No, sub distances are supposed to be longer than actual physical distance due to delays from filters in their circuits. There's v2.5 now and if you have dolby enabled speakers in the system, you should use that one. I suggest you listen to your system with the default options first before using any options.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Audyssey One - Nexus (Automated hi-res Denon/Marantz calibration with REW/Umik-1/2)
- Thread starter PHD
- Start date
emptygreat
Member
Got it! Thank you.
Currently got both uploaded to the AVR so will spend some time listening to them, on both music and movies.
Attached some after cal measurements
Nexus 2.4 unticked (1_LFE+Main.jpg)
Nexus 2.4 Force Small (2_Small.jpg) – Calculated crossover at 80hz
EDIT: Not sure if this is the right way to measure the results, but I did so with the same volume set on the AVR I had when making the pre calibration measurements.
Attachments
Last edited:
New amps arrived and I first did the stock Audyssey MultEQ App calibration. Results are FAR better compared to stock measurements on the old 26dB gain + 19dB gain amp setup. Now both are 31dB gain and it seems Audyssey really is made for this gain range. The bass and MultEQ are on par with the adaptive loudness module of E-APO. In the next days I will try Nexus again, but I am pretty sure the results will be great.
One question: Instead of the Atmos sweep, can't I just use the REW sweeps at 48 Khz and 0 to 24000 Hz range? I am asking because making many sweeps is a manual process with the provided dolby sweep and with REW I could just set my speaker to, let's say, 64 measurements, walk away, and come back to the job done. Wouldn't that be the same measurement data?
One question: Instead of the Atmos sweep, can't I just use the REW sweeps at 48 Khz and 0 to 24000 Hz range? I am asking because making many sweeps is a manual process with the provided dolby sweep and with REW I could just set my speaker to, let's say, 64 measurements, walk away, and come back to the job done. Wouldn't that be the same measurement data?
Last edited:
Yes, you can, if you don't use more than a total of 8 speakers that can be "seen" by REW (for example, via HDMI).
My setup, for example, LCR + surrounds + tops + sub channel = 8 channels, all addressable by REW. Anything more than that, though, requires using the provided sweeps and a compatible OS / player.
emptygreat
Member
@OCA
Hi again Serko. Couple of questions I'd like your advice on best procedures:
1. I have significant master volume variations when listening to music (mostly streamed, via Spotify/Tidal connect) and movies (source is android tv box).
Usually AVR is set between -40db to -35db for music, and between -24db to -19db for films. I've measured speaker responses at -19db. Is that the best practice or should I change it? Or Should I try making measurements at different volumes for Music and Films (sub only has a physical gain knob for volume, so probably would need to compromise one or another, or both)?
2. I've been getting inconsistent results regarding SW polarity across different versions of nexus. Version 1.8 asked to switch polarity, 2.4 to keep it as is, and 2.7 to switch again. Any reason for that?
Thank you!
EDIT: Dynamic EQ is on for both measurements and playback.
Hi again Serko. Couple of questions I'd like your advice on best procedures:
1. I have significant master volume variations when listening to music (mostly streamed, via Spotify/Tidal connect) and movies (source is android tv box).
Usually AVR is set between -40db to -35db for music, and between -24db to -19db for films. I've measured speaker responses at -19db. Is that the best practice or should I change it? Or Should I try making measurements at different volumes for Music and Films (sub only has a physical gain knob for volume, so probably would need to compromise one or another, or both)?
2. I've been getting inconsistent results regarding SW polarity across different versions of nexus. Version 1.8 asked to switch polarity, 2.4 to keep it as is, and 2.7 to switch again. Any reason for that?
Thank you!
EDIT: Dynamic EQ is on for both measurements and playback.
Denon and Marantz AVR's - Power amplifier channels
29dB gain
1V in, 100W out into 8 ohms.
Last edited:
I have many separates, run an amp per passive sub (and have been running subs since the late 1970's, one way or another), have frequently used EQ (sometimes analog parametric with a separate analog bass/sub-parametric) in processor loop before it leaves the pre-amp, X-over the frequencies before they get to the amps.
And I want stand alone DSP.
I've had quad & 5.1, one of which I abandoned when it became difficult to find source material and 5.1 (that I had doing 5.2), I ran for about 10 years and abandoned before 2000.
I no longer have space for multi-channel, so until (if ever, but I am hopeful that I will in a few years) I move to a larger place (similar to what I had before, much larger), I'll likely stick with 2 channel & a pair of downward firing (ported) 12"subs (built by me with an FR of 20 Hz to 80 Hz [yes, they are NOT small and take up a bit of space]).
If anything, I'd like to have 1 or 2 more subs.
And in recent times, they have come up with some DSPs that are suitable & it is said: relatively easy to implement in 2 channel with a pair (or perhaps more [hopefully more]) subs. When, (if) I do get a bigger place, I will be looking for (not an AVR) but that rare breed, a multi-channel processor. Yes, I know and understand why they are more expensive than an AVR but that is what I want, NOT an AVR.
I intend to personally have some of this (2 channel & subs DSP) in my system at some point to try next year.
It has always been that I march to my own drummer inside my head and it will always be so.
Tubes, never had them, never will! Perhaps (for me) more than their sound quality issues, (to me) they are just a pain in the butt to mess with.
But, to each their own
Some folks like continually messing with things. I like to get it "close enough", enjoy it and use my time to listen to my system (and/or) do other things.
I very rarely am not moving around doing something (why I haven't had cable or a TV since 2007). But I have an oPPo 205 UDP and a 32" monitor so that I can watch 30 or so minutes at a time of a DVD/Blu Ray or 4K disk (I can't sit still much longer than that) until I finish what I am playin).
Most times, it's a live concert and even being not in the room with it is not an issue (volume up, I can hear it in the yard or just use my (bought in 2010) Sennheiser TRS180 headphones (which have a 100 meter range [because they are on a robust set of Carrier frequencies from 2.4 – 2.48 GHz {good for when I am mowing the lawn})].
.
Okay, I tried it again with the new amp setup (2x NCX500 Stereoblocks) in my usual 4.2 surround setup.
First of all: Yes, the "weak" Dynamic EQ problem was really caused by prior low gain amps. With two 32dB gain amp Audyssey Dynamic EQ now applies noticable bass boost depending on volume but also still works at higher volumes, which is good.
With Nexus 2.7 I did 16 sweeps per speaker, correlated and vector aligned them. Then I launched Nexus and let it do its thing on all default settings. The results are there, there is no denying it. But even with the Harman10dB target curve and Dynamic EQ on 0dB the bass still feels weaker compared to Audyssey. This is measured with ASIO and the AVR on -25dB volume.
I am not sure why the bass is unsatisfying with Nexus but I thing it's the dip around 100 Hz, making everything lack punch. I will attach the log file so you guys can see maybe I did something wrong.
Edit: OHHH I set the LFE to 250 Hz instead of 150 Hz like instructed in the guide (since it was always 250 Hz before). I will try that now. I made sure to flip the polarity tho.
Edit 2: The above fix had no effect whatsever, but I noticed my latest calibration set all speakers to 2,90m distance, which is probably and error. I removed the measurements images to avoid confusion.
Log: https://pastebin.com/J9aVVByG
First of all: Yes, the "weak" Dynamic EQ problem was really caused by prior low gain amps. With two 32dB gain amp Audyssey Dynamic EQ now applies noticable bass boost depending on volume but also still works at higher volumes, which is good.
With Nexus 2.7 I did 16 sweeps per speaker, correlated and vector aligned them. Then I launched Nexus and let it do its thing on all default settings. The results are there, there is no denying it. But even with the Harman10dB target curve and Dynamic EQ on 0dB the bass still feels weaker compared to Audyssey. This is measured with ASIO and the AVR on -25dB volume.
I am not sure why the bass is unsatisfying with Nexus but I thing it's the dip around 100 Hz, making everything lack punch. I will attach the log file so you guys can see maybe I did something wrong.
Edit: OHHH I set the LFE to 250 Hz instead of 150 Hz like instructed in the guide (since it was always 250 Hz before). I will try that now. I made sure to flip the polarity tho.
Edit 2: The above fix had no effect whatsever, but I noticed my latest calibration set all speakers to 2,90m distance, which is probably and error. I removed the measurements images to avoid confusion.
Log: https://pastebin.com/J9aVVByG
Attachments
Last edited:
There's no way all speakers end up having the same distance. It'll sound pretty bad. There's a known bug with some AVRs not setting distances after the ady transfer and need a power cycle. Could be that.
Get ready for a pleasant surprise then. Without acoustic timing reference Nexus is just as good as random PEQI forgot the timing reference in REW. I used the built in sweep for more comfort during the measurements but I had "no timing reference" the whole time. Ugh. Gonna redo everything, sorry for the confusion.
Audyssey stock:
Nexus V2
The timing reference here seemed to work now since all speakers got reasonable distanced and levels but I am still not sure what Nexus is trying to do here. Both subs were set to 6.53m both and one was said to need polarity flip (blue control measurement). The orange one is the same but I flipped the sub back to 0°.
If you want to analyze this maybe I can provide you the Post-Nexus-REW mdat file or the logs.
Sub vs fronts is optimized for FL & FR together. You probably have asymmetry there. Also same distance for both subs with one polarity inverted during multiple sub optimization seems like a "required delay over limits" problem (or a measurement error). You need to turn off whatever processing is being done in the subs if any. They might be delayed too much or you may be not be measuring them properly with one turned off at each sweep. It's also imperative to use the REW measurement mode ady file and set sub mode to LFE to get the full sweep response or else measurement will be capped with a 120Hz lpf.
Sorry to bother you again, but I still want to get this working, now since I have the new amp that enables the use of Dynamic EQ even more so.
My process was:
- Upload .ady example file
- Download .ady measurement file with DynEQ
(note that for this I always kept using the same both files since my setup did not change, please tell if this is wrong)
- Set Sub to LFE only and 250 Hz so the "Large" speakers and subwoofers don't play sweeps at the same time
- Ensure suitable volume on AVR and gain on mic preamp so SNR and THD are in safe ranges, keep REW at -12dB and use FlexASIO at 48 Khz
- Measure stock REW sweep with acoustic time reference on FL, sweep range 0 to 24 Khz using 64k size and 16 sweeps
- Delete the first measurement of each speaker to avoid stuff like "auto mute" or me introducing noise while walking away from room during measurement
- Cross correlate and vector align all speakers into their final target curve and take Harman8dB as target curve
- Enable API in REW and hand it over to Nexus in default settings, clicking YES on the DynEQ question and inputting -25 Volume (which was my measurement level)
- Adjust the polarity and avr settings as told and upload the file via MultEQ to the receiver
I doubt that the sub processing is the cause since during my very first measurements everything was fine too.
But my hunches:
Wrong setting of acoustic timing reference (not only FL)?
64k size too low for Nexus to work with?
When Nexus tells me the measurement level was too loud I just add and SPL offset of -10dB to SW1 and SW2. Could this be wrong since "LFE -10dB" during measurement behaves differently? But it's just a level thing, not delay or phase, right? When using "LFE -10dB", the measurement SNR gets unneccesarly worse in my opinion.
No problems I'm here to help. Most your steps are correct, measurements should be 256k long though (64k could cause issues with some calculations and they have lower snr tatio), subs should be measured by lfe-10 sweeps and if that gets a too loud warning from Nexus then they should be dimmed down physically. PS you can choose delayed sweep start function since you're using REW'S own sweeps, you will not have to delete first measurememts.
emptygreat
Member
@OCA
Hi! Here to ask for some advice again. I'm remeasuring my system, starting from scratch. On the very fist measurements to fine tune microphone position between LR speakers, I've noticed what seems to be different polarity on L and R. I did check and recheck the wiring and all seems to be good (though I never exclude human error).
I'm attaching some screenshots of the impulse overlays of the sweeps (also mdat), red is Left, green is Right. Timing reference is set for Left speaker. Can you, or anyone help me understand what's going on, and how can I fix it? Thank you!
Hi! Here to ask for some advice again. I'm remeasuring my system, starting from scratch. On the very fist measurements to fine tune microphone position between LR speakers, I've noticed what seems to be different polarity on L and R. I did check and recheck the wiring and all seems to be good (though I never exclude human error).
I'm attaching some screenshots of the impulse overlays of the sweeps (also mdat), red is Left, green is Right. Timing reference is set for Left speaker. Can you, or anyone help me understand what's going on, and how can I fix it? Thank you!
Attachments
I think the impulse sound/graph isn't suitable for checking polarity since it's usually high frequency stuff where polarity doesn't matter really. If you want to check polarity, look at the phase behaviour in the bass range.
emptygreat
Member
I believe OCA mentions in the Centering Mic section of the tutorial video that if each speaker is reaching the 100 mark on different (positive/negative) parts of the chart, they have different polarity. I've anyway done the battery test on both, and both turned out correctly, so shouldn't be a problem. Thank you.
Impulse peak directions can be deceiving although rarer with front mains but still possible if one speaker is much closer to a wall surface than the other. A better way is to compare their phase responses in Overlays, preferably 1/1 smoothed and invert one of the fronts phase with REW (it's in most right click menus) and check which polarity matches the other better.
emptygreat
Member
Speakers (front ported) are roughly the same distance from backwall, though yes, FR is 60cm from side wall, while FL has a piece of furniture around 40cm from it. Anyway, I was wondering if this was the reason why I was getting inconsistent sub polarity settings through different version of Nexus, and now, on 2.9, it's again suggesting to keep polarity as is. Thank you again for your time and constant support of Nexus and its users!
Last edited: