TOPIC:

I am running a nanoAVD HD into a Blackmagic bidirectional SDI to HDMI converter from the HDMI side. That is. I am converting to SDI. This converter also reflects the SDI converted back to HDMI on a separate output if there is no SDI input. So: HDMI to SDI and HDMI.

If I drive it from a nanoAVR HD or nanoAVR DL I get horrible audio and video stuttering. I wonder if the nanoAVRs are not getting proper EDID from the converter.

I've checked that the output of the nano AVRs does not stutter, nor the output from the HDMI to SDI converter if the nanoAVR is not in the HDMI chain.

Anyone have any ideas? Can @devteam comment on what the nanoAVRs expect on the HDMI outputs? Maybe I just need an HDFury DrHdmi to give them EDID they like.

Are you able to drive it with different video resolutions / frame rates to see if some work while others don't?

I seem to recall there was once an issue with 24fps but I believe that was solved, but maybe it's something along those lines.

Just the standard output of a PC through an OPPO BDP-103D and out the secondary (audio) HDMI port. Supposed to be a hard 1080p. I tried from the PC directly, and changing HDMI cables to no avail. The combination of the nanoAVR (either model) and the minidsp HDMI to SDI side of a bidirectional converter does not play nice.

I also considered that the output audio sample rate of 96 khz from the nanoAVR HD might be the culprit but I have the same issue with the nanoAVR DL which has an output audio sample rate of 48 khz.

SDI audio is always 48 khz and I wondered if the HDMI audio sampling was an issue at 96 kHz.

Knowing how the nanoAVR manage EDID might give a clue.

I took my Ubuntu laptop and connected to the downstream Blackmagic bidirectional HDMI to SDI converter on the HDMI input and retrieved EDID data. It was very basic: one modeline of 1366 x 728. I know I can drive the device with a known HDMI signal. I suspect the nanoAVR is trying to negotiate EDID and is failing. I will try putting an EDID emulator between the two.

@devteam: being able to set a default EDID on the HDMI output of the nanoAVRs (or to capture EDID from a monitor/TV and remember it) would be a nice feature and save me the trouble of an in line EDID emulator. Would that be difficult? It appears you already negotiate it.

Well, this is odd.

This morning I captured a far more reasonable EDID from the Blackmagic bidirectional SDI to HDMI converter (on the HDMI input side, obviously). Yet, I STILL can't get the nanoAVR to send a stable signal to it. I tried capturing the EDID from a monitor (to which it was happy to send video) and put the EDID emulator between the nanoAVR and the Blackmagic converter, and still no love: I get a rolling display, and stuttering audio (The Blackmagic converter will decode the SDI output back to the HDMI output if there is no SDI input, so I can loop through the device). It almost looks like a 59.97 vs 60 Hz issue.

@devteam: what video do the nanoAVR units send on their HDMI outputs? Do they try to negotiate EDID? Knowing that would be helpful.

So I tried to capture the EDID of my display and force that between the nanoAVR and HDMI to SDI converter with an EDID emulator to no avail. I will try capturing the EDID of the converter, strip out all the 50 Hz settings, program that into the EDID emulator and see if that works between the nanoAVR and HDMI to SDI converter.

I SUSPECT that the nanoAVR is negotiating 50 Hz between it and the HDMI to SDI converter and 60 Hz between it and the upstream PC or Oppo. And, I don't expect it to convert vsync rates. But I can't be sure without knowing how the nanoAVR handles EDID. It reflects the EDID seen on its output (to the HDMI to SDI converter) at it's input (from the PC or Oppo), but that tells me nothing about whether it passes it through and lets the endpoints negotiate, or negotiates each one independently. I THINK it negotiates each end independently because if I take it out of the chain, it all works. Can @devteam confirm what it does? I started a support ticket, but no answer yet (not surprising over the holidays).

This is really frustrating as all the other equipment in my chain works together well. Just the link between the nanoAVR and HDMI to SDI converter is problematic. (And yes, if I run the output of the nanoAVR directly into a monitor, it works fine; and if I run the input directly to the HDMI to SDI converter, it also works fine.)

I will try a more fine tuned EDID on the emulator between the nanoAVR and the HDMI to SDI converter and hope it works.

Sorry, the above got duped.

Well minidsp sent me a logging app and I sent them back logs. We will see.

Grrrrrrrr!!!!

The nanoAVR HD and DL look for a damn HDCP chip (based on the logs and miniDSP's commentary) on the output.

Grrrrrrrrr!!!

WHY?

They don't DO HDCP.

The logs show they keep trying (with an unencrypted flickering display from the HDMI to SDI HDMI loop output).

They work if you feed through an HDCP complient HDMI splitter to TV but if you go into the HDMI to SDI converter the damn splitter encrypts.

WHY? There was no requirement to encrypt in the first place. (I tested this by putting the nanoAVR HD or DL into an OPPO HDMI input and taking the OPPO audio HDMI port output into the HDMI to SDI converter and it works fine -- the converter can't do HDCP so the audio OPPO HDMI port does not request it). The OPPO audio HDMI port normally feeds the nanoAVR so there is no need for HDCP anywhere.

The nanoAVRs try to talk to to an HDCP chip when there in no need and this apparently causes downstream HDMI splitters to go all paranoid and encrypt st their outputs.

I sure hope minidsp fixes this.

Hi rhollan,

I'm no HDMI expert by any stretch; I stumbled onto a setup that works reasonably well, can't say I know exactly how or why. But your last experiment above looks like a good clue to a workaround.

If you don't want to keep the Oppo in the loop, have you tried one of those HDMI boxes with buttons that let you set the EDID to various things? I have one feeding 2 nanoAVR-HDs (described here). The only thing I have to do is reboot it once per session to get all my audio channels working, but other than that no problems except for the IR volumes occasionally going out of sync between the 2 nano boxes, but that's a separate issue of course.

I know my situation isn't exactly the same as yours; I was just passing by and thought I'd toss this out there (you've probably already thought of it and/or tried it). Good luck.

- Jim

Thanks for the suggestion. I am Lready doing that to manage EDID as I thought it might be a 50/60 Hz thing. But, increasingly it looks like the nanoAVR insists on applying HDCP on its output even it it's input does not have it. I'd consider that a byg of that's the case.

I may seek to upgrade the HD to an HDA and use an Aphex 288 to level shift to professional audio levels. I was hoping to avoid long runs of analog audio though and carrying audio over SDI would have been ideal.

Wow, Jim: I hadn't realized you were "Gonzo" Jim. Thanks again for the advice but I nuked the idea of going through the Oppo for two reasons: (1) it's a silly waste of an Oppo. I'd have to get a replacement Bluray player (A 4k one is planned, but not yet). (2) and more importantly, volume control would be in the nanoAVR ahead of the Oppo -- if someone actually started to play a disk with it, well 0 dbFS at THX reference levels would not be pleasant.

Instead I will try one last time to deal with interfacing a nanoAVR to an HDMI to SDI converter: I have an HDFury AVRKey arriving today. This is supposed to split HDMI (up to 4k60 4:4:4 video) into audio and video streams, supporting HDCP 2.2 and 1.4 and, if necessary downconverting to the latter on the video output HDMI port. There is supposed to be no encryption on the audio HDMI port, and that's all I care about. If that fails, I will seek to upgrade one of the nanoAVRs (heck, both, so I can change the order if I want) to HDA analog outputs, and get a consumer to pro 8ch level shifter (Aphex 288 about $150 used, $400 new) and scrap the use of SDI to carry digital audio around.

Even though SDI carries video, I am only interested in using it to move audio around digitally. SDI to 4x 2ch AES audio deembedders are available and so are 2 x 2ch AES3 to 4 x 1 ch balanced audio converters.

A few caveats (besides the HDCP insanity):

1) SDI carries audio at 48 kHz. Period, end of sentence, full stop. I happen to think the room correction and speaker impulse response offered by the nanoAVRs far outweighs any higher audio sampling rate. Besides the DL only does 48 kHz anyway. Still it was a consideration since I would be locking in to a 48 kHz sampling rate. Where this matters is in analog reconstruction filters: they need to be a bit steeper and will have more phase distortion at the high end of the sampling rate if the sampling rate is close to only double the highest frequency of interest. I don't think this is audible.

2) No Atmos. While SDI can carry 16 channels of 24 bit audio at 48 kHz, I know of no HDMI to Atmos decoders that aren't full-fledged pre-amps. While I could go that route and analog sample all those outputs just to carry them over SDI, it strikes me as a pain. Once analog, might as well drive my amps.

3) Professional audio levels. Pro audio 0 dbFS is +24dbU (in the U.S.), whereas consumer 0 dbFS is around +10 dbV, some 11.79 dB lower. As I use three pro Crown (1x XLS 1502, 2x XLS 1002) amps (L+R, side surround, rear surround) and one prosumer Outlaw M2200 amp (center), I have to be wary of levels. I can drive the Crown amps at consumer levels, but have to turn the gain on them all the way up (well, the attenuation all the way down), and that is not good for the noise floor. Far better to pad the center channel down 11.79 dB on the nanoAVR output (or get a 12 dB analog pad).

Sheesh: why do I keep typing DHCP when I mean HDCP?

I ordered an HDFury AVRKey and will see if it solves the HDCP issues.

An interesting thought occurred to me: since I am going from HDMI to SDI to 4x 2ch AES digital audio (and then into two ADA-4 DACs and then Crown and Outlaw amps), if I got some ICE plate amps with AES3 digital inputs instead of the DACs and existing amps, I could go digital right up to the amp. Not something to do now (already spent way too much on this), but a possible future consideration.