Improving a HTPC audio chain.

[tuskenraider]

So I've been a very casual home theater/audio enthusiast over the past two decades, but beyond upgrading my AVR to keep up with new standards, and a two-channel amp swap here and there, I haven't gone that deep to consider it a hobby. I've also always built my own PC's during that time as well and have kept a HTPC in my main system to watch some movies with VLC, but mostly to use as a music player. Love to listen through Winamp to run Milkdrop visualization on the TV and chill with a drink. A recent speaker upgrade has dragged me deeper and I'm analyzing everything, just trying to create a better listening experience.

So I play music with Winamp software, through a WASAPI output plugin, send audio/video out of my GT 1030 video card via HDMI to my Yamaha RX-A3080, let YPAO do its thing or bypass with Direct if I choose, to a Parasound HCA-2200II amp connected to my Revel F208 mains. I've been considering a hi-res subscription service to run from the PC, or with Yamaha's Musicast software, as well. I keep reading/seeing info about DAC's, and how cheap a good quality USB model is nowadays. Since my audio is running through my AVR for room correction/eq and using its DAC(s) to decode the signal, is there a way to implement one that is superior to the Yamaha's and can bypass it? Are there any other avenues for improvement that seem more obvious to experienced users?

 

[JWAmerica]

The importance of the DAC is overrated. To benefit from a 'better' DAC you'd need a better amplifier as well. You could get a MiniDSP, use Dirac and get a nice two channel amp. Or you could upgrade to a better performing AVR.

 

[tuskenraider]

I'll have to slum it with the Parasound a bit longer. I also have an ATI 5-channel amp I recently acquired, but decided is just too much at 106 lbs. to integrate into my setup. I will likely be buying a Hypex or Purifi class D amp soon.

So while reading about, I saw that Dirac has a Windows software package and a plugin for JRiver. It just so happened I downloaded JRiver to try about a week ago. I've found it a bit cluttered compared to Winamp, and I was getting frustrated trying to simplify it to my liking, but it'll give me an avenue to try Dirac with the 2-channel trial. So I've been configuring it and got to the point I can take measurements, but.........I'm not quite sure how to make sure the Yamaha's speaker settings are in no way interfering with Dirac. I guess I'll take that to the Dirac PC thread.........thanks.

 

[JWAmerica]

I'm a big fan of keeping it simple. The easiest solution would be to switch to a Denon 3700 or 4700. Audessey XT32 room correction, still a considerable upgrade from your Yamaha, performed well on ASR measurements.

 

[tuskenraider]

Hmmm.........I happen to have a 4700 in my second system that gets little use. I was going to test it out in my main system, but my Yamaha is a bit larger, and would not fit in my secondary system center console, so I didn't. I've read the reviews here on both, and saw that Yamaha's flagship did not perform that well on measurements. Without an A/B comparison, I wonder if I could tell, so maybe it's time to find out.

 

[Sal1950]

I would agree that the weak link in your system is the Yamaha and being locked in its DRC software.
An upgrade to either a Denon/Marantz with Audyssey or one having Dirac software then allows you to do more detailed room corrections. Then you can go so far as to use REW to make measurements of their automated results and do more specific adjustment of the curves..
A pricey path but I believe a necessary one.

 

[tuskenraider]

I guess I've got to come to terms that the Yamaha is poo, even if it was a flagship. I guess I will first test Dirac on the HTPC through the Yamaha and see what that gets me. I've read up some more and caught a few videos that explain what to leave set in the AVR(speaker levels, distance and XO) when running Dirac on the PC. Then basically I can toggle YPAO and Dirac on/off to compare the difference audibly and through REW. Then I could swap in my 4700 and see how that performs, and then again with Dirac if I'm so inclined to keep chasing the rabbit.

 

[Sal1950]

I guess I've got to come to terms that the Yamaha is poo, even if it was a flagship

It's unfair to call it poo, Yamaha builds some excellent gear.
It's just that DRC has been advancing at a very fast pace and with 6-11 audio channels having excellent DRC makes all the difference for multich rigs. Sadly YPAO isn't exactly SOTA.
If you can get the Dirac software running on your PC for all available channels of the Yamaha, that might just be your end solution. I'm not at all versed in how that interface would function.

 

[Deleted member 46664]

Are you running your media players in Windows "Exclusive Mode"? If not you aren't getting the output you're probably expecting.

Some guidance here ... https://dbpages.ca/audio/9/

 

[tuskenraider]

I'm using a WASAPI driver for Winamp and JRiver Media. I can see the frequency change on the Yamaha's information display as I change through different resolutions of audio files. JRiver also has an informational tab that shows if anything in the audio signal is being modified, which it doesn't as I'm using a WASAPI output on that player as well. If I switch to a default DirectSound driver, I can see the output to the AVR shows the static bit-rate that is selected in the advanced sound properties tab regardless file resolution.

 

[Deleted member 46664]

I'm using a WASAPI driver for Winamp and JRiver Media. I can see the frequency change on the Yamaha's information display as I change through different resolutions of audio files. JRiver also has an informational tab that shows if anything in the audio signal is being modified, which it doesn't as I'm using a WASAPI output on that player as well. If I switch to a default DirectSound driver, I can see the output to the AVR shows the static bit-rate that is selected in the advanced sound properties tab regardless file resolution.

Okay that's good. However; using the Windows Audio Session API (WASAPI) does not automatically imply Exclusive Mode is activated. It would appear that your software is using the term incorrectly.

None the less that switching speeds song by song is from Exclusive Mode... which is what you want.

 

[Aerith Gainsborough]

Are there any other avenues for improvement that seem more obvious to experienced users?

As others have stated: the weak link is the lack pf proper Room Correction in your AVR.

What will make an audible difference:
- Get a subwoofer. Even with tower speakers, these often provide better bass extension. Get a model with built in DSP in order to be more independent of the AVR's Room Correction implementation.
- Get an AVR with Audyssey/Dirac Live. Especially when watching movies, this is FAR more convenient that messing around with multichannel USB DACs.

What will not make an audible difference:
- getting a "better" DAC (SINAD is hopelessly overrated)
- getting a "better" amp (see above unless you actually need more power or lower noise floor)
- messing around with windows playback protocols. People obsess over WASAPI, resampling, exclusive mode ect, yadda yadda. Ultimately, it doesn't make a lick of difference. You can measure a quality improvement but you cannot hear it in actual program material (unless some software seriously fucks up the audio stream that is).

Since you already use a PC s a source device, you could try out the 14 day demo for Dirac Live's Studio version to see if you like it, provided you have a measurement microphone. With the new standalone mode, it's easier to use than ever.

Dirac Live Room Correction Suite - Dirac

www.dirac.com

 

[Deleted member 46664]

- messing around with windows playback protocols. People obsess over WASAPI, resampling, exclusive mode ect, yadda yadda. Ultimately, it doesn't make a lick of difference. You can measure a quality improvement but you cannot hear it in actual program material (unless some software seriously fucks up the audio stream that is).

I can't vouch for your experience ... but I can for mine. Switching from Shared mode to Exclusive mode is about the same as taking a towel away from the front of the speakers. I'm no subjectivist. But in this case, on my system the difference is so night and day I don't need measurements to confirm it's real.

 

[Aerith Gainsborough]

I can't vouch for your experience ... but I can for mine. Switching from Shared mode to Exclusive mode is about the same as taking a towel away from the front of the speakers. I'm no subjectivist. But in this case, on my system the difference is so night and day so I don't need measurements to confirm it's real.

Then something is either wrong with the system settings (Windows EQ or "sonic augmentations" engaged?) or the software that you use. On my system there is no audible difference between the modes at all, I tested them extensively.

Which is why I ended up setting my Foobar to "direct sound" because I prefer the fade-in/out effects. Also, I do multitasking all the time, especially playing games and listening to music on the side. Exclusive mode simply is not practical to use for me.

 

[Deleted member 46664]

Then something is either wrong with the system settings (Windows EQ or "sonic augmentations" engaged?) or the software that you use. On my system there is no audible difference between the modes at all, I tested them extensively.

Nope ... nothing wrong with my settings. All enhancements are defeated, output is set to 48k/24b (for movies) and everything works just fine. The difference is most apparent when playing 96k Flac ... chopping it down to 48k just does not do it justice.

DB Pages : Windows Audio

 

[Aerith Gainsborough]

The difference is most apparent when playing 96k Flac ... chopping it down to 48k just does not do it justice.

Unless you do double blind ABX tests, I do not believe that.
If the resampling process is done properly, any mathematical changes to the signal should be imperceptible.

I tried to ABX it and failed miserably.

 

[Jinjuku]

I've found it a bit cluttered compared to Winamp

Winamp is a knife, Jriver is a swiss-army knife. You'll get out of JRiver what you put into it. It's that simple.

 

[Deleted member 46664]

Unless you do double blind ABX tests, I do not believe that.
If the resampling process is done properly, any mathematical changes to the signal should be imperceptible.

I tried to ABX it and failed miserably.

The active windows shared mode "audio funnel" is not simply resampling ... it also adds a "speaker protection" curve that reduces output below 50hz and incorporates soft clipping to limit output to 0dbfs, which can have a compressing effect on some content.

I'm so very happy that you are happy in shared mode.
I wasn't.

 

[Aerith Gainsborough]

The active windows shared mode "audio funnel" is not simply resampling ... it also adds a "speaker protection" curve that reduces output below 50hz and incorporates soft clipping to limit output to 0dbfs, which can have a compressing effect on some content.

I use Windows since the 98 days and this is the first I've ever head about these concerns. I know Windows CAN EQ but by default it doesn't touch the signal.

Do you have evidence for that?! Preferably per Microsoft's own documentation.
Trust me when I say, that the system performance sub 50Hz is NOT limited on my system in the slightest, nor anyone else's windows I know of. We would have noticed anything like that during the Dirac/REW measurement process. Also... my umm ... Neighbor can attest to that as well. *hides sub*
This does reek like a driver or setting issue (did you set speakers to small instead of full range?).

"Liliting to 0dBFS" Are you referring to volume normalization or actual limits?
Because you should know that there isn't anything beyond 0dB FS in the digital domain. Intersample peaks (the value goes beyond theoretical 0dBFS between two sample points) are handled by the DAC's output stage NOT Windows.

 

[Deleted member 46664]

I use Windows since the 98 days and this is the first I've ever head about these concerns. I know Windows CAN EQ but by default it doesn't touch the signal.

So far you've tried to endlessly debate every topic we've touched on...
We went through the whole clipping thing about 304,000 times when it should have been obvious the first time
I'm not going to do this again.