Overlapping corrections

[Ray Bell]

Derek

Some time ago you fixed a bug with overlapping (or very close) corrections

Version 3.1 release notes say:-
"When clusters of click corrections overlap, they are now processed in the order of their severity ranking. This overcomes a problem when editing corrections in such circumstances. In previous versions, edits were not always preserved after the 'Apply' button was selected. "

This effect seems to have returned in the latest 3.2 version. I noticed last night while manually trimming some corrections that the corrected waveform sometimes changed shape when the apply button was clicked. This appeared once again to happen when there were several corrections in close proximity. I could probably find a sample which showed this effect if that would help you fix it.

I wasn't sure where to post this, It's not really a support issue but strictly doesn't fit in beta comments either.

Ray

[Derek]

v3.2 should act the same as v3.1 in this regard. So yes, it would be helpful if you could send me a sample so I can investigate this.

[Glenn]

In previous versions, edits were not always preserved after the 'Apply' button was selected. "

This effect seems to have returned in the latest 3.2 version.

I've noticed this too, but it's been there since v3.1 for me. I always figured there were too many corrections in too small an area. My fix has been to remove a number of the smaller corrections. This usually does the trick. I've noticed also that too many splices placed too closely together don't always play back correctly, at least on my machine. The saved wave file seems to reflect what's in the corrections list, and that is what's important. While I'm at it, I've never been able to get the channel dependent corrections to work as advertised either, but this has always been a minor issue when reflected against the greater good. I have Windows 98; considering it's an old operating system, I don't bother too many people with it.
Glenn

[citguy]

I've never been able to get the channel dependent corrections to work

Glen. Do you mean you can select a channel and perform a correction on that channel only? I've tried but not been successful. Or do you mean something else?

Stan

[Glenn]

Hi Stan,
Yep, that's exactly what I mean. I don't know if I'm reading the help file correctly or not, but it doesn't bother me either way. The program works pretty well the way it is.
Glenn

[Derek]

Thanks Ray for the samples. They should help to track this down.

Regarding the channel dependent corrections, you can't globally make Wave Corrector only correct one channel. The stereo command is designed for the small number of cases where a click is present on one channel but not the other. In such a case, select it's correction and then select Stereo -Left or Right channel. This removes the correction from the channel where it's not wanted.

[Glenn]

Regarding the channel dependent corrections, you can't globally make Wave Corrector only correct one channel. The stereo command is designed for the small number of cases where a click is present on one channel but not the other. In such a case, select it's correction and then select Stereo -Left or Right channel. This removes the correction from the channel where it's not wanted.

Hi Derek,
Now that I'm clear on how it works, there are still places where this feature won't work properly on my system. When the channel dependant check-mark is removed, the correction is not updated to reflect this. Sometimes it works, and sometimes it doesn't. The most success I've had has been by removing the correction altogether, rescanning the click, then removing the one channel. But this is still hit or miss. Could it be that the programme 'sees' a bonafide click, however inaudible or otherwise unapparent it is, and is unable to remove it?
Glenn

[Derek]

Regarding the channel dependent corrections, you can't globally make Wave Corrector only correct one channel. The stereo command is designed for the small number of cases where a click is present on one channel but not the other. In such a case, select it's correction and then select Stereo -Left or Right channel. This removes the correction from the channel where it's not wanted.

Hi Derek,
There are still places where this feature won't work properly on my system. I've always wondered why, when the menu check-mark is removed, the correction is not updated to reflect this. Sometimes it does, and sometimes it doesn't. Any ideas?
Glenn

Are your sure it's not just because there is another correction very close by whose effect is swamping the one you're trying to remove. If you delete the correction entirely (instead of just on one channel), does the program behave as expected?

[Glenn]

If you delete the correction entirely (instead of just on one channel), does the program behave as expected?

If you mean 'does the program remove the correction entirely', -yes it does. This only occurs when I attempt to remove one channel only.
Glenn
PS I see we're both online at the moment; you may wan't to review my orginal post as there is a good chance it has changed (damned edit button!)

[Derek]

It does sound as though there must be a bug. Is there any chance of you getting this to happen consistently and then sending me a sample illustrating where it happens?

[Glenn]

Hi Derek, thanks for your interest in this.
I opened one of my later session files and scanned through my user corrections to find any instances that would help. It didn't take long to find one but I've discovered that the size of the correction makes a difference as to whether it can be split up. I use the narrow corrections by default, and when I set a correction at minimum width, the stereo selection doesn't appear to work. Interestingly, if I then resize the correction to a larger size, the trace will disappear indicating the correction has been removed from that channel. I played around with this a little while and tried to break the programme but was unable to.
I'd say in 99% of all cases where channel separation might be desirable, the resulting distortion is negligible. That said, I have experienced instances where a correction misapplied to a channel has caused noise in that channel, often where the waveform is uniform and symetrical. I can't immediately recall where these have occurred (and finding one now could possibly take hours or even days), but the next time I come across one I'll send it to you.
Glenn

[Derek]

Yes, it does appear there is a bug when the correction is at minimum width. I'll try to fix this in the next release but in the meantime, it is not going to cause anything more than a minuscule amount of additional distortion.

If you come across a more significant example, please let me know.