[pcasper]

Hello to all,

Preliminary searching has not revealed a solution to my problem...I hope someone here can help.

I use REW regularly to test and tune up my 5.1 system. I equalize the front L+R channels manually with the Audessey eq. function in my Integra DTC 9.8 processor. I inject the sweep from my laptop external soundcard via the analog Auxilliary input on the Integra front panel. This input is digitized, processed, equalized, converted to analog again and passed out to my external L+R amp. However, I am regularly frustrated trying to test the center, surround and sub channels, because there seems to be no way to inject the analog test signal into each channel such that it is processed in the same manner as the front L+R channels. I can use the analog multichannel inputs to drive each channel separately, but this bypasses the internal processing and equalization so all I can do is test but not equalize and retest.

I have gotten around this limitation so far, for instance, by driving the Left front channel, connecting the center speaker to the L channel amp, nulling the left channel equalization and readjusting the equalization for the center channel. Then I have to transfer the equalization settings to the Audessey center channel, restore the left channel settings, reconnect the speakers correctly, and hopefully I am done. The process is repeated for the surround and sub channels. It's a pain, it works, but there has to be an easier way.

Ideally, I envision a laptop-based software program that outputs a 5.1 encoded digital stream, with the test signal injected into the desired channel. It would be wonderful if REW did this.

I have a feeling that there is in fact an easy and fairly obvious way to do this, which I have been too dense to discover. If anyone here can enlighten me, I would be grateful and hopefully not too embarrassed!

I know I could just let Audessey do it all automatically, but the automatic function does not work well in my system, and I prefer to do it manually anyway, especially when I want to create a non-flat response, or I want to hear the effect of an adjustment immediately.


Thanks for any help,

Paul

 

[brucek]

It's a pain, it works, but there has to be an easier way.

Not that I'm aware of.........

brucek

 

[atledreier]

What's the technical reason why REW can't output a 5.1 bitstream?

 

[brucek]

What's the technical reason why REW can't output a 5.1 bitstream?

You would require a DD or DTS encoder licensing agreement, and that ain't cheap. On top of that, and the real show stopper, is that licenses are only available to companies that fulfil the licensing terms and whose products go through the verification and approvals processes. Licenses are not available to individuals.

This is an analog mono signal application for testing response. You have to get creative to do multichannel.......

brucek

 

[atledreier]

How about those that have a HDMI output in their computer that will also output audio. Many laptops have this already, and most newer ones will in the future. Could you output multichannel PCM over HDMI if the hardware allow it? raying:

 

[brucek]

And how would you calibrate the analog line-in channel that accepts the analog SPL meter?

brucek

 

[pcasper]

Well, at least I did not miss something obvious! Thank you for your reply. I will just keep doing it the hard way.

On a side note, is anyone aware of a test cd or dvd that generates sustained full-band pink noise in each channel? The Essentials and Avia disks don't do it as far as I know, except maybe for the front L+R channels. You could just use the spectrum analysis feature of REW with this disk, but you would have to negate the soundcard calibration unless you could figure out a way to calibrate only the receive side.

Anyway, I want to take this opportunity to thank brucek and anyone else that has contributed to the development of REW. This tool is absolutely fantastic and I just marvel at it's capabilities every time I use it. I am a grizzled old retired engineer and I remember when we had to do everything on a slide rule, so the technology available today boggles my mind. You guys deserve a special place in the audio hall of fame.

Paul

 

[thewire]

Well, at least I did not miss something obvious! Thank you for your reply. I will just keep doing it the hard way.

On a side note, is anyone aware of a test cd or dvd that generates sustained full-band pink noise in each channel? The Essentials and Avia disks don't do it as far as I know, except maybe for the front L+R channels. You could just use the spectrum analysis feature of REW with this disk, but you would have to negate the soundcard calibration unless you could figure out a way to calibrate only the receive side.

Anyway, I want to take this opportunity to thank brucek and anyone else that has contributed to the development of REW. This tool is absolutely fantastic and I just marvel at it's capabilities every time I use it. I am a grizzled old retired engineer and I remember when we had to do everything on a slide rule, so the technology available today boggles my mind. You guys deserve a special place in the audio hall of fame.

Paul

I used the blu sky test setup CD. After two days of calibration of volatages and levels I get this at the location next to ears.

I used Pink Noise fb to calibrate everything using the RTA in REW. After calibrating voltages, then changing them based on the difference between toslink (digital) and analog that I measured in my pre/pro, and based on the fact that my last AVR was damaged by full range analog input from my source.

This gave me a -10 setting on my analog source and 0 for the sub. I placed the mic at arms length between my left and right speakers and measured each channel at a time.

I played the full bandwidth pink noise for each channel without the sub and set them to 85dBc, ignoring that my RS Meter wanted to measure higher due to the strange looking peak around 3Khz. This lined it up above 70 in the RTA. I then calibrated the sub to +10dB higher in 1/3 octave using the RTA, which places it around 4dB higher on the dBFS measurement of the input that is near the RTA check box. The RS meter is not the best tool for such calibration by the way and you should be using a full range mic. I placed my surrounds at 82dBFS becuase I wanted to try it with Dolby Specs.

After all this I go back and measure with my RS meter each channel, this time turning on the sub output in my XA2 because the THX optimizer sub noise will not measure the crossover sub out in my pre/pro using it's internal crossover and full range signal from the XA2, and I measure 74dBc - 75dBc for the sub channel, and 74dBc -75dBc for the other channels. My pre/pro was at that time at -10.

Here are the test files I used.

http://www.abluesky.com/asp/catalogue/catalogue.asp?linkid=143

 

[au-734]

And how would you calibrate the analog line-in channel that accepts the analog SPL meter?

brucek

Wouldn't an analog pre-out on the rear of the AVR work assuming all signal processing in the AVR is off?

My Denon has 7.1 in but it does not go through an A/D/A loop so it is not easy to see how Audyssey adjusts the frequency response for the side and rear speakers. It is easy enough to work on the front left and right using stereo inputs, the side and rears are much more complicated.

Thanks for creating a great tool. Was this someone's graduate thesis project?

Cheers,

jr

 

[brucek]

Wouldn't an analog pre-out on the rear of the AVR work assuming all signal processing in the AVR is off?

The problem with using digital-out and analog-in to a soundcard is that the analog input requires its own calibration, which simply isn't possible. It's like one hand clapping - it doesn't work.

Either way, I don't think John is considering HDMI 5.1 PCM bitstream output for REW. The concentration is on analog I/O to test subwoofers and mains.....

brucek

 

[atledreier]

In my experience, soundcard calibration usually come very linear, with a slight drop in the extremes. My current soundcard calibration is down 1dB at 5Hz, and brickwalls at 20-something kilohertz. Plenty linear enough to be skipped, in my book. I get much more variation moving the microphone an inch.

 

[thewire]

Use analog instead of HDMI and it will work. It is just like being in the cinema. Use the voltage to calibrate voltage. It's that simple.

 

[JohnM]

On a side note, is anyone aware of a test cd or dvd that generates sustained full-band pink noise in each channel? The Essentials and Avia disks don't do it as far as I know, except maybe for the front L+R channels. You could just use the spectrum analysis feature of REW with this disk, but you would have to negate the soundcard calibration unless you could figure out a way to calibrate only the receive side.

For L+R you can use REW to do that, open the signal generator, select "Pink PN" as the signal type, set the length the same as the REW RTA FFT length and click the "Save PN to WAV File" button. That will generate a periodic pink noise sequence of 1 minute duration that you can burn to a CD or DVD. Periodic pink noise has the advantage that it contains all frequencies in a single sequence of the selected length, which means you get stable readings with far less averaging than using pink-weighted random noise. Make sure the RTA window is set to Rectangular.

Anyway, I want to take this opportunity to thank brucek and anyone else that has contributed to the development of REW. This tool is absolutely fantastic and I just marvel at it's capabilities every time I use it. I am a grizzled old retired engineer and I remember when we had to do everything on a slide rule, so the technology available today boggles my mind. You guys deserve a special place in the audio hall of fame.

You are very welcome

 

[T-Bone]

I'll piggy back on this thread since I have the same 5.1 (7.1 really) question. GREAT tool by the way. I have a separate amp, and an Onkyo SR702 receiver (a pre/pro in my system) that has 7.1 analog inputs. I've previously leveled my speakers using optical out and Avia... and 80 on my Onkyo dial equated to 75 SPL at the the listening position.

So, I've calibrated my sound card and SPL meter per the REW Help. For calibration, I plugged the analog out of my sound card into one of the Onkyo's L+R analog inputs (not the 7.1 inputs). As an FYI, 80 on my Onkyo equated to 75 SPL at the listening position.

So now I am wondering... since everything on my HTPC is calibrated and 80 on the Onkyo dial equates to 75 SPL at the listening position whether I use Avia or REW, I should be able to test 7.1 channels? I am thinking I can plug the sound card output into, say, the Onkyo 7.1 center channel input, run sweeps, etc. Then repeat for the remaining 7.1 inputs. Once I have the sweeps, I can tweak the Onkyo Equalizer manually to make corrections.

Am I missing something?

-T

 

[brucek]

Once I have the sweeps, I can tweak the Onkyo Equalizer manually to make corrections.

Am I missing something?

It would be fine if each speaker never played the same frequency as another speaker. You aren't considering the effect of interaction/cancellation between speakers....

brucek

 

[T-Bone]

It would be fine if each speaker never played the same frequency as another speaker. You aren't considering the effect of interaction/cancellation between speakers....

brucek

Yeah, each speaker will be tested/equalized in isolation, meaning no interaction with the other speakers. Thanks brucek.

-T

 

[audunth]

How about a feature that lets you select output channel also, not only input channel? That way you could at least do measurements of all speakers without having to fiddle around hooking up each channel separately, if you have the analog 7.1 outputs hooked up to the 7.1 inputs on the receiver. Also you could check EQ if you do the EQ on the computer with DRC/Convolver or similar. (I think it's possible to route the output of REW through Convolver using Virtual Audio Cable or something similar.)

Regards,
Audun

 

[kjgarrison]

I've been trying to figure out a way to do the same thing. This isn't exactly what we all want, but it's somewhat getting there.

My Denon 4308 has a surround setting which takes stereo input and outputs it to all of my (5.1) speakers. The Audyssey MultiEQ XT remains "ON". It is therefore possible to do the sub alone, the mains alone, mains plus sub (2.1), 5.0 and 5.1.

The surrounds alone isn't doable, but the effect of adding the surrounds to 2.1 and 2.0 should give some useful information on how well EQ'd they are, not to mention seeing how the overall EQ integrates everything.

At least this is what I'm puzzling over ...


Not that I've actually DONE it :dontknow:

 

[THX-UltraII]

Most processors have a MULTI CHANNEL STEREO mode. I think this will work for testing all signals. Just unplug the speakers you don t want to measure and only plug in eg. your center channel and run a REW measurement.

 

[loach71]

I used a Dolby Digital 5.1 / dts 5.1 encoder package to migrate a mono pink noise signal to the requisite format. :coocooUnfortunately my 64 year old brain can't remember its name!):nono:. I set the level at -20 dBm and recorded 20 minutes of signal. Following that I burned the signals to a DVD. It serves as my calibration source.