Suggestion: Repair Limits

[WeirDave]

It would really be helpful if the repair function would have an "automatic selection cursor" if you will to auto select a range of 128 samples if you'd like, so that I can take the guess work out of how much more I can grab to fix at a time.



WeirDave

[kozikowski]

What does it do now if you "run out."
Koz

[steve]

It would really be helpful if the repair function would have an "automatic selection cursor" if you will to auto select a range of 128 samples if you'd like, so that I can take the guess work out of how much more I can grab to fix at a time.

How about a slight variation on that.... To use the repair function, you don't select anything, you just click in the centre of the part that you want to repair and the effect acts on 64 samples either side of the cursor position?

[Gale Andrews]

It would really be helpful if the repair function would have an "automatic selection cursor" if you will to auto select a range of 128 samples if you'd like, so that I can take the guess work out of how much more I can grab to fix at a time.

How about a slight variation on that.... To use the repair function, you don't select anything, you just click in the centre of the part that you want to repair and the effect acts on 64 samples either side of the cursor position?

Moved to "Adding Features".

Repair would still have to do "something" with a selection. All users who read the fine Manual will know they must select audio to edit it, so would probably still select audio for Repair.

Might the subject idea encourage sloppy editing unless it retained the requirement to be zoomed in? If you were allowed to place the cursor or selection while zoomed out you will likely not have the selection starting at or centred on the "correct" 128 samples.

I don't think it's a bad idea but IMO the better aim is to have Repair working on longer selections and detecting and repairing the clipped samples as it goes along. There was a useful thread about that: http://forum.audacityteam.org/viewtopic ... 56&t=72712 .


Gale

[steve]

Might the subject idea encourage sloppy editing unless it retained the requirement to be zoomed in?

While "Repair" can only act on up to 128 samples I think it should retain that requirement. I don't see any way that a user could select accurately enough without being zoomed in very close.

Repair would still have to do "something" with a selection.

It could bring up a message box telling them how to use the effect.

IMO the better aim is to have Repair working on longer selections

There's clearly a lot of demand for that.
If it could be made to work effectively then that would be great.

and detecting and repairing the clipped samples as it goes along.

I'm not sure about that.
Detecting clicks that should not be present without falsely detecting clicks that should be present is very difficult. I hear that Brian Davies' Click Repair does this very well, but even that cannot hope to be 100% accurate. A click that needs to be repaired can be identical to a click that should not be there, with only the musical context differentiating between them.

There is demand for a good "detect and repair" effect, (like "Click Removal" but better), but I think there is also a need for a "dumb effect" that "repairs" the selection regardless of the content of the selection.

[waxcylinder]

I would NOT want Repair to always use 128 samples, or to auto select.

I think (from experience) that you need to be well zoomed in to make effective repairs. There are times when I want the full 128 sample range and others when a much shorter selection will give a more "accurate" repair.

Top-tip: Set your time format in the Audacity Selection Toolbar (normally at bottom right of window) to hh:mm:ss + samples that way you can see how many samples you are selecting.

WC

[steve]

There are times when I want the full 128 sample range and others when a much shorter selection will give a more "accurate" repair.

I don't often use "Repair" but I can well imagine that could be the case, for example if there is a click very close to a transient peak. So how about

If there is no selection: 64 samples either side of the cursor will be processed. If the selection is within 64 samples of the start or end of the audio clip (so that there are not 64 samples available), then as many samples as are available will be used.

If there is a selection and it is no more than 128 samples: the selection will be processed (as now).

If there is a selection > 128 samples: An error message is shown (as now).

The only problem that I see with this is that there is a potential conflict with the preference option to "Select all if no audio selected". Since, with "Repair", it is very likely that selecting all audio will result in an error, I don't think that it is unreasonable for "Repair" to be a special case.

Making "Repair" as special case, would become a little problematic if Repair was extended to be able to handle long selections. In this case there would be no need for an easy way to act on 128 samples and the Repair effect could behave in the same way as all other effects.

[waxcylinder]

That could be workable

Peter

[waxcylinder]

The only problem that I see with this is that there is a potential conflict with the preference option to "Select all if no audio selected". Since, with "Repair", it is very likely that selecting all audio will result in an error, I don't think that it is unreasonable for "Repair" to be a special case.

If all audio is selected and and the selection is greater than 128 samples then the repair should not take place

Making "Repair" as special case, would become a little problematic if Repair was extended to be able to handle long selections. In this case there would be no need for an easy way to act on 128 samples and the Repair effect could behave in the same way as all other effects.

My be a moot point as this is unlikely, I seem to recall that lengthening the repair selection exponentially increases the processing time

Peter

[steve]

I seem to recall that lengthening the repair selection exponentially increases the processing time

Yes - for selections greater than 128 samples the repair time rises very rapidly. However, there was a suggestion that larger sections could be repaired as a sequence of 128 sample length blocks. Initial tests indicated that this might work, but as yet it has not progressed much more than that.