V5.20.14 early access build

[JoinUs]

On Latest version, Ctrl+MouseRightClick and Drag is deleted? The error or drag does not work. (Used a lot to check DB deviation)

 

[John Mulcahy]

No, nothing deleted. Works for me on every graph.

 

[JoinUs]

No, nothing deleted. Works for me on every graph.

Re-run REW and it works now.
I delete the temp file periodically, but this beta version seems to be especially slow. There are only 8 messages, but something is lagging.
Downgrade back to the previous version...

 

[John Mulcahy]

Anything more than 7 days old in the temp folder is deleted automatically on shutdown. There's really nothing for me to work with to investigate "something is lagging", if you find any set of measurements that seems to cause issues or any particular action please let me know.

 

[JoinUs]

Anything more than 7 days old in the temp folder is deleted automatically on shutdown. There's really nothing for me to work with to investigate "something is lagging", if you find any set of measurements that seems to cause issues or any particular action please let me know.

It was a file that was testing Align, so I deleted it earlier, so there is no record.
What I did was to make two messurement of Subwoofer using alignment tool to make several Align Sums(7-8?).
And since then everything has slowed down. Everything from clicking on a graph to looking at another graph.

 

[John Mulcahy]

If there is some repeated calculation running you may see activity in the bottom right corner of the main window.

 

[John Mulcahy]

What I did was to make two messurement of Subwoofer using alignment tool to make several Align Sums(7-8?).

If the behaviour is repeatable I can send you a debug build to try and work out hat is happening.

 

[serko70]

If the behaviour is repeatable I can send you a debug build to try and work out hat is happening.

I have the same problem: CTRL + right click has stopped working. CTRL+Shift+left click still works. Restarting REW will probably fix it but just wanted to let you know.

 

[JoinUs]

If there is some repeated calculation running you may see activity in the bottom right corner of the main window.

Sometimes, Sometimes even if you don't perform any calculations, you can use a lot of CPU shares or slow down. (Even if the total number of messurement is only 3-4 from 256k basic sweap)
The problem is that you can't reproduce this lag. I don't know exactly what causes it. Restarting REW like I wrote above will solve it, but I think I need some more optimization.

 

[John Mulcahy]

Builds updated today to fix lag/interface issues:

• Fixed: Some actions could trigger repeated smoothing causing interface problems and lags

 

[JoinUs]

Builds updated today to fix lag/interface issues:

• Fixed: Some actions could trigger repeated smoothing causing interface problems and lags

Good! I'll try!
Also In previous version, I got message error many tabs(Almost 80?90?)
I forgot to report to you, and accidentally turned off the error window. So I couldn't copy the details. Can you look into this alone?

 

[John Mulcahy]

I forgot to report to you

In the Help menu please use "Generate diagnostic file" and either attach or send me the file that produces.

 

[John Mulcahy]

Builds updated today with these changes:

• Added: Noise and distortion levels are now calculated for sweep measurements and displayed in the Info window
• Added: Noise and distortion are checked when sweep measurements are made and a warning is shown if either is too high
• Added: When a measurement is selected its trace is highlighted on any displayed overlay graph
• Added: Clicking on the panel for the currently selected measurement toggles trace highlight on any displayed overlay graph
• Added: The Ctrl+H shortcut toggles trace highlight for the currently selected measurement on any displayed overlay graph
• Added: Measurements can be disabled or enabled by right clicking on the measurement's thumbnail. Disabled measurements cannot be selected, are skipped when using the next/previous measurement shortcuts (Alt+up/down and Ctrl+up/down) and are not shown in the legends of overlay graphs. All measurements can be re-enabled via an entry in the right-click menu of graph legends or using the Ctrl+E shortcut. All selected or all unselected measurements can be disabled via entries in the right-click menu of overlay graph legends. The Ctrl+D shortcut disables the currently selected measurement.
• Added: The installer has a 3 GB memory allocation option

The main focus is helping manage overlay graphs when there are many measurements. Measurements can be disabled, which will remove them from all overlay graphs and skip them when cycling through measurements with the next/previous shortcuts. That should make it easier to focus on the measurements of interest at the time. As the selected measurement changes it is highlighted on the overlay graphs, the highlight can be toggled with Ctrl+H.

The noise and distortion metrics are for the measurement as a whole and intended to help identify measurements which may be suffering from excessive distortion or level problems.

I have finally relented and added a larger memory allocation option, 3 GB, but there remains a risk that the OS may say no and REW will not launch, per the note shown on that installer page.

 

[sm52]

John, I tried zooming in on the frequency axis on the SPL plot to see the plot around 22-24kHz. But there the increase is limited. Between 23 and 24 kHz the distance on the screen is 2 cm. This is 1000 hertz. Is there any reason why that is?

 

[John Mulcahy]

Sorry, I can't understand what the issue is. Perhaps an image would help. The minimum span for the frequency axis is 5 Hz.

 

[sm52]

I'm talking about the zoom limit.

 

[John Mulcahy]

The button is disabled when the span for the log axis is about half an octave. Greater zoom is possible by switching the axis to linear or using the mousewheel on the axis.

 

[sm52]

using the mousewheel on the axis

It works. Thank you.
Another wish. Is it possible to make in all tabs where there is a zero value of the unit of the vertical axis, the axis of this zero value is 2 pixels? Slightly thicker than grid. For example, on the SPL tab, make the axis 0 degrees of phase 2 pixels. Or, on the impulse tab, make the axis of the zero value of the amplitude 2 pixels. Like that. So that this axis does not interfere, you can not bring it to the end on the right and left. But I don't think it will interfere.

 

[davide.bortolami-sb]

Hi John,

I would really appreciate a feature that allows to automatically take multiple sweep measurements stepping the amplitude of the signal.
I currently do this manually, but it is very tedious and time-consuming.

Here are two uses particularly dear to me:

- Test speakers at different power levels, understanding when they start to distort, and how. For speaker designers, this is particularly important to understand how to address the issues
- Test DRC filters
- Test amplifiers at different power levels (the stepped sine measurement is fine for this, but the sweep would be faster)
- Find at what SPL things start to rattle, rub&buzz

 

[Breeman]

Hi John,

It looks like the phase response is calculated differently between version 5.20.14 and 5.20.13 when applying smoothing. There are significant differencies between the phase responses calculated by the two versions at higher frequencies (>1 kHz). Is this due to the changes in phase smoothing that you have incorporated into the newer version? Do you advise that the use of version 5.20.13 be deprecated?

 

[Breeman]

Hi John,

It looks like the phase response is calculated differently between version 5.20.14 and 5.20.13 when applying smoothing. There are significant differencies between the phase responses calculated by the two versions at higher frequencies (>1 kHz). Is this due to the changes in phase smoothing that you have incorporated into the newer version? Do you advise that the use of version 5.20.13 be deprecated?

For instance, below is the phase response of the same speaker between the two versions when variable smoothing is applied but the differencies are evident across all smoothing options.
Version 5.20.14ea64

Version 5.20.13

 

[davide.bortolami-sb]

Another feature request, this one might be quite difficult!
Please support M-Noise measurement, according to AES75-2023

 

[Norixone]

Hello everyone. I am having an issue with the beta release. I am using an Umik-2 and the LFE channel seemds to be down by 10 dB. In other words, the output of the LFE channel is the same as all other channels, which was not the case with the latest offical version. I know about the possibility of selecting the -10dB option in the measurment menu, but I have not flagged it and still get the -10dB. The LFE channel is working correctly as I have tried Dolby discs and files that provide a pink noise for all channels including the LFE and all is well. Any ideas on what may be causing the issue?

@John Mulcahy congratulations for the latest REW version. It has a great UI and the auto EQ is incredible.

 

[John Mulcahy]

Is this due to the changes in phase smoothing that you have incorporated into the newer version?

Yes, that's why it was implemented.

Do you advise that the use of version 5.20.13 be deprecated?

Yes.

 

[John Mulcahy]

I am having an issue with the beta release. I am using an Umik-2 and the LFE channel seemds to be down by 10 dB

Are you using the latest build? It is updated frequently, the links in the first post always point to the latest build.