REW Beta Release REW API beta releases

[carl second]

In this case, the impulse colors repeat the frequency response colors.

Trace colors of the impulses in Alignment mode correctly displays in both Linux and Windows for me as well.

 

[sm52]

John, this behavior has been around for the past few months. Maybe this will help.

 

[John Mulcahy]

The problem of phase traces plotting in red occurs if the smoothing chosen on the Alignment tool dialog differs from the smoothing of the measurements. I have fixed it for the next build.

 

[sm52]

The problem of phase traces plotting in red occurs if the smoothing chosen on the Alignment tool dialog differs from the smoothing of the measurements.

That's right. Also, if the measurements were not smoothed, they remain unsmoothed when opened in the Alignment tool. But the colors of the phases are only red. After that, any change in smoothing in the Alignment tool or in the All SPL window does not change the red color.
Maybe this is also fixed, but it may be useful.

 

[HD MOVIE SOURCE]

When using Trace Arithmetic, and Magnitude Division, is there a way to set absolutely boost limits? I've tried using Regularization, but it doesn't work in the way I'd like it to.

Here's what I'd like. Add an ability to check a boost limit or positive gain limit when using Magnitude Division. I only want to a positive boost of +6db, and never want to go over that.
When using Magnitude Division you can see boosts of around +8db in parts of my EQ that you have to manually pull out. Adding a boost limit would really be helpful.

Am I missing something with Magnitude Division, or can something this be added?
Thank you.

 

[HD MOVIE SOURCE]

I tried using the +6db boost limit, and I still get boosts over +6db.
I tried using frequency limits, but they only apply to what is being EQ'd, and not what is being boosted.

If I could control this so the max gain would only use +6db boosts it would be perfect. But as it is, I don't see how the main gain setting works.

 

[John Mulcahy]

Images aren't much use for investigation, better to attach an mdat with the responses being divided.

 

[dcibel]

Is there a possibility for a higher resolution for burst decay waterfall or spectrogram, like 1/12 oct bandwidth?

 

[John Mulcahy]

Is there a possibility for a higher resolution for burst decay waterfall or spectrogram, like 1/12 oct bandwidth?

Sure, I'll add a wider range of choices.

 

[dcibel]

Thank you John. One other minor thing I noticed recently, is that the frequency limits noted on the measurement thumbnails have a limit of 20Hz min and 20kHz max, so they will show 20Hz-20kHz even if the measurement was completed from 0Hz to 48kHz for example.

 

[John Mulcahy]

One other minor thing I noticed recently, is that the frequency limits noted on the measurement thumbnails have a limit of 20Hz min and 20kHz max, so they will show 20Hz-20kHz even if the measurement was completed from 0Hz to 48kHz for example.

Yes, that's intentional.

 

[Hyacin]

Had REW working on this laptop previously, I've used it for all the cals I've ever done :-/

Today, a few months since I last used it I updated to B77 and ran a playback measurement - it gets right to the end, just before the final chirp, then throws a Java exception -

java.lang.NullPointerException

Searched the web, saw something about JRE8, installed JRE8 (though I had _nothing_ installed before and it was running, so it seemed unlikely to help), restarted, same. So I rolled back to B64 (the version OCA is using in the Neuron video, which is what I was hoping to generate samples for today), same thing. Not sure what version I was on prior ... I'm thinking as far back as maybe 45.

Surprised I'm not finding a thing from searching! Please let me know what additional details I can provide and how.

An aside John, as we discussed via email a while back, I've got a measurement script working now using measurement count as the blocker before moving on! Seems to be working really well (I'll know for sure when I can complete a measurement). Planning to throw it on GitHub when I'm done and call it "Neuron Companion" or something like that. May drum up a few more pro license sales for ya

Edit: Oh, and I tried ASIO in case it was related to the Java _drivers_ but got the exact same error!

Just tried again with B50, same thing, full message -

REW V5.40 Beta 50 running Azul Systems, Inc. JRE 1.8.0_422 64-bit on Windows 10 Locale: language en, keyboard CA, windows-1252 on 1707x1067 at 96 DPI Running in C:\Program Files\REW

Message:
java.lang.NullPointerException
Stack Trace:
roomeqwizard.hL.A(y:3382)
roomeqwizard.hL.A(y:2057)
roomeqwizard.hL.A(y:3293)
roomeqwizard.hL.run(y:3198)

Edit: My bad, measurement settings were wrong :embarassed:

 

[John Mulcahy]

Please let me know what additional details I can provide and how.

In the Help menu there's an option to generate a diagnostic file, if you attach that it will have details of the error.

 

[Hyacin]

In the Help menu there's an option to generate a diagnostic file, if you attach that it will have details of the error.

Bah, my bad, never mind ... it wasn't setting the reference file. It's working better now, but cutting out _before_ the final chirp! Surely just more issues with my settings! Thanks John!! :-)

 

[John Mulcahy]

I'd still like to see the diagnostic file, so that whatever led to the error can be handled in a more user friendly manner.

 

[Hyacin]

I'd still like to see the diagnostic file, so that whatever led to the error can be handled in a more user friendly manner.

Hrm, shoot, I clobbered the script version that caused it. Let me see if I can find it in my ChatGPT history and reproduce it while I've still got everything all hooked up! :-)

 

[Hyacin]

I'd still like to see the diagnostic file, so that whatever led to the error can be handled in a more user friendly manner.

Success! :-)

 

[danw1million]

Hi John,

Thanks for the most recent beta release, I just wanted to mention multitone appears to have a non linear response across now not sure if it is a bug or default setting change on the latest.

 

[John Mulcahy]

Looks like you have selected a pink profile on the generator instead of white.

 

[JoinUs]

Hello John, would you be willing to add “custom shortcuts”? For example, letting users assign their own hotkey for Export Wav, or applying a shortcut to generate minimum‑phase. Of course, this may not be necessary in most environments, but given the direction of many Beta improvements and with user‑friendliness in mind, I wanted to make this request.

 

[danw1million]

Thanks John, this solved my issue

Looks like you have selected a pink profile on the generator instead of white.

 

[brianvivolo]

I'm trying to pull FFT magnitudes through the API, but decoding the base 64 string and converting the bytes to a 32-bit float is yielding some wild numbers. I'm simply taking the encoded string, decoding as base 64, then taking the bytes and converting to float32 with numpy.

If anyone can take a look at my trace and accompanying code here, I'd appreciate the help. There must be some part of the implementation I'm missing.

Thank you!

 

[brianvivolo]

I'm trying to pull FFT magnitudes through the API, but decoding the base 64 string and converting the bytes to a 32-bit float is yielding some wild numbers. I'm simply taking the encoded string, decoding as base 64, then taking the bytes and converting to float32 with numpy.

If anyone can take a look at my trace and accompanying code here, I'd appreciate the help. There must be some part of the implementation I'm missing.

Thank you!

Nevermind!

After using numpy to convert to 32-bit floats, I had to add .byteswap(). I don't know enough to explain this; it was a copilot suggestion.

 

[JoinUs]

Hello John, this has actually been happening since a previous beta version (I’m not sure exactly when) – is it just me experiencing this?
My REW setting is 48,000 Hz. However, sometimes it keeps switching to 96,000 on the EQ tab.

Below are screenshots of the settings.

So to avoid this, force-set the sample rate once in the PEQ settings.

This outputs a single filter at 48,000 Hz, and then the Generic EQ sample rate also reverts to 48,000 Hz.


Of course, that workaround works, but I’m not sure why it’s changing like this.

 

[John Mulcahy]

My REW setting is 48,000 Hz. However, sometimes it keeps switching to 96,000 on the EQ tab.

Equalisers that run at the rate of the measurement default to 96k when there are no measurements, I can change that so they default to the current REW sample rate.