Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules. |
2nd December 2012, 22:49 | #15921 | Link |
Registered User
Join Date: Mar 2007
Posts: 934
|
I'm sure I've mentioned this before but auto-detection of deinterlacing seems to be broken for some HDTV (UK) streams in MKVs (maybe other types of files too, dunno). The same streams in TS files deinterlace fine, and the MKVs deinterlace fine in EVR too. Using LAV Splitter/Video/Audio.
EDIT: On second thoughts, I think this is probably an issue in MKV Merge. My older MKVs work fine, newer ones don't. If it is, Mosu needs to know why the interlacing isn't detected properly so can you help with this please madshi? Samples: TS - deinterlaced automatically MKV - not deinterlaced automatically
__________________
TV Setup: LG OLED55B7V; Onkyo TX-NR515; ODroid N2+; CoreElec 9.2.7 Last edited by DragonQ; 3rd December 2012 at 00:02. |
2nd December 2012, 22:55 | #15922 | Link | |
Audiophile
Join Date: Oct 2006
Posts: 353
|
Quote:
and a GTS 250 is not that weak. it was the highest end stuff a couple of years back(remember that it's a rebranded 9800 GTX) |
|
2nd December 2012, 23:10 | #15923 | Link |
Registered User
Join Date: Mar 2007
Posts: 934
|
Yeah but mine sucks, the fan's default setting is 100% and it seems to happily switch between 17-30% fan whilst idling in Windows, let alone playing videos, resulting in constant whining.
Thank Jebus for MSI Afterburner!
__________________
TV Setup: LG OLED55B7V; Onkyo TX-NR515; ODroid N2+; CoreElec 9.2.7 Last edited by DragonQ; 2nd December 2012 at 23:22. |
3rd December 2012, 00:04 | #15924 | Link | |
Kid for Today
Join Date: Aug 2004
Posts: 3,477
|
but there's also the story of some GPU's enabling their high performance speeds profile when using CUVID, isn't it? The GTX600 serie is not too power hungry but the 500 can quickly become a power hog...so in that case, only a wattmeter could really tell what your best option is I'm afraid.
Quote:
and talking about dead horse feature requests, that'd be really great if we could disable the "windowed/exclusive" OSD messages at some point |
|
3rd December 2012, 00:21 | #15925 | Link | |
Registered User
Join Date: Jan 2007
Posts: 530
|
Quote:
But for jRiver users (and probably MPC and others, haven't looked), it's not difficult to script-monkey AutoHotkey to do the same, albeit even more hacky as it changes the madVR scaler after the video starts playing... |
|
3rd December 2012, 03:19 | #15927 | Link |
Registered User
Join Date: Jan 2012
Posts: 106
|
I just installed PotPlayer build 34665 which has pixelshaders built-in but they don't appear to work with madvr. I use lavfilters for splitting and decoding, I am using hd4k for graphics atm (my amd 7750 is in rma...) in dxva2cb. Advice appreciated
Last edited by strumf666; 3rd December 2012 at 03:24. |
3rd December 2012, 07:19 | #15928 | Link | |
Playing with MKV files...
Join Date: Feb 2007
Posts: 278
|
Quote:
'Course they may have changed their views with recent performance improvements in both their respective projects. So I'd like to hear if they think differently. DXVA is more what I would turn to if my system was simply too underpowered to handle something on CPU for decoding. It's nice for low-end HTPCs and Netbook playback in other words.
__________________
Intel Core i7 920, 6 GB RAM | GeForce GTX670 | Win8.1 | LAV 0.64 / MadVR 0.87.10 | Zoom Player v10 / MPC-HC 1.7.8 |
|
3rd December 2012, 09:36 | #15929 | Link | |
Registered User
Join Date: Oct 2010
Posts: 131
|
Quote:
|
|
3rd December 2012, 11:26 | #15930 | Link | |
Registered User
Join Date: Mar 2007
Posts: 934
|
Quote:
By the way, this is all on my desktop and is nothing to do with the "MadVR reports deinterlacing failed" message on Arrandale. This is just failed auto-detection of interlaced material. A summary of the problems for me: 1) All TS files with AAC audio: - EVR: TS plays fine, MKV plays fine. - MadVR: TS plays fine, MKV is treated as progressive so not deinterlaced. 2) Older TS files (e.g. from 10/11/2012) with AC3 audio: - EVR: TS plays fine, MKV plays fine. - MadVR: TS plays fine, MKV plays fine. 3) Newer TS files (e.g. from 02/12/2012) with AC3 audio: - EVR: TS plays fine, MKV plays slowly (46-48 fps) with playback speed jumping all over the place and a really dodgy graph. - MadVR: TS plays fine, MKV is treated as progressive so not deinterlaced and also playback speed is all over the place.
__________________
TV Setup: LG OLED55B7V; Onkyo TX-NR515; ODroid N2+; CoreElec 9.2.7 Last edited by DragonQ; 3rd December 2012 at 11:53. |
|
3rd December 2012, 17:53 | #15932 | Link |
Registered User
Join Date: Nov 2012
Posts: 2
|
Don't know if this has been mentioned yet, but DXVA decoding with default settings crashes MPC-HC for me about 50% of the time when viewing 23.976p material. This only happens when skipping forward or backward, however if the player does not crash, I can skip forward safely as many times as I please. I am using LAV filters and a laptop with ATI 4670.
|
3rd December 2012, 21:21 | #15933 | Link |
Registered User
Join Date: Feb 2012
Posts: 18
|
I'm not sure if I understand completely yet. Can increasing or decreasing queue sizes help reduce dropped frames? Ever since I set up yCMS color calibration with madvr, I've had a good amount frame dropping, and I need to do something to make up for the performance loss. Is there anything I can do other than get another GPU (not an option, as I'm on a laptop).
|
4th December 2012, 02:07 | #15936 | Link |
Registered User
Join Date: Jun 2012
Posts: 16
|
Could anyone give me a step-by-step procedure to enable the "Save Image" option in MPC-HC? I'm using MadVR and it says that it's not supported... but how come I could see those people here have it working?
I got the official build 1.6.4.6052. |
4th December 2012, 02:23 | #15937 | Link | |
Registered User
Join Date: Jan 2009
Posts: 1,210
|
Quote:
Last edited by dansrfe; 4th December 2012 at 03:38. |
|
4th December 2012, 02:25 | #15938 | Link |
Registered User
Join Date: Apr 2011
Posts: 1,184
|
Official build doesnt support that, use nightly buildhttp://xhmikosr.1f0.de/mpc-hc/
|
4th December 2012, 03:50 | #15939 | Link |
Registered User
Join Date: Apr 2009
Posts: 1,019
|
Going back to the discussion about default scaling algorithms, since getting this GT610 (still waiting on my replacement GTX570) I am obviously right at the bottom end of what video cards will support madVR, which has had me re-evaluating things from a performance vs image quality perspective.
I've also had the opportunity to watch a few more films than I normally would recently, and having a larger sample group is always good to have to spot issues that show up in real-world content rather than test patterns or a small selection of specific samples. In particular, I noticed quite a few problems showing up from using Bilinear for Chroma upscaling, just from watching films. I really don't think that having Bilinear as the default for Chroma is a good choice. As Chroma is only ever a 2x scaling operation, the performance impact of using a better scaling algorithm seems to be minimal in most cases. While not related to selecting good defaults, in my testing, I also happened to come across a use-case for Jinc 8. There are some images I came across where using Jinc 8 AR for Chroma was the only option that looked really good, being the only choice that avoided aliasing and maintained the correct brightness/saturation for Chroma. This is not an endorsement of using it, as there are most likely problems using it with other images (as there tends to be with higher-tap filters) but I thought it was interesting. The only other algorithm that looked decent with this image was SoftCubic 100, which was too dull/desaturated, but avoided most of the aliasing.
But going back to selecting good defaults for madVR, here's an example where using the default Lanczos 3 AR Luma with Bilinear Chroma looks particularly bad: This had a rendering time of almost 90ms. In comparison, Bicubic 75 Chroma with SoftCubic 80 Luma (which requires no AR filter) has a rendering time of 32ms: Changing Chroma from Bicubic 75 to Bilinear with SoftCubic 80 Luma only drops rendering times by 2ms, to 30ms total. And as a "reference" image, this is Lanczos 3 AR Luma with Jinc 8 AR Chroma. (395ms!) Now I don't suggest that SoftCubic 80 be the default for Luma at all, but I worry that Lanczos 3 AR will be too taxing for a lot of hardware. I wish I had access to an Intel HD 4000 to test performance with, because that seems like a reasonable card to use as the "minimum spec" for the defaults. Here are some numbers from that particular test though: Bicubic Luma scaling:
While I didn't do testing on them individually, all the Bicubic variants (Mitchell-Netravali, Catmull-Rom, Bicubic, SoftCubic) should result in the same rendering times, and Lanczos/Spline tend to give roughly the same results as well. From looking at those numbers, I think that it would probably be best to use Bicubic 75 Chroma, which can be a significant visual improvement over Bilinear in many cases (often a close visual match to Lanczos or Jinc) for a minimal performance hit compared to Bilinear. While I know that some people prefer softer Chroma settings, Bicubic 75 still seems to be the best balance between ringing/aliasing and maintaining the correct brightness/saturation for Chroma with a minimal performance hit. I wouldn't use it for Luma, but I've found almost nothing where it has caused problems for Chroma. It may not be my personal preference, but I would then probably suggest using either Lanczos 3 or Spline 3 (probably Spline due to the reduced ringing) without the AR filter as the default for Luma upscaling. The rendering times I posted are for comparative purposes from upscaling Blu-ray beyond 1080p, which is not indicative of real-world use (typically DVD or 720p up to 1080p) so I haven't run into any performance issues with playback when testing those settings yet. (almost 50ms would imply dropping frames) If you're really wanting to reduce rendering times for your default, I would probably go with some level of SoftCubic for the Luma upscaling, but I know that's too soft for a lot of people. It does manage to avoid aliasing and ringing very well at higher levels though. EDIT: Or Mitchell-Netravali might be a better choice. The other Bicubic variants probably have too much ringing or aliasing to be suitable for the default Luma algorithm. And now that I've overclocked the card (1GHz GPU/720MHz RAM) I seem to be able to play Blu-ray downscaled in linear light using Catmull-Rom AR without any problems, though it is probably still too taxing on some systems. I still need to do a good performance and image quality test for downscaling algorithms. It's really not a priority for me though, as I use a 1080p native display, and the only time I downsample is when I have something playing at the side of a web browser, which is never a film, and therefore not particularly important to me as far as image quality is concerned. Last edited by 6233638; 4th December 2012 at 06:36. |
4th December 2012, 03:56 | #15940 | Link | ||
Registered User
Join Date: Jun 2012
Posts: 16
|
Quote:
Quote:
I've encountered that site and those builds before but I thought it was untrusted... looks like I'm wrong. |
||
Tags |
direct compute, dithering, error diffusion, madvr, ngu, nnedi3, quality, renderer, scaling, uhd upscaling, upsampling |
Thread Tools | Search this Thread |
Display Modes | |
|
|