Print

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

[hulkss]

Title: Hear music the way it was intended to be reproduced.

Five articles plus conclusions. Lots of measurements and data, not a bunch of audiophile BS. A good read.

Read it here:http://www.computeraudiophile.com/blogs/mitchco?page=1

I'm going to give it a try next week.

[candycane]

Very interesting.  Please report back on your experience and lessons learned!  I read the blog and would be interested in your views on:

- A $600+ sound card?  really?  is this a showstopper with anything less, or is this overkill?
- The blogger used Audiolense, but what version and what cost?  That company's website refers to a version 4.2 but there is no product cost that references 4.2 on the "shop" page
- How easy is it really to send the Audiolense results to the convolver, and how easy is it really to set up the convolver to work with JRiver?
- Did you use the B&K house curve as your target curve, or something else?  Could you tell the difference readily if you tried flat versus the B&K house curve?
- How many subwoofers did you use?  If one, could it have handled two from what you observed?

Looking forward to your results!

[hulkss]

I will be testing my new HTPC based system this weekend.

A $600+ sound card?  really?  is this a showstopper with anything less, or is this overkill?
I am using a 16 channel external USB DAQ. http://www.lynxstudio.com/nav/getFile.asp?i=65&t=productfile
ASIO is recommended. There are inexpensive solutions. ASIO4ALL can help http://www.asio4all.com/

The blogger used Audiolense, but what version and what cost?
I am using Audiolense XO version 4.2 (current release). Price here: http://www.juicehifi.com/show.asp?page=goxpage00000003dbp3.html

How easy is it really to send the Audiolense results to the convolver, and how easy is it really to set up the convolver to work with JRiver?
Seems easy to me so far. "ConvolverVST" is installed and appears to work but I have not fully tested it yet. http://convolver.sourceforge.net/index.html

Did you use the B&K house curve as your target curve, or something else?  Could you tell the difference readily if you tried flat versus the B&K house curve?
I will use the B&K house curve as a starting point. Roll off of high frequencies is essential when measuring at the listening position. A proper house curve is a HUGE factor. Read this: https://docs.google.com/viewer?a=v&pid=explorer&chrome=true&srcid=0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJiODQ1ZTUxMGQ4&hl=en_US

How many subwoofers did you use?
I have:
3 subwoofers with four 15" drivers in each
2 side surround speakers
2 back surround speakers
Front main speakers (LCR) are each tri-amplified.
That totals to the 16 channels of DAC that I am using to feed my power amps.

[Trumpetguy]

Interesting blog. I have the (almost) identical setup, and will read it with great interest. I know candycane asks questions to hulkss, but allow me to post some thoughts about my experiences.

- A $600+ sound card?  really?  is this a showstopper with anything less, or is this overkill?

Lynx produces really stable and low latency hardware and software, and the DACs are superb and of studio quality. In my setup I even upgraded to Lynx Aurora8. The main advantage for me was an improved clock and toroidal power supply (not noisy switch mode). Of particluar interest is the low latency between inputs and outputs when doing room response measurements, as this is a requirement for doing high quality frequency sweeps. I had less luck with that with my old m-audio.

You would need a calibrated microphone and mic pre-amp. Never use e.g. a non-calibrated emc8000 behringer and download some "calibration curve". Calibration is individual, hence the term....

- The blogger used Audiolense, but what version and what cost?  That company's website refers to a version 4.2 but there is no product cost that references 4.2 on the "shop" page

You do not buy a particular version x.y, but rather a 1 year upgrade license (choose amongst XO, 2.0 etc).

- How easy is it really to send the Audiolense results to the convolver, and how easy is it really to set up the convolver to work with JRiver?

Simple! Audiolense exports the wav filter files and the Convolver text setup files based on the speaker setup you configure prior to measurement. You would need ConvolverVST in DSP studio (not wrapper or the one named only convolver).
I have some issues regarding ConvolverVST forgetting the filter file, and keeps not changing filter file when I switch zones in JRiver (this may be because I do not know how to use ConvolverVST properly).

- Did you use the B&K house curve as your target curve, or something else?  Could you tell the difference readily if you tried flat versus the B&K house curve?

I know this question was directed to hulkss, but anyway: In Audiolense, the target curve is what you can spend infinite time and efforts on. You can even ask Mr Audiolense himself for help, but you will end up fiddling with it. My taste is a 1 to 2dB increase in the bass for music and 6 to 8dB increase for movies.

- How many subwoofers did you use?  If one, could it have handled two from what you observed?

You can have multiple subs. It all comes down to the number of output channels on your analog out. I use two subs on one channel using y-split.

[candycane]

Thanks to both of you for your thoughts - all are welcome!

You both have given this a great deal of effort, and thanks for sharing.  If you may indulge me further, a question about the target curve.  I'm curious if the preference of the B&K house curve and the Harman Kardon preferred curve, which seem similar in that they boost low freqencies and attenuate high frequencies, is related to or a function of assumptions the recording engineer made when mastering a recording.  For example, i think I've heard that many recording engineers assume that peoples' minimally treated or untreated rooms will reflect bass and absorb treble, and thus attenuate bass and boost treble in their recording to try for a flatter playback response at home.  If that's the case, then the B&K and Harmon Kardon curves make sense to me because such curves would actually result in a flat playback reponse of music.   I'm trying to make sense of why people would prefer rolled off high frequencies instead of flat, which seems counterintuitive, but could make sense if we differentiate between test signals versus recorded music and how the music was mastered.

If the above makes sense, then I would expect that there will be some small percentage of music that doesn't sound very good when played back in an environment corrected to a curve similar to the B&K house curve, in cases where the recording engineer assumed no room effects upon playback, and recorded in a heavily treated and neutral studio.

If none of that makes sense or is not what is driving the preference, any enlightenment is appreciated.

[hulkss]

A microphone connected to a computer in the listening position does not "hear" like an ear or process audio like a brain. If you calibrate a speaker in a laboratory anechoic chamber you want flat frequency response. A room is reflective to sound and speaker directivity and power response in addition to frequency response become major factors that must be taken into account.

Info here: http://www.hometheatershack.com/forums/rew-forum/96-house-curve-what-why-you-need-how-do.html#post448

There are also may ways to process the audio measurement, usually using the FFT: http://www.rationalacoustics.com/files/FFT_Fundamentals.pdf
This is an attempt to measure like we hear.

[BradC]

candycane,

you are correct, a speaker that has a flat frequency response will have rolled off highs at the listening position.

In a room not all frequencies are attenuated equally.