UPDATED: New images, more explanations.

Hi guys,

For long time I was puzzled and confused on this topic. I tried many methods and they all seemed to work fine but I just discovered a very realistic method (imo), the results of which are quite remarkable. This might not be new to some, but for many of us can be very useful and serves as a starting point.

A common method of creating depths is using three different rooms on three fx channels. The aim is to produce close, mid and far sense of perception. Rooms could be small hall (close), medium hall (mid) and large hall (far) using algorithmic or convolution reverbs. Instruments, depending on their respective positions in the hall, achieve depths using 'send' function. For instance, violin section 1 is 'sent' to small hall because they sit close to the conductor. Woodwinds are in the middle of the hall so they are fed by medium hall and so on. To create more realism predelay function of the reverb can be used with three different settings. To add more authenticity, all three reverbs can be treated by putting an Eq, e.g., instruments at the back of the hall will sound back of the stage by removing high frequencies (eventually lower). This is one method and there are lot of other tricks exits.

Few days back I had an online and offline conversation with the owner of this site http://www.beat-kaufmann.com/. Beat Kaufmann is a professional sound and recording engineer as well as composer. He, very generously, explained to me the details of how he uses reverb in his mock ups. I thought I will share it with you guys so that we can experiment with this technique. With his kind permission all preceding images and sound example are the courtesy of his site unless stated otherwise. According to him this technique is not a holly grail but results are very realistic.

Note: For best results, use dry samples only.

So, First and foremost thing which we usually forget is to practically draw a hall or at least imagine. If we have an imagination of hall, we can easily figure out how many depths we can create. Have a look at this pic:



There are four depths. The fourth depth is for choirs (very far), others are close, mid and far. Now going back to the first technique one can never achieve a realistic recreation of this hall because there are three halls in that reverb setup and each hall has it’s own acoustics. Mix can also sound muddy because of the different parameters used by three reverbs. We need to find some other solution. To achieve a realistic or a natural sounding hall, we need a strong and clean ER which is not that easy to find. Early reflections of algorithmic reverbs do not sound that real however convolution reverbs has this natural sounding early reflection contained in them.

Technique: As per the hall picture we need to create four depths. For this purpose you have to go through all IRs’ recorded very far away from the original spot of sound. Check IRs’ in your arsenal one by one that says ‘far’. Next step is very important. Look at the below picture. I am using Reverberate here.



The IR I found was more than 6s long. But we need only 300ms of it. Shorten the length of IR up to 300ms. It should start fading out at -60 db (hint: use ADSR to shape it). Why at -60db? We are using RT60 reverb formula here. Not going into the details of it but to understand just remember that RT60 (reverb time) is the interval during which sound pressure (acousticians say sound intensity) level decreases by 60db (see below image).


Image courtesy of http://www.sengpielaudio.com/calculator-RT60.htm

Now load a dry percussive instrument in Kontakt or other sampler. Play some notes and see how the decay sounds. It should not give snapping or abruptly ending sound. If it does, tweak the envelope. Once you are satisfied with the sound, save it as preset. Our test is complete. We used the percussive instrument to give the best possible decay which is not possible with mellow instrument say violin as it could have masked any decay artefacts. Now you have a natural sounding early reflection which we will later use to create all four depths. Yes all four depths from just one early reflection.

It’s time to load some instruments to create 4 depths; take violin (for close), flute (for mid), xylophone (for far) and choirs (for very far) for example. Put the ER we just created on each channel (instrument) and go to the mix knob and start experimenting with dry/wet signals. Hint: Keep the 'close' instrument as dry as possible. Again..goal is to reproduce the effect representing instruments being close, mid, far and very far from the conductor. When satisfied save four settings emulating the desired sound. We will use them in the routing step. Here is an example of dry to wet signals by Beat Kaufmann himself. He used an algo reverb as tail at 20% mix: http://www.musik-produktion-createc.ch/Tut_FX_Rev_Dif_Depth_Signal.mp3. Remove the samples and the sampler when you are done. Now we need to create the fx channels with all 4 depths that we have just created. Let's move to the routing for that.

Routing: Look at the below images.



It should look like this in the mixing board.



These pics are pretty self-explanatory.

Remember! We will NOT use ‘send’ in this reverb setup. ONLY outputs. We will NOT use predelay. VERY IMPORTANT!

I will explain routing for one group of instruments because rest is same. Create four fx channels. Load all four instances of your ER we have just created each on separate channel. Name them CLOSE, MID, FAR, VERY FAR. Now route the Output of the Strings group to the first fx channel (CLOSE = Depth 1 in pic) which is loaded with ‘close ER’. Route the Output of the close ER to either submix bus or master channel which has a tail from algorithmic reverb (any nice sounding hall from Valhalla, Breeze, B2, Lexicon or any unit of your choice). Remove any predelay and any trace of early reflection and other settings from it, only pure algorithmic tail. Mix it somewhere between 15%-20%. Use the same routing pattern for all instruments. In the end do not forget to save this setup as template. That’s it. You are done!

EQ: You can use EQ to carve the ERs’ to get more control over depths. This image will help to get started.



Panning: You can pan the instruments by using the readings in this image.


Image courtesy of http://www.audiorecording.me/symphony-orchestra-panning-and-reverb-settings.html

So, this is just one more technique to create depths. I hope it will be useful for some. Lastly, again all credit goes to Beat Kaufmann for his excellent easy to understand illustrations and of course above all - the reverb technique. If you are intrigued, share your method. Because this is the only way we can learn. Most important thing is in whatever way you do if it sounds good, it is good.

Best-

 

Last edited: Feb 19, 2016

Applying compression to lessen the amplitude of a transient will also make something sound farther away, not to the extent of a reverb but in conjunction with....it all add space.

 

aw, this is very interesting concspt and technic, I'll try it as soon as I'm home
thanks, and I will keep update this thread for more discussion

 

Thanks. Looking forward to your results. Please post here if you want.

 

Yes, compression can be used in fact it is used in professional mock ups but very intelligently. However, classical music is all about dynamics so many avoid it if they can.

 

Great article Ed and agree about classical orchestra and how easy it is to ruin the dynamics

 

so the point is

• Creating an ER (no tail) with ~300ms length
• Embed that "preset" in the 4 groups as insert, close depth = low dry/wet, far depth = high dry/wet
• On master channel (or bus for the instrument groups) a reverb with just tail (no ER) and a good sounding "hall" preset (long tail), mix in about ~20%

Did I miss something?

 

Yes, only one more thing to add which is very important is shaping that 300ms IR to fade out at -60db (should not look like a square shape but with gradual decaying shaped), otherwise it won't give you the decay of the notes and will end abruptly. See the Reverberate pic above. Find the cleanest possible (far, long prefixes) IR to create the ER from. You don't want to mask the dynamics behind an unclean IR which is the main point here.
You can mix the tail less than 20% too, say 15%, depending on the hall verb. Your ears will dictate you.

 

Last edited: Feb 19, 2016

I tried around, but I used several plugins for creating the ER space like Aether, breeze, Valhalla plugins etc. It's hart to match the requirement, because ER settings in algo reverbs do not include a length. at the end I used an oscilloscope to measure the ER length and try to match the 300ms. sounds pretty interesting.

I also tried around with several plugins. Hofa IQ Reverb seems to be interesting, it's possible to "Cut" or "Gate" the reverb sound to a defined length, but I don't know if it's according to RT60 law. Besides, Aether is using RT60, but it seems especially only for overall mixing. There is no ER length setting, but it has a good preset library for ERs. Hofa Reverb sounds totally different.

I'll try Liquidsonics out soon. Unfortunately there is no audio example for a "good sounding ER" from the tutorial.

On the other side:
Who is REALLY producing such classic, orchestral music? :D

I'd like to know how to adapt that mixing style to a modern hiphop or electronic music track (not edm).

 

@bigboobs you will certainly not get the desired results using algo reverbs for ER which is why I specifically mentioned to use convo reverb plugin as only an IR is capable of creating a natural sounding ER. Algo is best for tails.

Any convo plugin that deals with the alteration of the 'shape' of an IR would do. I know one more which is SIR. I also know the concept in VSL Hybrid reverb is same (convo ER + algo tail) but if you have Reverberate or SIR you can still replace the expensive VSL reverb with any of these cheap units.

I have posted one audio example above, will repost here http://www.musik-produktion-createc.ch/Tut_FX_Rev_Dif_Depth_Signal.mp3 Besides, a good sounding ER is up to you to find in your IRs or download some and experiment.

Regarding who is doing classical/orchestral music....well....orchestral music is not a cup of tea of everyone, however, the same genre is widely used in the movies, soundtracks, trailer and cinematic music so the other side is much greener than what you think you know or assume .

You can use this technique and tailor it to fit in the style you produce if you can (I believe), but results cannot be promised. Now it all depends on your creativity. This setup is basically for the genre I mentioned earlier but there is no hard and fast rule.

 

Last edited: Feb 20, 2016

When I was experimenting I created a very quick small track just to see this technique in action and results were pretty acceptable despite my terrible musicmanship, apologies in advance. I will post that track here, please use headphones if possible.
I created three depths and used Lexicon hall 17% for tail.

 

Last edited: Feb 20, 2016

this method is something ive been implementing lately.

 

Last edited: Feb 20, 2016

Dope