Axiom Home Page
Posted By: MarkSJohnson When plotting frequency Response....? - 01/03/05 07:00 PM
As some of you may know, I've promised a big report/posting when I finally get my HT finished, and I'm in the final stages of that now.

I created my own test-tone disc with various combinations of Left Only, Right Only, and Stereo tones. I'm using that disc at this stage to plot some FR graphs comprising an M-60 in combination with my 350 sub, and have a question that I need answered:

Across the full range of frequencies, I'm using the Left or Right only tones, including the bass. Real-world music typically has mono bass, though, with equal energy in both the right and left channels. When I'm done plotting my graphs and simply enjoying the music or movie soundtracks, I can certainly adjust the sub level to suit my tastes.

But when plotting frequencies, which is more accurate or an accepted standard, using a louder mono signal or a single-channel-only signal?

Posted By: SirQuack Re: When plotting frequency Response....? - 01/08/05 03:04 PM
to plot or not to plot, that is the question. can't help on this topic, but hey I'll BUMP you up on the list
Posted By: mwc Re: When plotting frequency Response....? - 01/08/05 04:21 PM
So, are you saying that you are making a separate plot for the right ch and one for the left ch and another one with both channels? Not sure I know what you mean by "louder mono signal".

Don't forget to adjust the values to compensate for the RS meter.
Posted By: bridgman Re: When plotting frequency Response....? - 01/08/05 05:56 PM
Mark is asking whether he should test with only one speaker running (separate tests for L&R) or with both speakers running for more realistic results.

I don't know the answer. I guess my first thought would be that if you plot them separately you can always combine the results but if you plot them together then you don't have any way of seeing "hmmm, the left speaker is OK but there's something scary happening with the right speaker's room reinforcement" and doing anything about it.

If you're happy and just curious plotting both together will take less time. If you think you're likely to change things as a result then plot 'em separately.

My 2 cents...
In reply to:

Mark is asking whether he should test with only one speaker running (separate tests for L&R) or with both speakers running for more realistic results.




Actually, mostly true. I was wondering this specifically in terms of bass frequency plotting.

I'm plotting the left and right speakers seperately, using "left only" and "right only" tones. If I then wanted to plot the Left speaker combined with the sub to see the full-bandwith results (say, to check the difference between "small" and "large" settings on my receiver), should I use a "left-only" bass signal?

My first thoughts would be "of course"...

Except, almost all bass on music (forgetting an independant LFE channel at the moment) is mono, meaning that I would theoretically get 3db more bass when both channels are on and playing. In other words, the right channels' bass will ADD to the left channels in real-world music listening.

Now, again, I know I can walk over to the sub and turn it up or down as I'm listening to various sources... but in terms of plotting an accurate frequency response, what signal should I send to the sub while checking tones? A one-channel Left or Right-only signal or a "stereo, "both-channel" signal that' more like real world bass?

Anyone else hope that Alan will write an authoritive bass management article in the next newsletter based (sic) on all the bass-management questions here of late?
Posted By: bridgman Re: When plotting frequency Response....? - 01/08/05 06:17 PM
>>Except, almost all bass on music (forgetting an independant LFE channel at the moment) is mono, meaning that I would theoretically get 3db more bass when both channels are on and playing. In other words, the right channels' bass will ADD to the left channels in real-world music listening.

Ahh, I think that one is easier to answer. You're not just going to get 3dB more bass, you're going to get 3dB more "everything" so there won't be any impact on frequency response, just the difference between a lumpy curve around 82 dB or a lumpy curve around 85 dB
In reply to:

You're not just going to get 3dB more bass, you're going to get 3dB more "everything"




I don't think that's the case.

In a typical stereo music mix, you have instruments typically spread out... one guitar left, another right , for example. Some signals, such as a lead vocal, is typically evenly mixed L and R to place it "center" in the soundfield. Vocals don't typically have any energy in the range that would matter to a sub, though. A Bass guitar is typically centered. The drums might be centered in the mix as well, meaning that as a drummer progresses through a range of drums/percussion from a trebly cymbal down to a bass drum or large Tom Tom, the bass would unnaturally increase. If the drums ARE indeed "centered", the bass energy from the left and the bass energy from the right combine equally. Now, the mixing engineer knows this and will EQ the lowest frequencies to take this into account. He or she is going to mix it so it sounds right to him or her. All's well and good in the real world.

But now along comes Mark. He wants to (in theory!!) EQ his system from 20Hz to 20KHz to be absolutely, perfectly flat. He does so for his right speaker and sub combined, and then does so for his left speaker and sub combined.

He puts on a recording of a bass-light flute and acoustic guitar, and it all sounds perfect. If he puts on a recording of the band outlined above, I believe his bass will be heavy sounding. Because the vocals in the recording stay out of the bass regions and the guitars will be mixed to the left or right channels, which don't "couple" with each other in the mids or highs. The bass that shares both channels, though, will "couple" and reinforce each other I believe.

So the recording was mixed to sound correct in the real world, and the speakers in Marks' living room were each EQ'd to be perfectly flat. Yet, I believe, the bass levels will be incorrectly loud when playing that recording.

Again, as this pertains to me, myself and I, I'll walk over and turn the sub down. But hypothetically, on paper, wouldn't the above scenerio show that it's incorrect to plot your Left speaker and your Right speaker without using both channels of bass... since that's the way almost all stereo recordings are mixed?

I know... this Bass Management stuff gets "heady" and is a pain in the A$$. I'm sure that's why it abbreviattes to B.M.!
Posted By: bridgman Re: When plotting frequency Response....? - 01/08/05 07:39 PM
I have to respectfully disagree. When a sound appears in the middle of a stereo soundscape it is not "in both channels" but is "divided between the channels" so the result is what appears to be a sound source physically located in the middle at the proper volume.

If you pan that instrument/voice to one side or the other you get more signal on one side and less signal on the other. Put differently, a bass signal in the middle is already taken down 3dB on each channel during the mix so the result is at the right volume.

Your sound reproduction system is expected to have a flat response across all frequencies, not to treat bass signals (or anything in the center) differently.

Same thing when you add a sub -- the sub isn't supposed to give you "more bass than flat", your system just tries to route signals to speakers which can accurately reproduce them. At the crossover frequency for the sub both main and sub signals should be down 3dB so the result is the same as if you just had a super-full-range speaker.

Some people leave their mains set to "large" AND route the same signal to their sub which gives them some EXTRA bass between the sub crossover signal and the mains rolloff but that is not "flat response" that's "their personal preference".
Posted By: bridgman Re: When plotting frequency Response....? - 01/08/05 07:44 PM
One point I should make -- if you play music through your system with only one channel + sub active and tweak THAT to sound correct then you will get exactly the phenomenon you are worried about. That's because you tweaked the system to compensate for the centered bass signal being down 3dB on each channel.

As long as you have all your channels powered up when testing it shouldn't matter whether your test tones are one-speaker or both-speaker because the level will be dropped 3dB for both-speaker.

The big difference IMO is that if you're testing with a tone going to both speakers it's like having two subs to even out your deep bass response -- the warts of one speaker placement can sometimes cancel out the warts of the other speaker placement resulting in a smoother overall response.

Now quit yakking about this and go finish your report
OK, I'm going to have to read this several times to wrap my head around it. I'll do so after the report!

Thanks for the info! I'm not exactly sure I agree but, like I said, I think I need to re-read it!
© Axiom Message Boards