Flat Frequency Response

B00nie

[H]F Junkie
Joined
Nov 1, 2012
Messages
9,330
You can get it in a few pieces of hardware too if you like. Emotiva uses it in their new XMC-1 processor. The MiniDSP guys that sell little general purpose DSP units have ones with Dirac Live on them too.

I'd love to see a pro sound card with room correction, but I don't think such a thing is likely to happen. Too many pros don't like it and you wouldn't want to waste the money on another processor in your unit for a feature many would hate on.

Thanks! I'll be probably getting one of those. This might be my holy grail of ESL response correction actually.
 

Sycraft

Supreme [H]ardness
Joined
Nov 9, 2006
Messages
5,090
Thanks! I'll be probably getting one of those. This might be my holy grail of ESL response correction actually.

If you do get one post which one and your impressions on it. I've been considering going that route with my system. I'd like to stop messing with HDMI and just get a pro card from RME, but of course that doesn't work well for doing room correction with a receiver. However those little DSP units look like a good solution possibly. Particularly since I might just do RC on the front channels, and not worry about the sides and surrounds.
 

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
There's another one not coming to my head about $1.5k and a bit and does all you need with room correction. Works great with the demo one I had.
 

|Tch0rT|

[H]ard|Gawd
Joined
Jan 11, 2007
Messages
1,245
I think Trinnov is still tops but they need it to come out on better receivers than the Sherwood Newcastle R-972...

I always forget about Trinnov cuz it's usually only found on the ultra crazy expensive stuff like $20k+. I wanted one of those Sherwoods, almost tried to track one down when I was ready for a new pre/pro but I read they can be very quirky and I can't do that, too much fuss when it has to work for my whole family who are less technical than I am with gear.

Thanks! I'll be probably getting one of those. This might be my holy grail of ESL response correction actually.

I've been curious to see how some of the higher end room correction work on my ESL's. I did try the Dirac Live trial about a year ago but it ended up being a shit week I ended up only listening to about 10 min of music based off some not that great mic placements when doing measurements. :/

Home Theater Shack has a review of some Dirac products with MartinLogan ElectroMotion ESL's (same model I have) and that should give you some insight on how it works with ESL's. I haven't really read them myself just cuz I'm not ready to drop ~$1000 on something like that yet but if I do I'll probably get a DDRC-88A.

miniDSP DDRC-22D Dirac Live 24/96 Room Correction Audio Processor Review
http://www.hometheatershack.com/for...6-room-correction-audio-processor-review.html

Audyssey XT32 vs. Dirac Live Listening Comparison
http://www.hometheatershack.com/for...-xt32-vs-dirac-live-listening-comparison.html
 

Meeho

Supreme [H]ardness
Joined
Aug 16, 2010
Messages
5,403
How useful are these room corrections for multiple listening positions or when moving around. Can they make matters worse when not in the measured sweet spot?
 

B00nie

[H]F Junkie
Joined
Nov 1, 2012
Messages
9,330
The more advanced correction systems have multi-point approximation that calculates a compromise that sounds the best in all positions.

If you're into sound stage and sweet spot listening you'll probably want to correct only for the sweet spot.
 

dandragonrage

[H]F Junkie
Joined
Jun 5, 2004
Messages
8,298
How useful are these room corrections for multiple listening positions or when moving around. Can they make matters worse when not in the measured sweet spot?

Yeah, typically you can get this feature if you want to spend at least $800 (if not more) on a receiver. But how well the feature works depends on your room and specific placement of your speakers and the listening positions themselves. But there is gear that can attempt to equalize multiple listening spots at once and sometimes it does work well.

But if you really care a lot about sound quality in several listening spots, you should at least buy a few things to help with room acoustics such as diffusers or bass traps.
 

B00nie

[H]F Junkie
Joined
Nov 1, 2012
Messages
9,330
Having a morbidly obese wife put to her place i.e. corner of the room makes for a good bass trap.
 

Nicholars

Gawd
Joined
Jan 10, 2012
Messages
788
"flat" sounds bright and a bit boring to me, the "Harman response curve" is the best way to judge frequency response (on headphones) imo.
 

paret0

Limp Gawd
Joined
Aug 29, 2012
Messages
368
That's the old way of EQing. The new way is a measurement mic, REW (Room EQ Wizard), and a PEQ. Also EQing should be used to subtract, never boost. Dips are phase and room problems, boosting makes them worse. ...


Thanks for that again. Gonna get uncharacteristically chatty here, so bear with me..

I had taken out that 8200hz we were talking about because I thought I was getting too much essing in vocals with some exaggerated high harmonics. I discovered it was OK to add 7-8khz back after checking the pink noise plot with SpecScope on my MotoX. If it esses, then it just esses on the recording, and I should be at peace with that. I haven't said, "Sibilance", just a little essing in some live recordings that I can now hear the sound men fix as the recording progresses :D

But after looking at the sliders on the eq,, some were boosted. Time to take what you said as fact and start over with the available tools: Internet Archive "10 Minutes of Pink Noise" @ 24/48>Deadbeef (Crunchbang) with no DSP>Alsa w/no sampling>ALC889>Peavey USB-P>Monoprice Eq>Monoprice 10"sub>Monoprice 5" monitors


I liked the way it sounded (not the pink noise... :p), except it was a little woolly on downsampled stuff with a lot of dynamic range, and there was still a lot of wall resonance and "boomy spots" if you moved around the room (the USB-P only supports max 24/48, so deadbeef has to downsample 24/96 Black Sabbath vinyl rips and 24/192 aspie Eagles rips :p).

So I used some throw pillows as bass traps to quiet a loud wall next to a hallway and went to town with SpecScope, only chopping freqs to flatten this time, never adding. I moved more sliders than I had previously, but it was a lot easier to flatten the entire curve.
Bye bye "Woolliness". The 4 microphones on a MotoX aren't entirely inadequate for this job on this day.

Fruhbeck conducting Albeniz' Suite Espanola= (blow out the windows ) :)
Baduzim @ 24/96? :) = (Level the housism wit no fuzzizm)
Essential Dixie Chicks @ 16/48 MP3 or 16/44.1 flac = big bass, but not too much :)
Boz Scaggs Dig 24/48 = HUGE
Diana Live In Paris = Piano sounds like a big Steinway with a light touch, and not like Hank Jones' big Yamaha with a cooler tone and an even lighter and more precise touch. I also think that double bass is the same model as Stanley Clarke plays, the tone is so similar... :)

Took about an hour of fiddling before I was satisfied I was not going to get any flatter, given the methods and instrumentation :D

Holy cow it sounds good. Big and fluid. About 3db extra below 100hz, but I made a preset to flatten it in Deadbeef's software eq in the off chance I decide to master anything for $$ on a $500 system I equalized with a phone*. :p

"The Old Way" is a buttload of work compared to a software solution, but demonstrably much better than I could do with my unaided ears in the interim.

Cheers! :cool:













* I did take it out of its case.
 

B00nie

[H]F Junkie
Joined
Nov 1, 2012
Messages
9,330
EQ:ing is not simple done by hand. The room response graphs don't tell you the truth unless they're time window based. WIth no time window all you get is an average of all the room reverbations - your brain doesn't handle audio like that.

Also when you eq, to reach neutral end result if you want to drop 1khz for example, you have to boost one octave higher and one octave lower just slightly to avoid a dip around the area you wanted to manipulate.
 

paret0

Limp Gawd
Joined
Aug 29, 2012
Messages
368
EQ:ing is not simple done by hand. The room response graphs don't tell you the truth unless they're time window based. WIth no time window all you get is an average of all the room reverbations - your brain doesn't handle audio like that.

Also when you eq, to reach neutral end result if you want to drop 1khz for example, you have to boost one octave higher and one octave lower just slightly to avoid a dip around the area you wanted to manipulate.


I'm looking at basically a realtime spectrum display and sliding olschool eq sliders to flatten the imaginary fit line to within a db or 2.

SpecScope:

https://play.google.com/store/apps/details?id=com.nfx.specscope&hl=en

https://www.youtube.com/watch?v=X1n7VRuTB34


I did everything I could to kill ambient noise and measured each speaker on-axis at 1m to get a baseline, and to match output. Then I put the phone in the prime listening position and started backing out resonances until pink noise was pretty damn flat on a linear display, and white noise is flat on a log display. It's not splitting the atom at the level I'm doing it, but it's better than my rookie ears can do and it's better than bypassing the eq and not bothering. :)

On Edit: I'm also a USB mic and a rainy afternoon away from a software solution like REW and that phase corrector-thingy, so don't look at my flailings as real audiophile stuffs. :D
 
Last edited:
Top