"Show Clipping" Options

[Gale Andrews]

I think that providing preference settings is only justified in cases where the default settings work well for one job and badly for another.

Which is possibly the case at least for the playback meter if its trigger level was reduced to one clipped sample.

Gale

[Gale Andrews]

I see no little value in reducing the opacity of the red lines or making them dotted. It's a warning, we need to see it. Opacity won't be so helpful when you are looking at Show Clipping zoomed in to sample level. Dotted could be confusing.

The meter clip indicator always comes on as soon as one sample touches or exceeds 0 dB. If only one sample does that, then the clip indicator will automatically reset after a couple of seconds and go off again.
In more than "N" sequential samples do that, then the meter clip indicator latches ON.

So this is the price of making the meters clip on one sample as well as the waveform? I think that's a good suggestion as long as there is a preference for once a single sample is clipped, the clip light stays on. I for one do not want fading of the clipping light for transient clipping.

And if we do that, I still feel it is important that the clipping light is made properly visible. Its visibility has been *considerably* reduced by the gradient meters.

We have Meter Preferences so I see no reason not to add an option to change the default number of samples threshold to > 1 (for the meters only).

For a serious overload condition, the flashers will pulse at you trying really hard to get your attention which, correct me, is their job

I don't mind if it does, but there is probably someone out there who won't like that.

the "bug" is the timeline red bars that appear with no associated sound damage.

The number of samples threshold for Show Clipping that displays audible clipping will vary with the material, the playback equipment and the listener.

The changes above for the meter might even let you use the meters to detect "audible" clipping and not use Show Clipping at all.

If you still want to use Show Clipping to display "audible" clipping then we need to add a separate threshold setting for that too.

I see the main "bug" that the threshold for number of clipped samples is not the same for Show Clipping and Meters.


Gale

[steve]

I see no little value in reducing the opacity of the red lines or making them dotted.

This is what a fairly typical (poor quality) downloaded from the Internet MP3 of "popular music" looks like:

firsttrack002.png (10.31 KiB) Viewed 1136 times

With reduced opacity, you can't see much more at this zoom level, but you do see more information - in particular you can see the rms level.

firsttrack004.png (13.27 KiB) Viewed 1136 times

As far as "warning" is concerned - I'd still be worried if my recording looked like the latter.

[steve]

I think that's a good suggestion as long as there is a preference for once a single sample is clipped, the clip light stays on.

I wrote:
"In more than "N" sequential samples do that, then the meter clip indicator latches ON.

The only thing that then requires configuring is the size of "N", which should be easy to get at because, as Edgar describes, the user may want to change it according to the job.
As this only affect the meters, I think it should be in the meter dropdown > preferences."

So to make the clip light stay on at one single sample, set "N=0".
"1" is greater than "0", so the light comes on and stays on from the first sample >= 0 dB.

[kozikowski]

I for one do not want fading of the clipping light for transient clipping.

All the objections are covered in the options panel. If you want the light to latch, select latch. I don't. I have never stopped a recording for one transient clip.

The only discussion is how to define the default sound damage — the "three samples over" example in my illustration. That will have to cover the majority of users expectations — and match between the two tools.

Remembering that the original posting was someone who was looking for sound damage by the presence of the red bars... and didn't find any because of the bug. Sending a user off to separate tools to reveal The Truth is just bad form. Moving the sample points at will is covered in the illustration.

Personally, I really like the idea of the flasher which comes on and fades away as you describe. We had an ancient board in the stable which did that and when I was using it for onstage monitor mixing (think of sitting in the pitch black darkness of the wings) it always caught my eye.

I don't mind if it does, but there is probably someone out there who won't like that.

Let me know when cards and letters pour in. No hardware I know latches on and makes the operator take some action to reset it. That's why I picked latching as a User Option.

Anybody spend quality time with a BBC PPM? The meter that peaks quickly to a carefully designed value and then slowly returns to the rest mark? Use that return rate for the flasher fade since it's already an established standard. I looked it up some time back; I have it here somewhere.

The number of samples threshold for Show Clipping that displays audible clipping will vary with the material, the playback equipment and the listener.

The Default Flasher indicates a probable failure of the digital system to follow the work.

The quality of the violin varnish is irrelevant. If it is required that the two tools activate at different levels because of specific user conditions, that's covered in the illustration.

The discussion thread is batting around a tiny fraction of a dB. The current red bars have the bad form to actually lie about the sound damage — and they get caught at it. I don't think that's a good thing and neither did the original poster.

Koz

[steve]

The Default Flasher indicates a probable failure of the digital system to follow the work.

and of course, if there is no audible damage, then that is a bug in your new tool - same goes for the light not coming on when there is audible damage. I look forward to see how you calibrate the bugs out.

[Gale Andrews]

I see no little value in reducing the opacity of the red lines or making them dotted.

This is what a fairly typical (poor quality) downloaded from the Internet MP3 of "popular music" looks like:

firsttrack002.png

With reduced opacity, you can't see much more at this zoom level, but you do see more information - in particular you can see the rms level.

firsttrack004.png

As far as "warning" is concerned - I'd still be worried if my recording looked like the latter.

I agree it "looks" better, but if we make a more opaque warning the default, what do users used to the harder slap in older Audacity think? Perhaps they think the clipping is not so bad, so they can ignore it?

In any case, transparency is off the agenda due to technical difficulties, as I understand it.

Are we getting any closer to agreeing that the meters and Show Clipping should default to clipping at the same level?


Gale

[Gale Andrews]

I think that's a good suggestion as long as there is a preference for once a single sample is clipped, the clip light stays on.

I wrote:
"In more than "N" sequential samples do that, then the meter clip indicator latches ON.

The only thing that then requires configuring is the size of "N", which should be easy to get at because, as Edgar describes, the user may want to change it according to the job.
As this only affect the meters, I think it should be in the meter dropdown > preferences."

So to make the clip light stay on at one single sample, set "N=0".
"1" is greater than "0", so the light comes on and stays on from the first sample >= 0 dB.

OK, sorry I did not know what Koz meant by a "Latch Flash" in his mockup. That did not sound to me like something that *stays* on as our current meter does.

If clip indicator fading is default (I hope not) then we'll need some clearer term than "Latch Flash".


Gale

[Gale Andrews]

The only discussion is how to define the default sound damage — the "three samples over" example in my illustration. That will have to cover the majority of users expectations — and match between the two tools.

Remembering that the original posting was someone who was looking for sound damage by the presence of the red bars... and didn't find any because of the bug. Sending a user off to separate tools to reveal The Truth is just bad form. Moving the sample points at will is covered in the illustration.

I think we are agreed the threshold for meters and Show Clipping should match by default.

If the number of samples threshold control also affects Show Clipping, then the control must be in Preferences.

If we let do let user separately configure threshold for meters and Show Clipping as in your mockup, there may be a case for those controls to be in Meter Preferences and View Menu and not Preferences.

I see no need compelling need for the thresholds to match if the user chooses a non-default setting.
It would probably be less confusing for users of Show Clipping if that stayed at unconfigurable one sample threshold, and only the meter threshold could change. If we decide Show Clipping is configurable I think it *should* be separately configurable from the meters threshold.

Personally, I really like the idea of the flasher which comes on and fades away as you describe. We had an ancient board in the stable which did that and when I was using it for onstage monitor mixing (think of sitting in the pitch black darkness of the wings) it always caught my eye.

I don't mind if it does, but there is probably someone out there who won't like that.

Let me know when cards and letters pour in. No hardware I know latches on and makes the operator take some action to reset it.

You're mixing quotes. I was talking about the pulsing flash that you asked for (as opposed to the clipping indicator showing a steady light then fading or not).

The Default Flasher indicates a probable failure of the digital system to follow the work.

I think Steve is saying that can always fail both ways, so no default of 3, 4 or 5 samples will do what you want infallibly. I agree with Steve if so.

I'm tending to think the meters should indeed default to clip (initially) on one sample. Do you object to that if the clip indicator has an option to fade?

Gale

[steve]

In any case, transparency is off the agenda due to technical difficulties, as I understand it.

The suggestion was an aside, so we can leave that out of this discussion.

the "bug" is the timeline red bars that appear with no associated sound damage.

It would be a much worse "bug" if there was audible sound damage but the red lines failed to indicate it.

The red lines do not indicate "there is sound damage".
On my old 1/4" tape machine I can happily drive it "into the red" with no obvious sound damage (perhaps a bit of "tape saturation" if I drive it too hard, but no clipping).
The wavetrack red line indicates that the sample value is >= 0 dB.
The red "peak" light on my 1/4" tape machine indicates that the signal level is >= 0 dB. The response of the peak indicator light is not "instant" (a limitation of old analogue electronics) but is very fast.
In neither case does red light = audible damage.
If our documentation suggests that "red lines = obvious sound damage", then that is a "bug" in the documentation.

In my opinion, indicating audible damage is impossible to do accurately. What is "obvious sound damage" when listening through my studio headphones may well not be "obvious sound damage" when listening on my laptop speakers, or even on my studio monitors. Some users will even claim that sample value >= 0 dB is too late and that audio is "clipped" when the reconstructed (inter-sample) waveform exceeds 0 dB, so that "clipping" should be indicated for sample values substantially lower than 0 dB.

Are we getting any closer to agreeing that the meters and Show Clipping should default to clipping at the same level?

Probably not

Personally I see no great need to change the current behaviour - I do see the reasons for why we might want to change the behaviour, but no pressing need to do so.

The meters show instantaneous peak level (the moving colour bar) and have a "peak indicator" that latches on when 4 samples in a row >= 0 dB are seen. I think this is a reasonable indication of when "obvious clipping" is likely to have occurred.

The lines on the waveform indicate where sample value are >= 0 dB. I think that many users (including myself) want to be able to see where a single sample reaches 0 dB.
The confusion here seems to be the idea that a "red line = obvious sound damage", which is an incorrect interpretation of what the red lines indicate.