HDMI to DisplayPort 38146 EDID problem

VGA - HDMI - Wall Plates - Speakers - Switches - Splitters
15 posts Page 1 of 1
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Wed Jan 24, 2018 5:17 pm
Hi,

I have a number of HDMI to DisplayPort adapters, model 38146. They all have the same issue. They appear to be overwriting one bit of the EDID data from my monitor (and correcting the checksum at the end). Unfortunately, the effect seems to be to corrupt an important part of the data.

Below, I have provided the raw EDID data read 1) directly from the monitor, and 2) via the Lindy adapter. You can see that byte 157 (a.k.a. byte 29 of the CEA extension block...) has changed from 6D to 65. I believe this byte is the header of a Vendor Specific Data Block, and the one bit change seems to have changed it's length. I have tried two EDID decoders, and neither can make sense of this changed EDID. The effect is seen on attached sources. For example, a UHD (4k/30) bluray player will play at 4k/30 direct to the monitor, but will only play at 1080p with the altered EDID.

We are considering standardizing on this adapter for use with our upcoming products, but this problem might prevent that. Can someone take a look at it?

Thanks,
Nick

EDID read directly from the monitor:

00,FF,FF,FF,FF,FF,FF,00,10,AC,EF,40,4D,37,33,31,29,1A,01,03,80,3E,22,78,EA,08,A5,A2,57,4F,A2,28,
0F,50,54,A5,4B,00,D1,C0,A9,40,81,80,81,00,71,4F,01,01,01,01,01,01,04,74,00,30,F2,70,5A,80,B0,58,
8A,00,6D,55,21,00,00,1E,00,00,00,FF,00,4A,34,32,4D,43,36,41,45,31,33,37,4D,0A,00,00,00,FC,00,44,
45,4C,4C,20,53,32,38,31,37,51,0A,20,00,00,00,FD,00,1D,4B,1F,8C,1E,00,0A,20,20,20,20,20,20,01,4C,
02,03,2B,F1,50,10,1F,20,05,14,04,13,12,11,03,02,16,15,07,06,01,23,09,07,07,83,01,00,00,6D,03,0C,
00,10,00,00,3C,20,00,60,01,02,03,04,74,00,30,F2,70,5A,80,B0,58,8A,00,6D,55,21,00,00,1E,56,5E,00,
A0,A0,A0,29,50,30,20,35,00,6D,55,21,00,00,1A,4D,6C,80,A0,70,70,3E,80,30,20,3A,00,6D,55,21,00,00,
1A,11,44,00,A0,80,00,25,50,30,20,36,00,6D,55,21,00,00,1A,00,00,00,00,00,00,00,00,00,00,00,00,32

EDID read from same monitor, via Lindy adapter:

00,FF,FF,FF,FF,FF,FF,00,10,AC,EF,40,4D,37,33,31,29,1A,01,03,80,3E,22,78,EA,08,A5,A2,57,4F,A2,28,
0F,50,54,A5,4B,00,D1,C0,A9,40,81,80,81,00,71,4F,01,01,01,01,01,01,04,74,00,30,F2,70,5A,80,B0,58,
8A,00,6D,55,21,00,00,1E,00,00,00,FF,00,4A,34,32,4D,43,36,41,45,31,33,37,4D,0A,00,00,00,FC,00,44,
45,4C,4C,20,53,32,38,31,37,51,0A,20,00,00,00,FD,00,1D,4B,1F,8C,1E,00,0A,20,20,20,20,20,20,01,4C,
02,03,2B,F1,50,10,1F,20,05,14,04,13,12,11,03,02,16,15,07,06,01,23,09,07,07,83,01,00,00,65,03,0C,
00,10,00,00,3C,20,00,60,01,02,03,04,74,00,30,F2,70,5A,80,B0,58,8A,00,6D,55,21,00,00,1E,56,5E,00,
A0,A0,A0,29,50,30,20,35,00,6D,55,21,00,00,1A,4D,6C,80,A0,70,70,3E,80,30,20,3A,00,6D,55,21,00,00,
1A,11,44,00,A0,80,00,25,50,30,20,36,00,6D,55,21,00,00,1A,00,00,00,00,00,00,00,00,00,00,00,00,3A
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Thu Jan 25, 2018 4:45 pm
Hi Nick,

The EDID should be passthrough on the adapter although some properties may slightly change as the adapter needs to inform the source device that it is only capable of converting resolutions up to 3840x2160 30Hz 8bpc.

If you could let me know what source device and display you are using I can have a further look into this for you. What software are you using to read the EDID also please?

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Fri Jan 26, 2018 9:40 am
Hi Guy,

We are testing the Lindy adapters on the displayport input of our own video processor. We can emulate any EDID and accept a very wide range of inputs.

In this case we are emulating the EDID of a Dell S2817Q. We do this because we are testing a 4k/30 source (Samsung upscaling blu-ray player) and we know that this source sends perfect 4k to the Dell S2817Q.

However, using the Lindy adapter prevents the blu-ray player from sending 4k. Yet the adapter is supposed to be capable of 4k, so I don't understand why it alters the EDID in this way. The result also seems to be 'illegal' EDID,

We have many ways to read EDID, including our own products. But you can use Moninfo (it can read any EDID attached to your PC, or can read EDID from a file, and provide a decode). We have tried different methods and got consistent results.

Thanks for your help,

Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Fri Jan 26, 2018 9:45 am
Hi Nick,

It could be an HDCP issue. Generally for a bluray player to send 4K video it would use HDCP 2.2 however if this version of HDCP is not available it will clock back to the HDMI 1.4 standard and send 1080p video.

The adapter supports up to HDCP 1.1, it may be worth checking what version your player uses.

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Mon Jan 29, 2018 4:17 pm
Hi Guy,

That is a very good point - the blu-ray player might be falling back to 1080p due to lack of HDCP2.2 rather than the damaged EDID. But the EDID is still broken, and (maybe by coincidence) the part that is broken is the vendor specific data block (VSDB) which tells the source about 4k support and TMDS max clock.

I'd still like to know if Lindy would be interested in addressing this issue? We would like to standardize on this product, but the damaged EDID is a negative factor for us..

Thanks & regards,
Nick
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Mon Jan 29, 2018 4:52 pm
Update: I just ran the bluray player into a device that does not support hdcp2.2 and it still plays 4k (it is an upscaling 4k player, so this is not too surprising), so I am now pretty confident that the damaged EDID is the root of the problem when the Lindy is in the signal path.

Cheers

Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Tue Jan 30, 2018 11:26 am
Hi Nick,

I've checked one of our blurays to a 4K 30HZ monitor today and confirmed that if I go direct HDMI I can achieve 4K 24Hz without issue, however when I go through the 38146 it drops down to 1080p 60Hz, which is consistent with your testing.

I think HDCP is the issue rather than the EDID (the EDID will vary slightly between your monitors HDMI port and DP port). When we connect direct our player outputs HDCP 2.2 even though our display is 4K30Hz under HDMI 1.4, however when we go through the DP converter it drops down to HDCP 1.4 which limits to 1080p.

Unfortunately if this is not suitable for your project I can only recommend returning it for a refund. Sorry I can't be of more help on this occasion.

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Fri Feb 02, 2018 3:21 pm
Hi Guy,

Don't worry about the refund, we are planning to buy hundreds of these adapters if/when we qualify them. But we know for sure that our issue here is caused by the adapter altering/damaging the EDID. For sure, this is not an HDCP2.2 issue. (See my message above - we already eliminated that possibility). In addition, we have forced the adapter to take 4k (at many frame rates and timings, including 30, 25, and 24 Hz) and it seems to work fine. So we think the damage to EDID (which removes 4k support in the CEA block) is not intended.

Is there some way I can escalate this inside Lindy, to make them aware of it?

Thanks for your help,
Regards,
Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Fri Feb 02, 2018 4:25 pm
Hi Nick,

I work in the Technical Support department here at Lindy UK so we are aware of this enquiry, if we require any further information we can talk directly to our suppliers and product managers.

I've had one of these out to check today. We have read the EDID of a display connected direct by DisplayPort, and then compared that to the EDID read through the 38146 adapter in the same set-up.

There are 3 hexadecimal entries that change when adding the adapter: Revision Number, DFP Compatibility and Checksum.

Revision Number - This is the revision of the EDID chipset used in the sink device, this changes because the source device will read the chip in the 38146 rather than the display, this would not affect any functionality.

DFP Compatibility 1.x - When we connect direct to our monitor we can see that it is supported as it reads "DFP Compatibility 1.x = 1". When we go through the 38146 this changes to "DFP Compatibility 1.x = 0" which means it is not supported by the converter chipset used inside the adapter. This is a feature of the VESA Enhanced EDID Standard and means that the interface signal is compatible with VESA DFP 1.x TMDS CRGB, if your player requires this feature then that would be why you can only achieve a lower resolution.

Checksum - A checksum is derived from the hexadecimal string used, because the revision number and DFP compatibility entries have changed, the checksum also changes so that as a whole the EDID Block is still valid.

We have checked our other HDMI to DisplayPort Converter (Part no. 38147) and unfortunately the result is the same.

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Mon Feb 05, 2018 10:06 pm
Hi Guy,

That is great, that you are able to do similar tests there. Thanks for spending time on this..

We did try this with a different EDID and got a different result. Like yours, it dropped revision 1.4 to 1.3. It always corrects the checksum.

But please also see our results (shown in the hex dump in my first post). The adapter does different things to different EDIDs. This particular EDID is already 1.3 so it does not overwrite that field. But would you agree with me that the length count in the vendor specific data block in the CEA block has been changed? And if so, would you agree that this makes an invalid EDID? I'm no expert on this but I have used two different decoders are both confused by the changed EDID. One gives some rubbish, and the other just gives blank fields after the changed byte. Neither report "TMDS clock = 300MHz", which is what the original EDID gave - but why would the adapter do that, when it is clearly capable of 300MHz?

My colleague will be at ISE where he hopes to discuss how we can cooperate on testing these devices, and on a couple of problems we have had.Will you be there?

Thanks again!
Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Tue Feb 06, 2018 9:53 am
Hi Nick,

I've tried reading your 2 EDID blocks this morning however it doesn't appear to work, are you able to save the EDID direct from the monitor and then through the 38146 as a .bin file please? If so, please could you email these to 'technical@lindy.co.uk' for my attention and I will have a look at them?

Did you also find out which HDCP version your player requires in order to support 4K?

As far as we can tell the only parts that change when the converter is put inline are the revision and checksum which is normal for any converter as the chipset will be read, DFP is an optional feature as part of the enhanced EDID and our adapter just doesn't support it so again this isn't really a problem with the product.

When we read the EDID both from our display and through the adapter the supported clock rate does not change so I cannot say why this is not being shown for you.

While LINDY will have a presence at ISE, none of our Technical Support teams will be attending. So for an issue like this you would be better emailing us directly at the above address.

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Tue Feb 06, 2018 3:26 pm
Hi Guy,

I emailed you two files now (each is an EDID that you can read into, for example, Moninfo), along with some explanation.

We have looked into the possibility of hdcp 2.2 issues, but eliminated that. This test set-up uses an upscaling blu-ray player, and it is not protecting upscaled 4k playback with hdcp 2.2. To prove this, I played it into a sink device that definitely does NOT support hdcp 2.2, and it plays fine at 4k. But if I use the EDID from the Lindy, it refuses to allow 4k.

Thanks & regards, Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Tue Feb 06, 2018 4:16 pm
Hi Nick,

I don't appear to have received any emails with EDID files attached yet unfortunately.

In regards to HDCP, when we use a bluray player set to upscale it will output HDCP 2.2 + 1.4 simultaneously when sending to our HDMI 1.4 panel on the HDMI input port and we can achieve 4K 24p despite this display being HDMI 1.4 (HDCP 1.4) only. However when we go through 38146 to the DisplayPort input, it will send HDCP 1.4 only and gives us a maximum of 1080p 24p or 60p.

So although our panel does not support HDCP 2.2 on any port (HDMI or DP), our bluray is outputting that signal and allowing 4K to be sent but only to the HDMI input. 38146 supports up to HDCP 1.1 whereas DisplayPort support was only added to the HDCP standard as of version 1.3 and most modern bluray players, especially new 4K or upscaling players will output HDCP 1.4 as a minimum.

If you can re-send the EDID files I'm happy to have a look at the but I think that HDCP is the reason why 4K is not possible from a bluray through this adapter, because only v1.1 is supported and you would need at minimum v1.3 or more likely v1.4.

Guy
NickG
New member
Posts: 8
Joined: Wed Jan 24, 2018 4:59 pm
by NickG » Tue Feb 06, 2018 7:59 pm
I sent the email earlier, with two .inf files attached. I wonder if your email system has blocked it because of the attachments. I'll resend with the attachments renamed to *.txt - pls change back to .inf before importing in the application.

The reason I know that the issue is not HDCP, is that I have emulated both EDIDs (the original Dell monitor EDID and the 'modified' EDID presented by the Lindy adapter). I emulated them onto the input of a video processor that supports HDCP 1.4. The blu-ray player would only play 4k when the un-modified EDID was presented.

Let me know if/when you get the re-sent email - thanks,

Regards,

Nick
LINDY Guy
LINDY Staff
Posts: 133
Joined: Thu Oct 30, 2014 1:10 pm
by LINDY Guy » Wed Feb 07, 2018 9:38 am
Hi Nick,

We've received the files and will have a look into it today. I'll get back to you via email.

Guy
15 posts Page 1 of 1