tineared

That's a pretty big if, but pretty exciting if it turns out to be possible...

Flak, could the technology be implemented over WiSA ?

For those curious, enjoy the music has a video interview with one of the guys at Dirac right here : . One of the interesting things that's brought up is the new technology being added to the software-only package.

Substantial enough processing requirements that a modern computer wouldn't have enough oomph to do the math and spit out the result to a receiver in real-time, or substantial enough that it makes more financial sense to license the technology to the OEMs before possibly trickling it down to end consumers through software ?

For the curious, here are two overlaid measurements. The dark one is from a denon calibration puck (DMS-305), the light one is from a UMIK with calibration file. The filter (and correction) from the UMIK measurements is better, especially in the highs. So, word of advice to everyone else : listen to what Flavio says, and make sure you've got a calibrated mic before you download the trial.

Well, there is a scientific explanation - a social science called psychology. Hard science would require hard numbers, which I haven't seen, anywhere (but that I'm genuinely curious in being shown). Thing is, the burden of proof is unfortunately on the claimants (namely, the cable companies and the reviewers that announce those night and day differences), which they're arguably strongly incentivised not to provide. So they pout, sulk and go all ad absurdum crying religious persecution instead. Which, I'd add, is quite understandable from a group catering to people whose hobby is, reduced ad absurdum, "buying expensive stuff".

IIRC, "flat earthers" was first used to describe Naim, which couldn't back up their subjective claims with science. It looks like it may now have been perverted by others that are unable to back their subjective claims with quantified empirical evidence but support a "forward-thinking" view of things, such as the idea that a USB cable will sound different than another, for an as-of-yet-unexplained reason-that-has-nothing-to-do-with-psychology.

And dare I ask what the name of that new player would be ? (couldn't find it on the page...)

Arjan, the decreased level of output in Audio Midi (whether I induced it, or it was induced by the Dirac installation I don't know - but, in my case, I'll merrily go for "issue between the screen and the keyboard" rather than blame people who actually know what they're doing) seemed to be part of the issue in my case. I tend to not listen at very loud levels (I haven't measured, but it's let's say something around the 8 o'clock position on the NAC82, which I realise doesn't mean much without voltage measurements from the source), and my feeling is that lowering the output from the computer is what degraded the dynamics so much. So, Audio Midi output set at 0, all attenuation with the analog, Naim pot. Sounds better now, especially on-axis (as a sidenote, I'm not sure where the guardians of online temple of the flat-earth believers from Salisbury stand as far as toe-in is concerned these days (it used to be "none at all"). If you have a bit of time to experiment, you might want to try that as well... if you don't like it, just please don't burn me at stake for suggesting it ) Flavio, would you care to let us know if dirac does indeed decrease the global system volume on macs upon installation ?

Flavio, thanks a lot for the input and the pointers ! the "puck" i referred to is a calibration mic (albeit without the accompanying file). So it's probably not completely dreadful, just a little bit (but good enough, as far as i'm concerned, to figure out that, yes, the software works, and in what I find to be a worthwhile way). The thing with airplay playback during calibration is a bit more of a bummer for people like myself who have bigger listening rooms - or input devices away from couches. It'd be nice to have the coders factor in an offset to allow everything to buffer when Airplay is used as the output device. The disconnection bug is just that - a bug, between the Dirac layer and the Os sending a signal out to the AE. One, the other, or both together, get confused when the connection disappears. No biggie, but a bit scary when you aren't expecting it.

So, I'll start with a little bit of context, to further encourage taking my observations with a grain of salt. I'm in my mid-30's, and, as such, would likely have some form of slight hearing loss from advanced age. I've also been shot at and shelled quite a bit, and been a touch too close to individuals with truly poor trigger discipline and big things that go boom, which probably hasn't helped in that regard. I also happen not to believe in the voodoo power of cables, but hey - to everyone their own. Now, my main system is a mac mini, airplay to an AE that feeds a first-gen DAC-1, going into a Naim system (NAC82-NAP250) that powers a pair of Spendor S8e's. My listening room is a sloped ceiling, with no acoustic treatment, and a large window behind the speakers. I'm certain there's better out there, but this tends to serve pretty well - the S8's are specced +-3dB, 44-20khz, -6 at 32Hz. So acceptably flat, all things considered. I also don't own a "proper" measuring mic - did the measurements with a Denon puck, because I'm cheap and didn't want to spring for a nice one before making sure what Dirac had to offer could be worth it. Now onto the stuff that could actually be of interest to you guys. The changes Dirac processing makes are NOT subtle. And in a rather good way. But they've come with a couple of issues, which I'd like to understand. First and foremost (and most scary, given the reports of blown tweeters) is that it seems to introduce a degree of harshness (or "graininess") in the highs. I'm not sure wether it's structural, or due to my measurement mic not properly picking up the higher end of the spectrum (therefore leading to some form of over-correction, or improper correction). Has anyone else had the problem ? One (albeit not necessarily totally satisfying) fix has been to not correct the EQ above the crossover point, but I'm not certain if the time-domain and EQ are linked, and if this would mean no time-domain correction above a certain point. The second (which I'm not sure is a consequence of the attenuation the output module brings into the system or of the correction itself) is that the processing seems to rob the system of quite a bit of dynamic scale. It's, clearly a lot more controlled, and the bass is there (well, comparatively MORE bass), including at lower listening volumes, but it also feels like there's less scale. Anyone know if this could be the amp running out of steam (I don't hear any clipping, though I'm worried the harshness in the treble could be just that) ? As a sidenote, it doesn't seem like the measurement app likes sending the sound out over AirPlay if you're measuring locally - most likely something to do with the buffering lag between the source and the airplay receiver. The error message it returns is somewhat ambiguous, so if you've been having the issue - just connect directly. In the weird bug category, I've had a weird issue with what appears to be the audio processor module incorrectly reconnecting to AirPlay in a case where the Airport base station crashed and had to be unplugged and re-plugged in. First, no sound, then intermittent, unattenuated static, with the DAC locking onto a signal from the AE. Not certain what it's due to, but another one of those cases where you might want to turn the volume down to avoid speaker (or hearing ) damage.