REW Beta Release REW V5.20 beta 30 to 34

[John Mulcahy]

By the way, seems that closing the distortion window during a sweep causes REW to stop producing any new data points, even though the sweep continues to run.

Doing that stops distortion calculation. I'll make a change to ignore the state of that button during stepped sine measurement.

 

[jtalden]

John,
A new feature that would be helpful is to allow setting changes in both the waterfall and spectrogram optionally apply to all measurements. Now when I have several measurement and want to evaluate the impact on each one with different settings, I need to make the all the same field changes in each measurement individually. It takes a lot of time and it's easy to get it wrong since there are several field changes for each measurement. I also think that it would be better not to update/regenerate on each field change. Maybe somehow indicate/remind the user to manually regenerate. I suppose one way is to clear or shade out the trace as soon as one field is changed.

 

[John Mulcahy]

John,
A new feature that would be helpful is to allow setting changes in both the waterfall and spectrogram optionally apply to all measurements. Now when I have several measurement and want to evaluate the impact on each one with different settings, I need to make the all the same field changes in each measurement individually. It takes a lot of time and it's easy to get it wrong since there are several field changes for each measurement. I also think that it would be better not to update/regenerate on each field change. Maybe somehow indicate/remind the user to manually regenerate. I suppose one way is to clear or shade out the trace as soon as one field is changed.

I have added "Auto generate" check boxes on the waterfall and spectrogram controls, if the box is not selected the Generate button will need to be pressed to apply settings changes.

Rather than a way to applying settings to all measurements I will look at implementing a right click menu to copy and paste settings from one measurement's spectrogram or waterfall to another, that seems a bit more flexible and saves adding another button to the controls.

 

[jtalden]

That's great, Thanks!

 

[Paul Kane]

Thanks Paul. What sample rate were you running at? Looks like the error may have been related to showing harmonic phases, was that selected from the start of the run?

Hi John,

Beta 33 fixed the sine sweep crash with the distortion harmonic phases -- thank you! I'm seeing something odd in the frequency stepped sweep. When I run the generator and measure distortions at the same settings as the frequency stepped sine sweep, the results for THD+N below about 300Hz don't seem to match, with the sweep showing significantly higher distortions.

Here are the RTA window distortions at 28Hz generator frequency (0-24kHz):

And here are the same settings used for a frequency sweep. Notice the large difference in H2 and THD+N, especially @ 28Hz. THD+N is nearly 30dB higher! Anything I should be doing differently?

Stepped sine sweep settings:

 

[John Mulcahy]

Can you attach the mdat? The difference looks to be in the noise trace, which isn't in the plot you attached. What looked different about the spectrum when making the stepped sine test?

 

[Paul Kane]

Can you attach the mdat? The difference looks to be in the noise trace, which isn't in the plot you attached. What looked different about the spectrum when making the stepped sine test?

Looked the same, but the THD+N result was consistently a lot worse (below 300Hz). Noise trace is the reason why, except that N(oise) on the RTA screen is measured at -108dBFS and not at -76dBFS as it is in the sweep?

Here's the mdat file (sorry, a new recording as I didn't save the original one, this shows the same THD+N difference).

 

[John Mulcahy]

Please try a run with the silence interval set to zero, and another with the silence interval back at 1 second but the distortion high pass set to 20 Hz.

 

[Paul Kane]

Please try a run with the silence interval set to zero, and another with the silence interval back at 1 second but the distortion high pass set to 20 Hz.

Better! At 0 silence, THD+N of -107dB at 28Hz:

1 second silence and 20Hz high-pass filter. Seems identical:

And here's the reason why, captured with 1 second silence but no high-pass filter as the sweep was progressing. My guess is that somehow the sweep isn't doing as much averaging as the RTA window (or the initial few captures are much more distorted after a switching from silence to full signal):

 

[John Mulcahy]

OK, thanks. A little extra settling time is required before starting to average when there is a silence period, averaging too soon captures some elevated low frequency content remaining from the transition. I'll fix that in the next beta. In the meantime you can either run without a silence period if appropriate or with the distortion HP filter active.

 

[Paul Kane]

OK, thanks. A little extra settling time is required before starting to average when there is a silence period, averaging too soon captures some elevated low frequency content remaining from the transition. I'll fix that in the next beta. In the meantime you can either run without a silence period if appropriate or with the distortion HP filter active.

Thanks, John!

 

[John Mulcahy]

Changes in V5.20 beta 34:

• Added a jitter test signal to the signal generator in the Tones group
• Added a right click menu for the waterfall and spectrogram controls to copy and paste settings between measurements
• Added check boxes on the waterfall and spectrogram controls to enable or disable the automatic regeneration of the plot when settings change
• Moved the ISO 3382 parameters on the RT60 graph data panel into a separate window
• Added an Analysis preference (selected by default) to use only the right hand IR window width when deciding the minimum valid frequency in the response
• Added a View preference to use dashed strokes on overlay graphs
• Added a check box on the merge operation to optionally blend the merge over 1/3 octave either side of the merge frequency
• RTA mic cal data was only updated when the RTA started, now updates if cal changes while RTA is running
• Stepped sine continues to measure distortion even if "Show distortion" button is deselected
• Improvement in RT60 Decay calculation
• Small improvement in RT60 calculation for longer decays
• Bug fix: Group delay trace could be incorrect for log spaced data with 1/48 octave smoothing
• Bug fix: ASIO driver selection could fail for a sample rate causing an exception
• Bug fix: Extra settling time may be required for stepped sine measurements when using a silence interval

 

[Paul Kane]

Changes in V5.20 beta 34:

• Added a jitter test signal to the signal generator in the Tones group
• Added a right click menu for the waterfall and spectrogram controls to copy and paste settings between measurements
• Added check boxes on the waterfall and spectrogram controls to enable or disable the automatic regeneration of the plot when settings change
• Moved the ISO 3382 parameters on the RT60 graph data panel into a separate window
• Added an Analysis preference (selected by default) to use only the right hand IR window width when deciding the minimum valid frequency in the response
• Added a View preference to use dashed strokes on overlay graphs
• Added a check box on the merge operation to optionally blend the merge over 1/3 octave either side of the merge frequency
• RTA mic cal data was only updated when the RTA started, now updates if cal changes while RTA is running
• Stepped sine continues to measure distortion even if "Show distortion" button is deselected
• Improvement in RT60 Decay calculation
• Small improvement in RT60 calculation for longer decays
• Bug fix: Group delay trace could be incorrect for log spaced data with 1/48 octave smoothing
• Bug fix: ASIO driver selection could fail for a sample rate causing an exception
• Bug fix: Extra settling time may be required for stepped sine measurements when using a silence interval

Thank you for adding J-Test to the generator, John! One small issue that confused me at first: the frequency for J-Test seems to be set once on start-up of the generator. Changing sampling rate does not modify the frequency, and there is no way to adjust it after the rate change. The only thing to do is to exit REW, restart it and restart the generator with the correct sampling rate already set.

 

[John Mulcahy]

Thank you for adding J-Test to the generator, John! One small issue that confused me at first: the frequency for J-Test seems to be set once on start-up of the generator. Changing sampling rate does not modify the frequency, and there is no way to adjust it after the rate change. The only thing to do is to exit REW, restart it and restart the generator with the correct sampling rate already set.

The signal does change according to the sample rate. The main component is always sample rate/4, but the display of frequency doesn't update unless you switch to another signal and back. I'll fix that for the next beta.

 

[POJoe]

Hello,

I've been using REW for some time, but never been able to use Asio for FlexASIO. When I select it from the dropdown I get an 'Unable to load driver for Flexasio' error message. Flexasio is rev 1.5 and appears properly installed, but not able to load. Any help?

Thanks,
Joe McQueen

 

[John Mulcahy]

That can happen if the input or output name in the FlexASIO.toml file is not available in the list for the selected backend.

 

[POJoe]

That can happen if the input or output name in the FlexASIO.toml file is not available in the list for the selected backend.

Thanks, you learn something everyday.

Regards

 

[POJoe]

Hi John,

Sorry to trouble you, but it works once then I get a Java error and the Preferences window is greyed out:


REW V5.20 Beta 38 running JRE 1.8.0_202 64-bit on Windows 10, en, windows-1252


Message:
java.lang.ArrayIndexOutOfBoundsException: 1
Stack Trace:
1
roomeqwizard.wD.<init>(y:1889)
roomeqwizard.nG.ǒ(y:2867)
roomeqwizard.nG.ƿ(y:3233)
roomeqwizard.H.F(y:2977)
roomeqwizard.E.F(y:1683)
roomeqwizard.nG$33.actionPerformed(y:2169)
javax.swing.JComboBox.fireActionEvent(Unknown Source)
javax.swing.JComboBox.setSelectedItem(Unknown Source)
javax.swing.JComboBox.setSelectedIndex(Unknown Source)
roomeqwizard.nG.B(y:6568)
roomeqwizard.nG.A(y:4065)
roomeqwizard.nG.B(y:5401)
roomeqwizard.E.C(y:2662)
roomeqwizard.E.C(y:1631)
roomeqwizard.nG$3.actionPerformed(y:1218)
javax.swing.JComboBox.fireActionEvent(Unknown Source)
javax.swing.JComboBox.setSelectedItem(Unknown Source)
roomeqwizard.nG.A(y:1846)
roomeqwizard.E$9.process(y:580)
javax.swing.SwingWorker$3.run(Unknown Source)
sun.swing.AccumulativeRunnable.run(Unknown Source)
javax.swing.SwingWorker$DoSubmitAccumulativeRunnable.run(Unknown Source)
sun.swing.AccumulativeRunnable.run(Unknown Source)
javax.swing.SwingWorker$DoSubmitAccumulativeRunnable.actionPerformed(Unknown Source)
javax.swing.Timer.fireActionPerformed(Unknown Source)
javax.swing.Timer$DoPostEvent.run(Unknown Source)
java.awt.event.InvocationEvent.dispatch(Unknown Source)
java.awt.EventQueue.dispatchEventImpl(Unknown Source)
java.awt.EventQueue.access$500(Unknown Source)
java.awt.EventQueue$3.run(Unknown Source)
java.awt.EventQueue$3.run(Unknown Source)
java.security.AccessController.doPrivileged(Native Method)
java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(Unknown Source)
java.awt.EventQueue.dispatchEvent(Unknown Source)
java.awt.EventDispatchThread.pumpOneEventForFilters(Unknown Source)
java.awt.EventDispatchThread.pumpEventsForFilter(Unknown Source)
java.awt.EventDispatchThread.pumpEventsForHierarchy(Unknown Source)
java.awt.EventDispatchThread.pumpEvents(Unknown Source)
java.awt.EventDispatchThread.pumpEvents(Unknown Source)
java.awt.EventDispatchThread.run(Unknown Source)

Attached is the Toml file

Thanks

 

[John Mulcahy]

Your [input] section is commented out even though an input is listed.

The REW config file handling is fairly basic, if you want to go beyond the basic backend, bufferSize, input and output selections you are best managing the file manually and not using the REW control panel.

 

[POJoe]

Well!

I have always been following the FlexASIO instructions and did not know the control panel existed. Previously when It was working I would hit the Panel button and the FlexASIO instructions would show. This makes it all a lot simpler - A simple job for simple people.

Many thanks,
Joe

 

[jmol]

Hi guys, I'm new here. Looking for the download link, where can I find it?

 

[John Mulcahy]

"REW DOWNLOADS" in the forum top banner.