Measuring with a timing reference

[FargateOne]

So...somebody?

 

[John Mulcahy]

HI,
@John Mulcahywhat happens around at 232 Hz (Center+sub) or worst at 112 Hz at 197 Hz with the rear left surround?

Contributions from room resonances and reflections.

 

[sm52]

John, when I want to see the minimum phase on the graph after the measurement, I press 'generate'. The minimum phase graph appears. But it's on a much larger scale. To make the scale readable, you need to press 'generate' again. Can this process be done with one click?

 

[John Mulcahy]

John, when I want to see the minimum phase on the graph after the measurement, I press 'generate'. The minimum phase graph appears. But it's on a much larger scale. To make the scale readable, you need to press 'generate' again. Can this process be done with one click?

Sorry, I don't know what you mean or which graph you are referring to.

 

[sm52]

Impulse Response.

 

[John Mulcahy]

That image is the SPL & Phase graph, not the impulse response graph. Generating minimum phase doesn't affect the graph scales, if you want the phase to cover a different span on the graph use the zoom buttons to zoom out on the phase axis or use the graph limits button to set the phase range you want the graph to show.

 

[sm52]

Yes. I was wrong in the title of the graph. SPL & Phase. I said that when REW drew this graph, there is no minimum phase by default. To obtain a graph of the minimum phase, press the button shown with a red arrow. If you click on it once, the graph of the minimum phase is generated in an enlarged scale. To get it on a normal scale, you need to press the same button again. Those. you need to press this button twice. Why not generate and display at normal scale right away with one click?

 

[John Mulcahy]

You must be doing something special. Pressing the button does not change the graph scales for me, however many times it is pressed. Nor should it. The graph scales are set by the controls I explained above.

 

[sm52]

Watch the video.

 

[John Mulcahy]

Shame you didn't start with that, or some screenshots. The issue was caused by not having any phase traces displayed, and hence no right hand phase axis displayed, before generating the minimum phase response. I have fixed it for the next beta.

 

[sm52]

John, I checked how the min phase is displayed after the update. Everything is as it should be. Thank you.
What is the difference between signal level and noise level for a measurement to be considered good? 40dB, 50dB, or more? For example, can such a tweeter graph be considered normal?

 

[John Mulcahy]

The noise floor on that measurement is almost 74 dB down, which is very good. Above 40 dB is generally fine.

 

[sm52]

I don't see -74dB before the noise floor Even 40 dB is difficult to achieve.

 

[John Mulcahy]

I was referring to the impulse response.

 

[sm52]

Is the impulse response noise level not equal to the acoustic noise level in the room?

 

[John Mulcahy]

No, not at all. Sweep measurement can see below the noise floor because of the nature of the sweep processing. The impulse response shows the signal-to-noise ratio achieved for the measurement.

 

[sm52]

Or is the 74 dB difference between the noise level and the signal level obtained because the noise level is considered the noise level in the feedback loop, and the signal level is the acoustic level of the signal in the room?

 

[John Mulcahy]

The question doesn't make sense to me, sorry. For more information on measurement with sweeps and how that compares to a variety of other methods of measuring the transfer function see http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.118.7473

 

[sm52]

Actually my question was about the noise level in the measurement room. In my measurement, this is only 39 dB below the signal level. Is this not a bad factor? Is it necessary to increase the signal level by at least 6-10 dB?

 

[John Mulcahy]

No, the measurement is already very good

 

[sm52]

There is a Length parameter in the measurement window. I tried to take measurements on all variants. The charts are 98% the same. Distortion and waterfall graphs change. Maybe John knows what the pros and cons of increasing-decreasing this parameter are.

 

[John Mulcahy]

From the help:

Each doubling of the sweep length improves S/N by almost 3 dB. However, the time required to perform the processing after each sweep will more than double. If REW is running on a computer which does not have a fast processor and a lot of memory, measurements will be much faster using the shortest sweep length (128k samples), at a small S/N penalty of about 3dB compared to the default. At least 4 GB of RAM and a fast processor are recommended if using the 1M sweep, at least 8 GB for the 4M sweep. Invalid measurements or out-of-memory errors may occur on computers which have insufficient RAM or processor speed.

 

[sm52]

In order for the computer to process the data without errors at Length = 4M, 8 GB of RAM and a fast processor are needed. Let's say this condition is met. But why are the frequency response graphs for Length from 128K to 4M the same? When I set this parameter to 4M, I expect a more detailed graph than when this parameter = 128K.

 

[John Mulcahy]

What do you mean by more detailed? A longer sweep lowers the noise floor, it shouldn't change the measured response which should already be far above the noise floor.

 

[sm52]

A longer sweep lowers the noise floor,

is this the only thing that distinguishes Length = 256K from Length = 4M?

What do you mean by more detailed?

If we have as a result of measurement 5 values of the measured parameter, as a result of interpolation we will get not a smooth curve, but a broken line. And if we have 100 values, the graph will be more consistent with reality.