Originally Posted by GeoJustGeo
I think you're doing too much. Keep this simple.
1- Place your subs where you want/feel gives you the best response.
2- start from scratch (remove all HPF, LPF, EQ, Room Correction such as Audyessey etc etc, set your Center channel crossover at max like 200hz, Sub distances to 0)
3- Gain match your subs first
4- then Phase match them (start both at 0 or however your subs have them labeled. In the Minidsp start adding delay to your nearest sub in increments of 1. Run a sweep in REW with Channel 3 (center wont play since its crossed high but both subs will) from say 10-100. Set delay to 2 and repeat sweep. Rinse and repeat until you've found the best FR you can get. it won't be pretty but get it as close to flat as it will go.
5- Run room correction
6- save that FR and post here.
7- use the EQ function in REW to drop the Peaks. Dont try to boost nulls in Minidsp.
8- post both FR in here.
As for BEQ on that movie in Amazon, most of the BEQ's are done for the discs. Chances are you might have used a BEQ that was not meant for the source you were using so it sounded horrible. Or it could be you have so many filters and EQ that it muddled the sound.
The advice i've always been given is the less EQ the better.
I'm not an expert but i've moved my subs to 3 different rooms in the last 2 years and have been able to get them working great in every room even tho they are 2 completely different subs.
Also you don't want your FR to be flat. Flat is the starting goal, but then you want to have about a 5-10db slope from say 15-80hz.
Here's an example of mine with 2 subs from 2 different companies with 2 different driver sizes, different power handling, different port tune, 1 far field and 1 near field behind my couch. No smoothing.
Sorry for not coming back in this thread and looking for post, but I've been in hot pursuit. To start this post I want to agree with the sentiment to not do what I've done here even though I now feel I've found success.
Thanks so much for your help and its very, very interesting to hear about different ways to skin the cat.
SVS Ted suggestest this just to add to the mix; get both subs flat in minidsp and then do the phase nob. I've done a variation of this where I just never touched the phase nobs and it came out quite well; perhaps the most musical settings I've done. I've also done throw both subs in (did not move from placement above) with no phase adjustment and just hit it with minidsp. And lately I've been doing adjust phase first and then adjust both subs. The SVS approach was rejected for this because I did not want the PB4000 running flat since its carrying the load for two subs from 14-20 Hz.
In my system their is no Auddyssey so minidsp is the only game in town. For those with preferrabley Auddsyey XT32 (and I parrot on Audyssey and can't spell it.
), I would almost always recommend a minidsp even over the SVS app parametrics (which don't allow adjustment below 20 Hz). I've not done it, but I'd do episode 7, 26.1, and this from HomeTheaterGurus (I needed control because my bass layer does have some odd output below and above 80 Hz):
Then you can try all the methods discussed in the thread and probably get great results fast.
For my setup I do them manually which takes a ton of time and much to my chagrin the settings between these two subs have been like a house of cards so when I listen and try to tweak them in all the way sometimes just a change of 0.1 db or even Q of 0.1 can make a big difference; that is not normal. For this combo I've come to realize that I must just go for visual perfection with no tweaking which has worked great in the past. So again I don't recommend the sub combo done here specifically because of this reason. I've seen a lot of experts like AVRant poo poo this forum and say its nonsense that you can't run different subs together. They are right to some extent, but if you really want things perfect I'm not so sure this is great advice because I'm hearing important differences in the sound which I think primarily have to do with the bass perhaps below 25 Hz or so and getting it right.
So here is my final graph for now:
The totally flat graph is a manufactured guess to the actual response as I'm using a stock umik-1 that has not been calibrated by cross spectrum and may have significant bass errors. When I adjust bass below 40 Hz I use a low_shelf filter with settings like 10 Hz 0.3 db and 0.5 Q (current settings, but the db can varied a lot and the Q massaged a bit to make kind of a "house" curve for a microphone.) The red EQ'd response is what is going to the SVS subs and this is averaged with the Base Layer response (5.0.4 Atmos system). The rest of the system has no auddsey like calibration so minidsp is the only game for me.
The Measured Final is not exactly the final (just what I have now) and does not reflect my little mic/house curve adjustment of 0.3 db (that small adjustment probably indistinguishable on the graph).
The Black no so good sound is a reasonable looking graph like the Measured Final that left me wanting better sound. I did give it a huge adjustment/boost in the bass with the mic/house curve and it helped, but just overall even with a similar slope it left me wanting.
So right now I'm in some form of sonic ecstacy with the current settings so allow me to cork sniff what I'm hearing for the first time before getting back to the whether to's and why for's of the lowest bass.
1. The bass has really, really locked in. I've always gone for very flat and just assumed it was not absolutely necessary when I went for this sub combo. So I've had pretty darn flat before, but not like what I have now. This flat sound can be just a tad clinical as sometimes we like colorations like my "chocolate" sound with my PB12's when I went from 21 Hz 3db down to 19 Hz, but at the expense of a little 2 db bass bump.
I'm sure other little errors in the bass can elicit other effects that are very pleasing. For now I'll just say that flatness (or perhaps I am doing a little house curve in the low bass) just is more revealing in a good way.
2. On one of my favorite reference tracks I ended up playing something like 20% lower volume. My guess is before with those colorations that I had to turn up the volume a bit to get all those hi-fi qualities to come out and this was at the expense of some harshness which I blamed on the soundtrack.
For me at this stage this is a priceless quality as I go for more and more bass. I listen at reference volume or maybe even a touch beyond so backing off a bit in level is huge for me. I suspect this effect is perhaps also from having better extension which compared to my previous dual PB12-NSD which is granting a bit of a loudness effect versus less extended systems.
3. For my Atmos trackings with this uber flat response all the little details are just popping out. I certainly have noticed this affect with prior tunings and its a big driver in going for perfection (I'm saying its worth going for the flattest possible response, and then applying your house curve.
4. Dialing the low end extension below 20 Hz just so also seems to really enhance the tactile feel to things. I'm not sure what this affect is but my little mic/house curve EQ that now stands at only 0.3 db really makes a difference. It seems silly when you've got massive BEQ settings swamping the same area, but nonetheless tuning that slam effect pays big dividends to my ears. You really want this tuned just so and once I have this adjusted for my graphs I never get benefit varying it from movie to movie so far. This last tuning I really tried to get my PEQing of the subs to come out with my preferred curve without this adjustment. This one you definitely tune by ear. And I'll add that because how we reproduce low frequencies impacts are perceptions of the highs, that this level of cork sniffing can really add to the natural perfect effect to the high frequencies so you're really dialing in both things and they appear to improve together.
This is all tricky stuff and if you look at the graphs the "bad" sounding black graph looks pretty darn flat. Why do they sound so different:
1. First the ear can easily hear pretty broad level changes so just 0.1 db shelfing up or down is easily audible.
2. From my tuning of the upper bass I'd say the main benefit from flattening is removing little peaks that inevitably are muddying your sound from the rest of your speakers. Getting rid of one little peak or tamping it down is a very subtle change and doubt it effects the bass character very much.
3. As you go lower in frequency every little dip and valley makes more of a difference. This does make sense as we see response graphs show these low frequencies as a greater percentage of the sound. For me this was most true below 25 Hz. I've spoke of the Chocolate sound of the little bass bump at 22-23 Hz. I suspect every little filling of a valley by 0.1 db is important in this range and this is responsible for much of the bass character of a system.
So right now I'm elated with the sound, but I still have work to do. Despite now having the PB4000 nominally get 9 db more requested from it I still am seeing limiter lights touch up on the big bass impact scenes on the PB12-NSD and not the PB4000. I do think I'm very close and may just leave it alone. I've got a couple big bass boosts on the PB12-NSD and one at 60 Hz I'm going to try to move over to the PB4000 which has more output in that range. Minidsp HD has 10 parametric type settings possible and what I've done is move three of those over to the SVS app to free up slots. I'm currently using 7 slots, but one will be taken up by my mic/house curve setting from the inputs (BEQ needs all those inputs free
). So I have two to spare and this will be my next trick to try to share the load. I also need to watch more bone crushing bass scenes as I've only tested this with explosion scenes at end of GI Joe Retaliation and that might be hitting upper bass a bit harder than the average BEQ heavy.
If I have to leave it as is that is fine because the PB12-NSD is not under much duress now and its probably better to have it be the sacrificial lamb as in the long run hammering subs with BEQ does shorten their lifespan.
Again don't do this kind of dual setup. I had a very, very hard time filling the hole at 20 Hz and many other issues in the response. I've had to do things like:
1. reduce bass on one PEQ to get the overall response to go up.
2. massage about five different PEQs around 20 Hz to get the valley to fill. The direct boost at 20 Hz could only do so much and eventually as I boosted it lost output and then causes waves in response beyond.
3. Because of phase issues in parts of the range sometimes you'd alter Q and on one side the response would move and on the other it would not budge at all.
4. You also have level differences between the subs so when out of PEQs on the PB4000 you'd have to do a lot more on the lesser sub to get the same impact and of course weird moves in FR because of this as well.
5. In general nothing you wanted to do directly in BEQ worked all that well because of 3 and 4 impacting. I also believe that their was special issue due to the different LF rolloffs which is why the 20 Hz valley was so, so hard to deal with.
6. All those warnings aside if you have tons of flexibility in your placement then doing two, three, and even four different subs in the same room may be dramatically easier.
I'm also quite confident minidsp HD has a bug with its PEQ as the response was always quite different if you used the slider to get to db whole numbers like 2.0, 4.0, 16.0, etc. If one uses the up and down arrows on the keyboard to get to these same numbers the response was usually different. What's worse as more times than not the slider activated whole number gave much better behavior than all the numbers so if I had to guess I'd say any db settings not down with the slider don't work right.