REW REQ

[PENG]

I have not touched the REW EQ for a while and now I have forgotten something that I hope someone can jog my memory.

The question is, why are some of the dips not getting any boost at all? Below is a screen shot. You can see that REW EQ did not even try to address the dips at 87 Hz and 438 Hz. I was going to post at the HT shack but there don't seem to have enough traffic to return a quick answer.

 

[Pogre]

Damn, that waterfall tho...

I'm not an expert, but just got done with a round of REW and DSP'ing. I always assumed room modes? I dunno. I'd like to hear a solid answer myself. Those dips are what I deal with manually in the input section.

 

[fuzz092888]

I'm pretty sure it's because your max boost is set to 6-9db and since those are much larger than it it skips those as unaddressable.

 

[PENG]

I'm pretty sure it's because your max boost is set to 6-9db and since those are much larger than it it skips those as unaddressable.

Got it, thank you for the quick help. I thought it would still try to boost it to the limit, instead of skipping it.

 

[TheWarrior]

Because dealing with dips below transition of the room cannot solely be dealt with EQ.... rather Positional EQ

If your subs are not in the right place boosting will not change the sound pressures recorded/heard at the listening position.

 

[PENG]

Because dealing with dips below transition of the room cannot solely be dealt with EQ.... rather Positional EQ

If your subs are not in the right place boosting will not change the sound pressures recorded/heard at the listening position.

I realized that, and that's why I limited the maximum boost. The interesting thing is, if I set the maximum boost to the highest allowable value, it would do it as fuzz predicted, though of course I will not allow 16 dB of boost to any frequency. Even Audyssey has a limit of 9 dB (IIRC) for boost.

 

[TheWarrior]

I realized that, and that's why I limited the maximum boost. The interesting thing is, if I set the maximum boost to the highest allowable value, it would do it as fuzz predicted, though of course I will not allow 16 dB of boost to any frequency. Even Audyssey has a limit of 9 dB (IIRC) for boost.

Totally! There are other reasons why you might be measuring dips, but for the safety of any users equipment it is good to have those presets.

87 hz could easily be related to the distance between a wall and the mic, as well as roughly 430 hz being a 5th order multiple of that.

Turn your resolution down, 1/48 is unnecessarily sensitive when recording in a reflective space!

 

[William Lemmerhirt]

Are you crossed at 80hz? Try adding a couple feet in the avr and see if it goes away. Or move the XO point and see if anything changes.

 

[PENG]

Are you crossed at 80hz? Try adding a couple feet in the avr and see if it goes away. Or move the XO point and see if anything changes.

Thanks, try all that already, and much more more, including minidsp. I may try Dirac live next, or Audyssey, at least those are less time consuming to experiment with.

 

[William Lemmerhirt]

I thought you may have thought of those but you know, I felt like maybe you needed a visit from captain obvious. Lol

 

[PENG]

I thought you may have thought of those but you know, I felt like maybe you needed a visit from captain obvious. Lol

My helps are welcome, so thanks again! I think I posted way back that I did do a lot of sub and speaker crawling (highly limited though) when trying out the 2XHD minidsp. I was able to flatten the bass from 20-80 Hz +/- 2 dB or better at 1/24 smoothing. There were 3 or 4 dips between mid 80's to 700 Hz visible at 1/24 could not be corrected no matter what, and that's the same with all the speakers I have tried on with the mini. The LS50 and R900 were particular worse because with them side by side and a piano in between, the sub had to be placed on the side wall at 90 degree as it had to be seated right next to the big hifi rack/cabinet.

I have since given up on the mini that is now back in the box doing nothing. I believe Audyssey or Dirac could do better because they both worked in the time domain but as TheWarrior mentioned, those kind of dips are not going to be solved by EQ (I disagree with him as usual, to a point). My question here is mainly on technically, and fuzz gave me the answer.

Despite those dips, I found that raising the crossover to 100 Hz allowed me to enjoy the LS50 more, and I just started to play with the JRiver parametric EQ and they sound even better. I actually prefer the LS50 plus sub more now, than I do with the R900 towers. I think for the money I paid, the LS50s are unbeatable. Having said that, Dennis Murphy's BMR will most likely do better even in that same room but they also cost much more.

 

[William Lemmerhirt]

I’m glad you found what you were looking for. I do recall you talking about “the crawl” and such.
I have to say though, I’m surprised at your preference to the ls50 over the r900. At the risk of a derail, I have to ask what makes you lean that way. I’m also daydreaming about upgrading my mains/center, and have looked at the q900, but wonder if the R series is that much better.(since they’re considerably more) Or if the Q series is possibly not that good. Idk, I’ve heard neither one. Just wondering your thoughts on the two.

 

[Pogre]

I routed my mains through the Mini instead, then put everything back. This is what I ended up with the other day.

I set limits at 9db boost individual, 6db overall. I didn't go more than 6db in the input sections either. This isn't the smoothest chart I've produced (first one is with no smoothing), but I went after it with a much lighter touch this time. I'm using lower max boost filters this time and it sounds really, really good now.

My waterfall graph (while not terrible) isn't nearly as purty as yours is tho. Did you achieve that through positional eq or did you tweak something else?

Another thing I've noticed too, is that Audyssey always puts whatever speakers I route through the Mini and apply filters about 5 ft further away than with no DSP. It did that for both my mains and subs.

 

[William Lemmerhirt]

Hey poges, pretty sure it’s sets all the distances further is because of the processing time.

 

[Pogre]

Hey poges, pretty sure it’s sets all the distances further is because of the processing time.

That makes total sense. Like, there was an area of my brain that was suddenly illuminated when I read that.

 

[fuzz092888]

Hey poges, pretty sure it’s sets all the distances further is because of the processing time.

It is. Distance is set not by physical distance, but by distance + processing time which is what most auto EQ’s are at least most useful for if nothing else.

 

[PENG]

I’m glad you found what you were looking for. I do recall you talking about “the crawl” and such.
I have to say though, I’m surprised at your preference to the ls50 over the r900. At the risk of a derail, I have to ask what makes you lean that way. I’m also daydreaming about upgrading my mains/center, and have looked at the q900, but wonder if the R series is that much better.(since they’re considerably more) Or if the Q series is possibly not that good. Idk, I’ve heard neither one. Just wondering your thoughts on the two.

Only because I thought the LS50 sounds more transparent, and the violin sound closer to what I heard from the original blades. It could all be due to their proximity. You and both know the effects on the FR if you move anything (mic, speakers, body) by just a few inches. the speakers are side by side but that's much more than a few inches of difference.

 

[PENG]

It is. Distance is set not by physical distance, but by distance + processing time which is what most auto EQ’s are at least most useful for if nothing else.

Well, my AV8801, AV7005, AVR-4308 all managed to set the distance to within an inch or two (if not better) of the measured distance. Even the sub distances were very close between Audyssey and the measuring tape. It sounds unbelievable, but true, and I realize it doesn't mean it is good either. Again, it depends on so many things.

 

[PENG]

Guys, I have about an hour right now so will try harder to do more crawling, no EQ at all, just keep on crawling and see what happens.

 

[PENG]

After hours of crawling, I did get a little better results. This time I tried EQing the left and right separately to hear what will happen to the sound stage. I was thinking that if there are a few points in the 1500-4000 Hz range where the levels between left and right are very different, wouldn't that cause the violin soloist's positioning to become evasive/vague?

I just finished implementing the REW filters into JRiver and so far so good. I did not use the boost, just the cuts, for now.