REW Beta Release REW API beta releases

[achugh]

You are measuring a subwoofer full range and REW detects very low output beyond 250Hz relative to noise floor and pops a warning. You can ignore it for subs or measure them with a limited frequency range.

Thank you for confirming my suspicion. I used to work with limited range but after watching your Youtube channel, I started doing full range for subwoofer sweeps. Since REW already knows that this is a subwoofer sweep as the channel selected is LFE, it will be great if REW applies some smart logic to check for distortion via some preference where one can provide the LFE range for distortion check to avoid this dialog box from showing up unless there is really a distortion to worry about. For now, I will continue to ignore this message for full range subwoofer sweeps.

This distortion graph is showing FUNDAMENTAL line plot as a straight line at 100 mark. In some other threads here I have seen this line to be drawn like other harmonic lines. I do not know if this is something I need to worry about or it is also because of the full sweeps on the LFE channel that I need to ignore. Does anyone knows why this 1st harmonic (Fundamental) line is a straight line at 100 value?

 

[John Mulcahy]

Does anyone knows why this 1st harmonic (Fundamental) line is a straight line at 100 value?

Because you are plotting the results as a percentage of the fundamental. If you change the Y axis type to SPL, for example, you can see absolute rather than relative levels.

 

[John Mulcahy]

Builds updated today (beta 60) with these changes:

• Added: API /alignment-tool/remove-delay endpoint to control the option to remove time delay
• Changed: Thiele-Small parameters "Write parameters to file" includes simplified model parameters
• Changed: Cross corr align improved for measurements with strong reflections
• Fixed: The alignment tool option to remove delay from the first measurement could be applied after using the alignment tool via the API
• Fixed: On the Impulse graph dragging the time axis during Offset t=0 was not working
• Fixed: STI panel was using internal test data

 

[JoinUs]

Hey John, Is it possible for REW to do this?

If there's a response like this

This way, it connects to the middle of the dip and the peak.
Smoothing can make this partly smooth, but I'm wondering if it's possible as if I drew a dotted line.

I can open the csv file in Excel and change the number from the desired frequency to the frequency to make it go on smoothly like that, but I was wondering if it would be possible in REW as well.


Making it soft with Python was helped by GPT, but I thought it would be better if these features were added to REW.
Yellow Original
Green Connect the calculated midpoints peak and dip

1/1, 1/3, PSY smoothing, etc. It's definitely different from what I intended.
I don't understand the exact action because it was made by GPT, but I also attached the py file.

 

[John Mulcahy]

No, it isn't possible for REW to do that. You could experiment with the effect of varying the right hand window on the Impulse graph, drag the "R" marker at the top of the graph.

 

[JoinUs]

No, it isn't possible for REW to do that. You could experiment with the effect of varying the right hand window on the Impulse graph, drag the "R" marker at the top of the graph.

It was a bit of a different story from gating, but I see.

Another suggestion (or question), is it possible to impulse the response from room simulation?
Currently, when I try to export it, I don't have Data. Of course, it's not an impulse, so I can't.

But if you generate mp and ep from room simulation and multiply it again, you get the same impulse as room simulation.
It would be nice to have this option (MP+EP with IR) for room simulation export as well.

Also, in a typical room simulation, 200 Hz is enough for observation frequency, but is it possible to provide an option to extend it to approximately 400 to 500 Hz?

 

[Mike0815]

Added: Multi-input capture has options to align the individual measurements to the first input's measurement using cross correlation alignment and to vector average the individual measurements, both are only available when saving inputs to individual measurements

Hey John,

thanks for adding this.
Would be nice if the resulting vector average would be automatically named like the individual inputs.
E.g: Measurement with 7 mics for left speaker: L-1, L-2, L-3, L-4, L-5, L-6, L-7 => L-VectorAverage (or L-VA)
Would prefer the shorter form.
And one more wish: Could there be an option to auto remove the individual L-1, L-2, etc. after VA was generated? Or maybe some button to remove these individual measurements from last measurement run?

 

[John Mulcahy]

Would be nice if the resulting vector average would be automatically named like the individual inputs.
E.g: Measurement with 7 mics for left speaker: L-1, L-2, L-3, L-4, L-5, L-6, L-7 => L-VectorAverage (or L-VA)

The "-VA" form will be used in the next build.

Could there be an option to auto remove the individual L-1, L-2, etc. after VA was generated?

I have changed the behaviour so that individual measurements do not need to be saved to generate a vector average.

 

[LastButNotLeast]

Can you make the measurements window resizable? I have to remember to change the resolution of my screen in order to get to the buttons at the bottom. And then, of course, I can't see anything else.

Thanks.
Michael

 

[John Mulcahy]

Can you make the measurements window resizable?

You can choose a smaller font in the View preferences.

 

[LastButNotLeast]

There's a reason I prefer a larger font. ;-)
Michael

 

[John Mulcahy]

Sure, but it's either reduce the font size slightly or use a larger screen.

 

[John Mulcahy]

Builds updated today (beta 61) with these changes:

• Added: API /measurements/{id}/command entries to Wrap phase and Unwrap phase for a measurement
• Changed: Multi-input vector average no longer requires saving inputs individually
• Changed: Improved FSAF band filters from MatLab FSAF V3.1 upload
• Changed: Flatten the later stages of the IR generated from sweep measurements by padding the captured data with data from the end of the capture
• Fixed: Sequential channels selection panel on the Measure dialog was limited to 4 rows

 

[dcibel]

Thanks John, I will try and find some time this weekend to test the improved FSAF routine.

As a simple feature request, since FSAF supports file format of .wav, .aiff, .flac, .mp3, are you able to also support these file formats in the RTA "open wave file" function?

 

[Mike0815]

Changed: Multi-input vector average no longer requires saving inputs individually

Hi John,

observed that in case "Save inputs to individual measurements" is not selected no distortion graph gets shown for the resulting -VA.

Michael

 

[Mike0815]

Another wish:
Could a button or so be implemented in every window or a global menu entry (e.g. in Preferences: "Save window positions" ) to save current pop up window positions and screens?(e.g."Make a measurement" )
Here this window always jumps back to screen 0 and every time I have to drag it to my larger screen for every new measurement. (macOS 11.7.10)

 

[John Mulcahy]

As a simple feature request, since FSAF supports file format of .wav, .aiff, .flac, .mp3, are you able to also support these file formats in the RTA "open wave file" function?

Yes, I have added that for the next build.

 

[John Mulcahy]

observed that in case "Save inputs to individual measurements" is not selected no distortion graph gets shown for the resulting -VA

I have fixed that for the next build.

 

[dcibel]

Thank you!

One more request would be to add M-noise to REW's signal generator and FSAF signal drop-down. Of course I can run the noise from file, it would just be a lot more convenient to have this built-in to REW.

 

[John Mulcahy]

Here this window always jumps back to screen 0 and every time I have to drag it to my larger screen for every new measurement.

The Measure dialog is owned by the main REW window. It first appears centred on the position of that window, but you can move it elsewhere and afterwards it should appear wherever you have moved it to. Is that not what happens?

 

[John Mulcahy]

I will try and find some time this weekend to test the improved FSAF routine.

Note that I have only added Michael's improved filters. He has made many other changes in the subband processing, but porting those changes to Java will take quite a bit more time.

 

[John Mulcahy]

One more request would be to add M-noise to REW's signal generator

There is no published algorithm to generate the M-noise so that would mean adding the M-noise WAV files to the REW distribution. I don't want to do that.

 

[dcibel]

That's unfortunate. I was hoping that something like pink noise with a pre-defined filter that matches the spectrum in my screenshot above would be enough to replicate M-noise functionality, perhaps there's more to it than that. No problem I will just use the freely available audio files.

 

[Mike0815]

It first appears centred on the position of that window, but you can move it elsewhere and afterwards it should appear wherever you have moved it to. Is that not what happens?

Nope, REWˋs main window starts up at my notebook's screen and when I drag it to my larger monitor the secondary windows appear every time at notebook's screen.

 

[Draki]

I haven't used this feature for a while, so I am obviously missing something: It used to be possible to adjust the FR gain of a measurement (in the Controls I think).
Very useful for comparison with imported measurements made on a different dB level etc.

Now I can't find that feature in the current version. Is it still possible?