Anthem D2/D2v/AVM50/AVM50v/ARC1 tweaking guide - Page 993

Originally Posted by Texas steve

Good work John, you save us a lot of time. So far i did .14 and am holding to re-ARC since there are no real sonic changes yet.

Originally Posted by Bob Pariseau;

MACCA350,
The problem with altering the levels is that if any portion of the mixed frequency tone you are hearing is participating in bass steering you will get an incorrect crossover for that portion, and the crossover may vary between speakers (certainly does with mine).

And by setting the levels for LPCM you lose the info about variation within LPCM, which is just what we're talking about regarding the other poster.

I know from my ARC charts that the sub should measure a little hot in SPL due to the Room Gain hump. And indeed that's what I see leaving ARC's levels unchanged. So my LPCM results make sense (no PLIIx of course in my 5.1 setup).

But if your sub was really 5dB or more too hot after ARC, that suggests something is wrong.

Originally Posted by jayray

. Not sure if any of the versions after 2.4.14 would make your sub 25 sound better. Wouldn't you like to know
John

Originally Posted by Bob Pariseau

Which version of ARC are you using? The "official" V2.4 or one of the recent flurry of "test" versions (the latest being V2.4.17 from last night)?

If you are using 2.4, I wouldn't spend too much time tweaking this as there is likely to be a new "official" version any day now that based on what we are seeing in the "test" versions may do a better job of handling your issue with RF. So you might just want to hold off for that.

Anyway, let's talk tweaking:

The thing to focus on is getting RF better corrected.

Your room is surprisingly "flat" for a room that has not been acoustically treated. ARC found a very small Room Gain, and indeed that's what the red Measured curves suggest by eye as well. I think you might enjoy the sound more if you pushed up the Room Gain a bit -- say to 2dB. And that alone will alter what ARC is trying to do with RF and may (or may not) yield a cleaner Calculated result. You can play around with different values quite quickly -- no need to re-Measure.

Another thing to play with is the Max EQ Frequency. Now since you already have insufficient correction in RF, you don't want to push up Max EQ Frequency trying to take advantage of additional correction at the high end because that will just divert resources from the low end. HOWEVER, the mystery processing going on inside ARC is sometimes sensitive to the range of frequencies you tell it to correct. Moving Max EQ Frequency around a bit will alter the way the correction resources map into that range of frequencies, and can sometimes have a surprisingly good result in getting a problem like your RF to suddenly snap back into proper correction.

So I'd try moving Max EQ Frequency up to 10KHz and then perhaps playing around with different values up and down from there for it to see if you can find a setting that clears up RF. Play around with Room Gain values at the same time. Your Measurements between 5 and 10KHz don't show you need a lot of correction in that range, so it should be safe to maneuver in there.

Keep in mind that both of these changes will affect ALL the speakers, so you need to check and make sure your candidate changes to improve RF don't screw things up for the other speakers.

So play around with that a bit and if you can find a way to clean up RF without messing up the other speakers then post your charts and Targets again and we can look at what else you might want to try to improve.

And again, if you are not using the "test" ARC version, plan on re-Measuring when you get that, perhaps as the next "official" version, as it should do good things for you. The necessary tweaks for that one will likely be different.
--Bob

Originally Posted by jayray

I did a new measurment with 2.4.17 and the volume levels are back to 75 dB. Nothing else changed as far as I can tell. Will test later.
John

Originally Posted by TboneofLA

My levels are all the way down to 70dB.

Originally Posted by JimP

When you say "pro audio calibration", what are you saying exactly?

What I'm not understanding is how without additional equipment, one can do much more than aiming speakers, adjusting levels and correct any timing errors. Things like frequency corrections for 7-9 channels can take some fairly expensive add on equipment.

Originally Posted by Bob Pariseau

Did you use your SPL meter to recheck your settings for Setup > Level Calibration > Test Level (and also the volume knob setting for your sub) prior to doing this new Measurement?

You may still have the additional 6dB reduction of Test Level in place that was required for some folks to get 75dB out of ARC in the "official" ARC V2.4.
--Bob

Originally Posted by The Bogg

pro audio calibration is getting a guy with specific certification and a bunch of expensive test gear to come and optimize the positioning and aiming of speakers, crossovers, etc... I understand that it's easy to simplify the "position and aim" part of it but I have experienced how much difference a good calibrator can make. If you look at just the measurements you don't necessarily see a huge difference but the difference in imaging, immersion and clarity are INCREDIBLE.

As far as the bass error, the ARC program corrected values looked great both pre and post calibration. However, if you put a full range signal into any single speaker with bass management engaged the bass level going to the sub was attenuated by 6-8db or so which is clearly an error that was reproducible. Correcting that error led to an astonishing increase in quality and quantity of bass. I don't know what would happen if you didn't use bass management at all but the fact is there was a measurable repeatable error at least in my situation.

Even with no adjustment of phase on the 4 subs I have the measured and corrected bass response of my subs was very good. However, it SOUNDS way better after calibration. There was repositioning of the front 2 subs. Sure you might have been able to come up with this result without hiring someone but not by random chance. The Sencore test gear etc... is not cheap and goes beyond what most enthusiasts would own. Like I said before, a good calibration is still worth it for some people (like me) but I understand that the ARC alone is "good enough" for most people.

Originally Posted by TboneofLA

Thanks Bob. I thought the same thing at first but went back and checked it after I saw the charts and I did have the test level set at 0.

When I measured the test noise with the SPL meter afterwards it was reading 75dB. The ARC test sweep certainly doesn't sound any lower than previsous tests either.

I may attempt to rerun the test again on the weekend.

Originally Posted by ninja12

Here are my measurements for ARC v2.4.17. I'm going to completely trust ARC on this one and not change anything. I have only been listening to music so far, and it sounds good. The imagining is really nice. I had to check my center channel, several times, to make sure it wasn't on.

Originally Posted by Bob Pariseau

Remember that you can't get a proper reading from Test Level unless the volume trim line for the LF speaker is at 0dB.

Also check the volume trims that ARC Uploaded. If you have any extreme values ARC may have had to shift the solution to get enough range to trim all the speakers. The usual culprit here is the sub.
--Bob

Originally Posted by The Bogg

pro audio calibration is getting a guy with specific certification and a bunch of expensive test gear to come and optimize the positioning and aiming of speakers, crossovers, etc... I understand that it's easy to simplify the "position and aim" part of it but I have experienced how much difference a good calibrator can make. If you look at just the measurements you don't necessarily see a huge difference but the difference in imaging, immersion and clarity are INCREDIBLE.

As far as the bass error, the ARC program corrected values looked great both pre and post calibration. However, if you put a full range signal into any single speaker with bass management engaged the bass level going to the sub was attenuated by 6-8db or so which is clearly an error that was reproducible. Correcting that error led to an astonishing increase in quality and quantity of bass. I don't know what would happen if you didn't use bass management at all but the fact is there was a measurable repeatable error at least in my situation.

Even with no adjustment of phase on the 4 subs I have the measured and corrected bass response of my subs was very good. However, it SOUNDS way better after calibration. There was repositioning of the front 2 subs. Sure you might have been able to come up with this result without hiring someone but not by random chance. The Sencore test gear etc... is not cheap and goes beyond what most enthusiasts would own. Like I said before, a good calibration is still worth it for some people (like me) but I understand that the ARC alone is "good enough" for most people.

Originally Posted by Bob Pariseau

Anthem's default for ARC is to correct only up to 5KHz. The *MAIN* reason for this is that getting quality Measured data in treble is tricky. What the mic hears is subject to all sorts of things including details of speaker pointing, mic placement, and even room humidity.

However, many of us believe we do have good Measured data above 5KHz, and ARC gives you the option of raising Max EQ Frequency. When you do that, ARC reassigns resources to apply the correction up there. Usually that works without untoward consequences, but some folks have found that applying correction higher up means they get poorer Calculated curves lower down -- more ripples and residual error.

Now some of that is undoubtedly due to simply mapping the correction tools across the wider frequency, but some of it appears to be lack of correction resources. That is, if a lot of correction is needed lower down, telling ARC to do work higher up means the lower frequencies can't be fully corrected.

The workaround of course is to pull back Max EQ Frequency to a smaller value. Often just a few KHz change is enough for everything to snap back in to full correction.

Folks who have setups that DON'T need a lot of correction lower down are more likely to have no problems like this when they raise Max EQ Frequency.

---------------------------------------------

ETA: It appears Anthem may have changed the way they allocate resources for the latest "test" ARC. If ARC can't do the full frequency range of correction you have asked for, it appears this new version may still give priority to the lower frequencies -- meaning you may not see as much correction higher up as you thought you might get. But we'll need to see more results charts from more posters before that gets nailed down. Also keep in mind that you can often change the amount of correction ARC needs to do lower down by judicious tweaking of Room Gain and the cutoff/crossover frequencies.
--Bob

Originally Posted by JimP

Bob...thanks for replying.

Bottom line....if one needs correction from subwoofer to way above 5K htz(meaning many corrections), ARC (or maybe the hardware) isn't quite there yet?

Originally Posted by Stevetd

What kind of problems might I run into running ARC on two separate units on one laptop?

Originally Posted by Bob Pariseau

Each unit will come with its own ARC mic (USB connection to the PC), and a pair of licensing/calibration files that are matched to the serial number of the unit and the serial number of the ARC mic.

You'll install ARC for either unit on your PC and the pair of files for that unit/mic will be installed with the application. Then you will manually copy the pair of files from the ARC install CD of the second unit into the same location (Windows > Program Files > Anthem > Anthem Room Correction) so you've now got two pair of files in there along with the one copy of the installed ARC application.

When you run ARC it will ask you to select which mic you are using. Pick the mic for the unit you intend to set up and just be sure the mic you actually use happens to be the mic you select. (ARC will warn you if you mistakenly try to use some random USB mic or the mic built into your PC, but ARC can not tell if you mistakenly use the WRONG ARC mic.)

That's basically it.

The result of doing an ARC run is a data file on the PC. There's no problem having data files for different Anthems. When the results in the data file are how you like them, you use the ARC application to Upload those results to that Anthem after which you no longer need the PC connected to the Anthem.

When you install new updates of the ARC application on your PC, the calibration/licensing file pairs will be left in place from the previous install, so no worries there either.

----------------------------------------------

Things are only slightly trickier if you also happen to have Paradigm subs with the PBK feature (the subwoofer portion of ARC). There's a PBK mic for those.
--Bob

Originally Posted by abc999

Just a clarification, is the sub level 6 to 8 dB below or just the crossover frequency as I believe I read in your previous post?

Alvin

Originally Posted by JimP

Thanks for responding.

What you're calling a good "Pro Calibration" is what many of us have already done with REW (or other software) and lots of time. Obviously it more efficient to hire someone to come in and do it...I just didn't know if there was some secret sauce involved.

Originally Posted by LesPaul

Hello Bob and everyone,

I updated all my ARC files as Bob suggested and I noticed that one of my two mic calibration files was only 40 bytes in size and when I updated the files it was iK Bytes in size. I thought that this would correct my problem but it did not. My curves look just as they did before. I asked Anthem for a new mic on Monday but they did not replay. I will need to call them again. Any other advice? My ARC version is Beta .16 and i am using the latest D2v software.

Originally Posted by LesPaul

Hello Bob and everyone,

I updated all my ARC files as Bob suggested and I noticed that one of my two mic calibration files was only 40 bytes in size and when I updated the files it was iK Bytes in size. I thought that this would correct my problem but it did not. My curves look just as they did before. I asked Anthem for a new mic on Monday but they did not replay. I will need to call them again. Any other advice? My ARC version is Beta .16 and i am using the latest D2v software.

Originally Posted by Bob Pariseau

The licensing file is actually only 40 bytes, but takes up the minimum record size for a Windows file when on disc -- 1K to 4K normally, depending on the file system.

Are you back to your original curves with the minor dip near 12KHz and everything else OK, or are you getting the not as nice curves you posted recently where all the treble rolled off significantly, with 12KHz just being a blip in that descent?
--Bob

Originally Posted by Bob Pariseau

Remember that you can't get a proper reading from Test Level unless the volume trim line for the LF speaker is at 0dB.

Also check the volume trims that ARC Uploaded. If you have any extreme values ARC may have had to shift the solution to get enough range to trim all the speakers. The usual culprit here is the sub.
--Bob

Originally Posted by TboneofLA


Wow! So I checked the sub level this time, not just the speaker level for the LF and the sub level was very high in comparison to the LF (75dB) so I turned the sub level way down to get 75dB. This caused the ARC results on the charts to be even lower, 65dB now.

Now if I increase the sub level to get the ARC charts at 75dB there is a huge difference in trim levels in the ARC solution, the speakers end up as high as +10dB while the sub is -10dB.

Anythoughts on this? Maybe my SPL meter is out of calibration?

Originally Posted by TboneofLA

So I increased the test level to 4db and turn the sub level down a little to get the following result.

I think I'm going to ignore my SPL meter for now.

Originally Posted by Bob Pariseau

Yes, it looks like your meter is off. A weak battery will do this. Also pick the meter range that puts 75dB in the middle of the range. The ranges above and below that will be wrong for 75 by a few dB. For the Radio Shack digital meter, that means use the 70dB range, slow response, and "C weighting, pointing straight up at ARC mic position #1.

It looks like you originally had Test Level reduced 6dB for the old bug -- thus raising it is now right -- but your meter isn't giving a good reading to confirm that.
--Bob