Page 2 of 3

Re: Vectorscope or equivalent

Posted: Wed Jan 29, 2014 6:19 am
by Robert J. H.
Thanks again, it seems to work.
I am not entirely certain what I have to do with the result...
A stereo width of 676.0 % is rather high, isn't it?
(after Pseudo Stereo and two times Channel Mixer --> extra wide)
However, it might be a good indicater how much the volume decreases when played over a mono device.
The dB values for the above sounds are after mixdown to mono:
-1.9 (Chirp)
-0.2 (after PS)
-8.2 (after CM)
-16.4 (after CM)
I assume that 100 % is equal to extending from one speaker to the other.
That's also the value that I get for uncorrelated white noise.
The width can somewhat safely be extended to about 120 %. 200 % is usually regarded as the maximum. It depends on the content though.

Re: Vectorscope or equivalent

Posted: Wed Jan 29, 2014 6:46 am
by kozikowski
OK, here's an artist representation of a diagonal scope showing "footroom" problems. The negative swing of the waves have become damaged. I'm trying to remember where I saw that and I think it was a misbehaving multiplex radio transmission system. If you were clever with a regular scope you could see the individual waves start oscillating on their negative cycles -- but you had to be a deity with a scope to see it; it wasn't there all the time. On selected loud audio peaks, the lower vectorscope display would break into ghostly loops, and all you had to do was wait for the Coca-Cola commercial.

"I'd like to teach the world to sing."

"There! See it?"

So, no, I don't think a single numeric value is going to do it.

Koz

Re: Vectorscope or equivalent

Posted: Wed Jan 29, 2014 3:01 pm
by steve
I must say Koz, I've never seen a jellyfish with a line and a loop at one end. If I did, then I'd have no idea what the fault was causing it. What it tells me is that the negative going response in the left channel is different to that in the right channel. If I'd used my recommended practice of using a mono track for mono recordings, then it shows there is an equipment fault. If the jellyfish is incorporated into the recording software (Audacity) then it tells me that the jellyfish is broken.

So, you get some weird response in the Jellyfish and call out the engineers and local deities to fix the problem. How does that help Audacity users?

One thing that I do agree about, is that jellyfish look really cool, but only when they are wiggling.

If the device is providing vital information, then how do we make that information available for visually impaired users?

Historically, the audio vectorscope was a natural progression of the CRT oscilloscope. "What happens if I plug the other channel in here...?" It pre-dates the television, and is often more entertaining.

It is not uncommon to see analogue instruments transit to the digital world. You can get pitch tuners that show a picture on an analogue meter on an LCD display, though these days tuners that show the pitch in plain English "A#" with LEDs to indicate "A bit sharp", "A bit flat", "Spot on", are more common and easier to use. I agree that a Vectorscope shows a lot of information at a glance, but only if the user has the knowledge and experience to interpret that glance. That'll be about 0.0001% of users. For the rest, it may be nice to watch that funny little wiggly thing, but it would probably be more useful to have the information spelled out in a form that does not require consulting the Oracle and casting bones. What information do you want?

Re: Vectorscope or equivalent

Posted: Thu Jan 30, 2014 11:42 pm
by kozikowski
I must say Koz, I've never seen a jellyfish with a line and a loop at one end.
Admittedly, that's an oddity, but you don't start out designing test equipment making it as restrictive as possible. There are exceptions, like making your Fluke test voltmeter not respond to radio waves. That's a pretty obvious desirable restriction.

I bet you've seen one with straight up-and-down or side-to-side segments from each tip. One channel clipping.
If I did, then I'd have no idea what the fault was causing it.
But you would certainly know there was something off. So the test succeeded.
What it tells me is that the negative going response in the left channel is different to that in the right channel.
It tells me a lot more than that. There is a very unusual error that only displays on some negative peaks.
If I'd used my recommended practice of using a mono track for mono recordings, then it shows there is an equipment fault.
Or more broadly, a system equipment fault which is what happened in this case.
If the jellyfish is incorporated into the recording software (Audacity) then it tells me that the jellyfish is broken.
It tells me to look at where the show came from, soundcard problems, etc, and more importantly, that the show probably arrived as two-track mono, not "real" mono.

When you're troubleshooting from nine time zones away, every little helps.
So, you get some weird response in the Jellyfish and call out the engineers and local deities to fix the problem. How does that help Audacity users?
It's not for the Audacity Users. Joe and Jane McUser have no idea what those Audacity blue lines are and what those red and green flashing lights on top do.
One thing that I do agree about, is that jellyfish look really cool, but only when they are wiggling.
Perfectly true, particularly in well-separated stereo. I did a very impressive Dog and Pony once by making a 21"TV do that full screen. I'd be stunned if there wasn't at least one Dr. Who with that in it.

However they maintain 80% of their diagnostic information if you freeze one. And it's fast. Good INFO density
If the device is providing vital information, then how do we make that information available for visually impaired users?
Useful, not vital. The same way we do it with the Spectrogram View?
Historically, the audio vectorscope was a natural progression of the CRT oscilloscope. "What happens if I plug the other channel in here...?" It pre-dates the television, and is often more entertaining.
I've often said I can get more entertainment value out of the back of a TV than the front.
it may be nice to watch that funny little wiggly thing, but it would probably be more useful to have the information spelled out in a form that does not require consulting the Oracle and casting bones. What information do you want?
No way to predict. Both? I don't see how you're going to avoid the My Eyes Glaze Over problem, but it's worth a shot. Until then, run this tool and attach a screen grab.

Koz

Re: Vectorscope or equivalent

Posted: Fri Jan 31, 2014 12:49 am
by steve
Koz, would you post that link to your mono/stereo/phase sample.

Re: Vectorscope or equivalent

Posted: Sat Feb 01, 2014 6:24 am
by kozikowski
How much is it worth to you?

http://www.kozco.com/tech/LRMonoPhase4.wav

Koz

Re: Vectorscope or equivalent

Posted: Sat Feb 01, 2014 3:10 pm
by steve
Is this what you want?

http://youtu.be/ZG_Vu32c0ZA

Re: Vectorscope or equivalent

Posted: Sun Feb 02, 2014 1:54 am
by kozikowski
Yes. I've never used one with a bar-graph on the bottom. What does that do besides get very upset when I went out of phase.

I think we've been here before. It's a lovely tool but it only works in Windows 2000 and Red Hat Linux before 1999.

I wonder if it has provision to capture a still without having to be really talented with the PrtSc key. Anyone diagnosing from LA is not going to be able to see the motion without the clip.

But still. If there is a clip.....

Koz

Re: Vectorscope or equivalent

Posted: Sun Feb 02, 2014 1:57 am
by kozikowski
And in a note from Hollywood, Everybody Knows radar has to go round and round, newsrooms have to sound like teletype machines and scopes have to be green. It is written.

Koz

Re: Vectorscope or equivalent

Posted: Sun Feb 02, 2014 1:03 pm
by steve
kozikowski wrote:I think we've been here before. It's a lovely tool but it only works in Windows 2000 and Red Hat Linux before 1999.
Works with any version of Linux using Jack Audio System.
Would presumably work with any computer using Jack Audio System provided that it was appropriately compiled.

The requirement for Jackd (Jack Audio System) is to allow the audio to be routed from Audacity to the sound card via the Vectorscope. For Windows to do that (without Jackd) you would probably need Virtual Audio Cable or similar. I don't know if Mac OS X is able to route audio signals between applications.

If the Vectorscope was an Audacity module, or built into the core code, then it could capture the audio signal directly, as the audio meters do. In fact the Vectorscpe could be an option in the audio meter drop down menu (http://manual.audacityteam.org/o/man/me ... .html#menu)
kozikowski wrote:I wonder if it has provision to capture a still without having to be really talented with the PrtSc key.
Help > Screenshot Tools > Capture Full Window:
fullwindow001.png
fullwindow001.png (134.22 KiB) Viewed 3206 times
kozikowski wrote: I've never used one with a bar-graph on the bottom. What does that do besides get very upset when I went out of phase.
Pretty useful. It shows phase correlation. The more "green", the better the mono compatibility (maximum green shows that it IS mono). The more red, the worse to mono compatibility (maximum red shows that it is 180 degrees out of phase and will produce silence if mixed to mono). Given that some of the audience will probably be listening on mono playback systems, red in that bar meter flags a warning. The more red (both time in the red and the amount of red) reflects the severity of the warning.

Also very useful for the use case that you gave previously. If consistently close to maximum (green or red) the Vocal Removal effect will be useless.
kozikowski wrote: And in a note from Hollywood, Everybody Knows radar has to go round and round, newsrooms have to sound like teletype machines and scopes have to be green. It is written.
but of course :D