TOPIC:

First of all, thank you to provide us such great products !

I am in progress to build a dedicated HT room and under investigation for the audio system.
I am very interrested by the NanaAVR HDA but would like to confirm the feasibility of the actual system i will design.

I will output the Video and Audio signal from my HTPC by the HDMI (PCM for audio) and i will not have a pre-amplifier. I will use the analog output directly connected on dedicated amplifiers for the 8 speakers.
I have 11 analog outputs to manage since i will have speakers LCR 2 way bi-amp with active filters made with the dsp.
(6LCR + LFE + 2 surround + 2 back)

I expect to use 2 NanoAVR HDA in paralleel for this purpose.
The HDMI output from the HTPC will be connected to a HDMI splitter (1 input to 4 outputs - HDMI 2.0 compliant)
2 outputs of the HDMI spiltter will be connected to both NanoAVR HDA.

The first NanoAVR HDA will be dedicated to LCR management (6 outputs used)
The second will be used for the bass management of the LFE and for the 4 surrounds (5 outputs)

I will used the video signal from the splitter for the monitor and the projector.


Does this configuration suitable ?

Hi

1. You can search for "gonzo" on this forum to find a similar type of project.

2. The issue with the arrangement you propose is synchronizing volume control between the two units. See #1, IIRC it worked OK but there's no guarantee.

3. I believe the control app can now be used to control volume of two units at the same time although I'm not sure it works with nanoAVRs.

4. There is another way to do the configuration you propose, which is to cascade the two units. Do the sub and surround from the first, pass LCR through unchanged, then do their crossovers in the second unit. Configure the first unit only for volume control. I did a quick test of this with an nanoAVR HD and AVR as the second and it worked (although I didn't use all channels). If you ask miniDSP directly about it they will probably tell you that they wouldn't recommend cascading too many HDMI units in series, since HDMI is such a bear.

5. Wouldn't it be great if miniDSP made a nanoAVR with 16 analog outputs

HTH

Thank you John for your quick 2 cents in my post.
i am going to check Gonzo project.

You are right for the volume control,
It was my second issue/question to synchronize the master volume of the both NanoAVR.
I was thinking that the IR code of master volume of each NanoAVR would be recorded and sent by GLOBAL CACHE iTach WF2IR (or other like IP2IR) but was not sure that some desynchronisation would not occur between the both NanoAVR....

as you wrote me, instead of using IR control for both NanoAVR, i have seen that some application is existing to control the NanoAVR via tablet and/or mobile phone using IP commands.
Probably the IP commands to control the volume on this app is not just "+" or "-" but the real level of sound.
Then, if the same level is sent to the both, i guess no desynchronisation will exist anymore and the information of the output will be known (and could be displayed) by the app.
It is just needed to know what is the IP script send by the app to the NanoAVR to set the correct volume and reproduced it with the HTPC. (or other app installed on a tablet that control all the HTPC system)



Cascade the both would be great and remove the master volume from one.
I was just afraid on the time calculation delta between the both ouputs but delays can be set to remove this problem.

Of course, if LCR are bi-amplified then minimum of 11 outputs would be great and 16 would be really great but maybe too much time consumming for the DSP.

Polopretress wrote:

Thank you John for your quick 2 cents in my post..

sorry, mistake of editing (cannot find how to clear..)

Hello,

If you haven't found it already, the latest installment of the Full Gonzo project can be found here. It's still working quite well for me (listening to it right now, in fact). I did toast a couple of tweeters attempting to integrate a Sony SACD player, as reported in the linked thread. Since then I've left that alone, and am simply getting by without SACDs for the time being. Apart from that, the HDMI part is behaving itself very well; I've only had one or two instances where part of the system would stop responding, requiring a reboot of one nanoAVR or the other. I should note however that the audio portion of my HDMI chain is pretty simple, consisting of a single line coming from the HTPC (computer), to the splitter, to the 2 nanoAVRs, and that's it. All of the video for this setup is connected to separate outputs on the PC'c video card. Prior to this, I was using a pair of Yamaha AVRs in the audio chain, and there were frequent HDMI handshaking headaches to deal with. As always, "your mileage may vary" depending on the particulars of your setup.

Meanwhile, the 2 IR volume controls continue to track essentially perfectly, just by pointing an old Sony remote at both of them on the rack.. I've gotten into the habit of just holding the volume down button for awhile, then running it back up, every time I power up the system, just to make sure they're both synced up, but I don't think this is even really necessary, as every time I've checked the level settings with the plugins, they were always at the same level. It still bugs me not having any sort of level indicator anywhere during normal use, but it hasn't caused any serious problems so far.

I did just now try again the Android app for level control. It had no trouble finding both of my nanos on the network, but I could find no means for controlling both of them simultaneously, so I guess it's back to the old Sony remote, which is OK by me.

Good luck, and holler if you have any questions.

- Jim

Great !
Yesterday, I have read with special care the summurize you have done with your step2 (NanoAVR HDA) and i am pleased to see that the result is good.
That convince me to go to this solution due to the fact that i do not have pre-ampli and was worried if i had to go with them....
I want my audio system very simple as yours and avoid multiple DA/AD convertion.

For the master volume, i will not use a control remote such as you So,ny because all the equipement is far from the HT room.
Then, i want to pilot the entire system with a low cost tablet under android.
Probably with the app IRules that is a slight complex to integrate but can manage all the equipment ( HTPC with video profile/Kodi/PotPlayer, Internet box, Projector, domotic such as climatize, humidity regulation, lights, power off sequence and so on...)

Then i can use IR of both NanoAVR with a specific module (Itach WF2IR or IP2IR) but maybe some desynchronisation might happened due to these additional control modules in the loop of control).
For this reason, i would prefer to use directly IP commands to control exactly the volume of both NanoAVR (and have also a feedback if possible on the tablet) and i am , of course, very interresting on your investigation to use this method if you continue to investigate.
If some shortcuts on PC can be set with the app that control by IP the NanoAVR, i guess it will be easy to implement using EVENGHOST.

John reekie wrote also that we can imagine to cascade the both NanoAVR instead of to use it in //.
Then only 1 control volume is necessary. I am just worried about the time delay betwwen the analog signals provided by the firt NanoAVR compared to the analog signal provided by the second one.
Didi you try such solution ?

Hi,

As far as I know, miniDSP has so far not provided us DIYers with any info for controlling the nanos over IP. I have asked about this here, but didn't get a response.

If / when I go to a different master volume control scheme, it will likely be some sort of downstream analog control, as described in my other posts. It still makes me squeamish having DAC outputs connected directly to power amp inputs, and ultimately drivers (particularly tweeters) with no means of attenuation or other filtering. I've just been too busy with other things to pursue this.

I haven't tried the nanoAVR cascade approach that John has suggested. Since all of my speakers are now configured as active 2-way, I need 12 parallel outputs, so the passthrough wouldn't work for me. I see no reason why it wouldn't work for a setup like yours, though. As for the latency, this would have to be measured somehow; you could then add that amount of delay to each output of the first unit.

- Jim

As far as I know, miniDSP has so far not provided us DIYers with any info for controlling the nanos over IP. I have asked about this here, but didn't get a response.

Strange, in your post, i have seen an answer from team dev who has proposed you to give you support for this purpose if needed.
Then nothing comes ?

It still makes me squeamish having DAC outputs connected directly to power amp inputs, and ultimately drivers (particularly tweeters) with no means of attenuation or other filtering.

Yes, thank for this. i will keep some securities on
First of all, i will add a capacitor with the compression driver at the beginning to avoid mistakes. Of course, if this problem occurs when the system is completly setted, it is not cool at all !
On my system, i will not have issue at 1Khz as you had since i will use compression driver that are ok until 700/800Hz.

But i have taken your point and will study how to be on the safe side.

Since all of my speakers are now configured as active 2-way, I need 12 parallel outputs, so the passthrough wouldn't work for me. I see no reason why it wouldn't work for a setup like yours, though

No, i do not follow you on this reason.
If you do not modify the PCM track of a given signal with the first NanoAVR, you can manage them with the second one.
For sure , the bass management must be done at the very beginning to have the complete signal of all the PCM tracks.

If cascade you have still 16 outputs available as in //.
Use the first NanoAVR to use the 7 analog signals for the bass management and LCR speakers
Use the second NanoAVR to use the 8 analog outputs for your 4 active surround.

Time delays will be necessary to adjust. Depending on the latency on the first NanoAVR, left speaker or left surround back speaker will have to be tacken as reference.

What do you mean by 12 analog outputs ? I need to to re-read your post ....

edit : My mistake....
If cascading the both NanoAVR , do you mean that managing the Sub and the LCR in the first NaoAVR (7 channels used), i will loose 3 channel of surround on the HDMI output of the first NanoAVR ?
Does the HDMI output is exactly the same tracks as the analog outputs ?
If yes, i understand now your comment....

Then , i understand also better the comment of John that gave me the information to manage the sub and the Surround (not 2 way active) at the first one (==> 5 tracks modifed and 3 tracks unchanged) and the LCR (2 way active) on the second NanoAVR

With this assumption, would you recommend me to use in cascade or in // ?
(sorry for my langage , i am French ...)

Polopretress wrote:

Strange, in your post, i have seen an answer from team dev who has proposed you to give you support for this purpose if needed.
Then nothing comes ?

In their defense, mine was just a quick request here in the forum; they might have missed it. I haven't had time to follow up on it with them.

...If cascading the both NanoAVR , do you mean that managing the Sub and the LCR in the first NaoAVR (7 channels used), i will loose 3 channel of surround on the HDMI output of the first NanoAVR ?
Does the HDMI output is exactly the same tracks as the analog outputs ?
If yes, i understand now your comment....

Yes, as far as I understand it, the HDMI output audio is processed the same as the analog outs; only the video is passthrough. And yes, that's how I have mine configured: nano #1 does LCR mids & tweets + 2 subs for a total of all 8 outputs used. nano #2 is L & R surround mids & tweets, using 4 channels, leaving 4 unused. I may someday use the remaining 4 channels for a separate Zone-2-style additional stereo pair.

Then , i understand also better the comment of John that gave me the information to manage the sub and the Surround (not 2 way active) at the first one (==> 5 tracks modifed and 3 tracks unchanged) and the LCR (2 way active) on the second NanoAVR

Yup, with your setup you could allow the 3 LCR channels to "squeak by" unchanged through the first nano. (Actually, you could do full-range things to the LCR channels, like additional EQ and/or compression. Crossover stuff would of course need to be done in the second unit.)

With this assumption, would you recommend me to use in cascade or in // ?

Hmm... Good question.

The 2 main drawbacks to the parallel approach are master volume sync and potential HDMI headaches. I consider myself really lucky that my own setup behaves very well these days in both areas. I fear that different setups will have varying results, as I've already experienced to a degree with a previous iteration (Gonzo v1). I guess if you can keep the audio portion of your HDMI chain as simple as possible, as I have, your odds will probably be better. Unfortunately I can't speak to the prospect of a dual IP2IR "ganged" approach for volume control, as I haven't tried this.

The cascade approach fixes the master volume dilemma. The resulting latency would need to be dealt with. I'm also a bit concerned with the fact that the second nano would be seeing input audio that has been scaled downward by the master volume attenuation of the previous unit, leaving less bits for it to work with (maybe a lot less). This particular concern may be unfounded; I'm afraid my maths simply aren't good enough to know for sure. Perhaps someone with more knowledge of this can chime in here. Also, I'm not sure if cascading 2 nanos makes the HDMI handshaking nonsense any better or worse. John, are you there? Have you tried the cascade idea, and if so, any thoughts on this?

(sorry for my langage , i am French ...)

You're doing fine; for what it's worth, I find you quite easy to follow.

- Jim

. I'm also a bit concerned with the fact that the second nano would be seeing input audio that has been scaled downward by the master volume attenuation of the previous unit, leaving less bits for it to work with (maybe a lot less)

Yes, i do not know also what could be the consequences.
In fact, i do not know how is managed the volume in each PCM track. Is it already taken into account in the signal (embedded) or is it a numeric data outside of the numeric signal and it is only tacken into account for D/A conversion.?

I will use a HTPC and can also managed the volume with my player but for the same reason (less bit in the input of the dsp), i try to push as far as possible the volume management but maybe it is useless.

Also, we do not know how the master volume is managed in the minidps since it is not really documented.
Does the master volume is active after or before the dsp effects ?

If it is applied prior to the effects then, the reduction of the bits for a given PCM track is always true and it is not only applied on the PCM tracks that will be managed by the second NanoAVR.
Then, this discussion is useless because there is no additional impact depending of the way to use the both NanoAVR (cascade or //)

Also, I'm not sure if cascading 2 nanos makes the HDMI handshaking nonsense any better or worse. John, are you there? Have you tried the cascade idea, and if so, any thoughts on this?

In fact, it is at this step that my English level is limited because i do not understand what you call "HDMI handshaking" and what is the consequence of the PCM tracks and Video track.
Is this some scratch/disturbance ? some disruption ?

John answered me :

I did a quick test of this with an nanoAVR HD and AVR as the second and it worked (although I didn't use all channels).

nanoAVRs use a SHARC processor, which is floating point. The I/O in the HDMI pipe is 24 bit integer (I believe/assume? don't really have a way to check it).

Back to front:

First, thanks for the quote from John - I missed that.

By "HDMI handshaking" I was referring to the ridiculous West Side Story-style square-off dance ritual performed by this equipment wherein upon initial connection & power-up, the capabilities, copy-protection and otherwise, are determined and negotiated between all the various pieces. It can be problematic, to say the least, and quite frustratingly random. (Sometimes it works, sometimes it don't.)

Regarding nanoAVR internal master volume: I believe this is handled as a final division (fractional multiplication?) on the processed PCM data before it "leaves the building." Sorry I can't cite a reference, but I remember reading something to that effect somewhere. (DevTeam, can you verify/clarify this?) This of course also reduces the bit depth, but my understanding (and subsequent experience) is that the sonic consequences are negligible, provided your system gain is set up properly.

The difference I see with the cascade approach is that nano #2 is seeing reduced bit depth at its input, ahead of a fair amount of processing yet to be done. I wonder if this increases the possibility of accumulated error on the signal as it is processed. As John mentioned, the SHARC processor uses floating point math internally, so I assume that accumulated errors are negligible. This leaves only the 24-bit integer "bottleneck" thru the HDMI connection between the 2 units as a possible source of degradation - which is probably also not a big deal, I suppose. I just don't know for sure.

- Jim

[EDIT] There would also be an additional asynchronous sample-rate conversion at the input of the second nanoAVR. (This is performed on all input sample rates, whether they "match" or not.) Taken by itself, this to me is no more cause for concern than any of the other items already mentioned. The question is: At what point do these things, when all added together, become a concern?

The simple way to find out would be to try it

Depends what you mean by "try it." If you mean, just hook it up and see how it sounds, I would have trouble doing this.

I've come to the conclusion in recent years that, although thankfully my ears are still pretty good, the less I trust them, the better my audio system sounds. By that I mean I've tried to do as much as I know how (learning more if necessary/possible), using careful testing & measurement, then relying on subjective listening for only the final tweaks. This is the approach I've taken to my current setup, and the resulting sound is head-and-shoulders above anything I've previously owned (not to mention several much more expensive systems I've heard!). So it would be difficult for me to try something like this (nanos in parallel vs cascade) without doing some tests to determine the effects on distortion, noise floor, etc. My guess in this particular case is that everything would look (and sound) fine, but it would bug me not to have some data to look at, even if only for the education.

- Jim