[joevt]

Good to know they fixed the HDCP 2.2 issue.

The firmware update instructions seem to have a mistake where button #6 is the same as button #2 when it should probably be the next button down.

I’m waiting for a reply from optomedia before I try this update (to make sure they use the same updater).

For my CMV-535, clicking the Switch buttons changes the USB product name from “HDMI2.0” (vendorID/productID/Vendor 0x5566/0x5566/Pana) to “USB-Serial Controller D” (0x67b/0x2303/Prolific Technology Inc.) to “WPM USB” (0x416/0xa316/NUVOTON).

[joevt]

Did the four EDID’s of the adapter change at all? Do any of the EDID’s mention 533 MHz? Maybe you need a firmware with an EDID that has a 533 MHz timing. Do they know what the update actually changes? Do they have a firmware bug list?

[joevt]

The gofanco amazon page notes: – “Supports 594MHz 4K/60Hz bandwidth (SMPTE 4K) displays, as SMPTE 4K is the format available for HDMI 2.0. If the displayport monitor does not recognize the 594Mhz bandwidth, it will not convert. Monitors operating in 533MHz 4K bandwidth also will not work, unfortunately.”

This doesn’t make sense because I can force my computer to use a 533 MHz timing with my adapter (CMV-535) (but I can’t verify that the adapter is not converting that to 594 MHz on the DisplayPort end – but doing that seems like a complication to be avoided by any sane adapter manufacturer). No one else here has tried the 533 MHz custom timing with a PC, their adapter, and their display. That seems like important information to have…

The SIIG and gofanco web pages don’t mention the 594 MHz limitation. I don’t see why this limitation would exist, since the adapters are made to support arbitrary resolutions and timings.

SIIG and gofanco have a firmware update to improve compatibility, but I don’t know if it fixes the problems people are having here. The firmware updates are the same, confirming that both are the same product. I might try it with my optomedia CMV-535. The only difference between the updaters is the model number mentioned in the instructions.

[joevt]

Archer333 reported that the Samsung U32D970 with the same 533.25 MHz timing as the Acer Predator XB321HK works with the SIIG adapter. It would be nice to know if either display can work with a 594 MHz timing from a PC’s DisplayPort.

[joevt]

The last we heard from you is that you were still trying to find an adapter.

Archer333 has the same LG31MU97 display. He tried the SIIG and a PS4 Pro and it didn’t work. I told him to do some tests with a PC. He hasn’t replied since. Maybe you can do the tests. In any case, if you get the SIIG adapter and it doesn’t work, then you may need to contact their tech support to get a firmware update.

[joevt]

To be clear, it’s reasonable that the SIIG adapter in it’s current state (uses 594 MHz) can’t support the XB321HK (which expects 533.25 MHz). But it is unreasonable that they haven’t tried a firmware update to enable the compatible timing of 533.25 MHz.

Their own product page says:

• Dual mode Micro USB port to power the adapter using the included power adapter, plus firmware update capability to ensure performance and compatibility

It seems they’re not trying very hard to “ensure compatibility.”

[joevt]

That’s a crappy explanation. The XB321HK’s DisplayPort EDID has a 533.250 MHz timing. Are they saying the PS4 Pro only outputs the 594 MHz timing reported by the adapter’s HDMI input and the XB321HK won’t support 594 MHz from it’s DisplayPort input? That would be reasonable since the XB321HK’s EDID indicates a max of 540 MHz.

It looks like when you tested with a PC or laptop that you did not try using a custom EDID for the PC’s HDMI port with a 533.25 MHz timing matching the XB321HK’s DisplayPort timing. If that works, then SIIG should give you a firmware update that replaces one of the four EDID’s with one that has the same 533.25 MHz timing.

I tried your XB321HK’s 533.25 MHz timing with my CMV-535 adapter and Dell display and it works fine. 594 MHz also works. 533.31 MHz too. I used SwitchResX in macOS to create the custom resolution. The SIIG adapter should also be able to use the 533.25 MHz timing. The problem is that you can’t create a custom resolution for the PS4 Pro, which is why you need a new firmware for your adapter to tell the PS4 Pro the correct timing. My Dell doesn’t display the pixel clock that it’s receiving (only the resolution and refresh rate), so I can’t be sure the adapter is faithfully reproducing the HDMI input signal on the DisplayPort output signal.

You said before that you didn’t get any image from any resolution from the PC or laptop through your adapter. That is bizarre. Did you try 1920×1080 or whatever? Custom resolutions? Did their tech support have an explanation for this?

[joevt]

Does the Philips BDM4065UC show the current resolution? The user manual says there’s a notification (that you can disable) when the display is powered on at a resolution that is not 3840 x 2160 @ 60 Hz but the manual doesn’t show what the notification looks like.

For linux, maybe you have to use the command line to create a new 4K 60Hz resolution? (randr or xrandr or whatever).

[joevt]

Yes, Titan Ridge TB3 will still be 40Gbps. It will probably still have two DisplayPort inputs, now v1.4, but both inputs can’t be using the max 25.92 Gbps on a single cable. Some configurations will require putting the second display on the other Thunderbolt 3 port.

The max for DisplayPort 1.2 is 17.28 Gbps so two of those fits fine on TB3 and should still work the same on Titan Ridge.

The most interesting thing about Titan Ridge is the variant of the chip that allows using a Thunderbolt 3 port as a USB-C input (JHL7440), so that a display using that chip can work as both a Thunderbolt 3 display and a USB-C display (same for docks). As a USB-C display, probably no PCIe device connected to the TB3 controller’s PCIe lanes will work. I’m not sure how USB devices will be connected (if the display is acting as a USB hub with internal USB devices such as audio, camera, etc. or has external USB ports). Will the display include a USB controller like current Thunderbolt 3 displays, which can switch to a USB hub, or the USB devices connected to the USB controller will switch to being connected to a hub from the TB3 controller? The TB3 controller could add a USB port, which the TB3 controller can connect via PCIe for Thunderbolt operation, or connect to the USB-C input for USB-C operation. Alpine Ridge has an integrated USB controller already, so Titan Ridge can build on that. In USB-C mode, JHL7440 would need pins for a USB connection or switch the output of one if it’s Thunderbolt 3 ports. Or maybe USB-C mode will only allow DisplayPort? – that would be sad. Intel said Thunderbolt 3 docks will also benefit from USB-C mode, so presumably USB devices will work, and not just the DisplayPort. Intel says “mobile employees could go into any shared office area or conference room with a JHL7440-based Thunderbolt 3 dock, and gain access to a larger monitor, faster Ethernet connectivity, and keyboard/mouse regardless of the laptop they have.” I assume that the laptop can be either Thunderbolt 3 or USB-C and that this an example of what Intel means by “basic compatibility” for USB-C mode.

[joevt]

Right. There are other USB-C DisplayPort 1.4 alt mode chipsets besides Titan Ridge. I don’t know if anything uses any of them yet.

When they exist, you’ll be able to test your new USB-C to HDM 2.0b adapter, but also your existing DisplayPort to HDMI 2.0b adapters (the CAC-1080 or CAC-1180) with a USB-C to DisplayPort 1.2 adapter (as it’s more likely just a USB-C to DisplayPort HBR2 adapter which is all the CAC-1×80 requires).

All these adapter use the MegaChips MCDP2900? That means they are stuck using four lanes HBR2 for 4K content. This makes them unusable with future USB-C docks that might support DisplayPort 1.4 since 2 lanes are used for USB 3.0/3.1 in a dock. Well, HBR3 is not twice as fast as HBR2, so 2 lanes of HBR3 can only do 4K 60Hz with reduced color (8bpc) or reduced chroma sampling (4:2:2 or 4:2:0) or Display Stream Compression.

[joevt]

Is there processing done (with a receiver or whatever) on the DisplayPort signals between the USB-C and DisplayPort connectors? The “documentation” at https://www.vesa.org/wp-content/uploads/2016/10/USB-DevDays-DisplayPort-Alternate-Mode-2016-final4.pdf does not show this.

What’s the difference between a USB-C to DisplayPort adapter that can do HBR3 and an adapter that can do HBR2? There is no documentation, so another question is: What can be different? Answer:
1) Wiring quality
2) USB 2.0 billboard device
3) Power Delivery (PD)
4) Something else?

#1 might not be a factor since the difference between 5.4 Gbps and 8.1 Gbps is not so great. You might get a lot of glitches if the quality isn’t good enough – but it should do something if nothing else blocks the alternate mode from working.
#2 doesn’t matter since the billboard device doesn’t do anything unless an alternate mode is not supported.
#3 might be the answer. There are vendor-defined messages (VDM) that describe the alternate modes supported by an adapter. The alternate mode specifications are not part of the USB specifications that are publicly available. The alternate mode specifications for Thunderbolt (Intel), DisplayPort (VESA) and MHL are not publicly available.

What could be different in the VDM’s between an adapter than can support DisplayPort 1.3/1.4 and one that can only support DisplayPort 1.2? I found a document at
http://www.ti.com.cn/cn/lit/an/slva844a/slva844a.pdf
which seems to imply the existence of a DisplayPort Capabilities Message that has a flag for DP v1.3 signaling rates but there’s not enough information to know if that is different for a DP 1.2 adapter. The actual DisplayPort alternate mode specification is required to know for sure.

Moshi has a USB-C to DisplayPort 1.4 Cable (bi-directional) (but the DisplayPort connector is male instead of female). I suppose a USB PD sniffer can be used to compare it with any of the other USB-C to DisplayPort adapters…

[joevt]

1) You have to test if the display can accept a DisplayPort signal using the converter’s HDMI timing. If not, then that’s probably the problem.
2) If that is the problem, then you should test if the display can accept a DisplayPort timing using the Display’s DisplayPort timing (of course that should work).
3) Then test if the display can accept a DisplayPort signal from the converter using the display’s DisplayPort timing.
If #3 doesn’t work, then the problem is not a timing issue but something else with the adapter.

[joevt]

The LG has a 4K timing that is slightly different than the Samsung or my Dell:
LG:
“3840×2160” 533.280 3840 3848 3992 4000 2160 2214 2219 2222 +hsync -vsync
Samsung:
“3840×2160” 533.250 3840 3888 3920 4000 2160 2163 2168 2222 +hsync -vsync
Dell:
“3840×2160” 533.250 3840 3902 3950 4000 2160 2163 2168 2222 +hsync -vsync

There are other differences but I don’t see anything strange. The LG, Samsung, and DP don’t have a VSDB.

[joevt]

The CAC-1504 should work fine with JHL7x40 (“Titan Ridge”) since JHL7x40 still supports DisplayPort 1.2 devices.

JHL7x40 adds DisplayPort 1.4 so it should support HDR from an adapter like the CAC-1080 or CAC-1180 (with an appropriate USB-C to DisplayPort adapter – any current USB-C to DisplayPort adapter should work since the CAC-1080 and CAC-1180 only use HBR2).

Should any USB-C to DisplayPort adapter be able to do HBR3? Or does that require a special adapter?

[joevt]

Thanks for the EDID. It doesn’t seem to do anything different/better than the BenQ
s EDID except that the BenQ marks 1920 x 1080p at 60Hz as Native in the CE video identifiers (VICs) even though the BenQ EDID also specifies preferred timing of 3840x2160p at 60Hz. The Acer doesn’t have a VIC marked as native so either the BenQ and Acer have different problems, or they have the same problem that is none of the above.

I guess the adapter doesn’t have a problem with the VSDB as I originally suspected since the Philips (which works) has a similar VSDB to the BenQ (except the BenQ uses CEC physical address of 2.0.0.0 instead of 1.0.0.0 like the Philips).

[Author]

Posts