Can't get this to work !

[Vaughan Odendaal]

Hi there,

I apologize if this is not appropriate but I am having a lot of troubles getting ETF (not room eq wizard) to work properly.

I have a Genuis value 5.1 card that I bought a few days ago and I bought the necessary cables. But when I run the measurments to test the sound card, there is a lot of noise.

The mixer settings are okay. And if I change the sampling rate from 44khz to 48khz, there is very little change. To make matters worse, the levels between left and right channels change quite a bit if I wiggle the cables a bit.

I have basically connected the left input to the left output and the right input to " " " just to test the sound card out. I have spoken to Doug and he is also a little unsure of what to do.

If I check the signal to noise ratio on the graph, it reads at best, 2 dB. Doug tells me that the reading should be around 20 dB.

Help !

--Sincerely,

[Danny]

Vaughan Odendaal wrote:

Hi there,

I apologize if this is not appropriate but I am having a lot of troubles getting ETF (not room eq wizard) to work properly.

If you have a question please ask, hopefully we'll be able to help you

The mixer settings are okay. And if I change the sampling rate from 44khz to 48khz, there is very little change. To make matters worse, the levels between left and right channels change quite a bit if I wiggle the cables a bit.

Depending on the speakers in your system the change from 44khz and 48khz may be hard to hear. RE the cables it sounds to me like you have a dodgy connection somewhere, possibly in the cables but it also could be elsewhere (sound card, source etc) maybe try replacing the cables then go from there.

Hope this helps

D.

[JohnM]

Have you selected the Genius card to be your default device for sound playback and sound recording? (Audio tab of the Sounds and Audio Devices properties in the Control Panel). Which output are you using on the card? Although you say the mixer settings are OK, that is where many problems occur especially with multi-channel cards - does the Genius card have its own mixer? Maybe worth posting screenshots of the settings. What sort of cables are you using the make the loopback connection?

[Vaughan Odendaal]

Alright. Thanks for asking these questions. Let me be more specific. I am just the front out and line in connections to do a loop thru. I take a duel RCA male to male and plug into a duel female to 3.5 mm computer jack.

I don't know if you know of my sound card, but it was a low cost, cheap sound card and Doug from ETF told me that it should work. A number of other people I corresponded with said the same thing.

It is a full duplex card. And to answer your questions, yes, my sound card does have it's own seperate mixer. I'm not sure what that must be set to but I am also using the windows mixer. That is the main mixer that I am currently using.

What worries me is that the levels between right and left channels jump by about 4-5 dB with just a little wiggle from the cable. Now I've replaced these cables with much longer duel RCA's and the problem remains the same.

Perhaps it is the duel female to 3.5mm computer jack which is somehow corrupt ? I'm not sure. But Doug from ETF told me that he has never seen the kind of problems that I'm experiencing in his entire life.

That isn't good. I will post some measurements. Again, I apologize because I know that this forum is for room eq wizard. Does anyone have experience with this acoustic program ?

BTW, I also downloaded room eq wizard too. I will post some measurements to show you what I'm talking about. Thanks for wanting to help.

--Sincerely,

[brucek]

I don't know if you know of my sound card, but it was a low cost, cheap sound card and Doug from ETF told me that it should work

Hard for us to know if you don't tell us what the sound card is?

What worries me is that the levels between right and left channels jump by about 4-5 dB with just a little wiggle from the cable.

Does your soundcard have the standard line-in and line-out stereo 1/8" jacks on it that either of the (1/8" stereo to RCA adapters) shown below can plug into?




If so, you would be better to use the wire type shown instead of the heavier adapters since they place much less strain on the soundcard connector. The connector on the soundcard is pitifully inadequate and prone to poor connectivity. The less strain you put on it the better.

As I remember, ETF use to require the loopback cable hooked up all the time during measurement (unlike REW that uses it once to obtain a compensation calibration file). Is it still this way?

BTW, I also downloaded room eq wizard too

It's a lot easier to use in my opinion.

brucek

[Vaughan Odendaal]

Thank for replying !

Hard for us to know if you don't tell us what the sound card is?

Sorry, it's called the "Genius Sound Maker Value 5.1".

Does your soundcard have the standard line-in and line-out stereo 1/8" jacks on it that either of the (1/8" stereo to RCA adapters) shown below can plug into?

Yes it does.

If so, you would be better to use the wire type shown instead of the heavier adapters since they place much less strain on the soundcard connector.

I have both of those adapters. The bottom adapter shown in the picture you posted I couldn't use to connect the duel RCA's because both my line in and front out are next to each other.

So it is impossible for me to use those connectors because of the width of the connector. So instead, a computer guy recommended that I use cables instead. So he made up a duel female to computer jack.

The connector on the soundcard is pitifully inadequate and prone to poor connectivity. The less strain you put on it the better.

Thanks for the advice. I have tried the duel female to computer jacks and I've also tried using duel RCA male to male cables (6 m long) to do a loop thru. Both cables give the same result.

As I remember, ETF use to require the loopback cable hooked up all the time during measurement (unlike REW that uses it once to obtain a compensation calibration file).

It is only required to test the sound card. Once sound card has been tested the loop thru is no longer needed.

It's a lot easier to use in my opinion.

Any advantages to ETF ? Can it measure to a better or similar resolution ? How do I do a sound card measurement with REW ? Is it the same way with ETF ? I'm not sure.

I conversed with Doug and he now thinks that my sound card could be at fault. According to him, "It just may be that this particular card will not work because it has separate oscillators for input & output. They may be using an old chipset to
keep the price low".

I really do not want to buy another card. I just bought this one a few days ago.

Thanks.

--Sincerely,

[JohnM]

Vaughan Odendaal wrote:

So instead, a computer guy recommended that I use cables instead. So he made up a duel female to computer jack.

Worth checking the workmanship, might have a bad joint in that cable.

Vaughan Odendaal wrote:

As I remember, ETF use to require the loopback cable hooked up all the time during measurement (unlike REW that uses it once to obtain a compensation calibration file).

It is only required to test the sound card. Once sound card has been tested the loop thru is no longer needed.

The left channel must always remain connected as a loopback with ETF.

Vaughan Odendaal wrote:

Any advantages to ETF ? Can it measure to a better or similar resolution ? How do I do a sound card measurement with REW ? Is it the same way with ETF ? I'm not sure.

REW measures to higher potential resolution, in both cases the resolution is dependent on the gate times you use - the longer the gate time, the higher the frequency resolution (simply reciprocal, 100ms gate time means 10Hz resolution, 1000ms gate means 1Hz etc. REW displays this figure for you). In most situations the measurement signal-to-noise cannot be made high enough for gate times much above 1 or 2 seconds. REW makes prettier waterfalls and makes it easier to read values from the graphs. ETF calculates many more acoustic parameters than REW. The help files explain how to make soundcard measurements etc, you can read them online here.

[Vaughan Odendaal]

Are you the man behind REW ?

--Sincerely ?

[Vaughan Odendaal]

I don't understand what you mean by higher "potential" resolution. I have been told that ETF can resolve information down to 0.7 hz.

I don't know what kind of resolution REW resolves to. The one thing I don't like about ETF is that the graphs like a bit ugly. On the other hand, the graphs and colours on REW and RTA look much, much better.

--Sincerely,

[Vaughan Odendaal]

Worth checking the workmanship, might have a bad joint in that cable.

How does one know if the cables are "okay" ? Thanks.

--Sincerely,

[brucek]

How does one know if the cables are "okay" ? Thanks

As long as they are not molded type connectors, you can simply do a visual inspection once you unscrew the body from the plug to expose the wires. Ensure no shorts and that the connections are well soldered. The tougher connection is with the 1/8" stereo plug. They're small and friggy to solder and easy to make a mistake.

You of course can simply check them with a meter if you have one also.

brucek

[Vaughan Odendaal]

Do you mean a multimeter ? How does one check these things out ? Sorry, I'm not that experienced when it comes to checking cable consistency.

Thanks. Because I would like to try and find out what could be the cause of the bad noise in my graphs. There are a few things :

Either it's the sound card, the cable that connects to the computer jack, the mixer settings haven't somehow been set up properly, or perhaps my own sound card mixer (the mixer that comes with the card) needs to be disabled ?

Currently it's enabled.

How does one rule out these things ? Trying to be all Sherlock Holmes. Heh.

--Sincerely,

[JohnM]

Vaughan Odendaal wrote:

Are you the man behind REW ?

Yes, that's me.

Vaughan Odendaal wrote:

I don't understand what you mean by higher "potential" resolution. I have been told that ETF can resolve information down to 0.7 hz

The internal processing resolution of REW is 0.366Hz when using a 48k sample rate. This is determined by the FFT lengths used when generating the frequency responses, 128K in REW's case. However, the true resolution of a frequency response plot is determined by the duration of the section of the impulse response analysed to produce it. For 0.366Hz resolution you need to analyse a 2.73s period of the impulse response (or 1.4s for 0.7Hz). If you only analyse a 100ms portion of the impulse response, the resulting frequency response has only 10Hz resolution, no matter how large an FFT you apply to the (padded) data. To put this another way, you would not be able to see individual peaks that were less than 10Hz apart, they would be merged together. To have more time resolution you need to analyse a longer time period of the impulse response.

If you look at impulse responses, they are initially large and then decay away until they reach the noise floor of your measurement. Once the impulse response has fallen below the noise floor you can no longer use it to provide information about the frequency response, including data which is only noise degrades your frequency response result. To get the best frequency response, the impulse response is "windowed" to select the portion from the start of the response until it decays below the noise - this window is also referred to as a gate, and its length is referred to as the gate time. If the noise level is high, it may only be a few hundred milliseconds until the response is lost in the noise, so that measurement can never offer frequency resolution better than a few Hz. If the noise level is very low the response will be discernable for much longer and so you can use more of it and get higher frequency resolution.

The noise level in the impulse response can be reduced by using a higher sound level for the measurement sweep, ensuring the background (acoustic) noise is low, using pre-amplifers and soundcards with high signal-to-noise ratios, using longer sweep times and carrying out synchronous pre-averaging of multiple measurements (those last two will be available in the next release of REW, coming soon )

[Vaughan Odendaal]

John, I'm glad I'm speaking to the guy behind the software. It helps matters greatly.

So with REW, I need to check sound card FR using my SPL meter ? Is there any other way of testing REW just for sound card problems without using an SPL meter ?

I have one but I don't think I have the necessary cables to do the test. I have two RCA male to male cables (to connect left and right front out and line in) and two female to computer jack cables that allow me to rig up the RCA's.

John, I need to ask you a question ; what is the cost of REW ? I downloaded the demo version I assume. Concerning the steps that I need to take to take the quality of the sound card out of the equation, what would I need to do ?

Thanks for your advice.

--Sincerely,

[brucek]

Do you mean a multimeter ? How does one check these things out ? Sorry, I'm not that experienced when it comes to checking cable consistency

Yes, a multimeter on the lowest resistance setting, checking for continuity and opens and shorts. Test that the sleeve of each RCA connects to the sleeve of the 1/8" stereo plug and isn't shorted to the center pin of either RCA plug or either the tip or ring of the 1/8" stereo plug. Check that the tip of the 1/8" stereo plug connects to the center pin of one RCA plug and no other and check that the ring of the of the 1/8" stereo plug connects to the center pin of the other RCA plug and no other......

So with REW, I need to check sound card FR using my SPL meter ?

No, you check the soundcard response by shorting one channel of the soundcards output to its input and saving a calibration file for the soundcard (after a special test) to be used every time you use REW. See the HELP FILES on calibrating the soundcard. In fact, read the entire HELP file and all your questions will be answered...

I need to ask you a question ; what is the cost of REW ? I downloaded the demo version I assume

The total cost is zero. You downloaded the full version.

brucek

[Sonnie]

brucek wrote:

The total cost is zero. You downloaded the full version.

brucek

Now why did you go and tell him it's free. Let me handle this please...


Vaughan... how much are you willing to pay? I'm sure we can work out an affordable payment schedule.

[RollsRoyce]

It's also possible that the receptacles on the sound card have lots of play and aren't making firm contact with the plug when it's inserted. Depending on the construction of the receptacles-some are open and some are enclosed-you may be able to carefully bend the side contacts in a little with a small screwdriver. Just enough to firmly grip the plug. It's easier to do this with the card removed, which will also make sure you don't short anything out. Look for cold solder joints where the receptacles are attached to the sound card. Good solder joints look smooth and shiny, while cold joints are usually dull and/or dark. Sometimes there will be a tiny ring in the solder around the pin.

[Vaughan Odendaal]

Can REW measure modal ringing too ?

I just wanted to ask you, or John something. How would I be able to rule out the sound card as the culprit for the bad measurements ? Because I may buy another card. The salesman that sold me the card told me that I could return it, so I might get the Sound Blaster Live Value 5.1 card.

But that card is 2.5 times the price. Please give me some steps to check so that I know for certain that my card is not up for the task. Thank you very much !

--Sincerely,

[brucek]

Can REW measure modal ringing too ?

REW calculates both Spectral decay and Waterfall Plots. Waterfall plots reveal modal ringing. Please see the HELP FILES.