Originally Posted by mik303

Looks very nice. Is it zero latency design?

Originally Posted by Gemylon

This is cool

Originally Posted by Gemylon

Why Fog ?

Originally Posted by joe_04_04

I never got into impulse stuff, so I never learned about it, but it sorta seems to do what a vocoder does in some way?

Correct me if I'm wrong, I probably am.

Anyway, this is cool and I'll keep my eyes on it.

Originally Posted by audiothing

It was quite hard to choose the right name for a convolution plugin. The convolution of two functions is often written as f * g and sometimes the tensor product symbol is used instead: f ⊗ g
So Fog looked like a nice name, if you also consider how fog can "blend" in a landscape

Originally Posted by phaded

Whyyyyyyyyyyyyoudothisss??????...

Originally Posted by Gemylon

Ok, I see you

Even so, fog is still associated with a dim, uncoloured kind of atmosphere to me,
by hey, thats just me.

I am seriously considering buying this just the same ...

Originally Posted by Sensorychaos

I'm interested...I like the concept and we need more of those...

Originally Posted by phaded

Whyyyyyyyyyyyyoudothisss??????

Oh man, I'm going to be so poor this summer. =]

Originally Posted by audiothing

It was quite hard to choose the right name for a convolution plugin. The convolution of two functions is often written as f * g and sometimes the tensor product symbol is used instead: f ⊗ g
So Fog looked like a nice name, if you also consider how fog can "blend" in a landscape


Phase Vocoding and Convolution uses the same principle: using one signal to modify another signal.
While phase vocoding takes two signals in real time, convolution takes one signal in real time (the input) and a static impulse response.
The main difference is that the length of the processed output is different. With convolution, the output length will be the sum of the input length and IR length, so you can create reverbs/tails.
With phase vocoding, the output length will be the same as the two inputs.

Originally Posted by frankymax

Thank you Audiothing for this plug-in! I've been wanting to use my old Tritone Digital Colortone (which only works for Macs) on my Windows machine for years and Fog Convolver allows me to use all those cool equipment IRs from Colortone and then edit them even more deeply in Fog Convolver! This is an amazing piece of software, and I am loving it! The provided IRs from Audiothing are really great too! Sorry to sound like a fanboy but I'm just really digging what this does to my audio.

Originally Posted by Beatworld

This is so cool, really nice addition to my reverb/effect arsenal.
Thanks Audio Thing, didn't bother with the demo because the pricing is very nice.
Fog Convolver and Relayer have just added a whole heap of cool options for me.

Originally Posted by EvilDragon

Would be awesome if you could make it true zero delay convolution. Some other vendors managed to do it, so...

Originally Posted by EvilDragon

I absolutely love the GUI of it!

Originally Posted by joe_04_04

1) this is awesome, very very creative tool

Originally Posted by joe_04_04

2) I'm noticing the GUI starts to lag when files are longer, the smoothness goes away as you dial in something and the action isn't occurring for a split second. For instance, adding a fade on a short file, as you dial the knob, the fade comes in smoothly, but on a long file, it doesn't even show up until you release the mouth click, almost like its starving the graphics

Originally Posted by joe_04_04

3) As I said before, I'm not super knowledgable on convolution, so my question is, how do these impulses work to faithfully recreate a space? I don't know convolution, but I always thought they sweeped a sine signal out of monitors in a room from 20 hz to 20 hHz and recorded it, then used intelligent processing to strip the original sine sweep to leave just the rooms 'reaction.' So I'm a little confused on how this works with an impulse response like a short blip. I'm not trying to come off in a negative manner at all, I just truly don't understand the process in layman's terms, so I'm wondering if anyone can help me.

Originally Posted by joe_04_04

4) I read up there earlier that you said it is not true zero latency, but almost zero latency, my question is, does the plugin report that small insignificant amount of latency to the host? I'm not seeing any latency on my end...

Originally Posted by joe_04_04

Edit: I just had an extremely loud blast of noise come after using the stretch function on an impulse... it was almost white noise-ish, but it was insanely loud and just about blew my head off since I was wearing headphones (clipped both my track and master fader). I'm afraid to play with it anymore, I would try to help narrow down what combination of parameters I played with to get that to that to happen, but I think I will hold off on playing with this anymore until that can be confirmed and fixed. I believe I was turning the stretch mode down and not up... that's about all I can remember, hope that helps figure out what it might be.

Pro Tools 10.3.8 on Mac OS 10.8.4.

Originally Posted by audiothing

Thank you!



Yes, unfortunately the convolution engine has to re-calculate all the parameters for the IR everytime you change even a small thing. With large IR this results in a lag in the interface. There's an option in More -> Settings where you can decide to update the parameters only when you release the mouse. This will help smooth it a little bit (so it won't recalculate everytime you move the mouse).
There's also an option you can disable Fast Load/Update that will improve working on large IRs.



There are basically two main methods for making reverb impulse responses. Sine sweep and then de-convolve the signal. Impulse.
The easiest way is definitely using an Impulse. Depending on the "system" you are using it can be a digital spike, a starter pistol, a ballon pop, or even wooden boards or hand clapping. Each one will have different results and different EQ profiles. The Sine sweep will be probably the most flat response, but sometimes is not possible to bring all the equipment in places like this one:

How this will translate into a reverb is explained in many articles online (for example here). The general explanation is that each sample of the incoming signal is multiplied with each sample of the IR. So if the IR is 10 sec long you will get a reverb with a 10 sec long tail.


Ok, I'll try to clarify this point which could be confusing even for the terms used.
Every plugin is subject to latency. In a modern system, if your host buffer size is 128 samples you should get a latency < 10ms which is more than acceptable. Fog Convolver won't create any additional latency to that, so basically the internal processing could be seen as "zero latency". The first part of the impulse is processed in real time with the input, while the tail is processed in the background. You can change buffer sizes in More -> Settings to improve the CPU load, but usually the head block size should always be the same size as the host buffer.


Sorry to hear about that. That's the first time I hear of this problem. I tried playing around with the stretch slider, with multiple IRs and it seems to work just fine. If you are afraid it may happen again you can put a limiter to the master, but it would really help a lot if you could replicate it.
I'll play with it again anyway and see if I missed something.

Originally Posted by joe_04_04

I re-downloaded the demo to give it a shot one more time and I'm not able to open the .dmg file, like the package is corrupted.. can anyone else confirm as of tonight?

Originally Posted by Gemylon

No problems opening file here.


I'm on Mac OSX 10.10.2.

Originally Posted by joe_04_04

I downloaded twice and got this: