Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mikkel]

I read the topic on convolution engine and wondered what exactly DRC achieves that using a spectrum analyzer, impulse response chart, delay-setup and the parametric equalizer cannot do?

Just curious...


Best regards,
Mikkel

[TheLion]

"Time Domain" / "Phase" / "Group Delay" Correction

Regarding frequency response: much more precision than possible with PEQ - a 65k tap FIR filter is like an 32k band EQ.

Sound Quality: Try both and decide for yourself. Audiolense has a free demo, DRC is genrally free, Acourate: two free test tracks

I myself came from consumer EQ (Audyssey (Pro)), went on to PEQ with REW, used Audiolense for more than a year now and Acourate for the last week. Each step was a BIG improvement for me. With Acourate I have finally found sonic bliss

PS: My room is very well treated ($ 4k of GIK panels), my speakers are very well behaved even without EQ (Danley SH-50 fronts, Genelec 1037C studio monitors as surrounds, dual Subwoofers) and still DRC is essential for me - has as much influence on the "sound quality" as the room itself.

[mark_h]

PS: My room is very well treated ($ 4k of GIK panels), my speakers are very well behaved even without EQ (Danley SH-50 fronts, Genelec 1037C studio monitors as surrounds, dual Subwoofers) and still DRC is essential for me - has as much influence on the "sound quality" as the room itself.

I too view digital room correction as essential.  There are some things that are not practical to treat physically and DRC is essential to get to that last level of performance.

[Trumpetguy]

I myself came from consumer EQ (Audyssey (Pro)), went on to PEQ with REW, used Audiolense for more than a year now and Acourate for the last week. Each step was a BIG improvement for me. With Acourate I have finally found sonic bliss

As far as I know (I am no guru on the fundamentals in DSP) Audiolense and Acourate are developed with different views on what is optimal DRC. How would you describe the two, what's different with respect to sonic results?

PS: My room is very well treated ($ 4k of GIK panels), my speakers are very well behaved even without EQ (Danley SH-50 fronts, Genelec 1037C studio monitors as surrounds, dual Subwoofers) and still DRC is essential for me - has as much influence on the "sound quality" as the room itself.

That may very well be the reason why you are so satisfied with DRC. Trying to DRC your way out of a hopeless room can give really nasty results. I think there are quite some that believes receivers with DRC give 'perfect' sound at all conditions.

[mark_h]

Clearly you can create a deliberately poor room in which DRC will have less effect, but in a normal room DRC has a lot to offer.

[TheLion]

As far as I know (I am no guru on the fundamentals in DSP) Audiolense and Acourate are developed with different views on what is optimal DRC. How would you describe the two, what's different with respect to sonic results?

That may very well be the reason why you are so satisfied with DRC. Trying to DRC your way out of a hopeless room can give really nasty results. I think there are quite some that believes receivers with DRC give 'perfect' sound at all conditions.

Exactly right - DRC is in no way a tool to fix a bad room. But to optimize a well thought out setup. Especially Time Domain Correction usually gives horrible results in "bad rooms". So you have to do your homeworks first - Treat the room as best as you can, find the optimal placement for speakers and multiple (!) subwoofers and your listening area. Only then the setup is ready for a final touch of DRC.

Audiolense is geared towards a turn key multichannel solution. Its user interface is great and very intuitive. It has all the features to implement a HT setup with any given channel configuration. Acourate is very different from that approach. It is like a scientific acoustic calculator which also can be used to generate filters. You don't work with multichannel setup but on a per speaker basis. The user interface is quite raw - geared towards the DIY crowd. But it shows much more information, better analyzing tools, much more options (eg. different XO filter types).

To my ears, in my room, with my setup I get much better filter quality with Acourate. I have been optimizing my Audiolense filters over the past year - Acourate with default parameters bettered the best I was able to achieve with Audiolense... Especially the preringing artifacts with Audiolense TTDC are not an issue with the latest Acourate version anymore (Bernt is working on Audiolense revision 4.3 which "is said" to make this much less of an issue as well). The basic psychoacoustic model behind those two apps seams to be very different - I get much different results when using the same measurement, same target curve, same correction parameters. For me Acourate works much better.

You can send Uli Stereo impuls measurements and two test music tracks of your choice and see for yourself how Acourate works for you!

[Trumpetguy]

Clearly you can create a deliberately poor room in which DRC will have less effect, but in a normal room DRC has a lot to offer.

I agree. For the record, I am an addicted Audiolense user myself, and cannot imagine my 2.0 or 7.1 setups without it. Such DRC offers so much in the XO integration of subs as well as in a traditional stereo setup. I have gained some experience over the past two-three years. One of them is that there is much to gain from optimizing the physical setup before applying DRC. Also, you can go really wrong if you are not careful in a room with large problems with nulls and broad frequency dips (valleys). If you are aware of the limitations, there is - as you state - much to gain.

[mojave]

I'm also curious about the differences between PEQ and DRC. Using REW and the PEQ in JRiver, I can get an almost perfectly flat frequency response. Below is a pair of JTR Captivators that I EQ'd. It took 3 minutes and 45 seconds to measure with REW, generate filters, enter the filters into JRiver's PEQ, and remeasure. I've also done EQ to some mains and was able to get them almost flat, too. When integrating mains and subs (or any separately amped drivers) you can see by the frequency response whether phase/group delay are an issue. By using the Real Time Analyzer in REW, you can adjust the distance setting and visually see exactly when the drivers are in phase. At the recent subwoofer GTG, the sealed subs needed about 2-3 ft of delay, ported subs needed 5 ft, and the horn subs needed 17 ft.

If both PEQ and DRC produce the same frequency response (within a dB or so), then what is better about DRC?

On a related note, why does an orchestra sound so good when there is obviously group delay among all the instruments?

[Trumpetguy]

On a related note, why does an orchestra sound so good when there is obviously group delay among all the instruments?

That is the difference between sound production and re-production. In the concert hall, the orchestra sound would reach your ear with all its grandeur, flaws and delays. You will listen and enjoy. The recording engieneer replaces your head, but during playback/reproduction your listening room adds frequency dips/peaks and group delay effects that would not be added in your head in the concert hall.

[candycane]

As far as I know (I am no guru on the fundamentals in DSP) Audiolense and Acourate are developed with different views on what is optimal DRC.

I'm no expert in this area either, but I thought they tried to do similar things.  What views differ between Audiolense and Acourate?

[Mikkel]

I'm also curious about the differences between PEQ and DRC. Using REW and the PEQ in JRiver, I can get an almost perfectly flat frequency response. Below is a pair of JTR Captivators that I EQ'd. It took 3 minutes and 45 seconds to measure with REW, generate filters, enter the filters into JRiver's PEQ, and remeasure. I've also done EQ to some mains and was able to get them almost flat, too. When integrating mains and subs (or any separately amped drivers) you can see by the frequency response whether phase/group delay are an issue. By using the Real Time Analyzer in REW, you can adjust the distance setting and visually see exactly when the drivers are in phase. At the recent subwoofer GTG, the sealed subs needed about 2-3 ft of delay, ported subs needed 5 ft, and the horn subs needed 17 ft.

If both PEQ and DRC produce the same frequency response (within a dB or so), then what is better about DRC?

@ Mojave: That is also my (current) position. But not having tried digital room correction yet, I wouldn't know. Audiolense is somewhat expensive and DRC, although free, is too complicated for me.

Maybe I'll give Audiolense a try one of these days to see if I can notice a difference.

[TheLion]

I'm also curious about the differences between PEQ and DRC. Using REW and the PEQ in JRiver, I can get an almost perfectly flat frequency response. Below is a pair of JTR Captivators that I EQ'd. It took 3 minutes and 45 seconds to measure with REW, generate filters, enter the filters into JRiver's PEQ, and remeasure. I've also done EQ to some mains and was able to get them almost flat, too. When integrating mains and subs (or any separately amped drivers) you can see by the frequency response whether phase/group delay are an issue. By using the Real Time Analyzer in REW, you can adjust the distance setting and visually see exactly when the drivers are in phase. At the recent subwoofer GTG, the sealed subs needed about 2-3 ft of delay, ported subs needed 5 ft, and the horn subs needed 17 ft.

If both PEQ and DRC produce the same frequency response (within a dB or so), then what is better about DRC?

On a related note, why does an orchestra sound so good when there is obviously group delay among all the instruments?

There was a time I thought the same thing - making the frequency response with REW reasonably flat and doing your real time subwoofer integration (finding a proper level, delay setting and placement for multiple subwoofers and integrate them the best you can with all 7 speakers is what I used to do for hours and hours looking at the RTA...)

Two setups with approx. the same frequency response or better yet one setup with two different methods (eg. PEQ and FIR) reaching that goal sound pretty much the same - aren't they? Well, in my experience, they can sound as radically different as a system in two different rooms can sound. In Audiolense you can use freq. correction only or phase+frequency = BIG difference and both have the same measured freq. response. Or compare your PEQ filtered "flat" (it is all in the smoothing ;-) response with a pure freq. correction with Audiolense using the same target curve. Even if they result in pretty much the same measured (and smoothed) freq. response the subjective difference between those two will be striking. Try it for yourself.

BUT - PEQ can sound awfully good. In my opinion as good or better than any automatic FIR/IIR approach IF used by an expert with experience and knowledge (about psycho-acoustics). Hand tuned systems in the right hands have great potential.

[Mikkel]

@TheLion: How extensive a smoothing does e.g. Audiolense apply before filtering? Or does it simply take the whole frequency response-range and invert it unfiltered (with a reasonable limit on overall boost)?

My experience with auto-eq in REW is mixed. In my experience it gives a much better result to use the RTA and then manually apply filters; both in regard to a more smooth cross-over, phasing, and not least sound pressure level between the individual speakers.

It takes a bit of time, though.


Best regards
Mikkel

[Trumpetguy]

One of the most essential steps in the Audiolense filter generation is the filtering of the measured impulse response. There is a number of user settings you can apply, such as number of taps, and number of cycles available for the True Time Domain (group delay) filtering. The result is a smoothed frequency and impulse response, corected for varying distance from speaker to listener. It is this filtered measurement and a user defined target curve that is the basis for the FIR filter generation. The user can also choose between minimum phase and linear phase correction; this is really outside my competence, but this is a fundamental choice and is much debated what is more correct/optimal. Choosing linear phase is argued by Audiolense to be superior, but can lead to audible low frequency pre-ringing if your DRC correction needs are extensive.