REW Beta Release V5.20 beta 29

[John Mulcahy]

There is something of a "live preview" theme to the V5.20 beta 29 release.

Firstly, the trace arithmetic parametric sum operation has been replaced by a more general "Alignment tool", which allows the gains, delays and polarities of two measurements to be adjusted while giving a preview of the SPL and phase of their sum. It makes most sense for this to be used with measurements that have been made with a timing reference, so the relative delays are meaningful. New responses can be generated from either of the aligned measurements or their sum by clicking the appropriate button.

The live preview theme continues with the addition of t=0 offset adjustment dialogs for the SPL & Phase, Impulse and Group Delay graphs. In each case a live preview of the effect the adjustment of the zero time will have on phase is shown.

The impulse graph also gets a new way of adjusting the impulse response windows, with indicators at the top of the graph showing the window reference position and the left and right window extents. Those indicators can be clicked and dragged to adjust the windows, with a live preview of the effect the change will have on the SPL response. Changes are applied when the mouse button is released.

The final live preview feature is for the effect of clock adjustments on the impulse response, again with a preview of the adjusted impulse and the effect on phase. Clock adjustment is fairly computationally intensive, so the previews are not as responsive as for the other dialogs.

This release also has noise criteria figures for the RTA, shown when the RTA is in one octave mode. The criteria available are Preferred Noise Criterion (PNC), Balanced Noise Criterion (NCB), Noise Criterion (NC) and Noise Rating (NR). The corresponding noise criteria curves can also be shown as an overlay on the graph. Note that self-noise of the microphone and its preamp will often be a limiting factor for noise criteria measurements.

Other changes in the release are:

• A button has been added in the SPL & Phase graph to invert polarity
• The control to normalise to a defined peak SPL has been moved from the Impulse graph to the SPL & Phase graph
• The "New" and "Avg" icons in the All SPL graph legend have been removed
• A menu entry has been added to export all impulse responses as text
• Bug fix: An exception could occur when checking levels
• Bug fix: Phase plots did not show data points when that option was selected
• Bug fix: NPE when applying clock rate adjustments to imported sweep recordings

 

[tankm]

Hi,
V5.20 beta 29
- Under "SPL & Phase" Controls, the "Add to Data" button is grey out for Measurement generated by "Average".
- Under "All SPL" Controls. Press "Align IR Start", it's now grey out. "Save All" and "Remove All" data. Load new measurements. "Align IR Start" is still grey out. Only way is to fully close REW and restart. This is now a new problem for this beta, happens in previous beta too.

Regards and Thanks for the great software.

 

[John Mulcahy]

- Under "SPL & Phase" Controls, the "Add to Data" button is grey out for Measurement generated by "Average".
- Under "All SPL" Controls. Press "Align IR Start", it's now grey out.

Thanks, I have fixed those for the next build.

 

[Breeman]

Thank you John, the new features in this version are really good

 

[jtalden]

John,
1. In the screen grab below the alignment tool signs are all negative for the offsets. The signs are all positive in the measurement panel notes. I assume these should agree? If we are thinking in terms of the needed REW zero shift I think the adjustment would be +7.037 ms and the distance would also be positive to have that same effect on the zero shift.

[Also note the negative sign in the tool for ft-in falls outside the parentheses.]

Comment:
It is confusing to me as I think in terms of what physical adjustment is needed to the audio system settings. In this example the actual delay change needed in a MiniDSP device is -7.37 ms. That is 7.37 ms less delay. If that adjustment was instead to be done in a receiver the distance would need to be +2.535 m.

I assume others will also be confused when making a physical adjustment to their system. I don't know a good way to address this issue. Possibly there could be something in the help that would help with this confusion? - just a thought. We will eventually get it right.

2. I use loopback timing and thus there is significant delay offset from zero depending on the buffer size and mic distance. 'Offset t=0' in the impulse chart will not allow impulse offsets greater than 10 ms even entered into the input field. The offset applied will be 10 ms instead of the larger number entered. Occasionally I need to shift zero a specific distance larger than 10. Can input field value made to accept larger numbers? The ±10 ms slider range is not a concern to me as I will only be entering specific changes.

 

[John Mulcahy]

1. In the screen grab below the alignment tool signs are all negative for the offsets. The signs are all positive in the measurement panel notes. I assume these should agree?

Yes, they should agree, the notes have the wrong sign. I'll fix that.

If we are thinking in terms of the needed REW zero shift I think the adjustment would be +7.037 ms and the distance would also be positive to have that same effect on the zero shift.

t=0 moves opposite to delay. If a measurement is delayed, so the IR peak appears later on the time axis, the t=0 point must necessarily be earlier.

[Also note the negative sign in the tool for ft-in falls outside the parentheses.]

That is to avoid having to put a sign in front of the feet and inch figures individually

2. I use loopback timing and thus there is significant delay offset from zero depending on the buffer size and mic distance. 'Offset t=0' in the impulse chart will not allow impulse offsets greater than 10 ms even entered into the input field.

Decoupling the field and the slider would be difficult to manage. If "Apply" is clicked with 10 ms of delay further delays can then be entered. A better way to deal with loopback timing offsets is to use the timing offset control on the Measure dialog.

 

[jtalden]

Thanks!

 

[Myles]

Wow John. You just keep on killin it man. I hope you find a good steward once you finally run out of energy. Ie, better than what happened to closed-source WinISD which is forever stuck right where it's at. Which isn't a horrible spot, but nothing like the growth REW keeps bringing.

What is your opinion on continuous sweeps while moving microphone for bass measurement. I think I would like to run quick sweeps on a loop as I move the mic around and have them all auto-averaged. Is that just a bad idea? If not, where is a good forum to discuss it? (I'll be dropping a $20 in your tip jar soon. I know, not much. But I hope you're getting a lot of those!)

 

[John Mulcahy]

What is your opinion on continuous sweeps while moving microphone for bass measurement. I think I would like to run quick sweeps on a loop as I move the mic around and have them all auto-averaged. Is that just a bad idea? If not, where is a good forum to discuss it?

The Moving Mic Method is worth looking at. There is something related on the way for REW, but taking a different approach.

I'll be dropping a $20 in your tip jar soon. I know, not much. But I hope you're getting a lot of those!

I wish! A donation would put you in a select group of about 1 in every thousand folk who download REW

 

[MatiasR]

I entered that select club too having donated my 10 usd to the cause.

 

[3ll3d00d]

I am probably missing something but this seems to make it *much* harder to simply shift an IR by x ms. This might be in the offset t=0 button in the IR window? but that limits you to +/- 10 and doesn't seem to shift anything either.

It's not obvious to me why the alignment window is presented as a replacement to the previous simple/straightforward move IR options. "Alignment" appears to service a particular use case that involves shifting an IR hence should be a new feature not a replacement.

EDIT: I see shifting a single IR is now in the multi response alignment tool, not an obvious place for it IMV

It would be nice to be able to add additional measurements to this tool btw (e.g. to handle >2 subs at once) as well as applying EQ at the same time

 

[John Mulcahy]

The Offset t=0 button on the Impulse graph brings up the replacement for the former t=0 offset controls. In the preview it shifts the time axis rather than the response, perhaps that is why it didn't seem to have an effect? Multiple shifts of up to 10 ms can be applied using the Apply button, but if there is a significant timing offset to accommodate for measurements that is probably better handled using the Timing Offset control on the Measure dialog.

 

[3ll3d00d]

OK I see what you mean now. Is there a reason to limit that to 10ms? I guess to make to slider increments offer a reasonable precision but I don't think a slider is a good tool for that from a UX point of view vs a spinner (and direct data entry) and/or copying a specific figure in from some other curve. I see this is pretty much the same point jtalden makes above so I guess we use it in a similar fashion

a minor bug report btw, the window line on the IR view does not redraw if you change the window (unless you uncheck the window curve and then check it again)

another feature suggestion..... the new window sliders are handy, would be useful to provide quick access to the window types too (perhaps a double or right click on the slider?)

 

[John Mulcahy]

OK I see what you mean now. Is there a reason to limit that to 10ms? I guess to make to slider increments offer a reasonable precision but I don't think a slider is a good tool for that from a UX point of view vs a spinner (and direct data entry) and/or copying a specific figure in from some other curve.

The field next to the spinner supports direct data entry. Arrow keys can be used to move the spinner with higher precision after clicking on it. The range was chosen to allow a pixel increment of 0.1 ms without making the control too wide. Time can also be adjusted using Set t=0 at cursor and Set t=0 at IR start. I do think a slider is a good control to use and better than a spinner for interactive adjustment

a minor bug report btw, the window line on the IR view does not redraw if you change the window (unless you uncheck the window curve and then check it again)

That has already been fixed for the next beta.

another feature suggestion..... the new window sliders are handy, would be useful to provide quick access to the window types too (perhaps a double or right click on the slider?)

The IR windows dialog already provides quick access to the window types.