V5.20.14 early access build

[John Mulcahy]

It is best to use some regularisation for division, and to specify the frequency span if your measurements have roll-offs. Those steps help to avoid low or zero levels in the B measurement from causing issues. Make sure you have the latest 5.20.14 build from the first post of the 5.20.14 early access thread. I'm away this week (camping) so don't have a way to look at measurements, sorry.

 

[gadzooks]

Thanks John. I will play around with that. Enjoy your camping sir!

 

[onlyoneme]

Regarding my issues to get valid results from stepped sine measurements when 2 independent devices are used - enabling "Get fundamental from generator" in RTA window had solved them.

 

[serko70]

Dear John,

If "Offset t=0" in SPL & Phase / Actions is used, a time value of "-10 ms" fills the box every time regardless of the value entered.

May I also kindly ask to have an option to keep "All SPL" / Actions window persistent please?

Happy holidays!

 

[serko70]

Another useful addition would be to add "smoothing" memory to "excess phase version" like the one "minimum phase version" has. Currently, it can be done with exporting smoothed EP version as text and importing it back but it loses its real "impulse" property in the process.

 

[Bentxu]

@John Mulcahy, thanks for making such a great tool. Loving every bit of it.
Not a bug report but more of a feature request for the Generic / Configurable PEQ.
Would it be possible to also include XOs X1 and X2 similar to the Generic / Generic implementation?
Love the ability to have more than 20 Filters (my Helix DSP supports 30) but would be great to also set XOs.

 

[John Mulcahy]

If "Offset t=0" in SPL & Phase / Actions is used, a time value of "-10 ms" fills the box every time regardless of the value entered

I can't reproduce that, was there some particular measurement that showed that behaviour?

May I also kindly ask to have an option to keep "All SPL" / Actions window persistent please?

It would be a bit risky to have a dialog with actions for selections that are not currently visible.

 

[John Mulcahy]

a feature request for the Generic / Configurable PEQ.
Would it be possible to also include XOs X1 and X2 similar to the Generic / Generic implementation?

I'll add it to the feature request list, but won't be soon.

 

[serko70]

I can't reproduce that, was there some particular measurement that showed that behaviour?

I just entered -25.776 to the box and pressed the tab key. Only happens if you use the Offset t=0 under "All SPL" tab!

 

[John Mulcahy]

That control has a range of +/- 10 ms

 

[serko70]

That control has a range of +/- 10 ms

The behaviour of the combo box is different in the "Impulse" tab.

 

[John Mulcahy]

Yes. Adjusting delay while looking at HF phase requires fine adjustment, the slider range can't be too large or the steps get too big. The impulse tab just has the numeric control. A similar issue affects the alignment tool, separate coarse and fine adjustments might be a way to handle it.

 

[serko70]

Yes. Adjusting delay while looking at HF phase requires fine adjustment, the slider range can't be too large or the steps get too big. The impulse tab just has the numeric control. A similar issue affects the alignment tool, separate coarse and fine adjustments might be a way to handle it.

Ok, understood!

 

[Bentxu]

I'll add it to the feature request list, but won't be soon.

Heroic, thank you @John Mulcahy

 

[serko70]

I'll add it to the feature request list, but won't be soon.

Can we please also add to the list something like a "A x B multiply all selected" to the All SPL actions menu?

To convolve multiple filters into one or to a response which is quite a frequent process, one has to create many AxB calculations which is not only time consuming but also eats space and is very error prone to find the correct AxB among others to re-multiply.

Pretty please with sugar on top

 

[sm52]

John, is it possible to use an equalizer to make a filter with the greatest steepness, which will lower the peak between 30 hertz and 75 hertz by 25 dB. If yes, what are the equalizer settings and how to apply it to an existing measurement and create a new one as a result?

 

[Dr.OP]

Yes. Adjusting delay while looking at HF phase requires fine adjustment, the slider range can't be too large or the steps get too big. The impulse tab just has the numeric control. A similar issue affects the alignment tool, separate coarse and fine adjustments might be a way to handle it.

May be some sort of control toggle?

Idea is to have a rotation knob(slider) with a multiplier toggle.

Two states scenario
Toggle red changes 22
Toggle green changes 25
Toggle can have more states to change value digit by digit.

Behavior could be as follows:

1. Default value is sensitive
2. Click on the toggle changes behavior to the rough mode
3. After dial is released toggle switches back to the sensitive mode

 

[John Mulcahy]

John, is it possible to use an equalizer to make a filter with the greatest steepness, which will lower the peak between 30 hertz and 75 hertz by 25 dB. If yes, what are the equalizer settings and how to apply it to an existing measurement and create a new one as a result?

I'm not sure I follow what you are after. If you want to lower a range of frequencies the d&b audiotechnik tuning filters EQ has an "Asymmetric" filter which does that.

The EQ window Filter tasks "Generate measurement from predicted" makes a new measurement with EQ applied.

 

[sm52]

I didn't get what I wanted. Here is the measurement. I need to remove the peaks at 40 and 50 hertz to the noise level, because this is network interference, the tweeter does not have such frequencies.

 

[John Mulcahy]

I need to remove the peaks at 40 and 50 hertz to the noise level

The use individual PK or notch filters. What does any of this have to do with the 5.20.14 early access build?

 

[serko70]

I didn't get what I wanted. Here is the measurement. I need to remove the peaks at 40 and 50 hertz to the noise level, because this is network interference, the tweeter does not have such frequencies.

I have a similar AC ground loop problem at my home but you have some addiitonal odd frequenices.

 

[Ghostknight]

Hello
i tried to use REW under Linux for the first time, unfortunaly my umik-1 isnt working properly anymore (it was working fine under windows before), it gets selected and i can use it but if i wanna load a calibration file there is a error that the sense factor cant be used for unrecognized mics

even if i do a manual SPL calibration the values shown seem off (overall too quite and somehow bassheavy under 20hz)

then i found another post of this issue pretty recently which suggested using the early access version but also this didnt help

this is the soundcard_debug file:

 

[sm52]

The use individual PK or notch filters

Thank you.

I have a similar AC ground loop problem at my home but you have some addiitonal odd frequenices.

Thank you. This is what I wanted. I tried it, but I still don't really understand how it works. Do you have a video about it?

 

[John Mulcahy]

i tried to use REW under Linux for the first time, unfortunaly my umik-1 isnt working properly anymore (it was working fine under windows before), it gets selected and i can use it but if i wanna load a calibration file there is a error that the sense factor cant be used for unrecognized mics

I'll try and improve that in the next 5.20.14 early access build.

 

[serko70]

Thank you.

Thank you. This is what I wanted. I tried it, but I still don't really understand how it works. Do you have a video about it?

I tried many things but couldn't sort it out, the source cause is probably not in my flat! Although they are not audible, it will cause switching phases between left and right speakers and measurements will show random peaks or dips at 50Hz which is annoying. You can get minimize them with longer measurements. 4M works best.