Audio feedback (also known as acoustic feedback, simply asfeedback, or the Larsen effect) is a special kind of positive feedbackwhich occurs when a sound loop exists between an audio input (for example, a microphone or guitar pickup) and an audio output (for example, a loudspeaker).
 
Audio feedback - Wikipedia,  
 
 

Anderton

Adq
And it was announced that it would be possible ( in pre-release description of track-to-track routing there was something about cool feedback effects).

I highly doubt there was ever any announcement that it would be possible to create feedback loops. Not creating feedback loops was part of the design spec that was baked months ago.
 

I can't say exactly right now, but there were some words about "interesting effects" or something like this, and there no other way to interpreter it, because this routing can't add any new effects without feedback loops, nothing that you couldn't do earlier.
 

Anderton
But, that's why external inserts exist - so those who want to pursue unusual "corner cases" that few others pursue have the ability to do so. As John correctly (IMHO) states, "It's not something you'd generally use in most of your mixing, but hey. Flexibility allows for interesting fringe cases." That flexibility is available for those who want it. I've taken advantage of it, but maybe only one or two times in the past five years. Still, when I needed it, I could do it.

Ok, we have discussed it earlier ( it is interesting that the week after that long discussion new routing possibilities have been announced). My argument was that your way introduce latency.

And again, auto-mute solves all possible problems. In fact, it would be cool to add auto-mute feature and without feedback loops, because there are hundreds way to achieve uncontrollable audio levels.

Nobody is saying that what Sonar currently provides is problematic. I'm saying that having the option to get a waveform of the click track (as provided by other DAWs) would be useful.

Man, you have to change the defensive attitude. Really. Sonar is great and the monthly updates are truly great. But it seems that any suggestion for improvements outside your comfort zone is met with all sorts of irrational replies, misrepresentations and disparaging remarks. Relax - Sonar is doing great, but there is always room for Improvement!

Adq

John

Adq
John, I can't understand what do you want to say. People use feedback in music for hundreds years, and nobody die (well maybe few guys actually died because of it, I don't know really). Have you ever seen guitarist shaking guitar in front of his cabinet?

Its feedback but its not a feedback loop. What I have in the post above from wikipedia is about a feedback loop. When you guys talk about using feedback in a delay it has been limited and controlled. Feedback loops are not that. If a guitar player uses feedback as many do its controlled. 

Feedback and feedback loop is the same thing. Loop in the routing creates feedback, that is why it is called feedback loop. https://en.wikipedia.org/wiki/Feedback . And nobody says that you shouldn't control it. It must be controlled by track settings, plug-ins and so on.

Thats why I'm saying we are talking about two different things. Because feedback is used does not mean its a feedback loop. What CW has tried to avoid is a feedback loop. Not feedback. And no one has been using feedback for hundreds of years.  

Adq

Noel Borthwick [Cakewalk]
Feedback loops in DSP is quite different from feedback in signal flow inside the DAW. 
I can't even imagine how that would be constructed and maintain delay compensation etc. Today a feedback loop in routing translates to infinite recursion or in other words a hang :) So sorry - feedback isn't feasible.
 

Here is how it works in ProTools:
http://www.youtube.com/wa...mp;v=MUb0Ln5GOCU#t=162

Great but that is not what a feedback loop is. The signal is not feeding back on itself. The commentator uses the term but it isn't feedback. The vid I listed has real feedback. Have you not heard an amp picking up a mics signal and a loud screech is the result?  That is feedback. What was shown in the vid you posted was a rerouting of the signal but it was not allowed to interact with the original signal. Thats why there was time between the signals.
 
Put what Noel said and what I'm trying to explain together and it may become clear.  

Doktor Avalanche
Must resist updating... must resist..

Resistance is futile!

...just my two cents.

First, thanks to the Bakers for patch points. I've always found other ways to achieve what I want, but my old school upbringing had me used to working with patchbays and mult bays as well the ability to find various patch points to tap...

Now it appears we are getting this equivalent in Sonar! Outstanding!

As to feedback? If I'm understanding the posts I've read here...

I'm not sure I see why it is stated that infinite loop setups are impossible in a DAW... It might not happen as automagically as in the analog world, but I have a feeling the physics could be modeled in some way

Not that I need it, and trying to use my iPad to run Amplitube is the exact nightmare of always having analog feedback problems due to the low grade analog I/o...

If I want to use feedback at this point in time, it's easy enough to mic an amp where I setup a controlled feedback situation as part of a performance...

That said, it sure would be fun if we could add controlled feedback to an amp sim with the ability to enable/disable it...

Noel Borthwick [Cakewalk]
Yes I see what you are getting at - it would appear that they are mixing in the output of one into the other.

You can't achieve this repeating sound this way, just mixing in the outputs.

John
You weren't wrong. It was because some people throw out terms without knowing what they really mean this is not aimed at you but the fellow in the vid you posted. In your context you had it right. I'm just talking about a true feedback loop. What CW has been trying to avoid. Graig's method was not creating a feedback loop. If I have it right he was inputing the output to a new input. That wont cause feedback. If he was connecting the output to the same input a loop would be created.  

There is no "true feedback loop". What you are talking about is feedback loop with such settings that produce increasing self-oscillations on some signals. There is the only difference between what you are talking, as I understand, and any other feedback loop. Self-oscillations aren't equal feedback loop.

jbow
Yaaaa, I have something to go do!! Thanks Bakers!
You guys should hire Karl to do a cumulative series of videos with a new one each month. A new SWA covering all the new stuff in each update. Create a job! I'd pay an extra 10 bucks a month to cover it.. maybe others would too but IDK. Everyone else seems to already know a LOT more than I do. LOL.
 
Thanks again!
J

I'd buy into that...

John
Graig's method was not creating a feedback loop. If I have it right he was inputing the output to a new input. 

Actually I was referring to my looper application where I stuck a bunch of delays (with delayed sound only) in series to get like 16-20 seconds of delay, and did indeed feed the output back to the input so the delay would have feedback and just keep repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and repeating and...you get the point. 
 
Yes, there is latency but as you can imagine, with a 20-second loop it didn't make a whole lot of difference to take on another 10-15 ms...the only other time I've used feedback is to drive things into oscillation, but musically speaking, it does me no good to have it slam into uncontrolled feedback - I have to be able to "coax" it into feedback, and "coax" it back out when it gets out of control. So again, even 15 ms of latency makes no difference as I can't move a control with 15 millisecond precision. I suppose there may be applications where someone would need a near-instantaneous application of feedback, but now we're talking about an even smaller subset of potential applications. 

Adq

Anderton
I wouldn't mind having the option to create feedback loops. That said, when I need feedback, I use the external insert plug-in and patch an output back into an input. That's how I did it before computers existed, and was the basis of my looper application for SONAR.
 
Then again, we ARE in the United States. The first time someone blew their speakers, they'd probably sue Cakewalk for a million dollars saying their hearing had been damaged permanently, and Cakewalk was negligent in not preventing feedback 
 

They have many ways to damage something, including your external scheme. And there is easiest way to prevent it: auto-mute if sound's level exceeds some threshold. And other DAWs do it, and nobody sue them. And nobody sue delay plug-ins developers, and some of them have crazy settings that could self-oscillate pretty easy. So your arguments don't work, in real life there is no problem with feedback loops, it is imaginary problems.

 
I added a smiley, which indicates something is said with humorous intent. Jeez, why does everybody take me so seriously? Did no one read the first paragraph of my latest "Tip of the Week"? 
 

And you always say that if some missing feature has a workaround, it should not be implemented primarily.

 
That's not true, you just quoted me saying "I wouldn't mind having the option to create feedback loops." 
 
I give opinions on what I think priorities should be. Suggestions need to be evaluated on their own merits. I am in basic agreement with someone like Doktor Avalanche that there are more important issues than being able to set up feedback loops.
 
I present workarounds not as an apology, but as a service to those who think something is not possible when it is. Most people realize I am trying to help them get the most out of their program. I also do this for programs other than SONAR, in the appropriate venues.
 
As to me, there are plenty of features I would like to see that would make creating sample libraries and loop libraries easier. But I am not sufficiently self-centered to think that most users would consider those kind of features relevant to what they do, so I do not ask for them. I can create sample and loop libraries within SONAR the way it is, because it's better suited for that task than any other DAW I've used. That's good enough for me.
 
Like most people who come to these forums, I acknowledge and appreciate what has been accomplished, not just by Cakewalk but by all the dedicated people who work on all the programs we use. If CupertoolsOneplitude has some cool feature, more power to them. Their not including features SONAR has is no more a reflection on them than SONAR not including all the features they have.
 
This is why I come to the defense of other companies on the rare occasions when they are dissed in these forums. I think it is equally rude to come here and knock the competition as it is to promote the competition. I take every opportunity I can to hang out with my many friends at PreSonus, Ableton, Propellerheads, Steinberg, Avid, MOTU, Magix, Native Instruments, Spectrasonics, IK Multimedia, Acoustica, etc.
 
All these companies are making incredible tools for us under difficult conditions in a tiny, low-margin market that's been devastated by rampant software theft. They can't even do a fraction of what they want to do, let alone what their customer base demands that they do. 
 
There is far more respect among competing software developers for each other than they get from the people who use their products. I believe that's because the software companies fully understand the realities of this business, and generally, the customers do not.
 
At least this forum is better than most, and with respect to many people who post here, much better than most.