fresponse

Unfortunately I don't know any streamer OS accepting third party plugins... You can still gain a lot by using convolution filters but more disruptive processing can only be done with plugin

Hi all, I just posted a new article on my blog to talk about a specific feature of my filters generation service. Stay tuned, something new is cooking on the plugins side...

All the matrix inversion methods listed above are focused on reducing as much as possible the contralateral signal : this is only useful for binaural audio sources which are really marginals. My objective is to have a better compatibility with the large majority of stereo recordings (sources placement done by pan potting) by preserving some phase properties b/w ipsilateral & contralateral signals.

As anyone tested my plugin and compared it with u-Bacch ? https://www.homeaudiofidelity.com/english/plugins/

"Old" algorithms like the RACE one are not well suited for the audiophile consumer : severe coloration (and increased speakers distorsion) unless you have a very damped room, highly directional speakers and sit at the exact same position. I have released the X-talk shaper plugin to bring an affordable and much more tonal neutral solution. Hope this will help more audiophiles (re)discover the magic of crosstalk reduction

I am used to generate crosstalk reduction filters as part of my room correction service since 2017. As there is some demand to find an alternative to the u-BACCH solution, I have now created an audio plugin available for Windows and MacOS. You can find download link and complementary info on my plugins page: https://www.homeaudiofidelity.com/english/plugins/

Hi all, I have added a new blog post & demo clip on phase correction here I remind that I have a permanent Christmas gift with this free headphones equalization tool Seasons greetings to all of you !

Hi, You can read a new article on my blog here, it is about crossover filters with steep slopes. I am still looking for beta testers, so don't hesitate to contact me if you have an active system with FIR convolution capability !

Hi @Sangster Thanks for you advice, would be good for me to understand what issues you are facing with my website... To not pollute this thread, please contact me at [email protected] I am still exploring the potential of signal processing and found a new solution to improve reproduction for "flat sounding" tracks or systems. Without changing the tonal balance and using some principles of analog reproduction, I created filters that give more relief to the reproduction. Instead of changing DAC you can now easily tune your system with such filters. As usual don't hesitate to contact me for a demo.

As there are some discussions regarding target curves, I'd like to share my thoughts and how I proceed when I generate filters for my customers. It might be worth reminding that a measurement at listening position is impacted by the loudspeaker on-axis and off-axis response (directivity) and the room itself. The ideal case of flat on-axis response, controlled directivity and well behaved room gives the famous slightly declining response curve. This is because the room and air damping increases with frequency, and loudspeakers generally have narrower radiation at high frequencies. Reality is different and the 2 main usual deviations are loudspeaker off axis response and room uneven damping. As most of the speakers have non coincident drivers there is always a polar response error in the crossover region : for a 2 way speaker it is very common to have a dip in the 2-3 kHz region when measuring in a room even if the on-axis response is flat. The other common issue is the lack of room damping in treble range which gives a non declining frequency response. The next question is how to handle those cases ? The easy way is to go with a target curve and generate related filters. This is not what I do, nor recommend. When the issue is related to room damping for high frequencies range, it is always better to add some passive treatment : addressing low range with passive treatment is a challenge but correcting a too bright room can easily be done with curtains, carpets, sofa, furniture. Reducing treble level by filters should be the last resort. For the main other issue which is non even off axis response, I already explained my approach on my website here Handling separately the direct sound and reverberated one is unique to my processing and helps with soundstaging. This was confirmed by an experiment done at B&O : they compared 2 configurations giving the same measurement at listening position for a 2 ways speaker : one configuration where the direct sound was corrected to compensate for the lack of energy in the crossover region and the other one with a flat direct sound and the off axis error corrected by addition of a dipole speaker (description and conclusions available here Finally, the direct sound can be equalized to implement loudness correction or personal preferences. Stay tuned, something new is coming...

Hello, I have made some new changes on Room Shaper (corrections, new feature) and you can find more info here Cheers, Thierry

Hello, Some new things on the plugin side, more details available here ! Thierry PS: all plugins can be tested w/o registration

Hi all, I have made further developments for my filters generation service. Check out my new blog page.... You can test new stuff even if you don't have performed acoustics measurements. Cheers, Thierry

Hello @taraba55 Yes Jriver dsp studio can handle convolution filters and Room shaper together. Your processing power should be sufficient and you can make a test as Room shaper is fully functionnal in demo mode.

Phase plays in important role for the "ambience" generation. The sources re-positioning part of the process is linear phase. You can keep a post EQ as long as it is common for left & right channels. I like the iso loudness approach (match headphones vs loudspeaker) promoted by David Griesinger. You can try it using his plugin or my own tool available here (my tool gives the option to correct your loudspeaker to have a fully flat frequency response when determining your iso loudness curve)