blogohl

A discussion at HeadFi about Head-Fit (thanks to Maxime for the info) informed me about an audio player that can load VST plugins. It seems to work from version 2010/03/18 but I have to check a bit more. Have a try, the download page is here , just be sure to get the latest one.

I think that for a media player purpose, I’ll have to make lighter plugins, maybe starting with Head-Fit.

Those seaking for the cheapest measurement microphones, go to John Conover’s website !

A very informative webpage : homepage of the very competent David Griesinger, full of papers and powerpoints, a must. 

DRC software of Denis Sbraggion, coupled with a convolution tool, is a wonderfull FIR Eq for loudspeakers. But it is not so easy to use because it’s only command lines… So I did a basic interface working in Windows. Download it here, try it and let me know you wishes.

Added informations : another GUI for DRC is available here , nice interface, but only in italian now ( so just a bit complicated for me !). There is a discussion forum about this interface.

un article “fondamental” de F. Toole est disponible gratuitement :

Loudspeakers and Rooms for Sound Reproduction, A Scientific Review

Pour ceux qui n’ont pas son livre, cet article est à lire absolument !

Coloration and binaural decoloration of sound due to reflections, la thèse est disponible ici : http://repository.tudelft.nl/assets/uuid:7f0331e3-bc1a-4d7f-8d2a-eb5d6cc04fbf/as_salomons_19951220.PDF

Un aspect que je n’ai jamais testé : la coloration due aux réflexions est plus prononcée quand on �coute avec une seule oreille, l’écoute binaurale étant facteur de décoloration. Déjà que c’est bien mieux d’écouter une seule enceinte (en mono) pour percevoir sa qualité, maintenant, il va falloir en plus écouter avec une seule oreille !

Notons quand même qu’il n’y a peut-être pas décoloration lors de l’écoute d’un signal mono au casque (écoute binaurale diotique), ouf…

J’en profite pour rappeler les termes :

• écoute monaurale : avec le signal présent à une seule oreille
• écoute binaurale : signal présent aux deux oreilles
• diotique : le même signal aux deux oreilles
• dichotique : un signal différent à chaque oreille

L’écoute d’une seule enceinte dans une pièce est quand même binaurale et dichotique.

Audio Musing is the new blog of Sean Olive : not many articles yet, but interesting ones.

See here

Denis Sbragion released a version 3.0 of DRC using a spectral envelope method to create the target curve. I’m not 100% convinced if this is the ideal method but it is a very interesting step toward best results. Now you may understand a bit why you had to create a frequency down slope to sound right. Try Align here.

The evaluation of "audible" dynamic range isn’t easy at all : should we measure between peaks and average, or between peaks and quiet passages, what time resolution ? Should the new BS.1770 loudness standard also be used for this kind of measurement ?

An interesting proposal from Richard Tollerton here. But it requires LabView runtime. Maybe I’ll do the same in a VST, it would easier to use.

Also have a look at Hydrogenaudio here and here.

A very interesting topic at diyAudio Forum with Earl Geddes and John Kreskovsky, comparing monopole, dipole and cardioid bass responses in a room.

Le local d’écoute est l’antépénultième maillon de la chaîne de reproduction sonore, avant l’oreille et le cerveau (ça en jette, non ?). Ceux qui s’y connaissent un peu se désolent du manque d’attention apporté au local. Mais est-ce vraiment nécessaire de traiter le local ? Et pire, mal traité, n’est-ce pas moins bien que pas traité du tout ?

Les études acoustiques ont presque toujours porté sur l’acoustique des salles (grandes) ou sur le traitement des lieux professionnels. Qu’en est-il des salons de Mr et Mme Tout-le-monde ?

Deux papiers importants écrits par des vrais spécialistes apportent un éclairage un peu à contre-courant :

• Floyd Toole a publié “Loudspeakers and Rooms for Sound Reproduction, A Scientific Review”, J.Audio Eng. Soc., vol. 54, pp. 451-476 (2006 June)” Extrait : “Any device inserted into a reflected sound path—reflector, absorber, or diffuser should perform uniformly well at all frequencies above the transition frequency region, say, 200 to 300 Hz. This is in order to preserve the spectral balance of the loudspeakers, to uniformly attenuate the full spectrum of reflections, and to ensure that the precedence effect is maximally effective”, à voir un avant-goût ici et ici. Mais je ne peux que vous conseiller de chercher le papier complet (empruntez-le, volez-le, mais surtout lisez-le)

• Siegfried Linkwitz a présenté une conf à l’AES à l’automne 2007 que chacun doit lire, c’est ici. Extrait : Reflections generated by the two loudspeakers should be delayed copies of the direct sound to the listener. The delay should be greater than 6 ms. The high frequency content of the reflections should not be intentionally attenuated……The requirement for full spectral content of the reflections rules out the use of frequency dependent absorbers on the room surfaces. The various commercially available foam or fiberglass panels absorb predominantly higher frequencies only and would color the room reflected sound dynamically to where it no longer can be cognitively separated from the direct sound. This then argues for relatively live room acoustics that are determined by the  normal stuff of life  with which the room is filled and decorated, acoustics with which we are intimately familiar as normal.

En résumé, plutôt plein de meubles que du traitement acoustique ou alors un traitement qui soit efficace sur une large bande (au-dessus de 200Hz) donc au moins 10cm d’épaisseur de laine de verre dense (plus de 50kg par m3), sinon rien. A lire et à méditer….

New page : 

is time alignement on loudspeakers important or not ? Try this new soft

First VST software of Jean Leung

a crossover with various types of filters, it’s now easy to listen and compare :

• 6 dB/oct
• 12dB Linkwitz-Riley
• 18dB Butterwoth with polarity inversion
• 18dB from JMLC (LeCleac’h)
• 18dB quasi-Linkwitz proposed by F Brooke
• 18dB proposed by P Sheinkin
• 24 dB Linkwitz-Riley

Go to the software page.

une autre initiative contre la peste envahissante du loudnesswar :

http://turnmeup.org/

et aussi justice for audio !

la dynamique vivra

Dans la foulée de Align, j’ai mis à jour The Final Cut avec les mêmes améliorations pour la correction DRC.

The Final Cut est évidemment bien plus puissant mais aussi plus complexe : il ne sert pas à corriger des enceintes et/ou le local mais à comprendre comment les enceintes et le local interagissent. C’est, je crois, un bon outil pour apprendre à paramétrer une correction. Maintenant que l’outil est là, il faut que j’apprenne à bien m’en servir.

No more reason not to have a perfect loudspeakers + room setup, just try this simple software called Align.

In 5 clicks, you measure and correct the amplitude and phase response of your speakers in your room.

Sorry the explanation page here is yet only in french but the soft should be easy enough to use, get it here.

Another update in Phase Audibility software : I added a triple tone, that consists of three sine waves at equal frequency distance, F2-F1=F3-F2.

Now we can easily do the Zwicker phase test, just center the filter frequency to same frequency as generator in triple tone position and listen to phase distortion. At 2kHz, I can hear and ABX differences between various filter types.

I’ve been asked if I could add absolute polarity test on phase audibility software, it’s done so try it.

From Wikipedia :, when two identical sounds (i.e. identical sound waves of the same perceived intensity) originate from two sources at different distances from the listener, the sound created at the closest location is heard (arrives) first. It is also known as "Haas effect" or "law of the first wavefront".

The drawback of this precedence effect for a stereo (or multichannel) setup, is that the phantom image collapses to the nearest speaker (the one heard before).
Can we avoid this ? Maybe…because a very interesting possibility exists : if the second sound has a higher level, the precedence effect can be avoided.

I wanted to check myself what is the needed level, for what frequencies, up to what delay, aso,.. so to know if we can really compensate the precedence effect in an audio system. For this, I did a software so you can play with delay and level and check, for different signals, if really the central image stays stable.
Download here

Crossvolver is a new soft that includes a 4 ways FIR crossover with room EQ through DRC :
– each FIR filter is calculated with Uli’s maXO and can be convoluted with your DRC correction, so the filter does both work in one convolution, crossover and DRC
It uses maXO and Convolver

You can download it here, it’s called crossvolver.
Right now, it’s very new and I had no time yet to do a page to explain how it works but you can have a look at its interface on this page