Anyone Else Notice Weird Distortion Issues with the Analog Delay?

[Craig Anderton]

I haven't found a recipe to reproduce, so I can't submit a bug report. But I wanted to check if I'm the only person experiencing this issue. Over time, the Analog Delay starts to distort. It almost sounds like there's a sample rate or block size problem, but that's not it. This becomes noticeable when set to delayed sound only. If the Mix control is set to 20-40% or so, the dry sound masks the distortion.

I use the Analog Delay a lot with vocals for a different kind of doubling effect so over the course of multiple vocals tracks, this distortion becomes obvious. Again, this problem seems to accumulate over time. Before exporting or rendering, I turn off all Analog Delays and turn them back on again. This ends the problem temporarliy.

This seems to have become an issue only since the Analog Delay sprouted a Resonance control (which I noticed doesn't go down to zero). But that may be coincidence. Is there any way to revert to the older Analog Delay so I can see if that solves the problem?

 

[Gerran]

I haven't noticed but will investigate later today. I did notice something weird sounding yesterday that got me puzzled. Might be the same thing. I'll report back later...

 

[Gerran]

I've been playing along for a while now, but the weird sound from yesterday I can attribute to a user error on my side (tempo synced analogue delay sounds weird when the tempo was mapped from a not so tightly played live performance). I've not been able to notice distorted output from the various instances of analogue delay otherwise.

 

[Ari]

The resonance control has always been there, just not on the surface. As it is a filter circuit, or EQ, you can compare it with a Q value. That's why you can't set it to 0.

It would help if I could hear the error once, in the form of audio or video file. Unfortunately, a link to a blog does not help. A screenshot of the plugin that is currently in the error state could also provide information about a repro.

 

[lokeyfly]

I'll try the Anolog Delay later Craig and report back. I use this delay at times for the same reason you do, ala the Roland Space Echo quality. Of course with a little degraded clarity from trying to mimic its tape like effect. I don't recall distortion or having it increase over the course of multiple layers, but I'll put on some clinical ears later. Thanks for the heads up as thats just as important, as it is verifying.
.....later.

 

[Craig Anderton]

It would help if I could hear the error once, in the form of audio or video file. Unfortunately, a link to a blog does not help. A screenshot of the plugin that is currently in the error state could also provide information about a repro.

Thank you for responding, and explaining what resonance does. Next time the sound happens, I'll bounce the audio. I think that would indeed help you identify potential reasons for why it happens.

I did try something last night. I turned Boost on for all the Feedback controls, even though there was no feedback applied. I made it through an entire lengthy mix without the problem surfacing. However it might be coincidence, given the evanescent nature of this possible bug. I'll keep investigating.

 

[lokeyfly]

so over the course of multiple vocals tracks, this distortion becomes obvious. Again, this problem seems to accumulate over time. Before exporting or rendering,

Perhaps the delay is recycling and building up some artifact from that loop that ends up now part of the delay. Increasing along the way. Weather that's below the 0.30 resonance setting, one would have to hear.
I'd be curious if setting the Sync button alters or removes it.

I turn off all Analog Delays and turn them back on again. This ends the problem temporarliy.

That will clear the anomoly. Or potentially dropping feedback to zero, wait, then return.

This seems to have become an issue only since the Analog Delay sprouted a Resonance control (which I noticed doesn't go down to zero).

Unless I hear some sample of that issue or can otherwise assist, I'll stay out of the comments here Craig.

Noting your last comment about boost, even though no feedback applied. BTW, I ran through a number of checks, also rendered the ano delay with mixdown using it heavily. No issues that I came across.

Keeping the lines open.

 

[Craig Anderton]

After not encountering the issue through several mixes, it happened tonight on a vocal track with analog delay as well as a synth track with analog delay. I was thrilled that I had something I could record, and exported the mix with the problematic vocal track soloed. But, when I played the exported wav file, it did not incorporate the anomalous analog delay problem . Furthermore, the mix played back properly afterward, without my having to do anything other than the export. Somehow, exporting a portion of the track "reset" the analog delays or something. Next time, I'll plug the audio interface output into a separate recording device to capture the audio that I'm hearing. The sound is not something subtle, I can pick it out in a full mix with lots of other tracks happening. I did capture a screen shot of the delay (attached).

I'll keep trying to figure out a way to reproduce the issue, but I don't see any value in posting any more about this unless I have something new to offer. There's no question I'm hearing a problem with the Analog Delay, and it's not a one-off but a (seemingly) randomly recurring issue. However, if it's such a corner case that it can't be isolated and fixed, it may be that simply exporting the file is all I need to do for a temporary fix instead of tracking down each Analog Delay and turning it off and on.

 

[Ari]

You have been asked for audio examples for several days and several postings before and by different people, so that you can classify the error. It would be helpful if you could follow this up so that everyone knows what it is all about.

 

[Craig Anderton]

You have been asked for audio examples for several days and several postings before and by different people, so that you can classify the error. Somehow you ignore this. And nobody knows why.

I thought my previous post was clear about why I have not been able to provide an audio example yet. But based on your response, obviously it wasn't clear enough, so let's try again.

To summarize what I said in the post, it's intermittent, so I have to wait for it to show up (FWIW days, or even more than a week, can elapse between occurrences. And I don't know yet if it's song-dependent, either). Second, when it did show up last night, I exported the track but the exported file did not exhibit the problem. Also, after exporting, the song played back correctly. It's almost like exporting the file solved the problem or reset something. I have no idea why. The problem did not reoccur for the rest of the evening.

So, I said that the next time it happens, I'll record the audio output from my interface with a separate recording device instead of exporting (I've already set up an audio recorder so I'll be ready when it happens.) I also said I'll keep trying to figure out a way to reproduce the issue, but that I didn't see any value in posting any more about this unless I have something new to offer (e.g., an audio example, or a way to reproduce it). The only reason I'm posting this is so you won't think you're being ignored.

I'm not "blaming" Studio One. For all I know it's an audio interface, SPDIF, or Windows interaction problem that's limited to the Analog Delay (although to be fair, I haven't used other delays to see if they're affected as well). I started this thread because I wondered if anyone else had experienced it. My hope was that someone had, and found a solution other than turning each Analog Delay off and on again. Or, that dev was aware of it. It's probably a corner case, which I also mentioned in the previous post. I also may have a workaround, which is to export a mix.

But as I said, I don't see any value in posting about this any more until I have something new to offer. I can't predict when the issue will appear, so I can't predict when I'll be able to record an example. I didn't even realize I would need to record the audio separately from an export until last night, but at least now I have a second recorder ready to go The problem doesn't happen often enough, or consistently enough, to allow the luxury of rigorous testing.

 

[lokeyfly]

Yeah Craig,
First, thanks for clarifying just how difficult reproducing it is.

It was clear right from the subject if others could verify any such weird distortion issues were present. Now that the post is here, any of us can add if it helps identify the issue.

if I see it, I'll capture a short audio clip of it. Then repeat the exact same clip with the non issue coming through.

We could then null it, and find the frequency of anomoly. I'll also record the anolog delay setting as to what they were at the time. Then it can be handed off to Ari & the big guns (musical group?).
They can likely take a deeper dive. Sounds like it takes just the right set of circumstances.
Cheers.

 

[Craig Anderton]

Maybe these artifacts aren't as rare as I thought, and I just didn't notice them until they became objectionable in the dense mixes I'm doing lately. As soon as I started up Studio One today it happened, again with Analog Delay on a vocal track. This anomaly seemed more time-related than distortion-related, although often in the past the time and distortion issues appeared together. That might be a function from the sound seemingly deteriorating more as time goes on. There's evidence that supports this later.

The scenario is:

• One original vocal track and one copy in parallel. Both have the DLX Mic plugin, Studio One Pro EQ3, and Waves RVox compressor (in that order).
• Both have sends to the Rare Signals Transatlantic reverb. Its track is muted in the audio examples.
• The COPIED vocal track has the analog delay (see attached screen shot) set to 1/32nd delay. That track is advanced by 1/32nd on the timeline compared to the original vocal, for the reasons explained in the blog post mentioned earlier.
• To hear the audio, here's a link from my personal OneDrive cloud storage: Analog Delay Problem.zip. I don't like exposing my OneDrive, but I'll leave the link up for a couple days before removing it. If you can't download from the alias, here's the full link: https://1drv.ms/u/c/0e145b2eb674536e/Ee_pbSG-SORIlBkZ9kBy5qsB6Y7xFnXKRpT5myMXn9lpfw?e=LPX3ba

Audio file descriptions:

• With artifact.wav is one of the objectionable artifacts. I couldn't capture the one that involves distortion, so please don't think I'm ignoring anyone who wanted to hear distortion. Often the two artifacts appear together. I don't know why it's only this artifact but I think it might be due to not enough elapsed time, or there haven't been enough changes introduced over the course of mixing. Perhaps something builds up that eventually leads to distortion.
• Turn off_on.wav is the sound immediately after turning the Analog Delay off and turning it on again.
• Turn off_on 1 minute later.wav is the sound after turning the Analog Delay off and turning it on again one minute later. You can hear there's already a very slight change.

I left Studio One in that state, unattended and unmodified, for about 30 minutes while I dealt with a client's phone call. When I returned, I simply hit playback and the With Artifact sound returned. This seemed to support what I thought about the problem getting worse over time. (FWIW Studio One is running on internal sync with an 1824c interface.) Additionally, I bypassed the other plugins and turned off the sends to the reverb, so the tracks were identical except for the one having the Analog Delay. This made no difference, the artifact persevered.

At this point, I'm so over this. I've presented proof. The bug itself has cost me many hours while mixing to turn the delays on and off (I eventually programmed the Komplete Kontrol switches to turn them on and off). Trying to track the issue down, setting up a separate audio recorder, and coming up with evidence so people wouldn't think I was imagining things took several more hours.

Besides, I don't think I have to worry about it anymore, because it seems there's a fix. I replaced half the Analog Delays in the project with the Sonitus:FX Delay and half with the Waves H-Delay to test them. I left the mix alone for 30 minutes, then started mixing again for 30 minutes, and there haven't been any problems with either delay. (I prefer the H-Delay, because its Analog control can add a slight coloration that further differentiates the two vocals.) If I do encounter any problems with the H-Delay, I'll post again but otherwise...peace out.

[Edit] Feel free to mark this solved, since I have a solution. Then if other people have this issue and they search on Solved, they can find how I fixed it.

 

[Ari]

Personally, I prefer a one-liner with a description of the error and an attached pic or audio file. That has to be enough for absolutely every bug.

Many many more words can only be a hindrance in identifying the problem. Also, many things can be irrelevant the more words are written. So please, please reduce problems to a one-liner. And then explain the problem via pic or audio.

 

[Craig Anderton]

The one-line description is: the Analog Delay creates unexpected artifacts that get worse over time. You have an audio file that I was FINALLY able to capture a few hours ago, and a screen shot. So, now you have everything needed to fix the bug.

I promise you I wasn't ignoring anyone, or trying to avoid submitting an audio file. It takes time to gather evidence with intermittent issues. Also, as I said, I don't blame Studio One. In all my consulting experience, companies ask for as many details as possible about context if I'm not sure the problem is with the program itself, but perhaps involves an interaction with something else (e.g., audio interface). I assumed, incorrectly, the same was true with PreSonus and I apologize for making that assumption.

As I said in the thread's first sentence, "I haven't found a recipe to reproduce, so I can't submit a bug report." I simply wanted to know whether other people experienced the issue. I hoped the community could assist me in defining the issue, so I could submit a bug report.

 

[Craig Anderton]

I can now reproduce the bug. It appears only under very specific conditions. This is why I had difficulty producing an audio example on demand:

• Pan two identical mono tracks to the same place in the stereo field (e.g., center).
• One has an Analog Delay with settings similar to previous screenshots, using the same temporal relationship to the other track described previously (this is crucial). If you listen to only the track with the Analog Delay, it won't sound like there's a problem.
• Add a short, major tempo track change (e.g., the "Time Trap" technique described in previous blog posts).
• Create a playback loop that incorporates the tempo change and the two tracks.
• Play the loop. The Analog Delay's timing accuracy deteriorates progressively with more loop iterations. It usually takes at least a dozen passes before the issue builds up enough to be obvious. My guess is that the Analog Delay loses sync with the host over time due to the tempo change.

I apologize in advance for the word count, but the conditions under which the bug occurs requires specificity or it can't be reproduced. Since Ari is already working on a fix, I won't waste the dev's time with a redundant bug report.

Coda: I came up with a different way to do the doubled vocal technique. It's based on phase decorrelation with a paralleled Pro EQ3 stage, so there's no delay needed and it works perfectly with tempo changes. The disadvantage compared to the original tip is that without the Analog Delay, you can't apply modulation for traditional ADT effects. The new FX Chain will be in a future blog post.

 

[Ari]

Very good that there is a reliable repro. Now I'm still missing the demonstrated issue to be able to decide whether it is one at all. May I ask for a song example? Very rudimentary and only a .song file. No audio! Take Mai Tai, with a simple Sawtooth and delay and the tempo changes you mentioned. The example should help you find the problem.

Thanks in advance!
Ari


ps: Just to mention it: To double a voice there is a double mode within the Chorus plug-in. It exists exactly for this task. Just saying.

 

[Craig Anderton]

I can send a file today that shows it, but it's a song with vocals. I'm very busy but I'll work on a version to your specs soon. It will need to have two Mai Tais with identical settings in parallel. When you get it, I recommend setting up the loop and going to have lunch. Or a beer When you come back, you'll hear the issue.

The good news is the problem doesn't render or export. It happens only during real-time playback.

 

[lokeyfly]

When you get it, I recommend setting up the loop and going to have lunch. Or a beer When you come back, you'll hear the issue.

This is the part that has my curiosity. I suspect any regeneration lasting for example, an hour is going to introduce......ehm..... stuff. Anyway, I could not replicate what hadn't been a part of the analog delay in the past. In fact, that's probably the check-point Charlie. What invokes this issue in version 7.2, but doesn't in 6.X.
Thanks Craig in all your efforts. You too, Ari!

 

[Craig Anderton]

That was easier than I thought, now that I know how to reproduce it

Ari, I noticed you can't attach files here so I sent it via email. Hopefully I have the right email, and gmail doesn't mind an 18 MB attachment. If you don't receive it, let me know and I'll post it to the cloud.

 

[Ari]

Unfortunately, we can't get any further here. Thanks Craig for the song files. Unfortunately they did not help to identify the problem as they contain too many irrelevant details, audiofiles, insert plugins etc. . A song cleanup would help alot to isolate issues. Maybe I'll write a sticky on how to prepare song files for a problem.

Since no one else or I have ever heard of the distortion issue and we have no idea what the actual problem is, we can leave the topic here as it is. If the problem can't be shown in the first place (by an audible file, wave, youtube etc.), then I'm in the dark here and can't judge whether it's by design due to a misconception of the normal natural plugin behavior, or a user error or actually a bug.

If anyone else comes back with a problem that they can demonstrate to me, I'll be right back on board.

best
Ari