What is better? HDMI audio extraction -> i2S or extra D->A->D step?

Hi everyone, I'm trying to determine the lesser of 2 evils.

I have an Onkyo AVR (RZ-820) that does a great job with connectivity, movies, network streaming, Tidal, etc but leaves a bit to be desired on the room correction/dsp front.

I’d like to add an external DIY box with a DSP (ADA1452/62 or similar), dac and a number of monoblock amps to drive my main speakers for music listening, while keeping the receiver setup for tv and movies.

I want to try my hand at building an amp based on the some of the boards that others have so carefully designed on here while using my existing hardware as a digital source. Essentially just use the onkyo as an overblown network streamer box (because i have it and already like the workflow using it)

However this receiver doesn’t have any traditional digital outputs, but it does have an HDMI to Zone 2 that can be used to output the digital audio/video.

Which do you think is the lesser evil:
1) use one of the myriad of HDMI audio extractors to get I2S from and pass that to the DSP and down the signal path... (with whatever jitter or other artifacts being introduced as the downside)

2) Use 2 channels of the analog preamp output (single ended, 1V) from the Onkyo and then use the A/D function of the DSP to redigitize.... the Onkyo has a nice DAC (AK4458) at least...

3) ????

I’m thinking that the combination of the 1V preout with redigitization is worse (it just FEELS wrong to have an extra D/A A/D in there. I know enough, however that sometimes the real-world benefit of one would surprise you.

What do you think?
 
Yea, well the key is that i need some better DSP to compensate my room. The onkyo is awesome but it doesn’t allow me to do the tweaks i want. So adding another amp to improve the sound would be basically pointless without the EQ ing i need. (I need per channel eq, oddly enough the onkyo groups both front channels together.

I’m used to the control i get in high end car dsp. Since my livingroom isn’t symmetrical, the EQ shouldn’t be either!
 
The first thing to do would be to set up the speakers properly in the room, unless perhaps something like WAF prevents it. After that would be the time to think about EQ.

Yea, i live in Silicon valley, we don’t really get to pick our housing based on room acoustics. My speakers are just about in the worst positions possible due to room/furniture/laws of physics getting in the way. Reflections off of furniture and a wall of mirrors (ugh!) with an open wall to a huge space off to the other side.

The nice thing is that having tuned a number of car stereos with sophisticated DSPs, i have a methodology that works great for me. But it requires being able to do independent EQ.

But aside from the room acoustics problem (my ‘official’ excuse), i want to build my first AB amplifier and would like to use the differential outputs from an adau1766 (or similar) as the input.