SDI vs HDMI 480i - Page 3  

Yes,I have compared sending Marantz DV-9600 SDI output to external scaler DVDO VP30 that comes with built in SDI/ABT102 cards vs sending it via HDMI output.
The result is that SDI output gives slightly better performance compare to 480i HDMI output.
Does any one know if Toshiba XA-1HD DVD player can be modified with HD SDI output ?

Kartono

whereas the virtues of SDI modded players are usually trumpeted, comparisons among sdi modded players are not always easily available. can anyone say for example, what the differences between the SDI modded panasonics rp-56, -52,-91 might be?

tnx.

The differences between SDI players are down to the quality of their MPEG decoders. In the Panny case my 82 is better than my 62 as it has a better MPEG decoder chip. The other big difference is in the audio quality.

Where do you get the players modified?

TWD,

PM Gary Murrell, he did my Denon 2900, which has a fantastic picture - as close to HD as I could hope for from SD DVD.

I feed my SDI 2900 in to a VP30 with ABT102 on to a 92" Firehawk. It does provide an incredible picture.

Steve

My 2 cents worth here if I may. I modded my OPPO 971H with the PM BT656 module. I posted the results in the OPPO modded SDI thread several weeks ago. But essentially, when comparing the OPPO's DVI output (at the same resolution, i.e 576p) with the Pixel Magic's SDI board (at 576i of course), and feeding both through the VP30, SDI wins out in resolution (sharpness) and noise. I make this statement objectively - using both DVE and AVIA test DVDs. The vertical and horizontal high frequency banding artifacts that are evident when viewing the DVI output, are completely absent on the SDI output. What really surprised me too, was the total absence of noise on SDI. You can clearly see this on a static, but repeating pattern, such as the "missing pixel" test on the ABT102 test DVD.

SDI and DVI are both available at the same time and the output resolution on the DVI port can be set to any of the 3 res, viz 480/576p, 720p and 1080i, without of course affecting the SDI output. A very neat setup to compare the OPPO's de-interlacing and scaling capabilities to those of the VP30/ABT102!!

The one caveat with the OPPO and PM SDI card, is that you will need the ABT102 with the VP30 to get a picture out on PAL. Without the ABT102, you will only get NTSC. The other thing to note is that the first 3 top lines are cropped when running PAL. I discovered this when I ran the ABT102 test disc. Otherwise, as I said, the picture is brilliant - the closest I have seen to HD.

Addendum:

To be able to see some of the differences between the 2 output sources, you will need a display that does NOT do any internal scaling, as this could mask some of the gains that SDI offers (at least in my setup). My display is a Professional Samsung 403T 40in LCD panel with native 1280 x 768 resolution and native 50 and 60 Hz Frame Rate support. The VP30 is pixel to pixel and FR matched to the display.

At this point we come back to the point already posted in this thread some feel its the same (or should be) others tested and found its not. At times its clear to all of us that dvi/hdmi is easer to add to a system than sdi so lets found out why. I mean i for one can't see why the dvi/hdmi should have noise .. it just doesn't make sense. Since you have a oppo and they are more open then most could you ask them why ?

Again i think we should try to find out what these players are doing to the dvi signal and start putting presure on them to atleast make these 'actions' selectable (because they don't have to be bad). If its the result of conversion parts that are bad we should pressure them to fix it. lets not forget the lessons Stacy/Dale and others have learned when they first encountered the CUE problems.

FYI, i also use a denon 2900vM (alot of changed) with a pms sdi mod for a few years now but would not mind to add a cheap dvi 480i model that would to the same.

Daniel.

There are actually processing chips in DVI/HDMI transmitters. It's not a simply the case of what goes in comes out. Now ideally the processing should be completely transparent and have no effect on picture quality, but I'll leave you to decide for yourself how ideal things really are...

DéjÃ* vu Â- I did the same thing Â- Oppo 971H + BT656PRO mod into a Lumagen HDQ. Got the same results Â- SDI wins by a large margin on image quality Â- sharper with less noise. I compared Oppo 480p DVI and Toshiba HD-A1 480p HDMI vs Oppo SDI Â- it is not subtle, it is significantly better via Oppo 480i SDI. I have the same top lines pixel cropping problem (non-issue for me Â- 99% of my disks are not 4:3 Â- so the missing pixels are part of the top black bar).

To be fair, this does not answer the question: Can 480i HDMI/DVI be the equal of 480i SDI? I see no reason that they canÂ't have equal image quality, it just takes someone doing it correctly. Something is being done incorrectly with HDMI/DVI outputs on DVD players, I donÂ't know what exactly is happening, but the image from HDMI/DVI is clearly inferior to SDI. The sad part is that it should be easy for a manufacturer to find out exactly what is going on. They should have all the diagnostic and analysis tools that they need to track this down, if they didnÂ't have those tools, they could not have built the players. Apparently they just donÂ't care Â- I donÂ't see any other explanation for why they are all still broken Â- lack of market pressure is my theory about why this is.

This is not a fair comparisson as you're using DVI's at 576p to SDI's 576i.

Same problem with usualsuspects' comparison - You need to compare 480i to 480i, not 480i to 480p.

The only way it could be done with an Oppo is to use the 970H, not the 971H. The 970H outputs 480i over HDMI and the 971H does not. So to do a fair comparison, one would need to SDI mod a 970H and then compare the SDI at 480i to the HDMI at 480i of the same player. In the 971H, I would certainly hope that the SDI output would beat out HDMI outputting 480p, as the deinterlacing is being done in the player.

One thing no one has mentioned up to this point and it really is surprising - The quality of the scaling/deinterlacing after the 480i out is far more important than whether the 480i is coming from HDMI or SDI. I mean, if you take the SDI 480i output of a player and then feed it into a video processor that is poorer in quality than the one built into the player, then what's the point?

I realize that, I tried to separate my comments into two sections. One was an agreement with a previous poster about Oppo 971H SDI output quality. The other was a general comment that HDMI/DVI outputs are broken. I realize that you need to compare 480i to 480i for a valid experiment. It is interesting to me that two very different players (HD-A1 and Oppo 971H) put out nearly identical 480p outputs that show the same flaws vs SDI. Again I realize that we are comparing apples to oranges, but it tends to point me in the direction of something happening to the data after it has been de-interlaced and not the de-interlacers as the problem. De-interlacing film with a good cadence is very easy.

To echo your comments Usualsuspects, you may have noticed that in my appraisal of SDI vs HDMI on my SDI modded OPPO, I did not comment on the difference between the two de-interlacing platforms. As you say, film material on PAL is a straight 2:2 cadence as it is played back at 25 frames/sec. So, as the cadence workload is minimal and the fact that the Faroudja Chipset is pretty good at adaptive deinterlacing of film anyway, I could not detect much/any difference between the SDI and DVI outputs wrt de-interlacing in the comparison tests that I did. Regardless, I only highlighted those differences that were not affected by de-interlacing.

My report was not so much to prove or disprove the differences between SDIi and HDMIi, as to confirm the reports that many have made that SDI appears to win out over the current implementations of 480/576i HDMI on most (or all so far?) players when outputting 480/576i.

BTW, I should correct an error I made earlier. The PM board crops the top 3 lines ONLY on NTSC, NOT PAL. I was using the ABT102 test disk when reporting this - but this disk is in 480i format. Using my DVE PAL test disc shows no cropping and indeed all PAL movies show none either. So, for once us PAL guys are out on top.

I didn't have time to do a full blown test, but SDI output of an RP56 had better PQ, IMHO than the OPPO 970H, but the difference was fairly marginal - not cardinal.

I agree that until an Oppo 970 (or other 480i capable player) SDI vs HDMI be available for observation, it would be just speculation at this point.

I own a 970H and would be willing to SDI mod mine...Does anyone know where I can get a DIY kit to do a SDI mod on the 970H? I am certainly more than capable to do the soldering, but I would need a schematic or better yet a diagram mapping out the connect points.

I have a SDI modded Panny RP-82, but we are still in the same position of not being able to compare apples to apples.

Bob, here's a couple from Digital Connections:

https://www.digitalconnection.com/st...t2_bt601pk.jpg
https://www.digitalconnection.com/st...t2_bt656pk.jpg

Thanks, Dave, but where do I get the pinouts? Like I said, I am a very capable solderer, but I don't have a clue where to connect what pin...

If you follow the links I believe it has some hyperlinks to that info for you.

Ooops, sent the wrong links, here ya go:\\

http://www.digitalconnection.com/pro...eo/bt601pk.asp
http://www.digitalconnection.com/pro...eo/sdikitp.asp

Ok, thanks, Dave...That might just be enough for me to figure it out!

I think I'll open up the 970H first and check things out to assess the work involved. Twelve wires might not be too bad at all depending on how everything is laid out in the 970H.

I offer some tips after doing my SDI mod to the 971H:

I apologize in advance if you are the master of soldering, I thought I was pretty good at soldering, but the small scale of the wires and pads threw me a little.

0) decide where to mount the SDI board and make sure that the wires will reach. In my opinion, the lock-nut on the BNC connector is sufficient as a mounting device - drill a hole in the player just large enough for the BCN connector to fit through. Watch those metal shavings! If you can use a punch to make the hole - that is much better.

1) Magnification and good lighting are needed - very small wires

2) My tack soldering technique: 15w iron max. Use the smallest point tip you can find for the iron and tin it. Strip about 1/32 of insulation off each wire end. Twist and tin each wire end just enough - no extra solder. Put a very small amount of solder on the end of the iron tip. Press the pre-tinned wire end to the post or pad you want it connected to. Press the iron tip with the small amount of solder onto both the post/pad and the wire end until the solder flows. Pull the iron tip away, but keep holding the wire in place until the solder solidifies.

3) When you are done - check with a magnifying glass! It is easy for stray wire strands to bridge pads. No need to re-assemble the player yet - plug it in and turn it on - if the LED on the SDI board does not come on - something is wrong - check everything.

FWIW: I have both the Oppo 970 (480i HDMI) and the 971 (480i sdi modded) feeding a VP30 with SDI/ABT102 but unfortunately to get close to 1:1 pixel matching (1366x768) I have to go analog to my display and HDCP makes that impossible.

Try it at non 1:1 pixel mapped then (for both SDI and HDMI), its still an interresting comparison and probably wouldn't mask as much as you'd think.

Cheers,
John.

Maybe I'm mis-reading what you're saying, but you didn't compare 480/576i HDMI/DVI to 480/576i SDI, you compared the SDI feed to a 480/576p feed on HDMI/DVI. As Bob and others already pointed out this isn't a useful comparison as you have invoked the players deinterlacing and processing for the HDMI feed...

John.

Talman, is that b/c your display is analog, or for another reason?

My display's native resolution is 1366x768 but will only accept 480i, 480p, 720p and 1080i over the HDMI port. So the closest I can get is 720p.


Wow--just made me think of something I hadn't before. I have compared 480i SDI into RGBHV @ 1360x768 and 480i HDMI to HDMI @ 720p but I haven't compared 480i SDI into HDMI @ 720p. I'll have to do that tonight and report back.

FWIW: In the comparison I did do above (analog @ 1:1 vs HDMI at 720p) 480i SDI was easily the better picture. I won't say it completely smoked it but the difference was considerable.

Thanks, usualsuspects! It never hurts to be reminded of good technique... One other tip I'll add to your already excellent list is to wear a grounding strap when working on digital ICs and do the work in a non static electricity environment. You never know when even a small static charge will take out a valuable IC and ruin your new SDI modification board or possibly even damage your player.

Forgot about ESD! Good point. One other thing - I put about 3 turns per inch of twist on the clock and ground wires (separated them first). This has been recommended by some.

John, see my edited posts above.

http://www.avsforum.com/avs-vb/showt...&&#post8062922
http://www.avsforum.com/avs-vb/showt...&&#post8070076

I completely agree that you can only do a true comparison when BOTH SDI and HDMI 480/576 streams have NO processing involved in their output stream once the signal leaves the decoder. We know SDI is clean, it's just that so far we haven't seen a better or identical performance from HDMI at 480/576i. So we are all assuming (hoping?) that that is due to the fact that players tested so far all somehow process the 480/576i stream (beyond just adding HDCP) before it gets to the HDMI connector. We all agree (I think) that in theory there ought to be zero difference between the two.

So, I guess this debate can be viewed from 2 angles. One is more focused on what players we can buy off the shelf that have equal (or even better?) performance on HDMI 480/576i than that SAME player (or one which has the same decoder chip) fitted with SDI. This I think might rule out the OPPO 970H with HDMI vs the 971H with SDI, comparison.

The other is whether SDI is inherently better than the equivalent HDMI (with same colour space and bits of resolution) and if so, why the hell, as we can't see any reason why it should be!! This really requires say an OPPO 970H to be modded with SDI - nowhere near as simple as modding a 971H the Oppo people say. The connection points are all over the place and may in some cases only be available from a via, or worse still, from a buried via.

To cap off this discussion, say you did prove that one or the other was better on this particular player, a brand with a better decoder could usurp the results on either or both output formats. Don't forget too, that the differences may only be evident on some displays. The more scaling the display does, the more likely it is that the results will be masked by scaling artifacts.

In my comparison between the 971H DVI and SDI output, I got as close as I could to get the 2 equal, bar de-interlacing. I set the DVI res to match the SDI output, so no scaling should in theory be taking place inside the Faroudja. All other enhancements were set to OFF or ZERO, so again, no processing should be taking place.
But, isn't this exactly what we're on about here? We are assuming that no processing is taking place either inside the Faroudja OR the HDMIi garnering of the signal from the decoder to the DVI/HDMI connectors. It ain't going to be easy to find the perfect platform to do a just/scientific comparsion.