Which HD **** would be sufficient for MadVR - Page 3

Well said.

Just to be clear I'm not trying to put anyone down, and I absolutely applaud what madshi is doing. Objectively madVR does produce better results. People with extremely good eyes and/or particular televisions (especially projectors) certainly may see a difference. I would love to see more front ends supporting madVR.

I will say that I'm a bit suspicious when people say the difference is dramatic, and I suspect that many people have their HTPCs configured incorrectly to begin with. From my experience this seems especially likely to occur with people using ATI cards, for a few reasons. First of all, Catalyst has numerous post-processing options turn on by default, and many people fail to turn all of them off. There are also some bugs in Catalyst that cause post-processing options to randomly turn back on, and I've seen the pixel format setting randomly change as well. As I understand it, madVR bypassing ATI's post-processing options. I also noticed that if the pixel format settings is incorrect, the adverse effects are somehow less severe when using madVR.

The differences are dramatic if you're upscaling lower resolution videos. They are far less noticeable if you're displaying 1080p video at 1080p, which is what assassin was looking for examples of.

In that scenario, the only things MadVR can do is chroma upsampling, which many TVs essentially throw away, if they process video in 4:2:2 which is particularly common with Plasmas, and 16-bit YCC > RGB conversion, which is only going to be noticeable on displays which are capable of actually showing true 8-bit gradation - which is a limited number of flat panels, despite what manufacturers may claim in their spec sheets, and those are mostly LCDs. Plasmas are typically much less than 8-bit and produce a very dithered/noisy image. (Kuros in particular) With projectors, DLPs should clearly show the effects of chroma upsampling, but probably won't show any improvement with gradation. SXRD/DILA should be capable of showing the gradation improvements but may not show chroma improvements due to convergence/sharpness issues.

For upscaling DVD content, there is nothing else out there that is comparable to MadVR's Jinc scaling, nothing I'm aware of that can scale in linear light (important when downscaling) or anything like the anti-ringing filter that Madshi has implemented.

Sitting on the fence? - that's not like you

Ah good point, and thank you for the information. The difference should be much greater when the luma component needs to be upscaled. I watch almost entirely 1080p content. Time to upgrade to blu-ray people :P

Maybe try telling the studios that. I have a lot of content that doesn’t have a Blu-ray release available anywhere, and likely never will.

And then there's general availability. You might be able to find content on Blu-ray in one region somewhere, but not in your own country, and importing is not as cheap as it once was. There are a large number of films that only have a Criterion release on Blu-ray, and those are not available here.

Still, with MadVR's non-ringing Jinc scaling option, 24p output with DVD and other improvements, DVD playback isn’t so bad..

Oh I agree with you. I was being somewhat sarcastic. Even some new movies don't have blu-ray releases (Cold Souls comes to mind). Very disappointing.

Thanks for the screenshots.

Have to say that I am not seeing what I would consider a "dramatic" improvement. Not sure how much difference it would make during the regular (meaning "non-text") part of a movie, either.

I also have a plasma HDTV so as you say the improvement may be lost on me. I do see where this would be helpful for less than HD quality content. Nothing wrong at all with using or trying Madvr. I am just still not convinced that it panacea for everyone that some touted it to be when it first came out. I still think getting a high quality display and properly calibrating it would have a much much bigger effect for most people.

I definitely notice the improved gradation when watching 1080p films, and I agree that when you are 1:1 mapped, chroma upsampling is not a major difference in many cases other than text when dealing with film content. But even then, if you really care about video quality, aliasing on text and other areas of strong colour will bother you, especially when you know it can be better. If you are watching animated content or game footage, the difference is usually much more apparent.

If you are upscaling, the difference that good chroma upsampling makes is definitely noticeable.

The difference MadVR makes when upscaling anything that's not 1080p (or 1080p to greater than 1080p) is also a much bigger improvement than chroma upsampling, especially now there is the anti-ringing option.


I'm not sure I see the point you're trying to make though. If you have a high quality display and have it calibrated, switching to MadVR is going to yield obvious improvements.
If you don't have a high quality display, or you do have one and it's not calibrated, well, MadVR is still going to improve image quality.

MadVR is a free product. It's not like you have to decide whether to spend $300 on calibration or MadVR, it's whether you want to spend a little time installing and configuring it. (which is barely any time if you know what you're doing)
MadVR can also aid in calibration with its 3DLUT support, gamma, and linear-light saturation controls, though 3DLUTs still need a bit of work.

I have tested it repeatedly. I disagree.

So do others as evidenced in this thread.

To each their own. I agree that its free to try so knock yourself out everyone and decide for yourself if you can notice the above screenshots in a real movie. I cannot.

I used MadVR with MPC-HC for months before I got a discrete card. It was playback of some 720p and 1080i content that I had (mkv's of TV shows and the like) that caused the problem with my i3-550 with HD2000 as it could not handle this fast enough. Ivy Bridge and HD4000 wasn't available yet at all. The cheapest solution to this was a discrete card at $60. Adding the GT-430 and additional control with the nVidia Control Panel caused a bigger improvement in picture quality from all sources, including Live/Recorded TV for me than did MadVR, but I was able to utilize more of MadVR's settings.

As I said, if you're only playing 1080p content on a 1080p display, particularly films rather than animated content, the improvements may not be huge, especially when a lot of displays out there are not even capable of showing them.

When you play anything that is not at a matched resolution, whether up or downscaling content, MadVR's differences in scaling algorithm should be immediately apparent on a properly calibrated display. (i.e. no video processing/sharpening options enabled)
Video card shouldn't have any influence over video quality if you're connected over DVI/HDMI.
Edited by Chronoptimist - 12/6/12 at 4:19pm

^^ I changed some of the contrast, brightness and color settings in the nVidia Control panel. The color temperature and color space are better.

Either that or it is all placebo.

After watching some movies, I'm going to have to contradict myself.

There's so much talk about the "24p bug"on Intel CPUs (which I don't see it, madVR shows 23.9724), but I have found a real show stopper for intel HD graphics:

- None of the Intel Graphics output Full RGB (0, 255) through HDMI.

I'm extremely surprised there are very few complains about this issue.

What people with Intel HD are doing to get decent picture?

In fear of getting attacked, I've never posted my experience with MadVR.  But, honestly, I've never been able to tell any difference in picture quality on any of my original BD movies and 1080i TV shows when using MadVR or not on either of my 55" and 46" 1080p displays; even when sitting 5-6 feet from TV displays.  It just doesn't make sense to use extra system resources when I dont need to.  I guess if most of the content I watched was low res material; and, wanted them to look at good as possible, it would be much more useful to me.

I would postulate that if you can't see the difference with madVR, it's your display that is not calibrated correctly or maybe your display is deficient.

Or you are playing HD content. Even madshi himself has stated that there is not much improvement when playing HD content. (This may be different now with his new deringing algorithms.)

The anti-ringing filter is part of the scaling algorithm. It's not going to remove ringing that is already encoded in the source.

I think the real test of MadVR, LAV and Re-clock is when you are watching a good movie and, despite being immersed in the story, you suddenly think - 'bloody hell, the picture quality is stunning'! Happened to me last night. So don't think of these enhancements as being a simple A/B comparison, it is all about synergy and the way everything comes together over a two hour film. I thought my picture quality was good before but having got these applications working it is now stunning. Incidentally I also have MadVR+LAV+Re-clock working on my i5 desktop PC (for testing) using the on-board GPU and it works for a lot of material. So if you are in any doubt and do not want to spring for a dedicated graphics card give it a go and see what you think. Can't break anything and easy to disable if it starts choking on some challenging upscaling.

Keep in mind that the difference in picture can be easyly eluminated by wrong video driver and display settings.

I had a handful of people PM me the same thing when I first made my observations about Madvr. Its interesting to still see some people tell you that you must be doing it wrong, your display is crap, not calibrated, etc because you SHOULD notice a dramatic difference even when Madshi, Nev, and others state there is little/minimal improvement with HD material.

I think it's a personality thing. In the past I would have been obsessed about max quality but I have learned that with HD it all looks pretty good.

Anything 720p or 1080p just looks fine to me. I don't watch much non HD stuff.

If you were an ultra critical viewer in search of perfection- then I think it might make sense. If for nothing other than insurance your getting good picture. But those are the type of people that looks for imperfections or pay attention to the quality more than they pay attention to the actual movie they are watching. Personality Thing.

Unless you have very high end theater- I don't see it being worth it for normal folks or non enthusiasts for sure.

But- on AVS I can see there is a crowd here that would want it because of the kind of enthusiasts they are.

I see nothing wrong with either.

Agreed. Just please don't tell others that they should notice a difference.

You may notice a difference. Or you may not. Only one way to know for sure and that's to try it out. But don't act like its the end all for utmost PQ on all sources no matter what because, well, its not.

That's all I'm sayin.

Note I don't bother using it and could.

IMO madVR is not for those who just want to watch some stuff with couple of beer. madVR is for those who are looking for perfection in picture (and able to understand and setup things properly). For last ones every single bit of improvement is a big step. Once again: not everyone can accept it.

I use a JVC RS45 on 120" screen and see a noticeable difference using madVR/LAV/MPC-HC verses XMBC. I wish there was a way to use the combo in place of the standard XMBC player! XMBC is cool but the video quality sucks.

+1

Quite often the improvement MadVR adds is taken for granted, I also use XBMC for my extenders, switching to that setup makes the difference that much more obvious. I have calibrated displays on all setups.

This is my feeling exactly. XBMC is so cool. I love the info screen and how you can see the plot and cast. I love that you can download subs automatically from the movie. I love everything about it except the PQ.

People can say whatever they want. I know what my setup is. It's the latest video card, the latest beta driver, every bit of processing turned off in Catalyst Control Center, 6500ub projector calibrated perfectly, almost every movie I own like (95% of 350 movies) ripped to the full video stream and the HD audio stream. I have 35 TBs on storage networked to 5 XBMC machines in my home. I know what I'm talking about. Period.

Somebody says they can't see the difference. That's fine. Don't insist that my setup is somehow below par because frankly that's just ludicrous.

I installed a Isotek mains power conditioner on my plasma and, against expectations, I noticed an improvement in picture quality. Thinking this was perhaps a "placebo effect" I got both my wife and my son (separately) to do an A/B comparison without knowing what I was doing. Each time they first watched some video without the power conditioner and I then connected it. Both claimed whatever I had done improved things noticeably but suggesting this to someone who was not there will get you eaten alive. Once you enter the realms of true Hi-Fi you find endless arguments about special mains cables, equipment supports and room treatments so this discussion is mild compared to that world. But then I think vinyl sounds a lot better than CD, something I can prove to disbelieving friends within minutes.

So whilst comments here suggest those who notice a difference with madVR do so because we want to, maybe it is worth considering that those who don't see a difference ............................. don't want to!

Wow. That's a pretty bold statement right there.