Possible DC Offset Issue

[billp]

I'm definitely out of my depth here.

One of my collaborators told me that a wav that I sent him was reported by Logic Pro to have a DC offset problem. This project started with an AIFF generated by Logic that he sent me, and which I imported into CbB.

After digging into this a bit, I can see a very small discrepancy between the original AIFF file as imported into Audacity, which looks correct, versus the same file imported into CbB, where there is a slight negative(?) offset from the center line on the right side of the stereo image. The first two screenshots below are of the right side of the stereo image, one from Audacity and one from CbB. The left side does not show this same characteristic.

The third screenshot is of the wave file that I sent him, which I reimported into CbB. It has a more dramatic offset issue and on both sides of the stereo image.

I tried various things, like changing the import bit depth to 32, but it did not change the outcome. I also tried using the DC Offset correction tool in CbB, but it didn't have any effect on the original AIFF. It DID have an effect on the reimported wav. It corrected the DC offset on the left side, but added that offset to the right, so that the right side was even more offset.

My projects are all 24-bit/48K, if that makes a difference. My audio interface is a BF Pro. I'm using the latest rev of CbB.

I'd really appreciate it if anyone has some insight into this. I had to expand the screenshots to see the detail.

 

 

 

[Will.]

Try cutting away bass frequencies by making slightly more headroom to correct the issue. Do this with the smallest slope curve your EQ has - around 10-30Hz. Do the same in with the highend side with the smallest slope curve and what i mean by smallest -  lean towards the linear line, as apposed to a brickwall cutoff. 

[billp]

Thanks a lot Will. I'll try that. 

[Glenn Stanton]

there is also this: Cakewalk - Cakewalk Documentation - Removing DC offset https://www.cakewalk.com/Documentation?product=Cakewalk&language=3&help=EditingAudio.22.html

    

[Will.]

I just ran into this topic that's currently open as well. 

https://discuss.cakewalk.com/index.php?/topic/49435-audio-track-view-asymmetry/

Edited September 19, 2022 by Will.

[billp]

Thanks Glenn. I did find this and eventually got it to work. My main concern was that an audio file that imported into Audacity (which I don't normally use), came in with no DC offset, but the same file imported into CbB had DC offset in the imported version.

I've since discovered that Ozone has a built-in filter for this. I haven't yet verified that it actually corrects it.

I'm actually starting to think that this could be a display issue. Check out the low res and hi res screenshots below. At hi res, the offset is almost not visible, but at low res there is a noticeable offset.

Low Res                    High Res

[Glenn Stanton]

i would try out the Ozone to remove it (in the past when preprocessing files i would use Sound Forge to kill the DC offset and set levels etc). now i mainly use RX-7 to perform my preprocessing on files.

[sjoens]

FWIW, There are areas where Sonar/CbB has trouble displaying things graphically accurate, but I don't see that happening here.

Edited September 19, 2022 by sjoens

[Byron Dickens]

My suggestion is to listen to it and see how it sounds before chasing ghosts.

[billp]

15 minutes ago, Glenn Stanton said:

i would try out the Ozone to remove it (in the past when preprocessing files i would use Sound Forge to kill the DC offset and set levels etc). now i mainly use RX-7 to perform my preprocessing on files.

I have Ozone and RX 10. I ran it through Ozone just now and it doesn't detect an offset. It appears to be doing essentially what Will suggested above, i.e., applying a filter.

33 minutes ago, sjoens said:

FWIW, There are areas where Sonar/CbB has trouble displaying things graphically accurate, but I don't see that happening here.

Don't know if you can see it in the video, but I think you are right about the graphic accuracy. You have to watch this fully expanded to see the detail. Visually, the offset at the beginning doesn't change while the resolution goes between low and high. I think it should be magnifying the deviation from the center line when it goes to high res. Instead it just varies somewhat randomly. It's pretty small in any case and I suspect that only my dog is going to detect it.