Is HDMI V1.4 mandatory for 3D images?

I believe you do need 1.4 as it's a mandatory spec for 3D in 1080p/24. 1.3 will allow for 1080i 3D.

IMHO w/ so much inertia already behind HDMI 1.3 with various TV and sources in everyone's home it wouldn't surprise me to see 3D support in legacy HDMI 1.3.

I believe that's right.

1.4 HDMI source and 1.4 HDMI displays are needed. 1.3 cabling is OK.

HDMI 1.3 cables support everything in 1.4 except the new ethernet channel.

To watch 3D Blu-Ray, HDMIs on source and on display must support "3D over HDMI", a set of protocols defined in HDMI 1.4 standard, but that can be supported by earlier HDMI chips (like PS3, for example).

There are not HDMI 1.3 or 1.4 cables, according to HDMI consortium. The cable you need for 3D Blu-Ray is "High Speed", the usual 1080p@60Hz cable most of us have in our home.

Many call "High Speed" cables "HDMI 1.3" and ""High Speed with Ethernet" cables "HDMI 1.4", but there is no difference for 3D or for bandwidth. The only difference is the HEC (HDMI Ethernet Channel) in the second.

There's also an audio return channel in the 1.4 spec.

Isn't it true though that a AVP which is designed for 1.3 will not appropriately pass a 1.4 3D signal, meaning that you must use a direct connection from the player to the display and at least so far, cannot use your video processor to switch between sources and formats? This renders useless the video portion of all of our video switch and processors, at least for 3D movies. You would need two HDMI outputs on your source, one to feed the audio to your processor/receiver, and one to feed your display.

Stepping out on a limb on this one . . .

According to Silicon Image's lineup of HDMI RX (Receiver) and TX (Transmitter) chips, they make a 1.3 TX chip that is 3D compliant, but they don't make a 1.3 RX chip that is 3D compliant - for those listed as being used in AV Receivers:

http://www.siliconimage.com/products...t.aspx?pid=109

So, IMO, the problem lies with a 1.4 3D signal not being understood by the HDMI Input chip - the 1.3 RX chip which means the signal (at least the 3D video portion) will stop at the RX chip and go no further, not allowing the 3D video to continue on to the Output side of the AV Receiver. The signal may stop altogether and not pass the HD Audio (just a guess)

I agree with as much as I know about it which is minimal. My other concern is that it seems a given the video would not pass through a 1.3 receiver, but would a 1.3 Receiver still recognize and decode the audio from a 1.4 output from a player if I use the other HDMI from source to the video player??

I just sent an Email to Silicon Image through one of their distributors asking this question. Hopefully they will respond.

No one has mentioned the reason for this issue; HDCP. The differences in handshaking between HDMI 1.3 and 1.4 due to the 3-D negotiation (does the display support 3-D and what protocols does it support) will not be understood by the receiver and SECURITY is paramount so video will not be passed.

I have found no information saying that HDCP has changed in HDMI 1.4. 3D negotiation is part of the "3D over HDMI" set of protocols, with a discovery part implemented in new VSDB fields in display EDID and a signaling part implemented in InfoFrame new fields.

But these changes have nothing to do with HDCP.

What, if anything, prevents using a 3D-PC for a non-HDMI 1.4- display? 3D sources, such as firmware-upgraded Sony PS3s, presumably would handle HDMI 1.4 3D signals. The PS3 3D Blu-ray output would feed the 3D-capable computer input (HDMI 1.4-capable, if needed). But, rather than HDMI from the computer, its 15-pin mini D-sub output would feed the 15-pin PC input of displays cable of accepting 120-Hz vertical scanning rates. I was thinking of my '09 VX100 pro plasma PC input, likely with HDMI 1.3 inputs.

BTW, if HDMI 1.4 electronics is required for 3D, how could all the ancient cable STBs used to deliver 3D from the Masters golf match recently have worked? They have, like my 8300HD STB, HDMI outputs, but mine's >4 years old and couldn't have HDMI 1.4 3D-capable electronics--unless 8300HD etc STBs nationwide were all firmware upgraded to 1.4. -- John

I haven't fully understood the first part of your post and don't have the time now to read the link (I'll do tomorrow), but for the second part, remember that Masters have used a frame compatible 3D format (Side-bySide half), that, from the STB point of view, is just a 2D frame that happens to contain two images; the STB doesn't care and send it to the TV. If you manually set the TV input to Side-by-Side, you can watch 3D without a HDMI 1.4.

In other words, frame compatible formats are a special case. Blu-Ray 3D are not.

There are no 2D HDTVs with HDMI 1.4 interfaces that will accept and display the content separatly for each eye using either 2 frame packed buffer or 1/2 R side by side format and use acfive shutter glases to insure that only the right eye sees the right eye content and only the left eye sees the left eye contents
This capability should not be confused with the small( less the 32") 120HZ PC monitors designed to use the Nvidia 3D video gaming application that can receive 120fps content over a dual link DVI interface for games.

Thanks for the details. That link about my VX100 plasma is a generalization from your earlier post (sublinked) suggesting 720p/1080i@120 Hz seems possible with my '09 VX100 pro/custom plasma (non-3D model) that has a 148.5-MHz dot-clock listed for several video HDMI formats. It's unclear, if 1.4 coding electronics is a 3D requirement, how to bypass VX100 HDMI 1.3 input electronics. I proposed--not too optimistically--using a computer's 15-pin output to the plasma's 15-pin input rather than using HDMI. [EDIT: Perhaps I'm mistakenly mixing RGB computer graphics and HDMI video concepts here.]

Yes, my 8300HD cable STB is connected with YPbPr and I watched a Masters golf snippet in the side-by-side format on my 65VX100 plasma (non-3D). Plasma was in its usual component input mode. So, presumably, if I feed that YPbPr side-by-side input (or HDMI?) into a 3D computer for sequential-frame conversion I could watch Masters video, bypassing HDMI. Perhaps you're indicating HDMI 1.4 from an up-to-date 3D computer to my HDMI 1.3 should work?

Spoke with a Matsushita rep (Panasonic) who insisted the VX100 can't display 3D. But the rep, who'd contacted their tech folks, wasn't that tech-acute, and we mostly discussed why their dual-link HD SDI board wouldn't help and the need for a HDMI 1.4 circuit in the plasma in order to pass 3D inputs.

(Wish I'd mentioned during the call that Panasonic needs an add-on 3D HDMI 1.4 plug-in board, with 3D-conversion chips and active-glasses signal, for all its pro 120-Hz-capable plasma monitors--akin to what Mitsubishi is releasing for its older DLP displays 3D-checkerboard-format.) -- John

It's still not clear what the display limitation is, assuming we're discussing the output of a separate 3D computer, complete with an IR emitter (or cable) for active glasses. Two weeks back you didn't seem to disagree with Mike5's conclusion that plasma monitors like mine had adequate bandwidth, as do most pro monitors. AIUI, from Mike5's post, 148.5 MHz bandwidth is needed for 720p/1080i@120 Hz, double that of standard HDTV.

Several "applicable input signals" in my VX100 manual (pp 50--51) indicate a dot-clock of 148.5 MHz. And one is speced at 175.5 MHz for a 1600X1200 (UXGA) component RGB input using the 15-pin PC input, although that's at a 65 Hz vertical rate, not 120, so UXGA downscaling would be necessary for 3D's 120 Hz vertical scanning frequency--(not to exceed ~175 MHz bandwidth). [EDIT: Perhaps I'm mistakenly mixing RGB computer graphics and HDMI video concepts here.]


Not sure, but the two pages of applicable input signals may represent a built-in lookup table for automatic conversion of these 63 different scanning formats to the plasma's native 1920X1080p. But the specification page 52 lists a range of scanning frequencies for PC input signals, such as 48--120 Hz vertical and 15--110 kHz horizontal.

Presumably my '09 2D pro plasma has HDMI 1.3, not 1.4. If computer 3D hardware/software prevents a downscaled UXGA 3D output (or similar) via a 15-pin mini-D-sub, from 3D sources, that clearly prevents 3D display. My plasma has no dual-link DVI options and the dual-link SDI plug-in card instructions don't show a 148 MHz or greater capability. -- John

John--You need to remember that refresh rate alone is not any guarantee of 3D compatibility. You can get shutterglasses to sync with just about any display technology, that's not the issue. It's crosstalk that is the real issue, and plasmas have phosphor trails that are going to ghost like crazy. High refresh rates serve to reduce or eliminate flicker, but in and of themselves can't eliminate ghosting.

The new generation of 3D plasmas have panels that are specifically designed for quick phosphor decay. The panel manufacturers had to go back to the drawing board on this. Older generation sets are simply not going to work well for 3D, if at all.

I have an 09 Samsung plasma that is specifically labeled as "3D Ready", and crosstalk on it is awful, simply because of the decay rate. The image looks great -- 120Hz and absolutely no flicker -- but the severe ghosting absolutely kills the 3D effect.

So even if Panasonic could give you all the things you wanted to ask for, they still could not change the decay rate of the phosphors in their old sets. Mistsubishi is able to provide backward compatibility to their customers simply because DLP technology, by its very nature, is more inherently 3D-friendly.

Thanks. Wonder if there's any difference with high-end pro monitor decay rates? Hard to understand why they'd spec them for up to 120 Hz vertical scan rates if they'll ghost badly. Lots of pro plasma AVSers but so far haven't read of anyone trying them for 3D. Also, AIUI, plasma phosphors are considerably faster than LCDs in monitors available for 120 Hz 3D gaming, etc. (BTW, read a recent post here, about hacking active glasses, from someone who added a microprocessor plus timing adjustment to his setup and apparently dialed out most of his crosstalk and/or ghosting.)

On thread topic, from Mike5's reply, it seems important to clearly distinguish 1.4's necessity for Blu-rays versus other 3D sources. -- John

Don't know. My Samsung does 120Hz in 3D mode. As I said, no flicker but lots of ghosting. Oh and I should also mention that the R,G, and B components of the plasma phosphors decay at different rates, so you get not only ghosts, but oddly colored ones at that.

^^^Sounds horrific. If marketed for 3D ability and everyone sees this, it's surprising there hasn't been a class-action lawsuit. -- John

The 3D capabilities on Samsung's PN*450 series plasmas were an afterthought at best. I suspect they simply re-purposed the 3D circuitry from their RPTV DLPs (where it did work well) simply to give these models another selling point. It was probably easy to do so they did it. It's not like many people were equipped to even try out the 3D mode on these sets at the time.

Interestingly, Samsung dropped 3D "readiness" on this same model in their current line, even though all other aspects of the sets are the same as the earlier "3D ready" models.

There's no basis for any kind of suit. It does what it is advertised to do -- it accepts and displays 3D signals. It just does it in a manner that you wouldn't want to watch for very long. They never said it was good 3D.

I watched the 3D Masters from Masters.com on my PC and displayed in on my 2009 2D LCD HDTV using analgraph Blue/Red overlayed content and Blu/Red glases
I was also able to receive the EDPN-3D Side.by-side content from my cable company and see that SBS content on my TV.
Certainly I would have preferred to have a new TV capable of displaying SBS 3D using active shutter glases controlled by the TV using an IR emitter that is built into the TV's Bezel.
Since all HDTVs can not receive anything higher then 1080p/60 resolution and they have some lag in image response time unless the active shutter glases are controlled by the TV itself then the glases can not be in sync
Since there is no legal defintion of what a 3D TV or a 3D optical disk is and the US Consumer Electronics Association has not deemed it fit or their responsibility o publish a glossary of terms for all of it's members to follow the american TV consumer will continue to be completly confused.
Bandwith is not the issue since both the 1.3b and 1.4 support support the same maximum bandwith. Signal content and ability sync left and tight eye screen images with glases is the issue.

John, it could be useful to read this description of what I call Type A and Type B displays.

Type B are definitively different form 2D displays. They have HDMI 1.4 and can understand new 3D formats like Frame Packing, managing the transformation in 3D presentation with their intenal electronics.

Type A are more similar to 2D displays. The main difference is they accept 120Hz in input. They generally are not able to synchronize the glasses, but a HTPC with nVidia card does it. To achieve 1920x1080@120Hz they need a Dual-Link DVI (or equivalent). These displays (especially LCD monitors) suffer a lot of ghosting.

A 2D display has many features different from a 3D display:
- it hasn't got a HDMI 1.4, so understating Frame Packing of Blu-Ray 3D is ruled out;
- it (generally) doesn't accept a 120Hz input, so systems like nVidia 3D Vison are ruled out.

Even if it accepts a 120Hz input (no idea of which 2D HDTVs have this feature):
- it hasn't got a 297MHz input (like a Dual-Link DVI), so at least 1080p@120Hz is ruled out;
- for 720p/1080i@120Hz it remains the problem of glasses synchronization; 2D HDTV don't have a VESA port or other connection for an emitter, so you should use anyway a HTPC to synchornize the glasses;
- a 2D HDTV, as has been pointed out, is not optimized for 3D (e.g.: black frame insertion for LCD, reduced phosphorus latency for plasma, ...); the result would be a ghosting worst then in Type A displays;
- some problem would probably exist for internal electronics to manage the process, too.

If we want a quality 3D we need a 3D display, otherwise we can always use anaglyph.

Thanks for reviewing all this. Relating to OP and need for 1.4, it's been my assumption that 1.4, when needed, would be part of any 3D-computer setup used for attempted 3D on a 2D display. So, for 720p@120 Hz, the 2D display must accept that 120-Hz vertical scan rate and up to 148.5 MHz bandwidth (per our earlier discussions).

From the applicable-signal tables (PC input section) for my Panasonic TH-65VX100U pro plasma, linked earlier above, it appears VX100s and likely other pro plasma monitors can accept such a frame-sequential 3D signal (3D-computer output)--perhaps, on VX100s at least, on the 15-pin mini D-sub jack. AIUI, frame-sequential 3D is progressive, so 1080i@120 Hz seemingly wouldn't work, assuming such an input bypasses the usual plasma deinterlacing. And if deinterlaced externally from the plasma 1080i@120 Hz would require downscaling to reduce the 297-MHz bandwidth need--or else the dual-link inputs you mention.

Clearly, the 3D computer must provide a sync signal for active glasses. And maybe, for sync effectiveness, built-in timing adjustability would help: Noticed in some software/hardware modification links , that someone adding a variable-delay microprocessor circuit benefited:

Perhaps 2D-plasma phosphor decay and ghosting is the main problem--although, from the VX100 specs noted above, its phosphors were designed to handle 2D up to 1600X1200@65 Hz with a 175.5-MHz dot clock; (vs., say, 3D's 1280X720@120 Hz and 148.5-MHz dot clock.).

Anaglyph? No thanks here. :-) But staying tuned for any pro plasma owners with 3D PCs who've experimented with anything but red-blue glasses. -- John

Mike...It seems to me there is yet another type, Type C: the "3D Ready" displays (Mitsubishi's and Samsung's DLPs and Samsung's '07-'09 plasmas).

These accept checkerboard input (and some also accept column interlaced and row interlaced) @60Hz at whatever screen resolution is native to the display. Internally, the display de-multiplexes the signal into left/right image pairs and displays them sequentially @120Hz (albeit at half resolution).

These displays also provide the sync pulse, either via IR emitter or DLP-link. In this case the source cannot provide the pulse, since the source is running at 60Hz but the pulse needs to run at 120Hz.

Yes, it's true. I didn't mention them because I know very little about them. They are not sold in Europe and, AFAIK, are sold only in North America (don't know why).