• Register
  • Help
Page 1 of 2 12 LastLast
Results 1 to 10 of 12

Topic: Convolvers, Reverb, and associated myths

Share/Bookmark
  1. #1

    Re: Convolvers, Reverb, and associated myths

    Oooh, heavy!

    Bruce, once again, a brilliant exposition of the underlying fundamentals, and why fundamentalism is not enough!

    I won a science fair award in 1957 or so for a similar demo with a pan of water and two \'beaters\' that struck the water in a rectangular cake tin at adjustable speeds. I could exchange a round knob for a flat plane or a convex or concave surface. The interactions were photographed and made quite a striking display, although at the time I couldn\'t come up with real purpose for it, only the formulae describing the pix. (Only got 2nd. place [img]images/icons/tongue.gif[/img] )

    But the true art, as you so eloquently state (no, I\'m not sucking up-some people write more clearly than others,) is to use this knowledge to make great music!

    Dasher

  2. #2

    Re: Convolvers, Reverb, and associated myths

    One thing I wonder is, dont you get \"distance\" in the impulses depending on where the speakers are placed, or gunshot is done?

    I mean this is obviously only in full wet sound. and, believe me I dont expect any convolver to give me an option to accurately recreate \"placemetn\" in a 3d sound space. Its jsut a question [img]images/icons/smile.gif[/img]

    I agree that its about using the knowledge to HELP you get a GOOD sound, not define what it takes to get a good sound. Use your ears, as well as basic physics.

  3. #3

    Re: Convolvers, Reverb, and associated myths

    I wasn\'t expecting to attend an Audio class on xmas day but thank you Professor Richardson for the enlightened discourse on Sound Waves.

    Having said all that I would like to know what is the \'best\' e.q. to use. I\'ve heard that the Sony Oxford E.Q.running on a Powercore Card or Pro Tools is the standard to measure all others by.

    What is the Best E.Q Plug-In out there? For smooth \'musical\' results. Has anyone done the research and hands on comparisons?

  4. #4
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Location
    Dallas, Texas
    Posts
    5,755

    Convolvers, Reverb, and associated myths

    Realism is overrated. Purity doesn\'t exist.

    You must use EQ to aid your imaging. If you want something to sound \"far away\" you will never make that happen with reverb alone, no matter if it\'s a convolver (which people seem to think was just invented, but which has existed for years) or a traditional algorithm-based reverb.

    In one of the many recent \"convolver\" threads, there is much wasted air concerning the dry unaffected signal. Well, hello. If you think that adding reverb to a dry signal alone is going to achieve pleasing imaging (convolver, reverb, chamber, or your daddy\'s spring), you should re-evaluate your grip on the physics of sound. Or, to put it another way, forget realism and try taking note of reality.

    Do yourself a favor, and find a friend who plays a wind instrument (or some sufficiently portable acoustic instrument). Find a quiet, large space. Have that person stand two feet from you, and play a note. Have that person stand five feet from you, and play a note. Ten feet, fifteen feet, twenty, thirty, forty. Have a pad and pen ready, and jot down how the frequency relationships change within the sound--which ones get absorbed by the air (and changed to heat) and which ones survive the distance. Make note of how much of the \"total air\" is filled by the sound at each position, how wide it is from left to right, top to bottom, and front to back. Also, jot down the way the reflective content changes in relationship to the source sound.

    Now, bring that player back to the five foot position, and repeat the trick with the person aligned 45 degrees to your left, moving back through each position. Repeat it again with the person moving back 45 degrees to your right. YOU face straight ahead, like an audience member sitting on the conductor podium. Again, take notes, all about the reflective alterations and frequencies that get burned up in the intervening air.

    Now, head outside to any reasonably large (10 x 10 or so) body of water. A pool is best, because it has nice \"reflective\" borders. Arm yourself with a handful of small stones. Drop them into the water at different relationships to the center and to the edges. Watch the ripples, especially how they interact once they\'ve reached the edge and begin to reflect back upon themselves.

    What you are seeing is EXACTLY what sound does in a room, only in slow motion and in a single plane. Close enough for a study. The first \"reflections\" are equivalent to the EARLY REFLECTION content of your reverb. In sound, we perceive these as the early \"echo.\" Note when they reach the point of the original \"splash.\" Note how dropping stones toward the edges (equivalent to say a timpani on the back row of the orchestra) results in a very lopsided \"arrival\" of first reflections at the conductor position. Note how the reflections \"pass\" the conductor position and bounce again, becoming more diffuse and plentiful as they exponentially increase in number while being reduced by the friction of the medium (water in this case, air in practice). The secondary, small ripples that echo again and again, colliding with each other and reflecting again--these are the REVERB TAIL.

    Observe. Learn. This is your education. Now, take those notes, organize yourself, and start working with single notes and reverb--AND EQ--to mimick each of those situations you studied, both in sound and with the \"ripples.\"

    Having accomplished that, you are now in possession of the basics of imaging.

    NOW....GO MAKE MUSIC WITH THESE PRINCIPLES.

    Next study: Having learned to do that, FORGET IT. Concentrate first and foremost upon making PLEASING SOUNDS based loosely on that acquired knowledge.

    Almost everyone in this group has better tools than were used to make some of the most amazing recordings in history. THE PROBLEM IS NOT THE TOOLS. This should be comforting. Conversely, even the thought of a \"one button\" solution should be troubling. Nothing good comes easy. Take comfort in that, and use your talent to prove the old truth...It\'s the witch, not the wand.

  5. #5
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Location
    Dallas, Texas
    Posts
    5,755

    Re: Convolvers, Reverb, and associated myths

    Originally posted by Mel Tron:
    What is the Best E.Q Plug-In out there? For smooth \'musical\' results. Has anyone done the research and hands on comparisons?
    <font size=\"2\" face=\"Verdana, Arial\">The one you own.

    Seriously, any of the better plugins are acceptable. The Waves Linear EQ is quite nice. I have heard people I trust say the Sony is lovely (but not available universally in terms of formats).

    Unfortunately, the EQ in GigaStudio is not one of the better ones. We can hope for change.

    Best regards,
    Bruce

  6. #6

    Re: Convolvers, Reverb, and associated myths

    I\'ve been using Altiverb for quite a while- my ears have changed so I brought up a Waves \'verb and was amazed how much smoother it felt. Using ALti the formula was to lower the high end frequency of the instrument as Bruce has mentioned with a smooth curve for distance while sending it to the Alti in careful amounts. eaxample DAn Dean Flute Vibrato- narrow the stereo filed, lower a little bottom, chop off the high and it still sounds better going through an artificial reverb so Bruce has an excellent point.

    I am still trying to figure out why the LOTR Soundtrack has a cetain \'sheen\' to the sound- oerhaps it was recorded in Pro Tools HD? Or they are using some kind of Mastering Process that smooths out the whole Orchestra and gently lifts the highs? Over- Compressing perhaps? Too bad I live in a small town if I were in L.A. Nashville wherever- I could hire a \'top gun\' to mix my stuff.And I wouldnt have to figure all this out on my own.

  7. #7

    Re: Convolvers, Reverb, and associated myths

    Bruce

    I\'m not sure I entirely agree with your convolver = pure so it must be good debunking.

    To my ears, Altiverb for example, simply sounds much MORE like a real acoustic environment compared to a single instance of any other plug-in. It\'s not the fact that it\'s a sample of a real, pure space that makes a difference to me, it\'s just that it simply sounds better in a \"stick-it-on-a- bus-and-get-good-results-quickly\" kind of way!

    I agree, one instance of any plug-in/outboard reverb will not accurately create a 3D (or more correctly 2D) rendering of a real space in terms of distance of source from mic or instrument placing within the space etc. However, trying to reproduce this accurately with multiple instances of a reverb plug-ins/eqs etc. would be impractical for many, including me, simply because of the amount of time it would take to set up successfully, not to mention the CPU hit involved using multiple verb plug-ins. Trying to recreate those \"ripples in the pool\" might also require some really fiddly send/bussing arrangements and most DAWs aren\'t really set up to do this kind of thing quickly and easily yet AFAIK. Some don\'t even have stereo sends for example!

    I would imagine a single instance of a plug-in might start to work in terms of creating a 3D space if it had four discrete inputs feeding a quad reverb or IR even if the outputs were in stereo. I mean discrete as opposed to four inputs that are summed before they hit the reverb processor of course. Again, this would be awkward to handle in terms of sends on most DAWs. Unfortunately no one has done this , probably because it would be just too CPU hungry to have a real quad reverb!

    I still think there is a strong argument for taking multiple stereo IR\'s from the positions of the different sections of the orchestra but keeping the mics in the same, optimum listening position.

  8. #8

    Re: Convolvers, Reverb, and associated myths

    Bruce, I may have missed some basics here. I\'ve associated \'convolver\' with \'process to generate a reverb space by calculating data from a sharp impulse,\' but it appears there\'s more to it. Where can I go for a good, reasonably technical but pre-calculus discussion of convolution technology?

    Thanks

    Dasher

  9. #9

    Re: Convolvers, Reverb, and associated myths

    The actual formula used to convolve one function with another isn\'t very easy to explain. I thought I\'d summarize what I know about them. I have a degree in electrical engineering and we actually had convolution problems in both my digital signal processing class and a linear systems class.

    Basically, any linear system (such as an actoustic space) can be described by an impulse response. Electrical circuits with linear elements can be described this way too. Theoretically an impulse is an infinitely thin spike with an area of one. Of course in the real world there is no such thing, so they approximate it using a very loud and short sound. The impulse reponse is the response of the system to that impulse. When getting the impulse response of a room, the response depends on where the impulse is played and where it is recorded. Basically the acoustic response of a room is really an infinite number of linear systems since there\'s an infinite number of places to place the impulse and to record its response. This is the problem that I think Bruce has touched on. You really need multiple impulse responses if you expect to get a realistic version of instruments in any given space.

    An interesting side note about impulse responses is that you can do a mathematical operation to it called the Fourier Transform and get the frequency response of the system. Frequency response is magnitude vs. frequency and phase vs. frequency. Impulse responses are vs. time. Also, the Fourier Transform of an impulse results in a flat curve, meaning that an impulse contains equal amounts of all frequencies. Do the convolution programs do Fourier Transforms, so you can see the frequency response you\'re getting? I don\'t know since I don\'t have one yet.

    Anthony Lombardi
    www.mp3.com/alombardi

  10. #10
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Location
    Dallas, Texas
    Posts
    5,755

    Re: Convolvers, Reverb, and associated myths

    Originally posted by KingIdiot:
    One thing I wonder is, dont you get \"distance\" in the impulses depending on where the speakers are placed, or gunshot is done?
    <font size=\"2\" face=\"Verdana, Arial\">Absolutely. The gun or speakers are the \"player\" and the mic position used for the capture is \"the listener.\"

    I think one cannot get too religious about convolvers. They are cool for lots of situations, not just reverb. But there is quite a bit you can\'t do with a convolver. Reverbs can usually sound much more massive in the end result, at least when used correctly. And a convolver is not as flexible in localizing sections within a mix. You can alter the panning of your input, and can do some general design modifications to the basic impulse, but for a single given impulse, you\'re filtering everything through exactly the same portal. Generally, they\'re better for \"glazing\" mixes or sections within mixes...and not for artificially constructing a soundstage from disparate sections. In an average hall, sound will be moving at about 1136 feet per second, give or take depending upon the temperature and humidity. With that knowledge you can compute \"virtual\" distances from say a bass section, to the back wall, and back to the listener position, and you can construct some very accurate reflective models-- by using multiple reverb instances to essentially design your own multi-reverb algorithm for an entire virtual ensemble.

    Basically, I just get the feeling that there\'s some sort of \"purist\" mentality at work here in the group which says Convolver=Pure, Reverb=Effect=Bad. It\'s not true. None of it\'s pure, and all of it requires skillful informed choices to get a good result. Convolvers are great, and so are reverbs, but there\'s no panacea in either one.

    BTW, I know I\'m preaching to the choir in your case, Ashif. I\'m just putting it out there for the group in general.

Go Back to forum

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •