TOPIC: NanoAVR Full-Gonzo Scenario Number 1:

Well, I finally spent some time last weekend with a UMIK-1 in front of all this stuff, and it's been a humbling experience to say the least.

After removing the crossovers in a couple of my speakers and setting them up active (by ear) with nanoAVR as previously described, my impressions of "a lot more energy in the midrange" were certainly true enough: +5 dB to be exact! Turns out I spent a whole evening enraptured by the 'clarity and detail' of a now-uncompensated baffle step. Sheesh.

I'm also thinking that the "transformation" in the highs was more a function of the different (better) power amp that was used in conjunction with the switch to active, and less about the active itself. What bugs me the most about this is that I know better than to change more than one variable at a time; I just got sloppy here. But I did discover that the freq. response on these 20-year-old tweets does indeed look fairly ragged, so that may be contributing as well. Next I'll measure some newer, supposedly better units I have here to see if it's the tweets themselves or my measurement setup that's to blame.

Anyway, back to the topic: I am still planning to go full 5.1 active. I still believe the overall benefits will be worth the effort, though the road is a little bumpier than expected. I now have 2 nanoAVRs, and I will hook them up together as soon as I get some of these other issues squared away, to see if Full Gonzo actually works. I'll post a summary as soon as that happens.

Curse this hobby!! Every door opens into a room full of doors. I am learning a lot though, and the ego bruises will heal.

-- Jim

Hi Jim, nothing particular to add but I am enjoying reading about your progress and thoughts

john.reekie wrote:
Thanks, John. I appreciate your moral support during these times.

I think I will learn more & more about how to 'get it right' with active, and will probably stick with it down the road. But this experience has given me new respect for those who can do all this stuff with passive networks! For example, when I go back now and measure one of my other speakers that still has the passive crossovers installed, I can see clearly how the baffle step has been perfectly compensated, and the tweeter output has been slightly attenuated on-axis so that it lines up better with the upper mids off-axis... Plus, when I remove the links and flip the polarity on the tweeter input, I can measure a perfect, deep notch at exactly 2 KHz, something I'm not even close to yet with my active setup. It's like looking at a piece of artwork.

-- Jim

Quick question about the nanoAVR. I need to confirm if the audio signals between the HDMI "processor" (as noted on the block diagram) and the SHARC DSP are I2S format. I'm scheming up a surgically altered version of the nanoAVR where I add a miniSHARC essentially in parallel to the nanoAVR SHARC DSP to provide FIR filter (and active tri-amp) support for the front channel speakers. I would still use the nanoAVR SHARC for the other channels. The miniSHARC only supports 2-channel input at this point (S/PDIF or I2S), so I need to meld the two SHARC's together in order to get the FIR (and zero-phase) capability I crave. I would then add Curryman DAC's to the I2S outputs of the SHARC's.

Hiya Non,

First off, I should apologize for abandoning this thread - my day job has just gone nuts lately, and I've had to put the Full Gonzo project on hold. I will get back to it, eventually...

Anyway. I'm afraid I can't answer your I2S question. I will admit to being a bit curious about hacking a nanoAVR in various ways, but haven't looked into it at all. Of course this is something that miniDSP couldn't support, and would certainly void the warranty, etc. etc.

I think our best hope along the lines of your idea is that they may release some sort of OpenDRC version of the nanoAVR, but now with the release of the DL model I doubt we'll be seeing anything like that, since the market for such a device would be relatively tiny. I could be wrong, though!

-- Jim

Thanks, Jim. I did get a reply back and it's officially a non-supported configuration. Since the audio signals will be uncompressed PCM, I'm confident I will find some way to make it work. Worst case is I have to send the front channels to a miniSHARC via S/PDIF. Good luck on your day job, mine's going to string out my little project as well.

UPDATE:

Yikes! It's been awhile, sorry this took so long. I've been really busy with other things lately, but I finally reached the point where I was tired of having the stereo in pieces all over the place (in 2 different locations even). So I finally set some time aside recently to start putting things back together.

I realized at some point that I was really attempting 2 separate projects here: The first was to find out if I was actually capable of making my cherished old Sony SS-M3 speakers sound better than they already did (which was pretty darn good) by converting them from passive to active. The second was (is) to determine if it's possible to use two nanoAVRs in parallel to achieve a full active surround sound setup (5.3 in my case). Turns out the short answer to both questions is yes! First, the speakers:

As I learned more about proper near-field and quasi-anechoic measurement techniques, I began to see that the original Peerless drivers in the Sonys are actually quite good; in fact, the replacement tweeters I was messing with were actually a little _too_ good. By that. I mean the dispersion at their low end (at the crossover region) was so much wider than that of the midranges, it was near impossible to obtain smooth off-axis response even at 24dB/oct, never mind 48. On the other hand, the original Peerless tweets have such nicely controlled dispersion at their low end that I was indeed able to go to 48dB slopes with only a minimal off-axis "step," which after careful level balancing isn't even audible. This was a stroke of luck for me, because I really prefer the 48 dB sound. It's more detailed and organic to my ears; the original 24 dB passives sound "blurry" by comparison. All the tweeters really needed was a notch to smooth a response bump in the 2KHz crossover region.

As for the 7" mid-bass drivers: I always thought these speakers sounded good, but were perhaps a bit too "forgiving" in certain respects. (I've heard them described as "elegant" by one reviewer; that describes them well.) While the midrange has always been nice, it's also been somewhat uninvolving, and with a bit of a "bloom" appearing from time to time in the 300 Hz range. I've come to the conclusion that the baffle step may have been slightly over-compensated in the original passive crossovers. I can only speculate, but I noticed 2 minor peaks in the midrange driver response, and I wonder if the designers might have tried to "catch" these with the baffle step comp. It measures about 5-6 dB in the transfer function of the passive midbass LP filter - quite a big step considering the relatively slight, trapezoidal baffles on these enclosures. Of course with miniDSP it was a simple matter to notch out the response peaks individually, then apply a duplicate of the original baffle comp shelf, but with a smaller overall step (more like 3 dB). As soon as I tried this, WHAM! Smooth, smooth midrange that not only measures like a calm fishing pond, but sounds wonderfully detailed and just plain _real._ The 300 Hz bump is a fading memory; even the low end seems to have improved.

To summarize, here's what I did to the 5 Sonys:

Yanked out the 24 dB L-R passive boards (got quite a pile of 'em here, heh), wired drivers direct to binding posts.
Applied a few, minor, somewhat low-Q notches to achieve smooth individual driver responses to >1 octave beyond crossover in both directions.
Applied a reduced version of original baffle step comp using HF_SHELF EQ.
Applied 48 dB crossover @ 2 KHz (original freq).
Flipped tweeter polarity, adjusted tweeter delay for nice notch at xover (measured at tweeter axis), flipped it back again.
Carefully adjusted tweeter level using both on/off axis measurement and listening tests.
After more careful listening, added one more small (-2 dB) notch at about 7500 Hz to tame a hint of aggressiveness in the sibilance region.

Okay: I certainly understand that anyone who actually knows what they're doing with this stuff is probably snorting with derision at the ham-fisted approach described above. I'm sure I still have a lot more to learn. (And of course any suggestions and advice are welcome!) But Good God, do these speakers ever sound _wonderful_ now! They don't even sound much like speakers anymore; the music just seems to appear in the room. Stereo soundstage is _huge,_ extending well beyond the L-R pair. Detail is startling at times - I can not only hear the instruments, but the space around them (whether real or artificial) much more clearly. And vocals are just plain emotional: To listen to The Beatles' "Love" album in 5.1, and to hear a young George Harrison singing "Something" in the center channel as if he's sitting right in front of me, damn near brings a tear to the eye of this old fan. And I haven't even started on subs or room correction yet!

Speaking of 5.1, that brings us to Project #2. The good news is: Full Gonzo lives! It is indeed possible to use 2 nanoAVR HDs, running in parallel off of an HDMI splitter into 2 AVRs, to achieve full-active, delicious multichannel surround goodness. The bad news is: It ain't easy! We all know it can be tricky enough to get 2 or 3 HDMI gizmos to play nice together; I'm trying to get EIGHT of them to stay on the same page. I'm still at the point where it refuses to work just about as often as it works, but when it all comes together and my turn-on/boot-up sequence is ju-ust right and the EDID gods smile down on me (it's probably more of a smirk, actually), it all sounds so ridiculously, insanely good that I have no choice - I just have to keep working on it. Right now I'm waiting for a new splitter to arrive; hopefully it will help. If I can get the whole setup to a state of reasonable stability, I'll start a new thread with a description of what I ended up doing. Meanwhile, I'm going to humbly request another hour's worth of glorious music from this cantankerous beast tonight before I shut everything down, because it's still a roll of the dice as to how much praying and cursing it will take to get it all working again tomorrow... :o)

-- Jim

@ Jim,

What an updated indeed! The Full Gonzo Lives for real..
You've got quite a setup here and I wonder if the EDID issue is something that can be simplified somehow...

Quick question: Did you insure that you upgrade the firmware to the latest version (see your user downloads). We did fix something with the 24Hz/deep color issue and I wonder if it's related here. Have a trial, doesn't hurt..

This is really neat to see this setup though, we look forward to seeing some pics of your setup along with some REW measurements if you get a chance. Our pic posting picture is broken with the new website but we're working behind the scenes to make it work! We can post pics if you email them to us first though..

DevTeam

devteam wrote:
Yes! I'm really enjoying it a lot.


As of this weekend I can add that I'm now using a different HDMI splitter that has reduced all the handshaking hassles down to virtually zero! I'll post a summary soon with lots more details.


I did switch to the rev. 1.12 firmware early on, when I was having all the HDMI trouble. It didn't seem to make a difference in my situation; then again everything was so confusing it was kind of hard to tell one way or the other. But now that things are working much more smoothly, I see that it does pass 24p with audio, no problems here. Unfortunately my old projector doesn't do 24p, so I've gone back to 60 Hz. But it did work!


Hmm... But that means I gotta clean up my apartment! Stay tuned, I'll see what I can do... :o)