ARC Genesis results

[ban25]

Hey all, just ran arc for the first time today on my AVM70.

I noticed it's set my sub to 30 feet? Is that normal?

I did have a message during setup saying my sub exceeded max target and arc will switch to relative delays, I selected ignore to continue.

Should manually adjust it?

I do have 2 sub but connected as one to sub1 on the AVM.

Are they wireless?

 

[Sideshowdad]

Yes wirelessly connected.

 

[ban25]

Yes wirelessly connected.

That's the issue. It's not measuring/correcting for distance, but time. It's a pet-peeve of mine when AVRs and room-correction software uses distance as a measure of time (and it even resulted in a speed-of-sound bug in Audyssey on D+M AVRs). When you use a wireless connection, what's really happening is the analog signal from your AVR is going into the wireless transceiver, being converted back to digital, buffered (for some time window, maybe 4-16ms), packetized, transmitted, re-assembled, converted back to analog, and passed to the line-level input on your subwoofer.

This process, as you may have surmised is slow. Generally on the order of 30-60ms. Sound travels at 1087 feet per second, so 30 feet would equate to 27.6 ms. Now, what ARC Genesis is saying is that it won't attempt to time align speakers that are delayed by more than 27.6 ms -- your real delay is probably worse than that -- I'd bet at least 50 ms. That's just the time it takes to transmit to the subwoofer -- if your sub is really 30 ft away from the Main Listening Position, then double it! Consider that your wired speakers are probably only 6-12 ms delayed and you can see why it's hard to time align when your subwoofer is delayed by 10x the rest of your system. ARC Genesis is looking at this and basically throwing its hands up, which is the right call. Dirac doesn't bother to correct beyond 20 ms, FWIW, even though some manufacturers have increased that value to 50 ms. That's already over 1 frame at 24 Hz and over 3 frames at 60 Hz -- something that's very noticeable in terms of late audio and lip sync.

TL; DR: switch to wired speakers across the board, or learn to live without room correction. Even the best engineered wireless solution is going to induce at least 4 ms of latency, but I have not seen anything on the market that is not significantly worse than that. In fact, I'm suspicious of pretty much all packetized digital audio transmission systems (ethernet-based AES67 or what-have-you) because they are all going to induce non-trivial amounts of latency.

 

[Sideshowdad]

That's the issue. It's not measuring/correcting for distance, but time. It's a pet-peeve of mine when AVRs and room-correction software uses distance as a measure of time (and it even resulted in a speed-of-sound bug in Audyssey on D+M AVRs). When you use a wireless connection, what's really happening is the analog signal from your AVR is going into the wireless transceiver, being converted back to digital, buffered (for some time window, maybe 4-16ms), packetized, transmitted, re-assembled, converted back to analog, and passed to the line-level input on your subwoofer.

This process, as you may have surmised is slow. Generally on the order of 30-60ms. Sound travels at 1087 feet per second, so 30 feet would equate to 27.6 ms. Now, what ARC Genesis is saying is that it won't attempt to time align speakers that are delayed by more than 27.6 ms -- your real delay is probably worse than that -- I'd bet at least 50 ms. That's just the time it takes to transmit to the subwoofer -- if your sub is really 30 ft away from the Main Listening Position, then double it! Consider that your wired speakers are probably only 6-12 ms delayed and you can see why it's hard to time align when your subwoofer is delayed by 10x the rest of your system. ARC Genesis is looking at this and basically throwing its hands up, which is the right call. Dirac doesn't bother to correct beyond 20 ms, FWIW, even though some manufacturers have increased that value to 50 ms. That's already over 1 frame at 24 Hz and over 3 frames at 60 Hz -- something that's very noticeable in terms of late audio and lip sync.

TL; DR: switch to wired speakers across the board, or learn to live without room correction. Even the best engineered wireless solution is going to induce at least 4 ms of latency, but I have not seen anything on the market that is not significantly worse than that. In fact, I'm suspicious of pretty much all packetized digital audio transmission systems (ethernet-based AES67 or what-have-you) because they are all going to induce non-trivial amounts of latency.

Wow thank you for the explanation!!!

My subs are about 12 feet away from main seating positon.

So unless I wire them there's not point using ARC because the delay to the sub will be off no matter what?

 

[ban25]

Wow thank you for the explanation!!!

My subs are about 12 feet away from main seating positon.

So unless I wire them there's not point using ARC because the delay to the sub will be off no matter what?

Yeah there's really only so much it can correct for. If you can hard-wire them, even if you have to run wire under a rug or something, you'll be surprised at how much it will help. Once you do that, ARC will report 12 ft as the true distance.

 

[peng]

Hey all, just ran arc for the first time today on my AVM70.

I noticed it's set my sub to 30 feet? Is that normal?

I did have a message during setup saying my sub exceeded max target and arc will switch to relative delays, I selected ignore to continue.

Should manually adjust it?

I do have 2 sub but connected as one to sub1 on the AVM.

It measures delay, not distance. You pronably should re-run arcg and follow instructions to the letter, but do not adjust the distance manually.

After that, if you feel bass response is weak, plat with the deep bass adjustment.

 

[Sideshowdad]

It measures delay, not distance. You pronably should re-run arcg and follow instructions to the letter, but do not adjust the distance manually.

After that, if you feel bass response is weak, plat with the deep bass adjustment.

Thanks will re-run but the only thing that came up was what i mentioned above asking me to abort/ignore/retry. I took a pic and see it read it at 38' away.

 

[peng]

Thanks will re-run but the only thing that came up was what i mentioned above asking me to abort/ignore/retry. I took a pic and see it read it at 38' away.

Did you disable all bass management features at the subs? If not, try that before re-runing arcg.

 

[Sideshowdad]

Did you disable all bass management features at the subs? If not, try that before re-runing arcg.

All I did was do a quick test on arc and adjusted my subs to hit 72db as I have two. Didn't mess around with any settings on my subs (svs3000).

 

[peng]

All I did was do a quick test on arc and adjusted my subs to hit 72db as I have two. Didn't mess around with any settings on my subs (svs3000).

Common practice is to not use the DSP of the subs and let the AVR/AVP does its own thing. For Anthem, they say the following in the manual:

If your subwoofer has a built-in crossover, set it to “bypass” or set its frequency control to the highest frequency since the MRX/AVM handles this.

Similarly, if your subwoofer has a phase option, set it to 0 since the MRX/AVM handles this as well. Highlighting Bass Management

If you follow the instructions above, the distance detected by ARCG might reduce somewhat, as more DSP activities would likely increase the processing delay, though that's not the main reason for following Anthem's recommendations.

 

[ad_fletch]

Might be wrong here as I'm finding it a bit difficult to pin down the info, but I think ARC Genesis can only correct phase/timing for multiple subs on MRX1140, AVM70/90 and STR models. Happy to be corrected, but surely with their twinned sub outputs, the MRX540 and 740 wouldn't be able to set multi-sub phase correctly?

UPDATE: Ran ARCG 1.7.4 on my MRX 310 and there's no option for phase, as suspected. Measurements and recommended targets were pretty consistent with those I obtained a year ago, although it decided to cross to sub 10Hz higher for whatever reason.

Still fine-tuning the bass - I have two custom sealed subs (one based on 12" Peerless XXLS and another based on 10.5" Wavecor) and ARCG recommends a super steep HPF at 28Hz. I did that for a bit, but think I'm going to try leaving it uncorrected below about 35Hz to see whether I can perceive any difference in movies (that would still deal with my primary room mode which is around 40Hz).

Even though these subs are modestly powered (esp compared with some of the monsters I read about on US forums!) I do find I have to turn them waaay down to measure them - even when they're each around 72dB in Quick Measure the full calibration tells me to turn them down further. It's a bit frustrating because I then feel the need to turn them back up afterwards....

 

[ad_fletch]

Just an update on the bass situation.

Annoyingly, having sold my UMIK-1, I've had to do a quite rudimentary analysis of the subs' output using a tone generator and SPL meter because those are the only tools I have to verify what ARC Genesis is telling me.

Using the handy Anthem remote control app, I played tones at a variety of bass frequencies and measured with ARC on/off. Even after setting ARC to 'flat' below 35Hz, I am losing around 5dB of output at 35Hz and 10 dB (!) at 30Hz compared with ARC switched off. This is with the sub target set to 0, and a deep bass boost of about 1dB. I'm also losing some output at 50 and 60Hz, even though those frequencies are actually meant to be boosted. It is really counterintuitive, because according to the Genesis graph, I should only being seeing a small cut at about 40Hz and boost everywhere else:

Any thoughts on what's really going on?

PS Yes, I know my subs' combined output is anaemic - unfortunately decent subs like SVS and Rhythmik cost about double in Australia compared with the US and are out of my price range.
PPS I wish ARC genesis would highlight specific frequencies on each graph when you move your mouse over them like REW.

 

[Beershaun]

Bump the deep bass boost to 3db and room gain to 6db.

 

[ad_fletch]

Thanks will try.

 

[OCA]

I have written an HTML script for a subscriber (a while ago) that can extract Arc Genesis .arc3 measurements to REW. I thought it could be relevant here (I don't own an Anthem). It will work directly in any web browser so it's platform free and doesn't require any installation (just save the text as "whatever.html" and double click to run).

 

[SeanAtx]

I've been playing around with ARC Genesis on my MRX 510 for a while now and could never find a bass setting that didn't feel like it was missing something. I have a PC hooked up to the analog input now, so I loaded up REW and measured what ARC was doing. It looks like it's not able to fill the dips in the bass region very well, compared to what I was able to do after disabling ARC and creating a set of correction filters using REW EQ. The setup is Emotiva T2+ towers powered by a 3 channel Emotiva amp and one SVS PB 2000 subwoofer. I'm running the sub a bit hot to help fill out a null (and I'm all about that bass).

Here's the baseline measurement with no corrections applied:

Here's a comparison between ARC filters and REW EQ (Ignore absolute dB levels here, the measurements have been moved so they overlap for easier comparison):

REW fills in the dips at 35-40 hz and 54 hz very well compared to ARC. ARC also has a rise between 60-90 hz that REW does not. I'm guessing ARC is unwilling to apply as much gain as REW, which I've set to 6 dB max boost. Or maybe it runs out of filters - REW is applying 10 corrections below 200 hz and 2 between 200 & 380.

This configuration requires applying a -5.5 dB preamp level in Equalizer APO in order to keep the overall max gain at 0 dB, which means the REW correction is noticably quieter than ARC during listening tests. I have to go from -25 to -19 on the volume knob to get the same output as measured on my phone's decibel meter. I'm still trying to A/B compare them to see if the differences in the graph actually represent something that can be heard reliably, but the REW EQ feels like it has just a bit more oophm on bass-heavy songs & action movie sequences.

 

[peng]

REW fills in the dips at 35-40 hz and 54 hz very well compared to ARC. ARC also has a rise between 60-90 hz that REW does not. I'm guessing ARC is unwilling to apply as much gain as REW, which I've set to 6 dB max boost. Or maybe it runs out of filters - REW is applying 10 corrections below 200 hz and 2 between 200 & 380.

This configuration requires applying a -5.5 dB preamp level in Equalizer APO in order to keep the overall max gain at 0 dB, which means the REW correction is noticably quieter than ARC during listening tests. I have to go from -25 to -19 on the volume knob to get the same output as measured on my phone's decibel meter. I'm still trying to A/B compare them to see if the differences in the graph actually represent something that can be heard reliably, but the REW EQ feels like it has just a bit more oophm on bass-heavy songs & action movie sequences.

With ARC, I assume you ran it for more than just the mmp, so when comparing the results did you make sure you were comparing apples to apples such as:

- Use the same number of mic positions, say 5?
- Are the REW filters based on measured results of the same number of mic positions or just the mlp?
- If yes to above, are you sure the mic positions, such as all 5, or more, were in the exact same spot in each case

It would be normal to have very different results the mic arrangements are different, obviously.

 

[SeanAtx]

With ARC, I assume you ran it for more than just the mmp, so when comparing the results did you make sure you were comparing apples to apples such as:

- Use the same number of mic positions, say 5?
- Are the REW filters based on measured results of the same number of mic positions or just the mlp?
- If yes to above, are you sure the mic positions, such as all 5, or more, were in the exact same spot in each case

It would be normal to have very different results the mic arrangements are different, obviously.

The main difference is that I used the moving mic method with a UMIK-1 in REW for the three measurements above. The ARC correction was done with 5 static positions centered around the same area that I swept the mic, but it was measured with the ARC mic that came with the MRX.

 

[SeanAtx]

Whoops, just realized I only uploaded one speaker measurement from ARC. Looks like it was the right side speaker, which is next to the hallway and does not get the corner gain that the left speaker does. I used the script that OCA provided above to import the ARC measurements into REW and did an RMS average:

That's better!

 

[peng]

Whoops, just realized I only uploaded one speaker measurement from ARC. Looks like it was the right side speaker, which is next to the hallway and does not get the corner gain that the left speaker does. I used the script that OCA provided above to import the ARC measurements into REW and did an RMS average:

View attachment 325233


That's better!

They look close enough, with some manual tweaks I am sure you can have the ARCG one looking almost as smooth as REW's. What are you using to implement REW's filters?