New DVDO iScan Duo [2.0 firmware released] - Page 63

Let's just say that MNR on the Duo is far from benign. If you have seen MNR on an HQV processor and then seen what ugliness the DVDO can produce in most content you would understand the wisdom of Josh's advice. I leave all the post processing on the Duo off myself. It just is not on the top of the list of things you will find useful on the Duo.. I speak only as somebody who started as an Edge Beta tester in the Edge prerelease days.

Everytime I though maybe it was worth another look it MNR and the other stuff all had problems, and it carries on beyond the current public release. If it worked well, it would be a welcome addition but nobody is going to mistake what is does for a Flea.

I have no credentials other than a consumer whose experience with the Duo has been a genuine positive learning experience.

I have recently got myself a PJ to get the 'larger picture' in a less expensive way and am amazed at the PQ I get.
All my 'wow' comments about a 42 inch Tosh LCD seem to be reproduced on the inexpensive Acer I now can use.

Perhaps I am not as technically aware as all other members but I have found grayscale, gamma, color gamut and other Duo PQ controls to be easy to use, predictable and something I would not be without.
It is unlikely that any other product would have educated me in such a way and all my improvements would have been made via new and expensive displays.

Sure, some of us including myself could find fault with certain areas on the Duo but as a piece of hardware with a dynamic and improving firmware it takes some beating.

Just an opinion.

Thats your opinion, and that's fine because it's a free site. That being said, I've not been around this thread long enough to know who he is, and you cannot take everything here at face value, because there are a lot of folks who will give you an answer, and truly have no idea. You can act that towards me if you like, it happens all the time on this site when I share my personal experience with calibration and televisions.

gtgray, thanks for the additional input. With the hints from you, Josh and DVDO, I see no need to use it in my situation.

Pretty abrasive/confrontational attitude for someone who admittedly has a limited data-set. I find it best to give the benefit of the doubt until someone gives a retarded answer to something (like, attacking someone's answer and knowledge before they even find out their background). I think that's why we don't see Dale Adams here much anymore (this is based on my understanding of his personality from years of working with him directly).


I second that, I saw how he backed down but didn't actually lay his cards out... Hey Don/Josh Interesting thread...

-T

Edit: See two posts below - I added a signature with my credentials...

Hi Tim

Hows married life treating you? Still get to play with toys?

Sending you a PM...

If what you are seeing from your DirecTV is really like the left half of the above image - you need MPEG block noise reduction more than MNR. For the record, I agree with Josh on the low setting being the limit of how I would use the control - and I would agree that the agrument that content-by-content would be how I would adjust the control not per-input/per-format. You would be best served in getting a direct access IR code to turn in on/off on a learning remote.

As for Blu-Ray, if it is encoded in 1080P the resolution should be high enough that you won't see much mosquito noise on a screen smaller than 100" at normal seating distances (and if you have a larger one, you should be runnig a higher resolution projector to avoid seeing pixels anyway). At 1080P, block noise becomes the most significant compression artifact you can see in an image. Duo does not have any MPEG block-noise-reduction that I know of (I helped develop and test the ABT2010, and it wasn't in there in HW... I was also the guy that sat in the THX labs for a few weeks to get VP50PRO certified).

MNR is typically a "softening/low-pass filter", so you are losing a bit of potential detail in the original signal. Unless you are using a MN-correction algorithm that actually corrects the compression error distribution in the block (what causes the visible MN artifact), you will lose quality with more filtering. It is the most destructive in a case like shown above where you have the high-frequency noise on top of actual high-frequency detail in a tight space. The space between the wrought-iron curves would get obliterated by Duo's MNR, you'd end up losing much of the horizontal detail from the wall behind the railing.

By the way - I see six zones of error in the right "corrected" image where in one case the Flea failed to correctly remove the MN without destroying detail and in the other five it failed to remove the MN at all... (noted though, it did a pretty darn good job).

-T

Tim, thanks for all the explanation, very helpful.

And to the above, it has nothing to do with backing down and my credentials. Yes, when it comes to the Duo, I am a newbie, but have plenty of experience with displays and calibration. A lot of the initial doubt problem stems from the diplsay forums, where misinformation is rampant, and one kinda gets into that "mode" sometimes. Do I excuse myself? No. But let's face it, Josh's first couple of answers were just the "cuz I said so", and like I explained, I always like to know more.

If one tells me not to use it, without facts, should I not question? I think what happened, is Josh, along with everyone else here, assumed that I knew who he was and went after him, which was not the case. Then, I learn who he is because he tell's me, then he gives a great explanation (like the one you gave).

Guys like the two of you are why most of us are here, to learn, but you can't get all pissy when you give 40% of the answer, and someone questions it, especially when they have no idea who you are.


To Josh, I'll say I'm sorry for comin at ya with the abrasive nature, and for making everyone have to sort thru the junk, for a little content.

calman4.1 version is supposed to be released soon.

does anyone know if there are new capabilities re: dvdo duo?
the limited press i've found talks about direct input of control values with
certain TV sets.........i wonder if direct input of values (or other controls)
are going to be changed/enhanced for the duo

From what I saw at CEDIA, the upgrade enables the control of Panasonic VT series of plasmas and Runco projectors. I didn't hear the mention anything about DUO enhancements.

I always use the Duo with MNR on Low with DirecTV. It cleans up scoreboard displays during sporting events. Since I have a Blu ray player with Dual HDMI outs I don't even use it for Blu ray.

Thats really where I see it. ESPN is horrible on my 58", and my local CBS has 2 subchannels, so its fracked too.

With ESPN, as you get close to the screen, the grass just looks horrible, and alone the field lines. To be honest, with it on low, it seems to offer a little more dimension.

Boy don't I know that one - right Josh? I remember a case where an Edge user got in to an argument with me because he disagreed with me over an LCD TV having native support for 24Hz at the glass (not to be confused with the HDMI input). He even argued the point after I posted the datasheet for the display's LCD module which clearly showed it was not supported... "But the guy I talked to at the manufacturer told me it does..."

Eh, that argument probably shouldn't have happened (in hind-sight it probably put me on the short list for down-sizing at ABT in '08), but I was operating for acuracy, not hearsay... I knew he didn't know what he was talking about - to the point where I could have made a legal case of it, but the reality was I didn't need to shove it in his face, or even tell him I knew he was full of it (just me knowing would have been enough for me to get my job done). Often, less is more.

I understand that you were skeptical about Josh's answer, but taking what I learned about the above argument with a customer - you can learn that it's best for yourself to dig a little for info about the knowledge being thrown at you on the front-end and keep the negative banter down to PM's only (or to yourself completely). When it goes public like this it doesn't end well. Ever.

I'm pretty down to earth, I don't think the episode reflects negatively on you - I'd just like to help.

I'm glad the answer I gave was more to what you were looking for.

Cheers.

-T

Thanks Tim.

To my comment above, I'd guess the CBS issue, with the overcompression, is more the "block noise" (watching this OTA;fast movement makes the screen look like rectangular blocks all over, but kinda different from macro blocking).

I spent a lot of time scrutinizing ESPN last night, and that type of noise in the picture above can be seen around field lines, players, especially helmet close ups.

I followed the link on the Blu-ray player (text with red highlight) and it took me to a Sony S570. It has 2 HDMI outs? That's pretty good on such an inexpensive player.

And I'll check my Duo Settings to see if MNR is on. Since I use mine as a hub more than anything else (meaning all sources are connected), most settings are flat/default, and my BDPs are both set to Auto output.

Skipping it on BR is a interesting idea, but only if you have the type of Display that can be dialed in via calibration. Many of them do not have a CMS, or have one that does not work well, and this is where the Duo shines.

As Stacey Spears can attest, there is a limit to how low you can go with bandwidth before the compression scheme basically breaks (quality drops off a cliff there).

So for CBS, it sounds like a bandwidth issue (either low fixed bit-rate causing not enough data to move into the decoder to describe the new blocks, or signal integrity to the receiver dropping key frame [I've been getting the latter on NBC at my place during "Chuck" - really irritating, I've had to watch it on Hulu instead of TV - You hear that Comcast of California?]).

For ESPN, it would be interesting to see what actual resolution they are broadcasting at. I recall that we at ABT had been looking into the broadcast resolution as a pixel-quality issue at ABT. Most relay-broadcasters (cable/satellite) cheat by cutting the resolution down to save bandwidth, so instead of 1920x1080i you may actually have a 960x540i signal that the MPEG decoder upscales in the chip to 1920x1080i before pushing it out to the HDMI jack on the back of the recevier. The other side is the broadcast itself. Not all of ESPN is really HD - as not all stadiums are HDTV wired. In some cases they upscale on their site-broadcast side and then it's HD from the transmitter truck out to the station. You may actually have an SD feed you were looking at. ESPN also has a habit of upscaling SD and leaving it as 1.33:1 but filling in the left and righ sides with banners that say ESPN-HD (true the station my be "HD", but the content in that case isn't...).

Most terribly, I recall HGTV using 1.33:1 SD content stretched out to 1.78:1 with a non-linear stretch, then squeezing the 1.78:1 image back into a 1.33:1 for their SD channel... It looks terrible no matter what channel you watch it on (SD/HD). Like the engineers at their station were told by some executive that all their content had to be wide screen no matter what the original source becase "black bars are bad" - it can be a truly dizzying experience to watch a show with a left or right pan ).

-T

Good info on ESPN. Their ESPN2 feed was pretty good last night, but when we switched over to ESPNU, it literally looked like an SD upcovnersion, I mean it was bad. So, that could be the issue. Speaking of VOD, it's funny that the SD downloads of AMC programming (Rubicon), looks better than the live content off the SD channel on Directv. If everyone would offer up 99 cent downloads, I'd just buy TV episodes off Amazon VOD. Their 1080i tv shows look as good or better than my OTA locals, which is not funny.

NBC has been bad in all 3 cities I've lived in, but CBS is actually worse in Mobile. I really hate this because many of the shows I record are off CBS OTA, as well as their Fball coverage. The Fball PQ is one of the best out there, when it does not have bandwidth problems. All of this is stuff a pre/pro cannot fix. S*** in = S*** out

Dish network does 1440x1080i on their 1080i feeds to customers, but as far as I know, Directv is still sending out 1920.

Tim was actually referring to DirecTV. The box that this was verified on was the HD-TiVo.

Hint: Always check someone's User Profile before you assume that they are full of nonsense. ;-)

Yeah, you are talking about the HR10-250, I actually had one. Those were the good ol HD lite days

http://en.wikipedia.org/wiki/HD_Lite

I actually went to Dish at that time. Once Directv launched D10 and D11, they have now been noted (by users with the proper equipment at DBSTalk) to be sending 1080i now as 1920x1080i. Dish has reduced all of their channels to 1440x1080i now, but I think their PQ issues stem from cramming more channels per transponder, to gain space. They are using variable bitrate encoders, but Directv is running around 5 on a TP, while Dish usually has 8. Lately, with Dish's issues, many subs are swapping to Directv, and people on both Satelliteguys and DBSTalk have noted the PQ improvement.

I also feel, on a larger screen, Directv takes the cake (other than FIOS), and Im generally happy with the HD I see on SyFy, FX, NatGeo and so on. My biggest issues with noise are ESPN (you can Google around, many folks have been complaining about ESPN for years), and the local CBS, but as Tim said, that's more bandwidth.

None of these bandwidth problems are an issue for me here in San Diego.
At 550 ft ASL every OTA DTV broadcast is received with excellent signal strength and SNR using an RS small UHF antenna on the roof.

In fact my HD Homerun receives virtually every one of these stations with an indoor UHF loop.
TSReader reports 18 Mb/s for CBS on VHF 8. CBS here does not transmit secondary channels, NBC has 2 of those but still shows 15 Mb/s for the HD channel.

Even my TWC channels to the Tivo Premiere are in the high bitrate category for HD shows as reported by KMTTG. A recent Nascar race on ESPNHD shows 18.8 Mb/s as reported by KMTTG from the Tivo.
It is all market dependent.


Rabbitears.info reports TSreader data although many may be dated, a sample here:
http://www.rabbitears.info/screencap.../42122-0_0.htm
although IMO the data are faulty

While I have not seen the need to use MNR for these sources, it is useful for my laserdiscs and VHS and S-VHS tapes, many of the latter recorded from DirectTV more than 10 years ago. Some Laserdiscs suffer from color gradient drift and the Duo eliminates that very effectively. The Duo is a winner in my book, not only as a video processor but as a pattern generator for TV calibration using ChromaPure.

There is a rumored 3D pass trough coming for Duo. Does this mean picture coming from a projector that supports (e.g RS40) 1.4a HDMI signal can ONLY be passed trough or can Duo process the picture normally when it is in 2D?

I ask because RS40 does not have CMS and I HOPE I can correct the Gamut with Duo.

Following up my previous reply, I checked and MNR is off. I put it on Low for the cable box input and will see if I notice anything different.

Consider yourself lucky then, my OTA channels here stink. They were good in the other markets I lived in, the subchannels just kill the PQ. Also, don't forget that bitrates are so high, because it's Mpeg2, so you can't really compare that to say a 7.5mb/s satellite feed.


I've been doing a lot of watching, with it on and off, and the issues I see around edges, are not helped with low on. I've just decided to leave it off.

I'm going to go out on a limb here and surmise, that yes - only pasthrough will be enabled - the scaler in the 2010 is not compatible with the pixel clock of CEA-861 format 63 (1080p @ 120Hz, scroll down a little bit to see the list) or dual streams.

That said, the comon 3D methods (when uncompressed) are technically compatible with the color management process so long as the pixel clock is low enough. While I'm not saying that ABT has any plans to (I'd be the last to know at this point), 720p120 and 1080i120 could be put through the CMS so long as the scaler and deinterlacer don't touch it at all (they are not designed for a dual stream, so can't split and recombine the video to do deinterlacing and scaling then output as a standard 3D stream).

But again, take my info with a grain of salt - I don't work there anymore and I don't know what they have up their sleeves... Maybe if ABT reads my post they'll be tricked into giving it a try or maybe they've already done it , otherwise you may be in for the old-school "Bang-Bang" control for passthrough (completely on or completely off - no middleground, no negotiation).

The reality is we probably won't find out until they release the FW publicly.

-T

Tangent:


This because of the time to live - with live TV, you need a lot of low-bit-error-bandwidth to get the full picture across. With internet downloads, the computer or Tivo or whatever takes as long as it needs to to get the whole file, and the time it takes to transfer a file is not directly attached to the playback experience (so you can use less compression, for a larger file size, which takes longer to download on less bandwidth). If a packet is lost or damaged, it can simply request a replacement - you can't do that with a live feed. When you lose a packet, your playback breaks because the data is delivered "just in time" for the decoder to use it. This is helped with a buffer, but if your uplink is pretty bad (like a poorly mounted satellite TV dish in a snow storm), there won't be enough time or bandwidth to fully rebroadcast any data before the picture is supposed to be drawn.

1Gbe fiber commercial free VOD IP-TV with local "metro" cache servers forever.

(my personnal opinion - not related to any company or their work) If the broadcasting companies got their heads out of their rears, and realized that a DVR with a network connection can download a TV episode (like a Tivo "season Pass") a week in advance using a network connection, but not show it in the playlist until the date of its release - the viewing quality/experience would be a lot more positive for their >>> paying customers <<<</u> - you could go "uncompressed" in some cases...

-T

This is a very good idea.

Did an initial CMS, grayscale and gamma calibration with my Duo last night on my FPJ1 / RS-2 clone. I have not had a chance to do the second or third iteration, so I'm sure I'm not spot on anymore based on the interaction of adjustments. I would note that shadow detail has increased a fair amount. However, I think my Eye1 LT colorimeter has a red bias (or rather has a harder time reading red), which has resulted in everything being a bit pinkish. Fleshtones are looking a bit pinkish and a bit posterized.

I have a second Eye1 LT that I can try, hopefully tonight.

Note, on a single iteration of setting the white point, setting the color gamut for the primaries and secondaries, setting grayscale and setting gamma took me about 2.5 hours. Subsequent iterations will, I am sure, take less time.

Does the EYE1 LT need the same recalibration every 9 minutes as the iOne Pro? If so, you may want to get a Chroma5 and Profile the EYE1 LT to it. May save you some calibrating time. Gives you the better accuracy of the EYE1 LT to the speed and Lower End Accuracy of the Chroma5

i1D2 LT are known to have issues reading RED. I have 2 of those (rev. C) and both are reading red low. i1D2 does not require constant calibration. I ended up gettnig a ColorMunki and profile against it.