Originally Posted by goneten
The response curves are never going to be exactly the same but then a gain matched approach won't yield similar response curves either if both are in dissimilar locations so you are knee deep in the same predicament.
This is the first place where you clearly don't get it. To "level match" the response curves have to be the same or the levels are not going to match. When gain matching, you do not care about the in room response at this step, you worry about it in later steps. Gain matching means matching the gain, simple, straight forward and achievable.
Matching sound pressure levels could fix problems within the response at the listening location while also offering higher peak SPL which one could then EQ.
You are not going to get higher peak SPL, it will be lower. With level matched subs one sub will hit max SPL before the other. Call it sub A. When sub A is at max clean output in the gain match approach, sub B will also be at max clean output. In the level matching approach, sub B will be at some lower level.
Unless you have set them up with destructive interference (stupid), if both subs are in the same locations in either scenario, both subs at max output will be more SPL than one at max and one at something less.
Proper positioning and measuring the results are still paramount to achieving decent results with either method.
With two gain matched subs, you can move them around and try different possitioning to get decent results.
How exactly are you doing this with level matching? Do you re-level match at every spot you try? Again, when level matching at each spot, how are you accounting for the peaks?
It's not like gain matching has some inherent superiority over level matching despite your so-so; the only advantage is that each sub theoretically will have equal drive levels (which could be less than 5% for all you know) which equates to equal work load per sub.
The fact that you have equal drive levels means you get more output. When one sub is at peak, so is the other, and Max+Max beats Max+Less. This is a big advantage if you push your subs, but if you are watchign chick flicks at -25db, I guess it is not a big deal.
The other advantage is much easier to get a good setup. See above on finding best location.
Or, alternatively, with a level matching method, slightly greater work load per subs but better in-room coupling which equates to 'not much worse of a situation if you ask me but, well if you had to ask, actually higher SPL as a byproduct which one could then equalize to their hearts content in a number of ways'.
You do not get better room coupling or higher SPL's. You cling to that fallacy to make your point, but it is false. See below.
Of course a level matching approach could also yield equal drive levels because the sub's could possibly be co-located in which case you gain full coupling and equal work load per sub in which case you have a level matched and gain matched set up. Harmony.
Ahhh, and here is the answer to one of my questions
Level matching is an OK shortcut when it gives results that are very close to gain matching. With co-located subs or in many cases subs that are fairly equidistant to the seats you can get similar if not equal results. That is not a big plus for level matching, it is simply gain matching in a more circuitous manner. AGain way off base on the coupling, see below.
Or....one sub has a hole at 40 Hz of 10 db's and another has a peak of 5 dB's at 40 Hz. Gain matching both in that scenario won't do jack to alleviate the 5 dB deficit. Even though you aren't really looking at two response curves but one combined curve, the combined gain at one location could yield piss poor results.
Again, you have a clear lack of understanding. In SPL levels -10 plus +5 does not equal -5. The fact that you don't even understand what a DB is explains why your understanding of coupling is so wrong, but that is another issue.
Lets assume the subs are not co-located but setup in phase and relatively equi distant to the seats so you are getting +5db coupling.
Say your test tone is at a target level of 75 db from each sub so when both subs contribute equally (not near 40 hz) the result is 80 db.
At 40 hz, one of your subs is outputting 75-10 = 65 db and the other is 75+5 = 80 db. What do you get when you add 65db to 80db? A summed signal that is barely over 80db. In other words, you example would give a perfectly flat response with gain matched subs. WTG proving my point
There is nothing level matching can do to make this situation better, only worse. If you turn down the gain on the +5 sub, then you need boost to get to flat, and you will have less output everywhere else. If you turn down the gain on the -10 db sub, it will have little audible impact at 40 hz, but result in less output everywhere else. I am not sure who is after less output, but you can achieve the same thing by simply turning down the LFE trim.
BTW, even if you assume the worst case and you only get +3db coupling. Now your summed target is 78db and you will have a +2db "peak" which if you even care about it, is pretty easy to EQ away.
You can then equalize at the listening position but comb-filtering will invariably mess up the results so you are actually correcting things which don't or shouldn't necessarily be corrected.
In either method, placement and phase need to be correct to minimize "comb filtering" which is not hard to do given the wave lengths involved. In real world rooms you will benefit from EQ at the seats. Your comment about correcting things that should not be corrected makes no sense. Over EQing is another issue that has nothing to do with gain/level setting.
A $19.99 Radio Shack meter when using the correction files for FR compensation can allow for accurate enough results, although it's not surprising that you would correlate price with quality. I don't use the Radio Shack SPL meter as my primary meter. I use the Behringer CM-8000 with phantom preamp.
again you simply did not get what I was saying, because you don't understand. I was not talking about measuring frequency response, I was talkign about measuring SPL levels. When used as a mic the RS meter can be corrected just like any mic. What I was asking is how do you set the SPL levels? if you are reading the display on the RS meter, than you are looking at the uncorrected mic responce of the device WITH C WEIGHTING applied. You could match the levels of the subs to give the same reading on the RS meter, but it will be heavily peak influenced especially toward the upper end of the range. If you are taking FR curves with any corrected mic you will have two dissimilar curves, so at what point to you match the levels? You fail to answer this rather basic question. Exactly HOW DOES ONE LEVEL MATCH in reality, not perfect response curve theory?
Of course a gain matching approach will give better max headroom, but then, no it won't. Because your understanding of wave coupling is sorely limited. Equal level and phase. Hence I said from the start that this discussion was way bigger than you.
You need to go back and learn what a db is and re-study wave coupling. If the waves are not equal level you will see a lower increase when measuring on a db scale but that is because the scale is log based and not linear. The energy still adds and there is no magic about equal levels that creates additional energy. Again, if X>Y adding X+X is going to be more than X+Y. This is not a hard concept, and on an absolute scale it is easy to see, on the log based DB scale perhaps you find it confusing.
Lets say your two gain matched subs at max are putting out 105db and 80db respectively at a certain frequency. The summed response will essentially be 105 db. Adding the two together gives virtually no increase on the db scale. If you level match, you will turn down one sub to 80db and the other will be at max at 80 db, and you you will get a max of 6 db in gain, to a summed level of 86 db. WHOOOOHOOOOO, look how much better the coupling is with level matching, you got +6 db instead of a small fraction of a db. Everyone knows plus 6 is better than +0.06 right? Well, one pair of subs is giving you 105+ db and the other is giving you 86db. Except for our wives, most people would prefer the 105. This will work with ANY difference in room response, the gain matched subs will always give more output, the level matched subs will show more gain on a db scale but that gain will be from a lower level and the result will always be lower. ALWAYS. Energy is not created.
Trust me, I am not the one whose understanding is "sorely limited."