Jorn, Thanks for the explanation.<br><br>I've found the wxyz wavs for a couple of fons's conf files, so i'm going to try those. My brain smells like a burning battery at the moment, trying to absorb enough ambisonics knowledge to be almost dangerous, and with the experiments i'm conducting.<br>
<br>I tried the split IR, and you're right, it sounded awful. I'm just not smart enough with this stuff yet.<br><br>Can't make LAC 2009, although i'd dearly like to turn up with a dictaphone, and wheelbarrow full of parchment for note taking. Perhaps 2010 might be on the cards though.<br>
<br>I'm having some success with more modest IRs, for position experimentation. There's three in my little conf file, aand i've truncated 2 of them. Not exactly scientific, but it's good for listening, and getting my head around emulating, however crudely, an orchestral spacial model.<br>
<br>I'm doing this for stereo at the moment, but don't rule out a near future multi channel recording model.<br><br>Not using firewire here, so that's ok.<br><br>I'm fond of pizza, with the very occasional sip of fine cognac.<br>
<br>Helps me think.<br><br>Honest.<br><br>Alex.<br><br><div class="gmail_quote">On Tue, Jan 20, 2009 at 8:34 PM, Jörn Nettingsmeier <span dir="ltr"><<a href="mailto:nettings@folkwang-hochschule.de">nettings@folkwang-hochschule.de</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div class="Ih2E3d">alex stone wrote:<br>
> I've done some crude experimenting and ask at this point, how do i<br>
> determine the format (i.e. short b-form ir or not) of a particular<br>
> IR? Is there a device or some sort of utility that i can use to do<br>
> this?<br>
<br>
</div>to clarify things, there are no "short" or "long" IRs. what fons means<br>
is: the early reflections are the most characteristic aspect of a room,<br>
and they affect localisation the most. therefore, if you want to have<br>
ultra-realistic reverb, use an IR that was measured with the speaker<br>
where your intended instrument is and a soundfield microphone where the<br>
listening spot is. of course, in practice this is not done.<br>
<br>
so instead you use one reverb IR instead. it can be short (tail<br>
truncated) to save CPU, because the tail is decorrelated (blurred) and<br>
does not provide localisation cues, hence it would be wasteful to render<br>
it in b-format.<br>
<br>
the problem is: you will probably not get such a two-stage reverb<br>
calibrated unless you have a lot of experiences with both b-format and<br>
convolvers.<br>
if you can make it to lac2009, let's talk this through over pizza<br>
(that's how i learned my first steps in ambisonics from fons, and it<br>
works suprisingly well).<br>
<br>
an IR can only be in b-format if it contains 4 channels (WXYZ) for full<br>
3D or 3 channels (WXY) for planar surround only.<br>
<div class="Ih2E3d"><br>
> The reason i ask is, i have a set of IR's for a hall, recorded 3<br>
> front stage, left middle right, and 3 back stage, left middle right.<br>
> I built a jconv .conf file with these IRs, and the result was near<br>
> 100% CPU, and a vertiable flood of Xruns. (AMD 64 Dual Core X2 5600+,<br>
> 64bit Ubstudio) I understand from this, i've used 'full' IRs, and<br>
> Jconv is attempting to process them all at once.<br>
<br>
</div>probably, and the cpu usage goes up quickly the longer the IRs are.<br>
also, jconv has the nasty tendency to starve ffado hardware threads (not<br>
really jconv's fault, but needs some tuning), so if you are using a<br>
firewire interface, you might be in for some fun. it can be done though,<br>
works nicely for me, but it requires some poking.<br>
<div class="Ih2E3d"><br>
> So, I gather from what you've written, that there is either a<br>
> mechanism for creating short b-form IRs from a single IR wav, or i'll<br>
> have to find a specific IR that is recorded in such a way that i can<br>
> use portions, or sections of the IR, without taking such a<br>
> performance hit, or there is a mathematical means of extrapolating<br>
> sections from a single ir, already present in Jconv, and i'm not<br>
> 'getting it', at this point.<br>
<br>
</div>as i said, i would recommend against attempting such a split reverb<br>
method, because very likely things will go haywire at some point.<br>
better bribe fons :)<br>
<br>
for now, i'd recommend to use one single reverb for all sources, and<br>
only vary the dry/wet ratio according to the distance to the virtual<br>
listening spot. should be really ok.<br>
<br>
what is your desired output format? plain old stereo, or 3d ambisonics?<br>
<br>
<br>
<br>
<br>
<br>
</blockquote></div><br>