REW Beta Release REW API beta releases

[dcibel]

Builds updated today (beta 72) with these changes:

• Changed: Adding an SPL offset to an FSAF measurement will offset the distortion SPL data
• Changed: Removing a measurement preserves the expanded state of measurement groups
• Fixed: Adding an SPL offset to an FSAF measurement shifted the distortion dBFS data
• Fixed: Generating minimum phase for imported magn and phase data could shift the SPL data

John, for FSAF data, I am seeing the dBFS level remain unchaned for SPL offsets which is great, however adjusting an SPL offset is not adjusting the SPL data in the distortion tab.

 

[John Mulcahy]

I agree, the drag and drop to re-order measurements is very awkward on the "ALL SPL" tab, I would prefer that it operates the same elsewhere.

All SPL (and other overlay tabs) allow multiple selections, so drag and drop is not as straightforward. It becomes easy if you first select the measurement(s) that you want to move, release the mouse, then click on one of the selected measurements for the drag. Alternatively make the selections, use Ctrl+X to cut them, click where you want them and use Ctrl+V to paste.

 

[John Mulcahy]

John, for FSAF data, I am seeing the dBFS level remain unchaned for SPL offsets which is great, however adjusting an SPL offset is not adjusting the SPL data in the distortion tab.

Are you using the "Add to data" button?

 

[dcibel]

I wasn't looking at the right measurement in the distortion tab, my bad. It's good, thank you John.

 

[sm52]

John, hello. If I press the sound card(Audient) levels check button in the Preferences/Soundcard tab, when there is a loop from the output to the input, ASIO drivers is selected and the level I set to -9.6 dB, the indicator shows -9.54 dB at the output. When I switch the selection to Java drivers, the indicator shows -9.6 dB.
Now I want to do a sound card calibration. On ASIO drivers. When the level indicators appear after a few clicks of the Next button, I see -9.6 dB at the output. I'm guessing that Java drivers turned on to calibrate the sound card. Because only for them the indicators showed -9.6 dB. But I need a calibration with ASIO drivers, for which the indicators showed -9.54 dB.

 

[sm52]

Yet. The calibration of the sound card was done in some way with acoustic synchronization. This is the setting of my last acoustic measurements.
Previously, the length was 256 during the calibration of the sound card. Now 1M. According to my observations, the least noise is at 512 and 2 passes.

 

[John Mulcahy]

The chosen driver is used for all audio input and output, it is not changed for soundcard calibration. Soundcard calibration uses a sweep that is long enough to provide accurate calibration data at the chosen minimum graph frequency. Soundcard calibration does not use a timing reference.

 

[dcibel]

Small ask, can the normalized waterfall plot change the y-axis scale to dBr, so 0dB is always at the normalized peak level?

 

[achugh]

All SPL (and other overlay tabs) allow multiple selections, so drag and drop is not as straightforward. It becomes easy if you first select the measurement(s) that you want to move, release the mouse, then click on one of the selected measurements for the drag. Alternatively make the selections, use Ctrl+X to cut them, click where you want them and use Ctrl+V to paste.

Hi @John Mulcahy can you please add more details to "if you first select the measurement(s)" statement? In the ALL SPL tab, is there any visual indication if a measurement(s) is selected? All I see is either WHITE or GREY background. Is it possible to visually show if measurement(s) are selected better and ready for dragging?

 

[John Mulcahy]

Selected measurements have a white background and their names appear in the graph legend and their data appears on the graph.

The complication with the overlay graphs, like All SPL, is that a click and drag that starts on an unselected entry is interpreted by the component as a range selection gesture. Click and drag on one of the selected entries to drag all the selected entries.

 

[sm52]

The chosen driver is used for all audio input and output, it is not changed for soundcard calibration.

Fine. Why has the output level changed when I turned on the sound card calibration?

Soundcard calibration uses a sweep that is long enough to provide accurate calibration data at the chosen minimum graph frequency.

Thank you.

Soundcard calibration does not use a timing reference

I assumed so too. But the measurement information contains an indication of synchronization.

 

[John Mulcahy]

Why has the output level changed when I turned on the sound card calibration?

No idea.

the measurement information contains an indication of synchronization

The excitation comment is wrong.

 

[achugh]

Selected measurements have a white background and their names appear in the graph legend and their data appears on the graph.

The complication with the overlay graphs, like All SPL, is that a click and drag that starts on an unselected entry is interpreted by the component as a range selection gesture. Click and drag on one of the selected entries to drag all the selected entries.

Thank you John for clarifying what constitutes as selected measurement(s). I understand why I am not able to rearrange among the WHITE background measurements. I have to first make the measurements I do not want to move into a GREY background.

Once you have good bit of measurements, if the sequence gets messed up, it gets really hard to reorder 1 or 2 measurements. Maybe you can think of some way to make it simply like a preference to disable multi-select measurement for movement/drag-n-drop. Moving individual measurement is required more frequently over moving a group of them; I think.

 

[John Mulcahy]

I have made a change for the next build so that an individual selected measurement on an overlay graph (including All SPL) can be moved to a new position by left click and hold on it until a move cursor appears.

 

[John Mulcahy]

Small ask, can the normalized waterfall plot change the y-axis scale to dBr, so 0dB is always at the normalized peak level?

That's more complicated than it might seem, since there are a lot of absolute SPL dependencies in the waterfall plotting, so that's a no.

 

[dcibel]

Ok, no problem. Main issue that it would address is that the normalized SPL may be quite different between two measurements, so I have to adjust SPL levels to match peak SPL in order to get a matching waterfall plot for each. It's not a big deal, but is a few extra steps.