V5.20.14 early access build

[Lowclock]

Sometimes being able to see or work with excess group delay or phase seems useful, and at least for me the impact of specific tails for every one in groups of similar measurements is not. I'm no expert though.

 

[Lowclock]

I had a good laugh at this one. Measurement actions > Minimum phase version produces a measurement dated Unix time 0.

 

[serko70]

I prefer to use REW with all windows simultaneously open to increase efficiency. An example below from my ultrawide monitor. Overlays will replace EQ window at times:

I have some suggestions regarding resizing all or some of Info / Measurement actions / Trace arithmetic windows to maximize available screen space to main REW window:

TA window needs half the vertical length for its 3 comboboxes
TA comboboxes are horizontally 1.5 times wider than even the longest allowed measurement name of 25 characters
MA buttons could be listed vertically in one row since there's still vertical screen space even with all windows placed vertically
Info window could be narrower if it didn't include full calibration file path
Lastly, Controls window could be made detachable at least for the All SPL tab

I am a heavy user of REW so these might not be important issues for many but these are just some suggestions I believe to improve user experience.


Thanks.

 

[Lowclock]

TA window needs half the vertical length for its 3 comboboxes

It needs room for all the options that appear when doing some of the other operations. I do kind of agree it might look better for those options to always be there but greyed out when they're irrelevant, but it's better than a popped out window that changes sizes while you're trying to use it.

 

[John Mulcahy]

I had a good laugh at this one. Measurement actions > Minimum phase version produces a measurement dated Unix time 0.

Can't get more minimum phase than that

 

[John Mulcahy]

I have some suggestions regarding resizing all or some of Info / Measurement actions / Trace arithmetic windows to maximize available screen space to main REW window

REW needs to accommodate a wide range of screen sizes, down to 1024x768, which necessitates some compromises. I'll make the Info panel resizable.

 

[Lowclock]

Accidentally found that I can consistently make REW freeze indefinitely by middle-click zooming in amplitude on the RT60 graph.

 

[John Mulcahy]

Accidentally found that I can consistently make REW freeze indefinitely by middle-click zooming in amplitude on the RT60 graph.

Thanks, fixed for the next build.

 

[staticV3]

When starting a Stepped Measurement, would it be possible to display the Distortion settings in that window as well?

So from this --> to this

 

[Lowclock]

The graph switching hotkey combinations work with either the left or right ctrl keys, but the measurement switching hotkey combinations only works with the left alt key. It's kind of annoying to be able to switch graphs with one hand, but not measurements. It would probably be even more convenient to put both sets on ctrl so it's less of a stretch, there's already nothing else on those combos, and ctrl can't trigger the title bar menu accidentally.

 

[John Mulcahy]

When starting a Stepped Measurement, would it be possible to display the Distortion settings in that window as well?

Yes.

 

[John Mulcahy]

The graph switching hotkey combinations work with either the left or right ctrl keys, but the measurement switching hotkey combinations only works with the left alt key. It's kind of annoying to be able to switch graphs with one hand, but not measurements. It would probably be even more convenient to put both sets on ctrl so it's less of a stretch, there's already nothing else on those combos, and ctrl can't trigger the title bar menu accidentally.

I can add the Ctrl+Up/Down shortcuts.

 

[Lowclock]

That would be cool, thanks. Just out of curiosity, is there some kind of java limitation that will let you bind right ctrl but not right alt?

 

[John Mulcahy]

Alt Gr should be recognised, and triggers the menu key binding highlights, but shortcuts using Alt Gr do not work with my keyboard so safer to use Ctrl.

 

[staticV3]

I have another request, but it's a lot more involved, and very specific..

When doing THD+N vs Level sweeps of Amplifiers, it would be awesome if REW could dynamically adjust the step size to selectively increase resolution where it's most needed.
Here's the sort of end result I have in mind:


I imagine the algorithm would first sweep the entire range in 10dB steps.
Then, from the first step that shows rising distortion, go one step back and add 5dB in-between steps.
Then, in the interval that first shows a drastic rise in distortion indicating clipping, add 1dB in-between steps.
Repeat, but with 0.1dB steps to get the precise point of clipping.

I'm aware that this is all very extra, and would probably require a lot of fine-tuning to work well with all sorts of Amps, so I understand fully if you'd rather pass on this.

 

[John Mulcahy]

from the first step that shows rising distortion, go one step back

That would be very prone to false triggering, distortion does not always fall so monotonically. I could add an option that uses the distortion limit. If the limit is exceeded the measurement could pause, go back one step and continue with a smaller step size, then repeat as long as the step is not at a lower bound (0.1 dB, say). That would find the distortion limit with greater precision.

 

[onlyoneme]

Would it be possible to store distortion analysis from FFT in the .mdat file the way which would allow to show it again in the FFT window after loading an existing .mdat file? Currently it's saved as a comment and not reflected in the "Show distortion" overlay.

 

[JoinUs]

When applying EQ, I hope the same values as something like 0.97 and 1.56 are applied on the EQ tab. Now it's rounded up and applied. Of course, it's not that important, but I think it'd be nice if it was applied.

 

[fcracer]

Hi John, thanks again for the amazing work you do with REW! With the latest version (5.20.14), I have a strange issue on an M2 MacBook Air running Ventura 13.2.1. I used to be able to run sweeps via Airplay2 to my NAD M33 and C3050LE. I have two strange things happening now:

1. The sweep starts with a timing signal, gets about 90% of the way through and then stops. Sometimes it stops after 20KHz, but before the final timing signal. Sometimes it stops prior to 20KHz and the graph gets cut off at that level.

2. With the C3050LE, when I run a sweep over AirPlay2, when I click "Start", it opens Apple Music and starts to play a song from the app. This will go along with the sweep. I can stop the music and the sweep will continue, but also gets cut off.

I should note that I've used AIrplay2 in the past without issues. I understand Airplay2 is not ideal so I use other methods such as USB drives with the REW sweeps saved onto them, which work without issue. However, I thought you'd like to know about this issue in case it's not just my machine.

 

[John Mulcahy]

I hope the same values as something like 0.97 and 1.56 are applied on the EQ tab

Applied where? What setting are you referring to?

 

[John Mulcahy]

1. The sweep starts with a timing signal, gets about 90% of the way through and then stops. Sometimes it stops after 20KHz, but before the final timing signal. Sometimes it stops prior to 20KHz and the graph gets cut off at that level.

2. With the C3050LE, when I run a sweep over AirPlay2, when I click "Start", it opens Apple Music and starts to play a song from the app. This will go along with the sweep. I can stop the music and the sweep will continue, but also gets cut off.

Point 1 sounds like a problem with high latency and the replay device stopping when the source signal ends even though it hasn't finished being transferred. I can look at sending zeroes for a period after the sweep ends, might help.

On point 2, REW has no connection to Apple music so I don't know what is prompting that to open.

 

[JoinUs]

Something like this. If i apply 1.56db , it change 1.6db every time.
Of course, This doesn't matter much, but I wanted to apply it more accurately because the PEQ provided by ASR's spinorama is often given in numbers such as 1.56.

 

[John Mulcahy]

Gains use the resolution defined for the equaliser selected. "Doesn't matter much" is an understatement I'd say.

 

[Lowclock]

Gains use the resolution defined for the equaliser selected.

If you set them all to 0.01, frequency snaps to 0.05 increments, and gain works in 0.01 increments when you type them in but displays only 0.1 and the + and - buttons don't work properly anymore. Might not really matter, but it doesn't exactly behave properly either, haha.

 

[John Mulcahy]

the + and - buttons don't work properly anymore

Works fine for me, can you be a bit more specific so I can try and reproduce what you are seeing?