V5.20.14 early access build

[John Mulcahy]

I cannot create a response smooth by default as the copy in the mdat.

You will need to select either Replicate data or add a tail to have the minimum phase conversion ignore the original IR. Your example response needs a LF tail at 2 dB/octave anyway if you want to match it.

 

[John Mulcahy]

I'm measuring the impedance. I calibrate before measuring. The levels are set to 8.94 dB. During 3 steps of calibration at 512k and up, headroom levels are fine. 8,9. At 256k the headroom level becomes 8.4 dB. Accordingly, an open circuit instead of 99.8 ohms shows 100.2 ohms.

What is your question?

Can REW remember the levels in the Preferences window that were at the time of checking, and then check them before measuring and correct if they drifted a little

No.

 

[sm52]

What is your question?

That's the way it should be?

 

[John Mulcahy]

It is typical for headroom to reduce as the sweep is shortened due to the transient response of the system being measured.

 

[sm52]

It's clear. Thank you.

 

[thothsong]

If I run the RTA with No smoothing, and save the data as a new measurement, the measurement looks unsmoothed. If I apply some smoothing to that measurement, and then apply No smoothing to try to revert it, I don't get the original unsmoothed measurement back, it appears to still be smoothed in some way. Is that expected?

 

[John Mulcahy]

If I run the RTA with No smoothing, and save the data as a new measurement, the measurement looks unsmoothed. If I apply some smoothing to that measurement, and then apply No smoothing to try to revert it, I don't get the original unsmoothed measurement back

I have fixed that for the next build.

 

[thothsong]

When I run the RTA in Spectrum mode with ERB smoothing, I get no visible plot.

 

[John Mulcahy]

When I run the RTA in Spectrum mode with ERB smoothing, I get no visible plot.

That has been fixed for the next build.

 

[thothsong]

I get a ClassCastException when using "Save the selected measurements" from right-click in the graph legend, when in a graph where the selected items aren't measurements. E.g., here's one from the Distortion graph.

 

[thothsong]

Is there any way to copy the Noise floor from a measurement, making it a measurement in its own right, so that it can be further manipulated?

 

[John Mulcahy]

I get a ClassCastException when using "Save the selected measurements" from right-click in the graph legend, when in a graph where the selected items aren't measurements

Those entries should only appear on overlay graphs, I have fixed that for the next build.

 

[John Mulcahy]

Is there any way to copy the Noise floor from a measurement, making it a measurement in its own right

You could use Export distortion data as text, exporting only the noise floor, then import that file.

 

[thothsong]

You could use Export distortion data as text, exporting only the noise floor, then import that file.

Thanks, that works.

 

[John Mulcahy]

Builds updated today to ea68 with these changes:

• Changed: If a measurement is disabled remove transient data such as waterfall, spectrogram, decay and RT60 decay and reclaim the associated memory
• Changed: Set sys.ext.installerUmask to false for Linux installers
• Changed: Application categories changed to Audio; Engineering; Java for Linux installers
• Changed: If a Java device reports supporting 10,000 channels (Linux plug plugin) test channel counts from 16 down to 2 to see what comes back as supported
• Changed: Improvements to RT60 estimation for short sweeps
• Changed: rePhase equaliser option runs at the sample rate of the measurement rather than being fixed at 96 kHz
• Fixed: The WASAPI Exclusive driver might not generate an output with some devices
• Fixed: The View preferences font size settings were not applied to all components
• Fixed: Improved UMIK-1 detection on Linux
• Fixed: Filters with zero gain were not removed after target match unless Drop small filters was selected
• Fixed: The wrong measurement name was shown during Sequential measurements
• Fixed: When merging B to A where A has an IR but B does not set suitably long IR windows for the result
• Fixed: Smoothing measurements saved from the RTA (or other measurements without an IR) could not be fully removed
• Fixed: RTA spectrum ERB smoothing did not work
• Fixed: Legend right click menu showed invalid entries for graphs that are not overlays
• Fixed: The result IR from multiplying responses with different sample rates might not be long enough

 

[serko70]

java.lang.NullPointerException trying to max out "configurable generic eq settings" as below:

REW V5.20.14 running Azul Systems, Inc. JRE 1.8.0_372 64-bit on Windows 10 Locale: language en, keyboard GB, windows-1252 at 96 DPI


Message:
java.lang.NullPointerException
Stack Trace:
roomeqwizard.NH.A(y:3195)
roomeqwizard.NH.C(y:2265)
roomeqwizard.GB.B(y:2193)
roomeqwizard.GB.A(y:1999)
roomeqwizard.GB$2.actionPerformed(y:1110)
javax.swing.JComboBox.fireActionEvent(JComboBox.java:1260)
javax.swing.JComboBox.setSelectedItem(JComboBox.java:588)
roomeqwizard.GB$1.actionPerformed(y:785)
javax.swing.JComboBox.fireActionEvent(JComboBox.java:1260)
javax.swing.JComboBox.setSelectedItem(JComboBox.java:588)
roomeqwizard.GB.B(y:3089)
roomeqwizard.GB.A(y:1573)
roomeqwizard.UB.A(y:194)
roomeqwizard.iA.E(y:2634)
roomeqwizard.GB.B(y:1859)
roomeqwizard.GB.A(y:1999)
roomeqwizard.gJ$2.run(y:1375)
java.awt.event.InvocationEvent.dispatch(InvocationEvent.java:311)
java.awt.EventQueue.dispatchEventImpl(EventQueue.java:758)
java.awt.EventQueue.access$500(EventQueue.java:97)
java.awt.EventQueue$3.run(EventQueue.java:709)
java.awt.EventQueue$3.run(EventQueue.java:703)
java.security.AccessController.doPrivileged(Native Method)
java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:74)
java.awt.EventQueue.dispatchEvent(EventQueue.java:728)
java.awt.EventDispatchThread.pumpOneEventForFilters(EventDispatchThread.java:205)
java.awt.EventDispatchThread.pumpEventsForFilter(EventDispatchThread.java:116)
java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:105)
java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:101)
java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:93)
java.awt.EventDispatchThread.run(EventDispatchThread.java:82)



It worked fine after restarting REW.

 

[John Mulcahy]

Thanks, I have fixed that for the next build.

 

[jeppekristoffer]

Hi John and thank you for an awesome piece of software. Its amazing.

Would it be possible to "load a single set off filters" to multiple measurements at once?

Regards Jesper

 

[sam_adams]

@John Mulcahy, below is an issue in v5.20.14ea68 in the distortion display when viewing an older .mdat file—from v5.19b6—where the noise floor is displayed as a vertical line to the left of the plot (highlighted):

It displays that way in all view modes for distortion on both Mac (10.14.6 - 14.0b2) and Win (10 - 11). I tried saving the old .mdat in the newest version but it did not change the behavior. The .mdat and diag files are attached.

 

[greatadrian]

Hi John,
Do you think it would make sense to add these two features?
- Having ability to manually set the SPL calibration constant (overriding the default 120 dB SPL). In my case I'm often extracting IR's created using "import sweep recordings", which were measured using audio chain with a particular SPL calibration.
- When "Scale IR data..." is used to adjust the gain of underlying impulse response this change is not saved in .mdat file (measurement doesn't turn blue) and upon reloading the project it goes back to the original value.

 

[John Mulcahy]

Having ability to manually set the SPL calibration constant (overriding the default 120 dB SPL). In my case I'm often extracting IR's created using "import sweep recordings", which were measured using audio chain with a particular SPL calibration.

There are two ways you can do that.

1. Select "Apply cal files" on the import sweep recordings dialog and have a mic cal file with a sensitivity entry, or
2. The cal offset used is from the currently selected input, so you could change that using the settings (cog icon) on the Cal files preferences

When "Scale IR data..." is used to adjust the gain of underlying impulse response this change is not saved in .mdat file (measurement doesn't turn blue) and upon reloading the project it goes back to the original value.

I have fixed the name not turning blue to show a modification has been made, but it saves as it should for me.

 

[John Mulcahy]

below is an issue in v5.20.14ea68 in the distortion display when viewing an older .mdat file—from v5.19b6—where the noise floor is displayed as a vertical line to the left of the plot (highlighted):

That file behaves the same in 5.20.13. 5.19 cannot load it. The noise floor is there but has a huge offset after the first point, you can see it by selecting % or dBr as the axis. I suspect the issue is with the file content for that beta release.

 

[John Mulcahy]

Would it be possible to "load a single set off filters" to multiple measurements at once?

There isn't a way to do that, a difficulty would be having a good way to choose the set of measurements. As is you can save a filter set for the current measurement using the File menu entry (or Ctrl+Alt+S) then select the measurements where you want to load it in turn and use Load filter set (Ctrl+Alt+O).

 

[greatadrian]

The cal offset used is from the currently selected input, so you could change that using the settings (cog icon) on the Cal files preferences

Wow, the cog icon suggestion works like a charm, thank you! Never been this deep in the settings

 

[jeppekristoffer]

There isn't a way to do that, a difficulty would be having a good way to choose the set of measurements. As is you can save a filter set for the current measurement using the File menu entry (or Ctrl+Alt+S) then select the measurements where you want to load it in turn and use Load filter set (Ctrl+Alt+O).

I was thinking that choosing the measurements in the "all measurements tab" and then apply the filters in the same way as as "cross correlate time align" which is a "multi measurement operation". But i have no idea how stuff is coded. I just thaught it would be very convenient when working with multiple location measurements and see how a set of filter would "influence" multiple measurements. Then the results could be viewed in the "overlay"-window and SPL-predicted tab as usual.