CoolEdit Nostalgia visualizer

[sveakul]

Thanks kamen for your work in 5.8!  A precision tool with great flexibility and a gorgeous interface.

[kamen]

Happy it works for you sveakul! Thank you for the confirmation.

[TF3RDL]

Maybe crossed too! So a MidSide view with freely selectable channels (except center & subwoofer) to form a pair could be useful. But that is just a thought for the moment.

Though with freely-selectable channels to form a pair (Mid/Side and phase correlation metering), it should be able to select center and LFE channels too for either two slots

BTW, with the LUFS meter part complete (w/ LRA measurement) and phase correlation and panorama meter implemented on my own multichannel peakmeter project, I think it is the right time to put these three meters from this CodePen project into CoolEdit Nostalgia visualizer because foo_loudness_peakmeter 2.x.x.x might very well be a vaporware right? considering Crossover (dev of foo_enhanced_spectrum_analyzer for foobar2000) on HydrogenAudio forums wasn't active since last 6 months and MusicBee finally got a new API to get raw samples data

BTW, here's a mockup of this new loudness peakmeter part:

[kamen]

BTW, here's a mockup of this new loudness peakmeter part:

I have been pushing away the idea of a peak meter tool inside CEN for months, so it might take months to actually do something on that.

However your idea is more than welcome!
It goes a lot beyound what SoundForge's peak meter V2 offers together with its mono compatibility meter (as an example of an advanced peak meter). It implements the upper half of your mock, at best.
On the other hand CoolEdit has always been very simplistic in those terms. So there is some potential to expand a bit on the standard functionality.

[sveakul]

Hi kamen:  while testing an unrelated bug , I noticed some interesting behavior from CEN.

Apparently, unless a file has been "scanned" first within MusicBee, when played CEN only shows the message "No signal from RawAPI source" if the Source has been set to "RawAPI Bridge." Changing real-time to "MusicBee Native Fourier 2K" brings the spectrum back to life.

If "Rescan" is invoked (which would also be the case anytime a file is offically "Added to Library),"  the RawAPI  spectrum immediately returns.

FWIW, this also affects the VUMeter Plugin 2.6, which is "dead" until the file is scanned.

[BoringName]

FWIW, this also affects the VUMeter Plugin 2.6, which is "dead" until the file is scanned.

I would assume this must be due to GetPCMRawData not providing any data to the plugins unless the track has been "scanned".

[kamen]

"No signal from RawAPI source" means technically that GetPCMRawData is not being triggered by MB. Maybe GetStreamInformation too, but CEN is a bit immune to that.

I couldn't reproduce it myself.

Native and DSP sources should be working as they are independent of the above Raw API functions.

[sveakul]

"No signal from RawAPI source" means technically that GetPCMRawData is not being triggered by MB. Maybe GetStreamInformation too, but CEN is a bit immune to that.

I couldn't reproduce it myself.

1.  Find, download, whatever, a music file you are sure you have never played in MusicBee before or added to its library.  Not seeing it listed in your library is not good enough;  it must not appear at all inside the MusicBeeLibrary.mbl file due to the bug being discussed here .

2.  Right-click the file in Windows File Explorer (not in any MB window like "Computer" etc.), and choose Open With-> MusicBee

3.  MusicBee opens and starts playing the file; while the file is playing, on CEN select Source as "RawAPI Bridge"; the spectrum will cease with the message "No signal from RawAPI source".

4. While the song is still playing, change source to "MusicBee Native Fourier 2K"--the spectrum will return.

5.  Run a "File Rescan" on the file (OR add it to the Library), and RawAPI Bridge will show normally again.

Be sure you don't accidently "scan" the file first by moving it within MusicBee, etc.

Note:  this isn't being represnted as a "bug" per-se, just as a behavior eccentricity, demonstrating that files need to be scanned first manually or automatically by MusicBee before "GetPCMRawData" is triggered by MusicBee.

[kamen]

Thanks for the description, I got it right then yesterday, with the information from the discussion. But I couldnt reproduce it with MB 3.6.109 and 146 too.


I would make anyway a guess that the File Scan and Rescan might be unrelated to the issue.

Steven did make a similar fix to the RawApi back then:

2) You close the player while playing a song with this option selected . After starting the player again it continues to play, but GetPCMRawData and GetStreamInformation does not get called, unless another song is selected.

The idea is that when playing a song at startup (as in your case when using "Open with"), the RawApi functions aren't triggered.

After the fix from Steven, GetPCMRawData gets called at startup, but GetStreamInformation doesn't (this is status quo). And here is the trick - you have to double click the file and start playing again. Then you are all right.

Maybe this helps narrowing down the issue.

[sveakul]

Yeah.  Like I said, no bug, the circumstance I was using was not "standard" anyway, but the sudden dead spectrum and VUMeter just surprised me.  Obviously part of the Raw API design as you say.

[kamen]

Version 6.0 is uploaded

Not many visible changes for the new version, but many small things here and there.

Most important changes are:

-visualzation/audio synchronization for various combinations of settings, outputs and devices
-added a peak meter tool, with input(without preAmp and replayGain applied) and output peak levels visulalization
-usability adjustments to the colors dialog and added "fit lines" bar size option (mainly useful for linear and note scales)



In the end, the peak meter turned out a bit more on the simplistic side. It supports levels of up to +12dbfs (if the extension option from scale menu is enabled). It also adapts to the selected channels mode (similar to the waveform tool).

Greetings to TF3RDL for his part in the brainstorming phase of the peak meter with his mock and to all who may use the end result in CEN.

[sveakul]

Thanks kamen (and TF3RDL!)!

The peakmeter is cool, but I'm not sure how its readings compare to "peak" readings on a VU meter.  I understand that a VUMeter normally measures average loudness (RMS) not peaks but there are also separate LEDs on some for "Peak."  The screenshot below shows MB with a calbrated VUMeter set to have its LEDs follow Peak readings, and they are quite different than the CEN peak meter in the same shot.  CEN is using RawAPI source).

[kamen]

The confusion with the readings is understandable.

The VuMeter uses VU units, while CEN uses DB units for the spectrum(above) and DBFS units for the peak meter scale(below). A good explanation is here.

Independet of the units, there could be peak and rms measurements.

CEN's peak meter shows at the moment only the momentary peaks. RMS is also calculated, but not visualized for the moment, as I didn't find it especially useful.

[TF3RDL]

While I'm not active at MB forum as used to be at the moment, I've updated my own filter-bank spectrum analyzer project to display one channel at the time and Mid (L+R) or Side (L-R) part of a user-selectable channel pair (which should allow M/S of front vs back channels or even center and LFEs)

BTW, is stereo balance/phase correlation spectrum useful? Like this one

[TF3RDL]

CEN's peak meter shows at the moment only the momentary peaks. RMS is also calculated, but not visualized for the moment, as I didn't find it especially useful.

Displaying peak and RMS measurements at the same time probably gives you a very rough idea of a dynamic range thing in music that a peak-only measurement doesn't