Poll: How do YOU set your gain structure???? - Page 3

my understanding is that the goal is to maximize dynamic range. And if the amp cannot be driven into clipping at AVR calibrated "reference" level than the dynamic range is truncated by definition. Does it sound about right?
Granted one can allow some padding for headroom if power is plentiful. But I suspect most of us do not have more than 3dB or 1/2 of the max power to spare.

Bass Addict,

Just wondering here, if once you have finished setting the gain structure in the system, are you than going back and setting the individual speaker and sub levels with the internal test tone of receiver-processor? I just cannot see how you are getting such a difference in between the speakers and sub(s) once you have set up the gain.

James

I don't use Audyssey's level settings, especially for the sub channel.

And, Yes, I always set gain structure after Audyssey and outboard EQ is complete. I don't know how one would do it before proper, not knowing what boost filters may or may not be applied?

yeah, setting gain structure after Audyssey makes perfect sense. How do you do the level matching? SPL meter?

Aint nothin wrong with that! I don't manually EQ my mains(yet) but I hope to when I grow up.........

FWIW, Ive read both ways work: Audyssey first then manual EQ and manual EQ first then Audyssey. IDK which is the definitive best way, I just enjoy the control of the final say on FR.

Sure, meter works.

I use REW though.....

Well that certainly clears things up. I was thinking there was no way this thing was cranking out 20v. 7 sounds much more likely. Thank you. That being said, the Onkyo does not show clipping at full tilt.

So in my other post when showing gain structure, my SMS1 is clipping the output signal at 1.77 volts not 5. Now what I'm not sure of is if this is an absolute value, or a value correlating to boost over incoming voltage via the volume knob.


Once again, the way I understand it; gain structure and calibrated level are two totally separate things.

I setup the gains as I should. I didn't use full boost on Onkyo as it's not required. Onkyo level was set to 0, I then opened the ep4000 gains to wide open and turned up the volume on the SMS1. I know the SMS clips the output at 28, so as long as I'm below that level I'm fine. Clip lights came on at volume 26, so I backed it down to 24.

So initial gain settings before any eq'ing were Onkyo 0, SMS1 at 24, and EP4000 wide open. Here's where I have the issue:

Both Audessey and the SMS see some major differences when calibrating at this point. And it makes perfect sense. Setting up a gain structure is not the same as calibrating output. When I went to EQ my SMS curve, at my gain structure level, there was no way to match the main level and sub level. I had the volume so far down on the receiver the mains sweep wasn't even registering. I had to turn the sub level to around -10, and raise the volume, to get the sub level and speaker level at around the same 80db level. So I know according to the ECM mic that the sub level was crazy hot at an output of 0.

Audyssey also confirms the same thing. When calibrating speakers, they are set around -2, and the sub level is set at -12; and this is with the gains on the 4000 backed down to about a 1/4 and the SMS volume at 13. When I turn them up any further, the sub level can't take enough out. Now I haven't verified this against an SPL meter as it's loaned out to a friend, but I trust that my SMS mic is pretty accurate and it's showing the same thing.

So here's the nutshell. Assuming the SMS1 is boosting 1db per volume unit increase; I have a difference of 22 db's and 3/4 gain between gain structure setup and calibrated setup?????????????????????

"Once again, the way I understand it; gain structure and calibrated level are two totally separate things." They are seperate but I guess what I stated - was asking is I just cannot see how the speakers and subs are so far apart after doing your steps.

To set the gain I did the steps below. I used the files provided by Wayne over at HTS.

1) Trims for L-R and sub were set to +12db (highest they could go using an Onkyo 885) with Onkyo set to stereo

2) Speakers set to an 80HZ crossover (I measured them with a volt meter and received the same voltage with a 1KHZ 0dbfs signal with this crossover and at full range).

3) I increased the master volume until my signal went into clipping and then backed it down until it was out of clipping.

For the mains with the 1KHZ signal I was clean until 0db on the master volume.

For the sub signal I was clean until -2.5db on the master volume

I than (with speakers disconnected) played the 1KHZ signal at 0db on master volume and increased the gain until I barely lit the clip light. The mains use a QSC GX5.

I than played the 60HZ signal for the sub channel at -2.5db on the master volume and increased the gain until I barley lit up the clip light. The subs use a Marathon MA-5050.

I than played the internal test tone to set the trims to read 75db on my SPL meter.

I have attached some screen shots of the o’scope readings for both the 1KHZ and 60HZ signal with no clipping and clipping.
* If you want you can see the scope reading at the thread below
http://www.avsforum.com/t/1330718/how-to-set-gain-structure-properly/30


James
Edited by exojam - 2/20/13 at 1:39pm

Not really, in regard to where I think the confusion may lie. If, at calibrated reference level, only 250 watts are required to achieve ref level, you're not going to be driving a 1,000 watt amp to clipping.

More or less, the only time I'd think you'd expect to see amp clipping at calibrated reference level is if the amp has only enough (or not enough) power to drive reference level volume in your system. You can't just adjust your gains to make the amp clip at reference level if the power output at (verge of) amp clipping isn't necessary to achieve ref level.
Edited by whoaru99 - 2/20/13 at 11:22am

Well, this does not work in my system. My sub is nowhere near required SPL with Audyssey calibrated ref sub signal level. And how could it be, if with that input signal and gain settings my amp can only produce around 150W into 4 ohm load?

Hi James,

Thank you for sharing and for the link. you obviously went through the troubles of setting up gain structure as proscribed by the pros. respect.

Can you check how far below clipping your subwoofer amp is during the WoW tripod scene at ref level? That is assuming you can tell what settings on MVC correspond to the ref level, which may be wrong assumption given that you did not use Audyssey for level matching.

thank you

Zheka,

I have to be honest and say I just do not have time on my plate to go through the process you requested. Sorry.

James

Sure, I understand.

I'm not clear what you're saying.

If you forget about trying to tweak the gain structure and just run an ordinary, every day straight up Audyssey cal, what are the resulting channel trims?
Edited by whoaru99 - 2/20/13 at 5:04pm

right, that's exactly what I did. I was prompted to lower the sub volume during the calibration. So I got the trim on the amp around 12 o'clock position to get it within the range Audyssey accepted. After the calibration my sub channel level was set at -7, surrounds - at -11.5 and the fronts are around -3.

I arrived at the 150W number based on the assumption that at 0dBFS my amp is supposed to do around 2kW bridged into 4ohm and the max output during the tripod scene was at -12dBFS which is roughly 1/15th of the max power or 133W

Where does 0 dBFS or -12 dBFS come from? What are these with reference to?

0dBFS - max output before clipping lights on the amp meter turn on

-12dBFS observed on the the amp meter during the tripod scene. It was confirmed by the fact that I had to boost the signal by roughly this amount to get the amp into clipping territory. In fact it was slightly more than 12dB boost: 7db on AVR sub channel and 6db input gain on the amp.

I must be missing something...

If you ran Audyssey cal and made the necessary adjustments for the cal to successfully complete, then your amp is doing what it's supposed to be doing. You can't force the amp to put out 2kW / indicate 0dB on the meters and still keep the calibrated channel/system balance if it only takes 133W from that amp to do the work.

I do not know what to say. I understand your point that the amp should not necessarily be at full capacity for ref. level playback. And there are guys here with systems that have plenty of headroom that would support your point. But mine is not one of them, especially with this much power.

actually, maybe it can get close to ref level with 130W, depending on the frequency of the test signal. wow.
I'll post the sims later.

CHT SS18.2 - dual 18" pro woofers, approx 90dB efficient each.

Without EQ the drivers begin to roll off relatively high but w/ 130w, predicted SPL is above 110dB from approx 40Hz and up.
Obviously the output would be lower at 10ft(my LP) but then the room may play role as well. perhaps there is a mode at LP that boosts the test tone.



do you guys know the frequency range used for the subwoofer PN test tones?


i guess the lesson is that I have much more room for the LF boost than I thought. e.g. I can get a curve like this with a few filters and still be well within power and excursion limits.







I may have to rethink power requirements for my upcoming setups.

whoaru99, thank you for making me look at this more closely.

yeah, i still think the levels are off in my case but maybe not by as much as I thought, especially in the mid/upper bass.

A good house curve eats up headroom pretty quick...... Audyssey applies a pretty good bit of LF boost for me out of the box. Im using the MulteqXT variety.

Z,

it really sounds like to me, that your ready for a full measurement setup. Many of the questions your asking beg measurement to resolve completely.

I had thoughts on what 0 dBFS means to me but I was looking for more clarity of the context it was being used in this case.

That's in part why I was surprised by such low power consumption in my case. I use a low shelf filter plus whatever Audyssey/DEQ does.
With Audyssey disabled my curve is fairly flat to 20Hz at the LP. But I measure at relatively low SPL which may be my problem.

I will definitely be investigating more. My new subs are coming soon, I want to make sure the are set up properly.
I am using REW with calibrated Galaxy meter and Edirol USB SC . should be good enough for the task (as long as I know what I am doing which may not be the case)

No problem.