my HiFi Journey after Year-1 with ASR =)

[khaliss]

Here's a full writeup/review of the Douk Audio T8pro, in case anyone's curious of the circuitry inside & how it sounds <LINK>

 

[Oddball]

to anyone following this thread, I just updated the original post to also show the HT Atmos side of the setup in case you're curious about it as well =)

I spend countless hours trying to set up HT with Audy and REW, albeit I engaged if even more complex setup. One of the worst graphs from ART is really great and sounds amazing. It has this hump on the back, but I actually love it few times it ever kicks in. All these hours spent in the past were not really waisted altogether, but none of that legacy helps in current times.

 

[khaliss]

It's all about how you want to allocate your resources and budget. You do seem to have some not so cheap components in it. Overall planning of the system and budget allocation is important. If one can not do try-and return play for whatever reason, it is always helpful to audition showrooms that might have a newer tech and might provide some clues in that respect.

Agreed! At the end of the day, this is just one of my other hobbies (I got a couple others LOL)... and I try to stick to a certain spending limit. As much as I want a real Accuphase amp, spending $10K is just not something I'm willing to do on one piece of equipment (at least for the time being LOL)

Also, what not-so-cheap components? If you are referring to the Marantz AV-7702 Atmos processor, I actually got it for cheap (used unit, think around $350). My Martin Logan speakers are also all used, from FB marketplace... all at 25-40% of their original prices =)

 

[Oddball]

Agreed! At the end of the day, this is just one of my other hobbies (I got a couple others LOL)... and I try to stick to a certain spending limit. As much as I want a real Accuphase amp, spending $10K is just not something I'm willing to do on one piece of equipment (at least for the time being LOL)

Also, what not-so-cheap components? If you are referring to the Marantz AV-7702 Atmos processor, I actually got it for cheap (used unit, think around $350). My Martin Logan speakers are also all used, from FB marketplace... all at 25-40% of their original prices =)

You should definitively not spend much on amps. They are mostly the same within their specs.

Good speakers and subs are always a smart spend. New D&M gear with ART is amazing, but not cheap either. It's more that life than a budget. It just needs to be experienced so that one finds a way to break the budget if experience is worth it. I have invented a thousand shortcuts to breaking the budget over the years. After all, Excel worksheet can whatever you want it to do.

 

[Bland_username42069]

it's actually only 2 preamps, but the T8pro is mainly serving as an EQ+tube stage & setting a hard-cap for the PC's volume (so I don't accidently blow-up my ears LOL)

The topping also serves as a preamp

 

[khaliss]

The topping also serves as a preamp

I know it does have a Preamp mode, but in my use-case... it's just better for me to put it in DAC mode only (in DAC mode, the display shows the sampling rate instead of the volume). So in Tidal, when I'm playing a high-res track, say at 96Khz... then I know it's a high-res track =)

 

[Ruediger]

Yes of course, any 2-channel audio content (like Youtube concerts & "Tiny Desk") I watch it from my PC. I updated my original post to now also show the other side of the setup (which is the HT 6.3.4 Atmos part).

Correct, HDMI is purely video only, while all Audio gets redirected to the E70 at all times =) There is NO delay... video & audio are always in sync... my PC has plenty of horsepower to play lightweight apps (like Youtube & Tidal) & no latency at all when it comes to playing back Video & Audio. My PC does have a discreet GPU (Intel Arc A750) which has full hardware-accelerated AV1 decoding built-in... YouTube & other video streaming platforms can take advantage of the A750's AV1 codec because it offers better video quality at lower bandwidths.. and thus freeing up the CPU for other tasks, no stuttering in HD/4K.

Had no doubt the PC can do it.

More like the audio chain adding milliseconds here and there, possibly amounting to a real delay.

 

[khaliss]

I spend countless hours trying to set up HT with Audy and REW, albeit I engaged if even more complex setup. One of the worst graphs from ART is really great and sounds amazing. It has this hump on the back, but I actually love it few times it ever kicks in. All these hours spent in the past were not really waisted altogether, but none of that legacy helps in current times.

View attachment 520293

Here's my 20-200Hz measurements for my LEFT channel:
1.) Blue = Martin Logan ESL speaker only (no Klipsch sub)
2.) Yellow = Klipsch sub only (@40Hz crossover knob setting), ESL is muted out
3.) Red = ESL + Klipsch sub working together

 

[Oddball]

It's a nice graph - so don't get me wrong, and sure it took a lots of time and effort. So great job.

Only making a point that it can get tighter with newer tools. Also, this is just FQ response. Much more to see in decay or phase charts. We can't really stop the progress - it just keeps coming uninvited.

 

[khaliss]

Had no doubt the PC can do it.

More like the audio chain adding milliseconds here and there, possibly amounting to a real delay.

even though there are a couple of devices in the audio chain from the source (T8pro + C2860 pre) instead of the DAC directly plugged in to the amplifier... I observe zero delay. Both the T8pro & C2860 are full analog devices... so any latency in the audio chain would be negligible.

Now, that wouldn't be the case if say I added a DSP type of device that has an ADC-to-DAC component (ex. DBX Venu360 speaker management system, which is essentially a sophisticated multi-crossover device with room correction), for sure that has real latency, since it has to convert analog to digital, then back to analog.

 

[khaliss]

It's a nice graph - so don't get me wrong, and sure it took a lots of time and effort. So great job.

Only making a point that it can get tighter with newer tools. Also, this is just FQ response. Much more to see in decay or phase charts. We can't really stop the progress - it just keeps coming uninvited.

actually when I took this measurement (after a closer look)... I think my sub is actually out of phase LOL, I gotta check again when I get home

 

[khaliss]

UPDATE: So, I did some more tuning work with my HiFi setup, adding some acoustic wall panels (spent like $200 on Amazon, nothing crazy LOL). I heavily experimented with a hybrid approach to room correction —using traditional REW measurement techniques paired with AI analysis to help identify stubborn acoustic issues. I wanted to share the results of my latest tuning session and get the community’s thoughts on the performance of my current setup.
The Goal: Achieving a seamless handoff between the subwoofers and the mains while keeping the decay times under control in a small & cluttered listening room =X (the graphs you see posted below are just for the left channel, to keep things simple/straightforward)

1. Frequency Response & Integration ​

The crossover integration looks good (IMHO). By isolating the Sub only (Blue) and ESL only (Red), the summation (Green) shows very little phase cancellation at the 80Hz crossover point. There is a slight "house curve" preference here, with a gentle lift in the sub-bass and a very smooth, linear response through the mid-range and treble:

2. Time Domain & Energy Decay (Waterfall & Spectrogram)​

This is where the AI analysis was most helpful. The Waterfall plot shows that while I have some modal activity around 35Hz and 55Hz, the ringing is being suppressed effectively.

The Spectrogram confirms that the peak energy time is well-aligned across the spectrum, with the low-end energy dissipating quickly enough to avoid "muddy" bass(?)

3. Group Delay & Phase
Group delay is often the "hidden" culprit for slow-sounding bass. My measurements show peaks of about 100ms in the sub-30Hz region, which is expected, but it settles down significantly by the time we hit the crossover.

The phase wrap (the blue dotted line in the SPL+phase graph) is consistent through the vocal range, which has significantly improved the center image focus.

4. Impulse Response & RT60
The Impulse Response shows a very clean initial spike with minimal early reflections, suggesting the room treatment is doing its job.

The RT60 decay times (averaging around 400ms-500ms in the lower frequencies) indicate a room that is "alive" but not "cavernous"—striking a nice balance:

AI Analysis Takeaway: The AI pointed out that while my frequency response was "flat," my group delay at 85Hz was causing a slight smear that I hadn't initially prioritized. After a few more PEQ iterations, the "snap" in the mid-bass has improved noticeably.

Having a PC-based system (for streaming & REW + EQ+APO software-based DSP) is really convenient. I can do changes on the fly, & I don't have to stick my hands on different devices to do one thing or another... everything is done via just one system... if that makes sense. Thoughts?

 

[Multicore]

thanks, it took many many hours of swapping, tweaking, measuring, etc to get to this point... it's been a fun learning experience though, I gotta admit

Can you or anyone else say what difference these many hours made in your subjective experience of music, movies, games or whatever it is you listen to?

 

[khaliss]

Can you or anyone else say what difference these many hours made in your subjective experience of music, movies, games or whatever it is you listen to?

it does make a whole lot of difference, at least for me (and tinkering is fun too, part of the hobby, right?). Here's a comparison graph of the LEFT channel (before vs after several hours/days of working the system haha):

 

[Multicore]

it does make a whole lot of difference, at least for me (and tinkering is fun too, part of the hobby, right?). Here's a comparison graph of the LEFT channel (before vs after several hours/days of working the system haha):
View attachment 523314

I saw how good your charts look but that's not what I was asking about. Can you describe the subjective experience of the difference these changes made in the sound?

 

[khaliss]

I saw how good your charts look but that's not what I was asking about. Can you describe the subjective experience of the difference these changes made in the sound?

Ahh, right on! So, comparing before all the REW measurements, physical adjustments to the system, applying PEQ via EQ-APO... the system sounded already good to my ears just by adding the subwoofers & applying equalization from the T8pro... note that the size of my listening room (~350sqft) and given how cluttered it is (& no acoustic treatments at all), I was definitely experiencing audible issues.

After all the tuning work was done (& adding a modest amount of acoustic panels)... here are the three major (subjective) differences I’m noticing in my system's current state:

*The 'Disappearing' Subwoofer: Before tuning, I could 'point' to where my subs were. There was a slight thickness or 'chewiness' around 80Hz that made the bass feel like it was sitting just behind the main soundstage (my subs were also positioned about a foot behind my ESLs, now they are side by side with the subs about 2 inches forward). Now, the integration is tight that the bass feels like it’s actually coming from the ESL panels themselves... also seems time-aligned with the ESLs =)

*Soundstage Focus & Imaging: Previously, the center image was broad but a bit 'blurry'—vocals felt like they were 4 feet wide. By using the AI to help me minimize the group delay and align the phase wraps (as seen in the SPL+phase graph), the phantom center has locked in. Vocals are now pinpoint-accurate and sit at a specific height and depth in the room.

*Clarity in the 'Mud' Zone: I used to have a bit of a 'one-note bass' issue where certain bass guitar lines would bloom and overpower the lower-mids of the panels. Looking at the Waterfall and Spectrogram, I was able to identify exactly which room modes were ringing (with the help of AI interpreting the data for me in a way that I could understand). Cleaning those up with EQ APO has also cleaned up the midrange. I’m hearing textures in voices... a good example is listening to a video recording by Depeche Mode ("Heaven", Live Studio Session... look for it on Tidal), felt like I was there in the studio listening to them in person. The sound from Martin Gore's guitar work is just out of this world... I suggest you check it out =)

The AI was particularly helpful in moving me past just flattening the EQ curve and actually focusing on the timing issues that I was hearing but couldn't quite put my finger on.

Sub & ESL panel side-by-side placement:

 

[khaliss]

by the way, here is my Right channel (ESL+sub, PEQ tuned + acoustic panels):