target" alone does not make any sense. However, doing "original->big->postProcessing->target" is a different matter. Depending on the post processing algorithms, doing the post processing in a higher than target resolution can be beneficial.

I disagree. With SoftCubic100, the lower resolution chroma is blurred much more than any luma information, so red/blue lines will be thicker and less saturated than green for example.

From my extensive testing, Bicubic75 produces the best results when chroma detail is compared to luma.
You have to decide whether chroma detail is more important or lack of chroma aliasing artifacts. If you pick Bicubic75, you'll have more detail, but also more artifacts and more aliasing.

One more thing to consider: Currently madVR resizes the chroma channel directly from source -> target resolution, using the chroma upsampling algorithm. In a future version I plan to change that. I plan to upsample chroma from chroma source -> luma source resolution (2x in X and Y direction). Then I plan to use the luma upsampling algorithm for the remaining scaling. I hope that this will allow a better compromise between chroma detail and lack of artifacts/aliasing. So once I've changed that, you'll have to recheck which algorithms work best for you.

spline has an 'artificial' look to it due to everything being drawn as curves.
Sorry, but that's totally incorrect. All the resampling algorithms used by madVR are doing the same math. They simply, for every target pixel, build a weighted average of the surrounding source pixels. The only difference between the various resampling filters is which weights are applied to which pixels. Spline does not try to draw everything as curves. Spline just has different weights than Lanczos. And Lanczos has different weights to Mitchell-Netravali etc. But in their core, they all do exactly the same math.

madshi is always full of mysteries, but he recently said that maybe the most faithful to the source chroma wasn't the softest...need to chew on this, as he didn't seem too willing to back up this statement :D
What would be the fun, if I spilled all the beans? I want to keep you guys excited about the future. There are some things planned which you don't know yet and which you'll probably like. Isn't it much more fun if you don't know what exactly I've planned? At least it's more fun for me. :)

I just noticed that madVR upscales videos if display width (1280px in my case) is the same as video width (1280px).
Is there any need for upscaling in this kind of situation?
The media player decides whether any upscaling is done. madVR does what the media player says. Anyway, how did you come to the conclusion that madVR upscales in that specific situation? What did the madVR OSD say? If you switch your media player to "100%" output, madVR should not scale.

FYI there a a group of us using JR Media Center pushing for the abilility to use madVR as the Video Renderer but it looks like it is not going to happen as one of the devs over there have quoted:

http://yabb.jriver.com/interact/index.php?topic=60810.0

I don't know if madshi is interested in following up with the JR Folk on this.
I'm generally interested, but I don't really feel like registering for their forum right now. Can you post a message there for me? If the JR Media Center programmer(s) want to contact me to work out whatever problem there is, they can directly email me at madshi (at) gmail (dot) com.

a question regarding bitdepth of LCD TVs: I came across some TVs which have 8-bit and some with 10-bit bitdepth. now I wonder is that actually enough to display video put out by madVR without any loss of quality?
A TV with native 8bit support should be good enough, if it can really show those 8bit natively, without banding. True 10bit could be useful for a future madVR version, but I doubt you would see any difference in real life. The only difference you could possibly see would be a slightly lower dithering noise.

Are you effing crazy??? madshi spent countless MONTHS perfecting this renderer to output in 16-bit all in the name of quality. 16-bit output is default in madVR then you can change that to 10-bit if your computer is too slow to output 16-bit.
Math is done in 32bit+ per component (96bit+ per pixel). Temp buffers are 16bit per component (48bit per pixel). The video data is dithered down for final output to 8bit per component (24bit per pixel).

On a TV (Sony LCD), i'm NOT supposed to see the black bars below 16, right?
Make sure you create custom resolutions for all output modes you want to use on your TV. Only this way you can make your nVidia card output untouched video data.

madVR noobie here. Could someone please tell me what I have done wrong to produce these artifacts
You're probably using the MPC HC internal decoders, right? These are known to be buggy. They work with other renderers, because other renderers don't force the decoders to output the YV12 format (which madVR does). Use a different decoder, that should fix the problem.

That artifact across the bottom is a bug in madVR that happens sometimes when swapping from exclusive mode to windowed mode when the slider is used.
I've never seen that yet. Nor do I remember to have read about it. Is the problem reproduceable? How?

I was just blown away by the quality. Nothing even comes close, nvidia purevideo, Haali or HQ VMR, even with very carefully tuned postprocessing.. there are just ZERO artifacts, no matter what you resize to. Amazing.
Glad you like it!

EVERY time I pause, it crashes.

I've tried different decoders (2 internal decoders, external ffdshow), different audio renderers, disabled audio completely, different media types, super/high speed mode, different source filters, different video renderer settings, older kmplayer versions. It just dies on me every time I pause, no matter what I do.

I've searched the forums and teh interwebs high and low, I don't see any references to crashes on pause.
Yeah, I also don't remember anyone having posted about a similar problem. I don't really know what's causing it. Unfortunately your logs don't help in this case. For crashes the logs are often not much use.

One suggestion I have would be to try a completely different media player, e.g. MPC HC. Does the problem also occur with that?

The final solution would be for me to implement some better crash analyzation methods. I plan to do that sooner or later, but I don't know how long it will take...

Any advice on how to fix this would be appreciated.
Hmmmm... Does it help if you disable the exclusive mode?

Another thing worth trying would be to use a lower refresh rate, just a test, only, of course.

Do any of you actually get playback of a movie with the osd not showing a single dropped frame? Even with my monitor at like 23.976hz and my movie at like the same exact fps i still get the occasional dropped frame and just starting/stopping the video or opening the osd will report dropped frames.
A couple of frame drops at the start of the movie, or after a seek are to be expected. With a perfectly optimized HTPC setup it should be possible to get along without any additional frame drops during normal movie playing, though. So if you have occasional frame drops, there's still room for improvement. The first thing to do would be to find out *why* there are drops. For that I'd recommend that you check the madVR OSD to see how closely your GPU refresh rate matches the movie frame rate.

My rendering times on the gt220 are over 40ms and as a result the video comes in behind the audio. Its so annoying.
For exclusive mode try replacing all "flush & wait" options in the madVR settings with a simple "flush". Does that help?

[Razoola]

I've never seen that yet. Nor do I remember to have read about it. Is the problem reproduceable? How?

Good to hear from you madshi, have a great christmas and new year incase I don't get a chance again.

The curruption I was talking about was the yellow strip across the bottom of the video and not the incomplete taskbar. I can get it to happen by simply going quickly from exclusive mode to window mode after a video is opened in MPC-HC (straight to fullscreen exclusive mode). It does not happen every time however.

[madshi]

have a great christmas and new year
Thanks, I wish the same to you - and to everyone else, of course.

The curruption I was talking about was the yellow strip across the bottom of the video and not the incomplete taskbar. I can get it to happen by simply going quickly from exclusive mode to window mode after a video is opened in MPC-HC (straight to fullscreen exclusive mode). It does not happen every time however.
Does it happen with all decoders or only with one specific? I've never seen this happen on my PC yet.

[Razoola]

Thanks, I wish the same to you - and to everyone else, of course.


Does it happen with all decoders or only with one specific? I've never seen this happen on my PC yet.

I have just tried for the last 20 minutes to get it to happen and I can't. I take it this means it some other issue not cause causing it.

[madshi]

Does anybody have a Radeon 6xxx card? Could you please check if the latest PowerStrip version fully supports it?

[madshi]

I have some problems with madvr.
When i open it starts with black screen and then it crashed.
I have YV12 for output.
i tried the debug version and i got that log.
Unfortunately I can't see the reason for the crash in the log, so it's hard for me to help. You could try different splitters, decoders or a different (or more recent version of your) media player.

[madshi]

I have not seen the list of "known problems", so I'll write here.
If the path to madVR are non-Latin characters, the renderer does not work. :(
I can't reproduce this problem. I've put madVR into a folder with some russian chars in it, installed it from there, and it works just fine for me.

[leeperry]

Results?
I can't get 0.32 to lock up anymore(and I tried really hard :D), I guess it was a PotPlayer problem indeed.

Careful, it's easy to confuse people. YUY2 is 16-bit per pixel, YV12 is 12-bit per pixel. That's all correct. However, dithering is done to 8bit *per component* RGB. Without saying that it's "per component" people might think that YV12 has a higher bitdepth than madVR's output. So:

YV12: 12bit per pixel (native DVD/Blu-Ray/broadcast format)
YUY2: 16bit per pixel
madVR math: 96bit+ (floating point) per pixel
madVR buffers: 48bit (integer) per pixel
madVR output: dithered 24bit (integer) per pixel
Yep, the TMDS transmitter that encodes to DVI/HDMI <1.3 outputs RGB24 anyway.

I've no idea, ask the Reclock programmer...
It was a mix up w/ an EVR screenshot, Reclock doesn't say YUY2 when using mVR...for very obvious reasons ;)

[janos666]

Does anybody have a Radeon 6xxx card? Could you please check if the latest PowerStrip version fully supports it?

I don't think it works.
I have a HD5850. It wasn't supported until the last PowerStrip release but this latest version doesn't seem to work neither. I tried to force some custom refresh rates but nothing happened.
May be it works with older drivers (<10.12) or other outputs (I am using DP right now).

[121sea]

YV12: 12bit per pixel (native DVD/Blu-Ray/broadcast format)
YUY2: 16bit per pixel
madVR math: 96bit+ (floating point) per pixel
madVR buffers: 48bit (integer) per pixel
madVR output: dithered 24bit (integer) per pixel

If you increase madvr's bit per pixel(double it for example) will that result in a visible improvement in video quality or is that a waste of computer resources?

[trainspotter]

Hmmmm... Does it help if you disable the exclusive mode?

Another thing worth trying would be to use a lower refresh rate, just a test, only, of course.

In order to edit the madVR settings I have to wait for the timeout / D3D error. Any settings I make once that timeout has occurred, only seem to be saved for the session - after restarting and waiting for the timeout to edit again, the defaults are back.

Does madVR still save its settings to an ini file, so I could manipulate that file manually? Or is there any other possibility to get to the config editor without having to run madVR first?

Same behavior (as I described in my initial post) with a lower refresh rate (60hz) on a single monitor setup.
Would logs of these tests be of any use to you?


€dit: Strange observation - I am getting more and more confused by the minute..
If I wait for the timeout / D3D error and try to play the same file with the already opened mpc-hc, all madVR features (including exclusive mode) seem to be working.
Any other file opened in that instance of mpc-hc afterward works flawlessly.

I have created a log ("http://rapidshare.com/files/438933788/madVR_-_log__automatic_playback_after_timeout_.rar) of that test. This time I did not get the D3D error but the movie started playing right away after said timeout.
Please let me know if there is anything else I could do to help debugging.

[madshi]

I can't get 0.32 to lock up anymore(and I tried really hard :D), I guess it was a PotPlayer problem indeed.

It was a mix up w/ an EVR screenshot, Reclock doesn't say YUY2 when using mVR...for very obvious reasons ;)
Good, two problems less I need to look into.

I don't think it works.
I have a HD5850. It wasn't supported until the last PowerStrip release but this latest version doesn't seem to work neither. I tried to force some custom refresh rates but nothing happened.
May be it works with older drivers (<10.12) or other outputs (I am using DP right now).
Too bad... :(

If you increase madvr's bit per pixel(double it for example) will that result in a visible improvement in video quality or is that a waste of computer resources?
Which bitdepth? The output bitdepth? You mean from 24bit per pixel up to e.g. 48bit per pixel? As I said earlier, increasing the output bitdepth will lower the dithering noise floor. Apart from there there should be no change. If you don't notice the dithering noise with the current madVR 24bit output, then any higher output bitdepth will likely not bring any visible improvements for you.

€dit: Strange observation - I am getting more and more confused by the minute..
If I wait for the timeout / D3D error and try to play the same file with the already opened mpc-hc, all madVR features (including exclusive mode) seem to be working.
Any other file opened in that instance of mpc-hc afterward works flawlessly.

I have created a log ("http://rapidshare.com/files/438933788/madVR_-_log__automatic_playback_after_timeout_.rar) of that test. This time I did not get the D3D error but the movie started playing right away after said timeout.
Please let me know if there is anything else I could do to help debugging.
Hmmmm... According to the log the timeout occurs during one of the following actions (not sure which):

(1) loading of "madHcNet.dll"
(2) enumeration of monitors
(3) initialization of madHcNet network
(4) start of "madHcCtrl.exe"

One thing you could try is to start madHcCtrl.exe manually. Wait until the madVR tray icon is visible (in case you haven't disabled it). Then start madVR. Does the timeout still occur? If yes, the problem is likely not (4). Otherwise it's probably (4). Also try disabling your firewall as a quick test.

It the problem is (2) or (3), it might be difficult to find the real cause for the timeout. Because no other user seems to have this problem, I fear it might be some kind of OS problem, which may only be cured by reinstalling the OS. :eek: Hopefully not.

[121sea]

Which bitdepth? The output bitdepth? You mean from 24bit per pixel up to e.g. 48bit per pixel? As I said earlier, increasing the output bitdepth will lower the dithering noise floor. Apart from there there should be no change. If you don't notice the dithering noise with the current madVR 24bit output, then any higher output bitdepth will likely not bring any visible improvements for you.

the math bitdepth if that's even posible whit current video cards

[trainspotter]

Hmmmm... According to the log the timeout occurs during one of the following actions (not sure which):

(1) loading of "madHcNet.dll"
(2) enumeration of monitors
(3) initialization of madHcNet network
(4) start of "madHcCtrl.exe"

One thing you could try is to start madHcCtrl.exe manually. Wait until the madVR tray icon is visible (in case you haven't disabled it). Then start madVR. Does the timeout still occur? If yes, the problem is likely not (4). Otherwise it's probably (4). Also try disabling your firewall as a quick test.

It the problem is (2) or (3), it might be difficult to find the real cause for the timeout. Because no other user seems to have this problem, I fear it might be some kind of OS problem, which may only be cured by reinstalling the OS. :eek: Hopefully not.

It has to be something that was introduced together with the exclusive mode, as all releases I tried since 0.24+ seem to show the same behavior.
So I doubt it's because I have my firewall blocking the network function, as version 0.24 did not have that function and did not work either, but I will try and test to rule out all possibilities.
Another thing I don't get is why I don't seem to be able to save my madVR settings(?).

And you are right - it could be one of those random OS mess ups. Even though I can't recall any major changes I have made that would cause madVR to glitch, but as an IT professional I know that hardly anyone ever recalls those kind of changes until they are pointed on them specifically ;)
When I get a chance I'll try and see if I get the same behavior with the same machine under another OS / a live OS.

€dit: I went through your list of possible causes.

(1) Any advice on how to rule this possibility out?
(2) Just did a couple of tests using a single monitor setup. No change.
(3) Even if I disable both my fw and av, I get the same behavior :/
(4) If I start madHcCtrl.exe manually it indeed takes forever until the task bar icon appears (even though the process starts immediately and does not cause any load). But madVR still shows the same behavior (timeout and afterward flawless playback).

[madshi]

the math bitdepth if that's even posible whit current video cards
The math is done at the internal bitdepth of the GPU pixel shaders. That's 32bit per component (96bit per pixel) with most newer DX9 GPUs, or even higher with some of the latest GPUs. Increasing this bitdepth any further would surve no purpose, because 96bit per pixel is already pretty much overkill.

It has to be something that was introduced together with the exclusive mode, as all releases I tried since 0.24+ seem to show the same behavior.
Are you 100% sure that the problem occurs with 0.24, and does not occur with 0.23? If so, that would be interesting...

Another thing I don't get is why I don't seem to be able to save my madVR settings(?).
Not sure on this one. Saving depends on madVR still working ok in the media player. Otherwise saving won't work.

(1) Any advice on how to rule this possibility out?
I'll add some more log outputs to the next build. That should help figuring out which step exactly causes the freeze for you.

(4) If I start madHcCtrl.exe manually it indeed takes forever until the task bar icon appears
That's already a good indication! On all my PCs the tray icon appears *instantly*, right when starting the exe, without any noticeable delay. madHcCtrl.exe, before showing the tray icon, does some of the same network initialization stuff that madHcNet.dll is doing. So my current guess is that one of those network calls hangs on your PC, for whatever reason...

[Snowknight26]

Does the corruption always occur at the same frame? Or is it somewhere else every time? Are you sure that your RAM is alright? :p

I can't reproduce it anymore on the same system, too, so I'll consider it a non-issue.

[madshi]

I can't reproduce it anymore on the same system, too, so I'll consider it a non-issue.
Great!

[janos666]

Too bad... :(

I tried the HDMI connection. The custom refresh rate forcing works with this output! (At least on the output side, because my display does internal FRC to feed the fixed RR panel. :()

[6233638]

The log file says that switching to 1080p72 failed with the Direct3D error information "invalid call". That's not a very helpful error message from Direct3D, I don't really know what went wrong. Are you sure that your display actually supports 1080p72? I'm not sure if single link DVI/HDMI can do that, maybe you need double link DVI for that?The display does support 1080p72 however the nvidia drivers typically limit you to 1080p60 over HDMI. For some reason with that specific build I was able to create custom resolutions above that. (perhaps they broke the check?) It seems like this is probably an nvidia/windows issue rather than madvr. I no longer have that display now so it is not a problem for me any more.

One more thing to consider: Currently madVR resizes the chroma channel directly from source -> target resolution, using the chroma upsampling algorithm. In a future version I plan to change that. I plan to upsample chroma from chroma source -> luma source resolution (2x in X and Y direction). Then I plan to use the luma upsampling algorithm for the remaining scaling. I hope that this will allow a better compromise between chroma detail and lack of artifacts/aliasing. So once I've changed that, you'll have to recheck which algorithms work best for you.I actually thought this was how madVR was working currently. I did not test for it, I just assumed that to be the case.

It certainly seems like this is the way that it should be done. I would expect chroma upsampling to be done using the algorithm that has chroma matching luma as best as possible, and then both being scaled to the display resolution using the selected 'luma' scaling algorithm.

My original testing for chroma upsampling that resulted in me choosing Bicubic75 was displaying everything at the native luma resolution on CRT (720x480 for dvd and so on) so perhaps Bicubic75 is not the best choice for chroma if you are going to a 1080p native display, at least with the current method. I know you don't really like commenting on the future of madVR, but do you plan on implementing this change soon, or should I start testing chroma when going from SD to 1080p?


For what it's worth, I recently did another round of testing luma scaling going to a 1080p native LCD and after several tests my conclusion was the same: SoftCubic50 produces results that are truest to the source and the best approximation of what an SD source looks like natively on a high end CRT. (the best SD can look in my opinion)

Mitchell Netravali does look sharper, with less ringing than SoftCubic50, but suffers from a considerable amount of aliasing unfortunately. If it were not for this, it would be my choice.

8-tap Lanczos does an amazing job with aliasing and some detail reconstruction, but unfortunately there is far too much ringing with any variant of it. I wish there was an option for scaling with a high tap count that did not introduce such high levels of ringing into the image.

Sorry, but that's totally incorrect. All the resampling algorithms used by madVR are doing the same math. They simply, for every target pixel, build a weighted average of the surrounding source pixels. The only difference between the various resampling filters is which weights are applied to which pixels. Spline does not try to draw everything as curves. Spline just has different weights than Lanczos. And Lanczos has different weights to Mitchell-Netravali etc. But in their core, they all do exactly the same math.My mistake, I thought splines were curves, and that spline was trying to use the data from the surrounding area to calculate them. (number of taps)

Whatever it is doing, looks very artifiical to me.


Thank you for taking the time to go back through this topic and answer all these questions!

[pankov]

madshi,
I'm so glad that you're back.
:)
Could you, please, look at this error report
http://forum.doom9.org/showthread.php?p=1449086#post1449086
This problem is from a while back but it's still present and as far as I can tell independent on sourice filters/splitter and decoders and forced me to go back to v 0.30 to avoid it.
If you want I can provide more recent .log file?

Edit: Actually it was v0.31 I was using not v0.30

[madshi]

I tried the HDMI connection. The custom refresh rate forcing works with this output!
Ok, so it works with digital output, but not with analog output?

I know you don't really like commenting on the future of madVR, but do you plan on implementing this change soon, or should I start testing chroma when going from SD to 1080p?
I'd refrain from doing further chroma tests now, because they will be invalid sooner or later, anyway. Don't know how long it will take, though.

For what it's worth, I recently did another round of testing luma scaling going to a 1080p native LCD and after several tests my conclusion was the same: SoftCubic50 produces results that are truest to the source and the best approximation of what an SD source looks like natively on a high end CRT. (the best SD can look in my opinion)

Mitchell Netravali does look sharper, with less ringing than SoftCubic50, but suffers from a considerable amount of aliasing unfortunately. If it were not for this, it would be my choice.

8-tap Lanczos does an amazing job with aliasing and some detail reconstruction, but unfortunately there is far too much ringing with any variant of it. I wish there was an option for scaling with a high tap count that did not introduce such high levels of ringing into the image.
http://forum.doom9.org/showthread.php?t=145358

Don't hold your breath too much, though, I've *loads* of other things on my to do list for madVR, first.

My mistake, I thought splines were curves, and that spline was trying to use the data from the surrounding area to calculate them. (number of taps)
Splines are curves, they're used to calculate the weights used for the weighted average. But weights/curves are *not* image content adaptive (which is I think what you thought). The exact same weights are applied to every destination pixel. FWIW, Bicubic and SoftCubic weights are also calculated by using curves... ;)

Whatever it is doing, looks very artifiical to me.
Is it eventually possible that you were a bit prejudiced when testing spline, because you thought the image would be converted into curves? :p You're the only person whom I have heard saying that Spline resampling looks artificial.

[madshi]

Could you, please, look at this error report
http://forum.doom9.org/showthread.php?p=1449086#post1449086
This problem is from a while back but it's still present and as far as I can tell independent on sourice filters/splitter and decoders and forced me to go back to v 0.30 to avoid it.
If you want I can provide more recent .log file?
I have this problem in my "to look at before next build" list.

It doesn't occur with 0.30? Which is the exact first version it occurs with? 0.31? I need an exact answer to that, that would be helpful, thanks.

[leeperry]

Doing "original->big->target" alone does not make any sense. However, doing "original->big->postProcessing->target" is a different matter. Depending on the post processing algorithms, doing the post processing in a higher than target resolution can be beneficial.

You have to decide whether chroma detail is more important or lack of chroma aliasing artifacts. If you pick Bicubic75, you'll have more detail, but also more artifacts and more aliasing.

One more thing to consider: Currently madVR resizes the chroma channel directly from source -> target resolution, using the chroma upsampling algorithm. In a future version I plan to change that. I plan to upsample chroma from chroma source -> luma source resolution (2x in X and Y direction). Then I plan to use the luma upsampling algorithm for the remaining scaling. I hope that this will allow a better compromise between chroma detail and lack of artifacts/aliasing. So once I've changed that, you'll have to recheck which algorithms work best for you.
my Avisynth scripts(LSF/GrainF3/SmoothL) look much better when I upscale before applying them(spline for luma/bicubic 0.0 for chroma) and then let mVR downscale using spline/softcubic100. I guess that mVR tweak you just mentioned wouldn't change anything for me? What do you use for luma/chroma up/down on your own rig? I find spline more natural and less noisy/oversharpened than lanczos(3/4) and softcubic100 perfect to clean chroma noise.

Chroma is encoded at 1/2 res because the brain supposedly isn't too sensitive to chroma res? So who cares if it's too soft? At least its noise doesn't interfere w/ luma anymore? But maybe that mVR tweak is what you meant when you were implying that the softest chroma was't necessarily the truest to the source? Ah well, PQ is eye popping anyway :cool:

[6233638]

I'd refrain from doing further chroma tests now, because they will be invalid sooner or later, anyway. Don't know how long it will take, though.Fair enough. I've been here long enough that I know it can be quite some time between releases so I don't mind about that.

The more I think about it, the more it makes sense to do it the other way. (which is why I assumed it was being done like that before) What might be the best chroma algorithm for DVD could be completely different for 720p sources, which would be different for 1080p the way things are being done now.

Only using chroma upsampling to bring chroma resolution to match the native luma resolution, and then scaling both using the luma resampler chosen should produce consistent results regardless of input video resolution. (because then you are always scaling chroma by the same amount)


http://forum.doom9.org/showthread.php?t=145358

Don't hold your breath too much, though, I've *loads* of other things on my to do list for madVR, first.Sounds good, I seem to remember reading about it a while back actually. Of course it remains to be seen how the end result compares to other algorithms in the real world.

Did anything ever come of that iNEDI / ICBI or other more advanced resamplers?

Is it eventually possible that you were a bit prejudiced when testing spline, because you thought the image would be converted into curves? :p You're the only person whom I have heard saying that Spline resampling looks artificial.That's what I thought the cause for things looking artificial was, it's not why I thought it looked artificial. I went into it 'blind' without any assumptions. If I have any free time I might see if I can come up with some examples.



Something I have noticed recently is that when bringing up the madVR preferences, it is quitting fullscreen mode in MPC-HC, which isn't very helpful when you're trying to compare scaling algorithms. (it resizes the window so that video is displayed at 100%)


my Avisynth scripts(LSF/GrainF3/SmoothL) look much better when I upscale before applying them(spline for luma/bicubic 0.0 for chroma) and then let mVR downscale using spline/softcubic100. I guess that mVR tweak you just mentioned wouldn't change anything for me? What do you use for luma/chroma up/down on your own rig? I find spline more natural and less noisy/oversharpened than lanczos(3/4) and softcubic100 perfect to clean chroma noise.By oversampling with ffdshow you are losing most of the benefit with madVR in my opinion. However, I don't know what the solution would be, other than madshi implementing his own oversampling + sharpening + downsampling solution. With a good sharpening algorithm, I don't think there should be any need to initially oversample and then downsample though. It is probably a lack of precision with the sharpening that requires you to do this.

Chroma is encoded at 1/2 res because the brain supposedly isn't too sensitive to chroma res? So who cares if it's too soft? At least its noise doesn't interfere w/ luma anymore? But maybe that mVR tweak is what you meant when you were implying that the softest chroma was't necessarily the truest to the source? Ah well, PQ is eye popping anyway :cool:I think this is just an excuse people in the industry have come up with to cut corners. Chroma resolution can be very obvious at times. With games or PC use, it is immediately obvious on anything. With Video it is less obvious but certainly does show up. With video it's not the resolution difference that is most obvious (unless it happens to be red on black) but rather the desaturation of coloured objects when chroma is blurred too much with something like SoftCubic100.

Quick example taken from a very poor DVD scaled to 1080p, shown at 200%.

http://imgur.com/EXkpys.jpg (http://imgur.com/EXkpy.jpg) http://imgur.com/2wkDss.jpg (http://imgur.com/2wkDs.jpg)

If you open each in its own tab and switch between the two, the difference should be obvious. It has very little effect on reducing chroma noise, but a significant effect on reducing saturation.

[leeperry]

I'll give you that SoftCubic100 might look blurry on very sharp 1080p...70 looks too noisy to my eyes, maybe 80 is a good compromise. What remains to be seen is what it'll do on noisy SD, but it still looks like a good compromise between noise and bluriness.

[pankov]

I have this problem in my "to look at before next build" list.

It doesn't occur with 0.30? Which is the exact first version it occurs with? 0.31? I need an exact answer to that, that would be helpful, thanks.

I did more testing and found that the problem begins with version 0.32. In fact when I though I reverted to 0.30 I actually used v0.31. Sorry for that mistake. I hate Mircrosoft and their ability to break simple stuff like showing file version of .ax files. I had to manually compare by file date and size to the older versions and I made a mistake and dubbed the file as v0.30 when in fact it was v0.31.

I hope this is more helpful information now.

[janos666]

Ok, so it works with digital output, but not with analog output?

It works with HDMI but not with DisplayPort. I didn't try with any analog connection.