V5.20.14 early access build

[Johan Gustavsson]

December 2018. Probably better you ask them as a customer.

i have been in contact with Storm and they will take a look and comeback to you

 

[Breeman]

Hi John,

Can you please consider adding an option to toggle between log and a linear scale for the horizontal axis of the SPL & Phase and All SPL tabs? Thanks.

 

[John Mulcahy]

Hi John,

Can you please consider adding an option to toggle between log and a linear scale for the horizontal axis of the SPL & Phase and All SPL tabs? Thanks.

The frequency axis button does that.

 

[Breeman]

The frequency axis button does that.

Oh my! Don't know how I missed that, thanks.

 

[onlyoneme]

Hi @John Mulcahy !
I'm trying to make a stepped measurement on the external device, so I capture the output when executing the stepped sine and later I replay it on the external device synchronously with the measurement run in REW. For some reason I cannot get the level based distortion panel in REW when the measurement ends although it looks good in RTA window. Only when I stop it manually it brings me the correct panel. What's the reason?

 

[John Mulcahy]

I'm trying to make a stepped measurement on the external device, so I capture the output when executing the stepped sine and later I replay it on the external device synchronously with the measurement run in REW. For some reason I cannot get the level based distortion panel in REW when the measurement ends although it looks good in RTA window. Only when I stop it manually it brings me the correct panel. What's the reason?

No idea, sorry. Stepped sine measurements need tight synchronisation between replay and capture to ensure the results are correct and there are some stability checks that run during measurement to work out when it is safe to move to the next point, so timing can vary run to run depending on the behaviour of the device. There isn't an offline option for that due to the timing complications.

 

[onlyoneme]

No idea, sorry. Stepped sine measurements need tight synchronisation between replay and capture to ensure the results are correct and there are some stability checks that run during measurement to work out when it is safe to move to the next point, so timing can vary run to run depending on the behaviour of the device. There isn't an offline option for that due to the timing complications.

I can imagine that the timing is critical here, but why it works when I stop it manually just one step before the end, and it does not work when it goes to the end itself?
Maybe there is a way to relax a kind of restriction here to let it work in such case?

 

[John Mulcahy]

why it works when I stop it manually just one step before the end, and it does not work when it goes to the end itself?

As I said above, I don't know and I think it is risky to try an offline measurement.

 

[onlyoneme]

As I said above, I don't know and I think it is risky to try an offline measurement.

I understand what you've said and I can see your point, but such situation can happen also for online measurements. Plots in RTA windows seems to look fine but no correct distortion panel. No warning, so I've no idea what should I change to get proper results. When I stop the process manually I can see the level related panel.
I'm testing it with flexASIO and 2 separate devices, DAC and ADC.

 

[phofman]

@onlyoneme : can you check wasapi exclusive instead of ASIO? ASIO assumes that capture and playback devices run synchronously (the bufferSwitch method passes equally-sized buffers for each direction at one call). Using two independently-clocked devices is basically a hack allowed by FlexASIO/Asio4All, but not designed into the protocol by Steinberg.

In WASAPI capture and playback devices are accessed independently (like in linux alsa)

 

[onlyoneme]

@onlyoneme : can you check wasapi exclusive instead of ASIO? ASIO assumes that capture and playback devices run synchronously (the bufferSwitch method passes equally-sized buffers for each direction at one call). Using two independently-clocked devices is basically a hack allowed by FlexASIO/Asio4All, but not designed into the protocol by Steinberg.

In WASAPI capture and playback devices are accessed independently (like in linux alsa)

I've tried WASAPI also, results are similar to flexASIO here.

 

[sm52]

There is a question about calibration before measuring the impedance. From help:

Short circuit calibration

The second step is to calibrate the measurement rig with the test leads shorted. They should still be connected to the rig, but shorted together at the load end. The measurement compensates for the series impedance of the test leads, it particularly improves high frequency accuracy. If this measurement is not carried out the RLEADS box can be used to compensate for the test lead resistance.

Does the last sentence mean that you can do the first calibration step, then enter the resistance of the test leads without the second step, and then do the third step? I can't do it that way. REW prevents you from taking the third step without the second.

 

[John Mulcahy]

Does the last sentence mean that you can do the first calibration step, then enter the resistance of the test leads without the second step, and then do the third step?

No, it means if you only do the first step you can use the lead resistance entry instead. For the most accurate results go through the three calibration steps.

 

[sm52]

Thanks for the answer. If the measurement circuit uses a tube amplifier and there is no desire to short the test leads, what options can be used?

 

[John Mulcahy]

The amp would not see a load lower than the sense resistor, if it can't drive that it wouldn't be suitable for impedance measurement.

 

[sm52]

Thanks for the answer.

 

[serko70]

Header is still there despite the tick:

* Measurement data measured by REW V5.20.14
* Source: Trace Arithmetic result A * B
* Format: Trace Arithmetic result A * B
* Dated: 27-Jul-2023 17:37:01
* REW Settings:
* C-weighting compensation: Off
* Target level: 114.8 dB
* Note: ; Trace Arithmetic A times B A = A times B B = Filters EQ C
* Measurement: C
* Smoothing: None
* Frequency Step: 1/48 octave
* Start Frequency: 10.000 Hz
*
* Freq(Hz), SPL(dB), Phase(degrees)
10.000000, -0.010, -5.4892
10.150000, -0.010, -5.5762
10.300000, -0.011, -5.6635

 

[serko70]

View attachment 63223

Header is still there despite the tick:

* Measurement data measured by REW V5.20.14
* Source: Trace Arithmetic result A * B
* Format: Trace Arithmetic result A * B
* Dated: 27-Jul-2023 17:37:01
* REW Settings:
* C-weighting compensation: Off
* Target level: 114.8 dB
* Note: ; Trace Arithmetic A times B A = A times B B = Filters EQ C
* Measurement: C
* Smoothing: None
* Frequency Step: 1/48 octave
* Start Frequency: 10.000 Hz
*
* Freq(Hz), SPL(dB), Phase(degrees)
10.000000, -0.010, -5.4892
10.150000, -0.010, -5.5762
10.300000, -0.011, -5.6635

Sorry, I was using an older ea version. ea69 does not have the problem.

 

[gadzooks]

Hi all,

Hope this is the right place for this post. I scanned the thread but did not see any other complaints along these lines.

I am following the Room Correction Guide provided by OCA.

So this may be a bug, or it may be a PEBCAK error (Problem Exists Between Chair and Keyboard). Or hopefully a setting I have screwed up somewhere. I have tried to align my setting with OCAs videos and the guide.

I have wiped out and re-installed REW and the result has not changed.

Version I am using:
REW V5.20.14 EA69
JRE 1.8.0_372 64-bit on Windows 10

When executing the "Trace Arithmetic" step "A/B", I get a blank result or, sometimes (not consistently) I get a drop off curve after 20khz (as seen in the screen shot).

If I use the "|A| / |B|" function it works as I would expect it to producing a response as seen in the screen shot.

Thank you!

(edited for obvious grammar mistakes and removed a mistaken conclusion regarding phase)

 

[serko70]

Hi all,

Hope this is the right place for this post. I scanned the thread but did not see any other complaints along these lines.

I am following the Room Correction Guide provided by OCA.

So this may be a bug, or it may be a PEBCAK error (Problem Exists Between Chair and Keyboard). Or hopefully a setting I have screwed up somewhere. I have tried to align my setting with OCAs videos and the guide.

I have wiped out and re-installed REW and the result has not changed.

Version I am using:
REW V5.20.14 EA69
JRE 1.8.0_372 64-bit on Windows 10

When executing the "Trace Arithmetic" step "A/B", I get a blank result or, sometimes (not consistently) I get a drop off curve after 20khz (as seen in the screen shot).
View attachment 63242

View attachment 63238

If I use the "|A| / |B|" function it works as I would expect it to producing a response as seen in the screen shot.

Thank you!

(edited for obvious grammar mistakes and removed a mistaken conclusion regarding phase)

Is it possible you measured A and B at different sampling rates?

 

[gadzooks]

Is it possible you measured A and B at different sampling rates?

Looks like the same sampling rate throughout.

A - Both Right and Left look like 48k

I then Derived "B" from the RMS average

Resulting in the Target curve "B"

I included the mdat incase it is helpful

 

[serko70]

Looks like the same sampling rate throughout.

A - Both Right and Left look like 48k
View attachment 63245
View attachment 63246
I then Derived "B" from the RMS average
View attachment 63247
Resulting in the Target curve "B"
View attachment 63248

I included the mdat incase it is helpful

mdat is not attached, must be over file size limit.

I can't see anything wrong in the info windows. 4M measurements might be causing some memory issues (wild guess). You'll need to wait for John .

 

[gadzooks]

mdat is not attached, must be over file size limit.

I can't see anything wrong in the info windows. 4M measurements might be causing some memory issues (wild guess). You'll need to wait for John .

Thank you sir!

During install I selected max memory at 3 GB for what it is worth.

 

[jeppekristoffer]

Hi all,

Hope this is the right place for this post. I scanned the thread but did not see any other complaints along these lines.

I am following the Room Correction Guide provided by OCA.

So this may be a bug, or it may be a PEBCAK error (Problem Exists Between Chair and Keyboard). Or hopefully a setting I have screwed up somewhere. I have tried to align my setting with OCAs videos and the guide.

I have wiped out and re-installed REW and the result has not changed.

Version I am using:
REW V5.20.14 EA69
JRE 1.8.0_372 64-bit on Windows 10

When executing the "Trace Arithmetic" step "A/B", I get a blank result or, sometimes (not consistently) I get a drop off curve after 20khz (as seen in the screen shot).
View attachment 63242

View attachment 63238

If I use the "|A| / |B|" function it works as I would expect it to producing a response as seen in the screen shot.

Thank you!

(edited for obvious grammar mistakes and removed a mistaken conclusion regarding phase)

I think that you have to use "Vector average" and not "Average" for all the averages when following OCAs guide. Then your B-average also has phase included and then you can use "trace Arithmetic" step "A/B".

 

[gadzooks]

Tried it with Vector Average and RMS +Phase and same issue. I must have a setting horked, or maybe it is a memory issue as suggested by Serko70.

The first A/B measure is vector, second is rms+phase.