I think the time has come when I can consider my “own” sound to have taken shape.
My deeper dive into this topic began when the parents of one elementary school student gave me a UMIK-1 measurement microphone for helping their child make geometric shapes for school.
Before that moment, I was quite happy using a “zoo” of the IEMs I had as my main listening option, and simple Edifier R19bt as desktop speakers.
However, after starting to try REW, I became convinced of how much better the sound could be even with the Edifier R19bt…
So, in the end, I bought Edifier MR4.5, which are an order of magnitude superior to the Edifier R19bt. After measurements in REW and “polishing” the result in DecayCore, I got a much (no, MUCH) better sound. This also affected my listening preferences – I began spending more time listening specifically through speakers.
After a very short time, I also bought an Edifier T5s subwoofer. Not that it was strictly necessary, but I was interested in the result… Here things were no longer so simple, and specifically for my conditions and listening position, I had to plug the bass reflex port with a foam plug. And despite the fact that REW immediately registered some increase in distortion in the low-frequency range, to the ear the sound began to be perceived as better – without excessive boom and “sluggishness”.
This result satisfied me quite well, and I kept the Edifier T5s as part of my system.
But some software problems began to appear that had not been noticed before because I wasn’t using FIR filters.
I’m talking about EasyEffects, which uses the zita-convolver library. And this is the cause of rather unpleasant clicks at the beginning of the next track. Moreover, it’s not constant – it can run for a day without clicks, and the next day “click” on every track. At least that’s how it is in my configuration.
CamillaDSP, however, is free of this drawback because it uses its own signal processing chain.
However, launching and configuring CamillaDSP is a whole different story… Everything is so simple with EasyEffects, and so unclear and illogical (for me) with CamillaDSP. But, calling DeepSeek in as an assistant, I solved this task too. It took a lot of time and cost me a huge number of nerve cells, but I solved it.
But a new inconvenience arose – the need to reconfigure both mpd and the local player to work with EasyEffects (using IEMs) and with CamillaDSP (listening through speakers). In the end, for the mpd use case, I wrote “on” and “off” scripts. The “on” script creates a loopback, starts CamillaDSP, its interface, opens the browser page with it, and switches mpd to use the CamillaDSP processing chain. When using the “off” script, all of this is shut down and mpd is switched to using ALSA.
So the final software chain looks like this:
· Listening with IEMs using mpd and Fooyin through EasyEffects.
· Listening through speakers using mpd and Audacious (configured to work with CamillaDSP) through CamillaDSP with the “on” script activated.
I don’t switch between IEMs and speakers every five minutes, so I don’t see any significant difficulties in using this setup.
I’m satisfied with the result. My preferences regarding the listening “scheme” have changed somewhat – now I use the speakers considerably more. But for critical listening, I still prefer IEMs…