I’ve noted that with some filters, you get a different result when applied to a single track individually versus when the same filter is applied to multiple tracks simultaneously. And with some filters, it doesn’t matter if the filter is applied to one track or multiple selected tracks, because you always get the same results.
For those filters where you get different results depending on whether a single track is selected or multiple tracks are selected, the first track always gets the same results and then the later tracks get different results.
It looks like with these particular filters, the result on the next track is altered by what is done to the previous tracks, which is undesirable to me. I would like to be able to select multiple tracks, apply a filter and get the same results regardless of track order or position, but with some filters this is not happening.
Is this something that is well known to occur and is there any way to avoid this behavior?
I’ve tried a few that have done this, but the most recent was the Melda MSpectralDynamics. Note that I’m using version 3.3.3, not the most recent.
I didn’t do this test by ear but rather looked at the results in a hex editor and did a side by side comparison. I ran the filter on two tracks simultaneously, then I ran the filter on each track individually. Then I compared the results side by side. When the filter was run on the two tracks simultaneously, I compared the results of the one on the track at the top with the one run individually and they were identical.
Then I took the one on the bottom track and compared it to the one individually, and they were totally different.
So, when the filter is run on multiple tracks simultaneously, only the top track gets identical results to when it is run individually. Everything below the top track has different results, based on what is shown in a hex editor. So, it seems that the results are affected by track position.
I’ve done this with many other filters where the results are identical regardless of track position.
It sounds like a bug, but no idea if it is a bug in the plugin, or in your old version of Audacity, or just a compatibility issue. Do Melda say that Audacity is a supported host? (I’m not on Windows so I can’t test it).
I don’t see it listed as supported, however it works fine. The fact is that I don’t even know if running it simultaneously on multiple tracks actually sounds different. All I can say is that the actual data looks different. Sometimes, the data outputted can look very different and no one can discern any difference in the sound. So, I can’t necessarily say it’s a bug. It’s just more for verification that it ran correctly. I want the data output to look the same.
Some filters produce different outputs any time you reset them. Meaning that every time I reset the filter then apply the exact same settings, the output looks totally different, even if the sound might be the same. They might be introducing some sort of randomized data into the audio.
My question is with the way Audacity works, is the filter actually aware of what track it’s being run on or are they all run identically. Let’s say a filter is run on track 4. Is the filter able to access the data from prior tracks and modify its output accordingly or is each track run completely independently of the others? Because it seems the filter, when run on multiple tracks simultaneously is modifying its output based on either the track number or the data in prior tracks.