Spec an Audiophile PC With Me Using CAPS 3.0 as a Starting Point, Adding HDMI for Multi-Channel Hi Rez Audio!!! - Page 7 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #181 of 790 Old 07-30-2013, 04:37 PM - Thread Starter
AVS Addicted Member
 
Steve Bruzonsky's Avatar
 
Join Date: Jan 1999
Location: Gilbert, Arizona
Posts: 17,612
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by Chronoptimist View Post

I thought the ALC1150 on your motherboard should already support S/PDIF out, so I would not have expected there to be any need for another output device. Most people buying sound cards are using them for their analog output capabilities rather than digital. (which is why I recommend a USB DAC)
That's very strange indeed. I wonder why 176.4kHz support is relatively rare now. My USB DAC does have support for 176.4kHz, but it still doesn't show up as an option in the Windows control panel - but if I output 176.4kHz from JRiver it plays correctly.

That said, high res audio is typically 96kHz or 192kHz. I don't think I've seen anything that is 176.4kHz PCM native - only DSD conversions. And you should not need 176.4kHz with DSD conversions, because anything above 30kHz or so is just ultrasonic noise that should have been filtered out. And if you have filtered out everything above 30kHz, there's no benefit to using 176.4kHz instead of 88.2kHz.

That's not what some of the top grade audio engineers are saying. Regarding the technical expertise behind their arguments, I leave that to them. I will tell you that I have demod converting DSD to 192 vs 176 with 176 the clear winner!##biggrin.gif

"Doug Winsor" used to troll at some AV Forums as "Steve Bruzonsky"! My home theater at:
http://www.avsforum.com/avs-vb/showthread.php?t=1158431
Steve Bruzonsky is offline  
Sponsored Links
Advertisement
 
post #182 of 790 Old 07-30-2013, 05:03 PM
RUR
Innocent Bystander
 
RUR's Avatar
 
Join Date: Feb 2007
Location: California Republic
Posts: 2,216
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 33 Post(s)
Liked: 49
Quote:
Originally Posted by Chronoptimist View Post

I don't think I've seen anything that is 176.4kHz PCM native - only DSD conversions.
Reference Recordings HRx.
RUR is offline  
post #183 of 790 Old 07-30-2013, 05:20 PM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,559
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 203
Quote:
Originally Posted by stevekale View Post

Over what path? HDMI?
No - it's a USB DAC. 24/176 is not an option in the Windows control panel, but works when I use JRiver and WASAPI Exclusive or ASIO.

CUjy3fus.png

I'm not saying that it is applicable to your ALC1150, but I thought it was a strange omission, and makes me think there's probably some reason unknown to us, that 176.4 support is missing on so much hardware. But I would not be concerned about 176.4kHz support.
Quote:
Originally Posted by stevekale View Post

That is another can of worms. Oppo would no doubt agree with you and that's why they capped their SACD output at 88.2kHz. Others will vehemently disagree.
I posted examples of this here already: http://www.avsforum.com/t/1473076/spec-an-audiophile-pc-with-me-using-caps-3-0-as-a-starting-point-adding-hdmi-for-multi-channel-hi-rez-audio/0_100#post_23376588

There's nothing up there but noise, and almost all PCM recordings will be 96kHz or 192kHz.
I have since used another program for analysis which can measure down to -144dB (24-bit) rather than the -120dB limit (20-bit) in those examples, and even at 30kHz there's low-level ultrasonic noise. There is definitely no signal above 44.1kHz with DSD. (44.1kHz being the maximum frequency 88.2kHz can handle without aliasing)

Quote:
Originally Posted by stevekale View Post

Your comment re the lack of need for an SPDIF sound card when the mobo already has an SPDIF header is much like why have a SOTM USB card when you can run USB straight off the motherboard. In both cases (and with USB I'm specifically excluding the situation where a very good USB to SPDIF converter is deployed) the additional cards have been proven to add a lot of value.
Unless the SOTM USB card is fixing a ground loop problem, I would be very surprised that it's making a difference - unless you are talking about the subjective impressions of people that have just spent $300 on a USB2 card.

As long as your digital connection works - whether it's USB, Optical, Coax, or anything else - it cannot be improved by a "cleaner" signal. It doesn't matter how good your 1's and 0's are, as long as they get there. It's your DAC which affects sound quality.
Quote:
Originally Posted by Steve Bruzonsky View Post

I will tell you that I have demod converting DSD to 192 vs 176 with 176 the clear winner!##biggrin.gif
But have you compared 192 to 88.2? There are a number of reasons why 192kHz may sound worse.
Quote:
Originally Posted by RUR View Post

Reference Recordings HRx.
Thanks, I was not aware of these. I guess it must be because these masters were created for HDCD (20/44.1) that they decided to use 176.4kHz for the masters. Still, I would put money on no-one being able to hear the difference between 176.4kHz, and playback at 88.2kHz. 88.2kHz is already more than twice the upper limit of human hearing.
Chronoptimist is offline  
post #184 of 790 Old 07-30-2013, 05:46 PM
RUR
Innocent Bystander
 
RUR's Avatar
 
Join Date: Feb 2007
Location: California Republic
Posts: 2,216
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 33 Post(s)
Liked: 49
Quote:
Originally Posted by Chronoptimist View Post

Still, I would put money on no-one being able to hear the difference between 176.4kHz, and playback at 88.2kHz. 88.2kHz is already more than twice the upper limit of human hearing.
No argument from me. Merely responding to your post on the absence of 176.4 native.
RUR is offline  
post #185 of 790 Old 07-31-2013, 03:59 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by Chronoptimist View Post

No - it's a USB DAC. 24/176 is not an option in the Windows control panel, but works when I use JRiver and WASAPI Exclusive or ASIO.

CUjy3fus.png

I'm not saying that it is applicable to your ALC1150, but I thought it was a strange omission, and makes me think there's probably some reason unknown to us, that 176.4 support is missing on so much hardware. But I would not be concerned about 176.4kHz support.

That may merely be a property of your DAC. As I said, I've not had the time to test it but 176.4kHz is not listed as a supported sample rate for the Realtek codec used by Intel over HDMI for the boards we've been discussing. Remember, Steve (and I) want multichannel over HDMI from the audio server as well as stereo over SPDIF or USB as the case may be.

Quote:
Originally Posted by Chronoptimist View Post

I posted examples of this here already: http://www.avsforum.com/t/1473076/spec-an-audiophile-pc-with-me-using-caps-3-0-as-a-starting-point-adding-hdmi-for-multi-channel-hi-rez-audio/0_100#post_23376588

There's nothing up there but noise, and almost all PCM recordings will be 96kHz or 192kHz.
I have since used another program for analysis which can measure down to -144dB (24-bit) rather than the -120dB limit (20-bit) in those examples, and even at 30kHz there's low-level ultrasonic noise. There is definitely no signal above 44.1kHz with DSD. (44.1kHz being the maximum frequency 88.2kHz can handle without aliasing)

I'm not arguing this one. In my mind there's a lot of confusion and superstition in this arena
Quote:
Originally Posted by Chronoptimist View Post

As long as your digital connection works - whether it's USB, Optical, Coax, or anything else - it cannot be improved by a "cleaner" signal. It doesn't matter how good your 1's and 0's are, as long as they get there. It's your DAC which affects sound quality.

1s, 0s and timing thereof. The latter being called jitter. What makes the Alpha USB good is (a) isolation from noise from the audio server and (b) reclocking of the 1s and 0s. With SPDIF the source is the clock. People tend to forget that digital is analogue.

Quote:
Originally Posted by Chronoptimist View Post

But have you compared 192 to 88.2? There are a number of reasons why 192kHz may sound worse.
.......... I would put money on no-one being able to hear the difference between 176.4kHz, and playback at 88.2kHz. 88.2kHz is already more than twice the upper limit of human hearing.

I agree with the first part. There are reasonable arguments for doing a DSD to PCM confusion with sample rates that divide evenly into the DSD's 2.822MHz sample rate (i.e. 44.1, 88.2 and 176.4). In my view, Steve should be comparing conversion at 88.1kHz with conversion at 176.4kHz (and not 192 versus 176.4).

Your last statement is a restatement of the Nyquist theorem (albeit a little sloppy). I guess the better question to ask is whether 88.2kHz provides enough headroom for high frequency filters without ringing within the audible range. I don't know. It's a complex arena and the workings of the inner ear are still not well understood.

But we are getting way off-track. If Steve wants 176.4kHz sample rates over HDMI he's entitled to search for a way to get there.

PS: there was a similar discussion at one stage on Whatsbestforum in Bruce Brown's section. Bruce was railing against Saracon. The core of the claim was that a spectrum analysis showed no content above 35kHz and therefore they weren't high res files. On the other hand, the only conclusion I felt one could draw from the spectrum charts was that above 35kHz was filtered away (which indeed Saracon does). I didn't feel that this meant that the underlying sample rate wasn't indeed high-res. It is also impossible to draw comparisons of the content within the audible range from the spectrum chart. (FYI you can use the freeware SOX to produce these spectrum charts.)

Which sounds better? A spectrogram of the first 30 seconds of a DSD file processed to 24/176 PCM first by Saracon and second by Audiogate (soft roll-off) (no dither in either, 24/176). The effect of the differing filters above 35k is clear but I don't think one can draw any conclusions with respect to the quality of the audible content from such charts.

Weiss Saracon



Korg Audiogate



In similar vein, were I to post spectrograms of 88.1kHz and 176.4kHz processed by the same DSD->PCM conversion, I don't think you could draw any conclusions about the quality of the sound within the audible range from the spectrograms.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #186 of 790 Old 07-31-2013, 07:32 AM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,559
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 203
Quote:
Originally Posted by stevekale View Post

That may merely be a property of your DAC. As I said, I've not had the time to test it but 176.4kHz is not listed as a supported sample rate for the Realtek codec used by Intel over HDMI for the boards we've been discussing. Remember, Steve (and I) want multichannel over HDMI from the audio server as well as stereo over SPDIF or USB as the case may be.
Yes, I know - I was just pointing out that 176.4kHz does not seem to be well supported, even on devices that have it. This makes me wonder if there's a reason why 176.4kHz support is missing on so much hardware.
Quote:
Originally Posted by stevekale View Post

1s, 0s and timing thereof. The latter being called jitter. What makes the Alpha USB good is (a) isolation from noise from the audio server and (b) reclocking of the 1s and 0s. With SPDIF the source is the clock. People tend to forget that digital is analogue.
Modern DAC designs only take the 1's and 0's and discard the timing, replacing it with their own internal clock. Source clock jitter shouldn't matter.
Quote:
Originally Posted by stevekale View Post

Your last statement is a restatement of the Nyquist theorem (albeit a little sloppy). I guess the better question to ask is whether 88.2kHz provides enough headroom for high frequency filters without ringing within the audible range.
It does. 44.1kHz is enough to filter at 20kHz and still avoid ringing/aliasing. This was really only a problem back in the 80s with the first CD players.
Quote:
Originally Posted by stevekale View Post

But we are getting way off-track. If Steve wants 176.4kHz sample rates over HDMI he's entitled to search for a way to get there.
Certainly, if that's what Steve wants. I'm just saying that 176.4kHz support isn't really all that important, and depending on your needs it may not be worth compromising and using an AMD card to get it.
The SACD specification itself requires that players have a 50kHz filter in place, because they found that most amplifiers and other audio equipment don't like signals that high frequency and it will actually introduce distortion. (it's also potentially damaging to some equipment)
So if the spec calls for a 50kHz filter, and anything above 40kHz is just noise anyway, does it really matter?
The reason that JRiver has selectable filters, with the 24kHz filter being the default, is that they found that some well regarded high-end DACs had audible distortion when playing back DSD without the ultrasonic noise filtered out.
Quote:
Originally Posted by stevekale View Post

The core of the claim was that a spectrum analysis showed no content above 35kHz and therefore they weren't high res files.
Well I don't agree with that. Anything with information above 22.05kHz is a "high res" file.

High resolution audio is largely marketing though. What matters more than anything, is the quality of the mastering. The format is far less important. It just happens that the people producing high res files, are the type of person that put a lot of care into the mastering of their recordings.
Humans can only hear up to 20kHz anyway, and most adults' hearing will be much lower than that, as we lose our higher frequency hearing with age.
Quote:
Originally Posted by stevekale View Post

FYI you can use the freeware SOX to produce these spectrum charts.
Yes, I switched to SOX which lets you set the lower limit at -144dB rather than -120dB. Thanks for the recommendation anyway.
Quote:
Originally Posted by stevekale View Post

Which sounds better? A spectrogram of the first 30 seconds of a DSD file processed to 24/176 PCM first by Saracon and second by Audiogate (soft roll-off) (no dither in either, 24/176). The effect of the differing filters above 35k is clear but I don't think one can draw any conclusions with respect to the quality of the audible content from such charts.
You are not seeing dither - you are seeing DSD's ultrasonic noise. The signal is the same in both, but Saracon has filtered out the ultrasonic noise - as it should. You will also see that there is nothing above 40kHz after filtering out the noise, which means that 88.2kHz can fully represent the signal in any DSD file.
Chronoptimist is offline  
post #187 of 790 Old 07-31-2013, 08:08 AM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by Chronoptimist View Post

Modern DAC designs only take the 1's and 0's and discard the timing, replacing it with their own internal clock. Source clock jitter shouldn't matter.

As far as I know, the majority of DAC (including "modern" designs) operate in synchronous mode for S/PDIF and AES/EBU sources and derive the master clock from the source.
edorr is offline  
post #188 of 790 Old 07-31-2013, 08:08 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by Chronoptimist View Post


Modern DAC designs only take the 1's and 0's and discard the timing, replacing it with their own internal clock. Source clock jitter shouldn't matter.

Some are better at it than others but SPDIF-input DACs are still dependent on deriving timing from the source.

Quote:
Originally Posted by Chronoptimist View Post

Well I don't agree with that. Anything with information above 22.05kHz is a "high res" file.

Don't confuse sample density with the range those samples can be.
Quote:
Originally Posted by Chronoptimist View Post

Humans can only hear up to 20kHz anyway, and most adults' hearing will be much lower than that, as we lose our higher frequency hearing with age.

Our hearing is still not particularly well understood.

This is an interesting watch. From 23.30 you will hear interesting anecdotes of engineering's understanding of hearing versus real hearing ability.

http://www.youtube.com/watch?v=1CkyrDIGzOE
Quote:
Originally Posted by Chronoptimist View Post

You are not seeing dither - you are seeing DSD's ultrasonic noise.

My reference to dither was in relation to the settings I used in Audiogate/Saracon when making the conversions.

But, again, we stray way off-topic and so I will end my contribution to this part of the discussion here.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #189 of 790 Old 07-31-2013, 12:01 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Surprise surprise surprise!!!!

So I finally got around to converting some DSD to 176.4kHz and playing it in JRiver over HDMI to the Casablanca. I should have done it much earlier rather than letting Steve upset my apple cart with claims of needing a video card to play 176.4. It plays just fine with Intel HD Audio over HDMI with Realtek ALC892. Despite not being listed as a supported sample rate, my suspicions/hopes after reading the data sheet for ALC898 which refers to 176.4kHz even though it's not listed in features proved correct. So no need for a whopping great big, power sucking video card at all. Admittedly I was playing a stereo file but the sample rate is supported and so I see no issue with multichannel. Just in case, I checked the audio path to confirm that I hadn't inadvertently turned on JRiver's DSP and downsampled on-the-fly.

To quote Steve: Ha! Ha! biggrin.gifbiggrin.gif @@@!

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #190 of 790 Old 07-31-2013, 12:12 PM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

Surprise surprise surprise!!!!

So I finally got around to converting some DSD to 176.4kHz and playing it in JRiver over HDMI to the Casablanca. I should have done it much earlier rather than letting Steve upset my apple cart with claims of needing a video card to play 176.4. It plays just fine with Intel HD Audio over HDMI with Realtek ALC892. Despite not being listed as a supported sample rate, my suspicions/hopes after reading the data sheet for ALC898 which refers to 176.4kHz even though it's not listed in features proved correct. So no need for a whopping great big, power sucking video card at all. Admittedly I was playing a stereo file but the sample rate is supported and so I see no issue with multichannel. Just in case, I checked the audio path to confirm that I hadn't inadvertently turned on JRiver's DSP and downsampled on-the-fly.

To quote Steve: Ha! Ha! biggrin.gifbiggrin.gif @@@!

You may be laughing too soon. As far as I remember, Steve had no issues with 2 channel, but could not get MCH to work - at least this is why he bought the card, not to get 176/24. So back to the lab and let us know how you fare with MCH....
edorr is offline  
post #191 of 790 Old 07-31-2013, 12:28 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
I doubt it. This was over HDMI (not USB which is presumably where Steve directed stereo files). If the features list 8 channels up to 192kHz and 176.4 works for two channels I see no reason why they'd implement the sample rate for just stereo rather than 8. But you never know until it's properly tested. I'm confident though. I can't test multichannel because of the Casablanca EDID issue.... John Baloff said the "mad scientist" in their software department may have a work around soon. I'm hoping to hear more by the end of this week.

PS: I now have to hope it works on my ASRock board which uses ALC1150...

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #192 of 790 Old 07-31-2013, 12:31 PM - Thread Starter
AVS Addicted Member
 
Steve Bruzonsky's Avatar
 
Join Date: Jan 1999
Location: Gilbert, Arizona
Posts: 17,612
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by edorr View Post

You may be laughing too soon. As far as I remember, Steve had no issues with 2 channel, but could not get MCH to work - at least this is why he bought the card, not to get 176/24. So back to the lab and let us know how you fare with MCH....

You are correct. I couldn't get multi-channel via 176-24 to work over HDMI.Tried it with my Toshiba laptop which had a NVIdia card and Intel board with Realtek. Tried it with a Dell 8500 XPS and found that its AMD video card HDMI did do multi-channel via 176.24.

"Doug Winsor" used to troll at some AV Forums as "Steve Bruzonsky"! My home theater at:
http://www.avsforum.com/avs-vb/showthread.php?t=1158431
Steve Bruzonsky is offline  
post #193 of 790 Old 07-31-2013, 12:37 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
But did you try stereo 176.4kHz over HDMI? If that worked but multichannel didn't then I haven't shown hope. But I would find it extraordinarily odd that the codec implemented 8 channels with sample rates of 44.1, 48, 88.2, 96 and 192 and implemented 176.4 for only two channels.

For me to test multichannel I either need the workaround from Theta, a Gefen HDMI Detective or loan of another processor.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #194 of 790 Old 07-31-2013, 12:39 PM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

I doubt it. This was over HDMI (not USB which is presumably where Steve directed stereo files). If the features list 8 channels up to 192kHz and 176.4 works for two channels I see no reason why they'd implement the sample rate for just stereo rather than 8. But you never know until it's properly tested. I'm confident though. I can't test multichannel because of the Casablanca EDID issue.... John Baloff said the "mad scientist" in their software department may have a work around soon. I'm hoping to hear more by the end of this week.

PS: I now have to hope it works on my ASRock board which uses ALC1150...

Clearly, if the fact that MCH over HDMI does not work is caused by a CBIII EDID issue as opposed to an Intel HDMI mobo issue, Steve (theoretically) can get the CBIII fix once available and sell his HDMI card.

On a different note, I personally have not gotten my lynx card nor Dirac live to work properly, so no matter where you look the HTPC route spells trouble and unpredictable system behavior, and endless hours of trouble shooting.
edorr is offline  
post #195 of 790 Old 07-31-2013, 12:42 PM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

But did you try stereo 176.4kHz over HDMI?

176/24 MCH is completely useless running into the CBIII, which runs at 48/24 internally. The issue was failure to do MCH all along.
edorr is offline  
post #196 of 790 Old 07-31-2013, 12:54 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
I think it's quite possible/probable that Steve's tests of multichannel with the Intel board were not because of the sample rate but the CB's EDID issue. I guess we'll find out in good time.

(You mean sell the video card, albeit it won't be worth much used. He could also sell the SOTM biggrin.gif)

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #197 of 790 Old 07-31-2013, 12:58 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by edorr View Post


On a different note, I personally have not gotten my lynx card nor Dirac live to work properly, so no matter where you look the HTPC route spells trouble and unpredictable system behavior, and endless hours of trouble shooting.

The one good thing about the Bryston was that there was a team trouble-shooting for you.

Rather than trying Dirac Live, I'd start with the convolution available in JRiver. Just use REQ to create your filters.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #198 of 790 Old 07-31-2013, 01:08 PM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

The one good thing about the Bryston was that there was a team trouble-shooting for you.

Rather than trying Dirac Live, I'd start with the convolution available in JRiver. Just use REQ to create your filters.

Just running Dirac is infinitely more simple than getting REW up and running (which is a bitch), figuring out this program, doing measurement, and then creating and exporting filters for Jriver, which is precisely why I went the Dirac route. Unfortunately, I ran into mysterious problems. Without a functioning Lynx card it is all moot anyway.

I am currently just playing the modded Oppo into Trinnov like I have been for a while, so thankfully I still have phenomenal MCH sound.
edorr is offline  
post #199 of 790 Old 07-31-2013, 11:30 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
REW is pretty easy and worth the effort to understand room acoustics better

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #200 of 790 Old 08-01-2013, 05:11 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by edorr View Post

176/24 MCH is completely useless running into the CBIII, which runs at 48/24 internally.

Yeah and this is one of the reasons why I am annoyed there isn't even a simple, proper Stereo mode. To get to 'normal stereo' we need to use Matrix mode (with other speakers set to off) which presumably runs through the DSP at 48/24. So even though the Xtreme DAC is 96kHz capable you can't play 96kHz material cleanly on Theta without adding the digital out card and buying a Gen VIII.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #201 of 790 Old 08-06-2013, 10:43 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
I thought I would post here a collection of comments made in the Theta thread that are better placed here. Hopefully the conversation can continue here.

Quote:
Originally Posted by stevekale View Post

......Not questioning the Alpha but rather why Berkeley would recommend the SOTM in conjunction with it. Berkeley have either done a good job isolating or re-clocking or they've not. From all reports, they have done a good job. It's hard to see what the SOTM adds in this case.

Quote:
Originally Posted by tyree91 View Post

Serious noise rejection for one. The SOTM has low noise Filters, and ultra low noise Regulators. The SOTM can be powered independently of the Motherboard eliminating MB hash, and the SOTM has an ultra low Jitter Clock. Now because the Berkeley re-clocks the output, you might think this doesn't matter, but think if the jitter coming out is already ultra low, how much easier it is for the Berkeley to get the lowest possible jitter, and as noise free as possible. I think Steve's answer from Berkeley says that they agree with this. Regards, Norm

Quote:
Originally Posted by stevekale View Post

They have either achieved galvanic isolation or not. They have either re-clocked the data or not. If they have re-clocked then the quality of the source clocks doesn't matter at all. While they use the source USB 5V to power the USB bus at entry, they claim to have gone to great pains to have isolated this (and hence any noise from the source) from the rest of the unit. This is their claim to fame, their raison d'être. In any event, if there is more detail from Berkeley I'd like to read about it (on the other thread).

Quote:
Originally Posted by edorr View Post

Steve, the Offramp 5 does NOT use the 5v USB power, is completely asynchronous and comes with an $800 clock upgrade and optional battery power. Yet designer Steve Nugent is very finicky about quality of the source, and even recommends mega bucks USB cables. I have heard many different explanations from digital design luminaries as to why, but the overwhelming consensus is source matters.

Quote:
Originally Posted by Empiricalaudio View Post

Megabux cables do make a difference if the interface is not galvanically isolated. They should not if it is.

Also, SOTM interface should not make any difference if galvanically isolated. Something funny there.

Lots of anecdotal evidence that the computer power supply still makes a difference in all cases though.

Even if the interface is galvanically isolated, it is more important that the clock have good regulation. Clocks specs are rarely met with typical regulators. It takes a very special regulator to come close to clock specs on jitter. 3-terminal regulators need not apply.

Steve N.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #202 of 790 Old 08-06-2013, 11:09 AM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by Empiricalaudio View Post

Megabux cables do make a difference if the interface is not galvanically isolated. They should not if it is.

Also, SOTM interface should not make any difference if galvanically isolated. Something funny there.

Lots of anecdotal evidence that the computer power supply still makes a difference in all cases though.

Even if the interface is galvanically isolated, it is more important that the clock have good regulation. Clocks specs are rarely met with typical regulators. It takes a very special regulator to come close to clock specs on jitter. 3-erminal regulators need not apply.

Steve N.

Steve, I'm a little confused here. Your OR5 is galvanically isolated, so you're saying the USB cable should make not difference. However, unless I misunderstand, you have been advocating usage of a high grade USB cable with your OR5.

I also recall you saying no converter or DAC is completely immune to jitter of incoming signal, even if asynchronous. If the SotM is a lower jitter source than the mobo USB, shouldn't the card sound better?
edorr is offline  
post #203 of 790 Old 08-06-2013, 02:15 PM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
I am also interested in Steve's response. In the interim, some of my own....

Define jitter in a USB domain. It is packeted data. If the source is not responsible for the clock or timing of the outbound S/PDIF signal (because the S/PDIF converter uses its own clock) I don't see how the concept of source jitter even applies (except that generated by the S/PDIF converter itself). The greater problem with USB has, as I understand it, been achieving true galvanic isolation between the source device and the USB converter (or DAC) particularly when the converter uses source power to run the USB input bus. (This is much easier to achieve with S/PDIF. On the Juli@ card, the Hanrun part basically serves this purpose.) Achieve it well and one is left to focus on minimising converter-generated jitter which demands high quality clocks driven by very stable power (in turn requiring very high quality regulators).

So, as I see it, you can skin the cat in two basic ways. Either pile a bunch of attention on the audio server's power supply and voltage regulators, sound card, sound card clocks (the Juli@ card clocks are pretty woeful +/- 10 to 50ppm) and sound card power supply or forget all that and focus on USB galvanic isolation and the power/clocks in the USB converter (and DAC if relevant). The first is hard to generate a solution for without controlling the full build (although some companies, e.g. SOTM, have found niches they can exploit). The second, however, provides a useful platform for providing a "solution in a box". If you get it right (and I'm sure there a lot of progress going to be made over the next few years) the source device doesn't matter. Voila, Berkeley bridges a plethora of source device configurations (all different types of computer builds of varying componentry and quality but each with the ubiquitous USB port) to the broad range of high quality DACs that exist without any or as good a USB input board. Theoretically, one likely ought to be able to get as good quality from either path. One, however, is easily to implement and, as I said, easier to productise into a solution in a box.

(I presume a bad USB cable, e.g. one collecting additional noise from other system components, might well provide a greater challenge for the USB converter components responsible for isolation and that's why one would generally prefer a better cable as a very general rule.)

FYI these are the oscillators on the ESI Juli@ card which is the backbone of Steve B's formerly preferred solution.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #204 of 790 Old 08-06-2013, 02:30 PM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

I am also interested in Steve's response. In the interim, some of my own....

Define jitter in a USB domain. It is packeted data. If the source is not responsible for the clock or timing of the outbound S/PDIF signal (because the S/PDIF converter uses its own clock) I don't see how the concept of source jitter even applies (except that generated by the S/PDIF converter itself). The greater problem with USB has, as I understand it, been achieving true galvanic isolation between the source device and the USB converter (or DAC) particularly when the converter uses source power to run the USB input bus. (This is much easier to achieve with S/PDIF. On the Juli@ card, the Hanrun part basically serves this purpose.) Achieve it well and one is left to focus on minimising converter-generated jitter which demands high quality clocks driven by very stable power (in turn requiring very high quality regulators).

So, as I see it, you can skin the cat in two basic ways. Either pile a bunch of attention on the audio server's power supply and voltage regulators, sound card, sound card clocks (the Juli@ card clocks are pretty woeful +/- 10 to 50ppm) and sound card power supply or forget all that and focus on USB galvanic isolation and the power/clocks in the USB converter (and DAC if relevant). The first is hard to generate a solution for without controlling the full build (although some companies, e.g. SOTM, have found niches they can exploit). The second, however, provides a useful platform for providing a "solution in a box". If you get it right (and I'm sure there a lot of progress going to be made over the next few years) the source device doesn't matter. Voila, Berkeley bridges a plethora of source device configurations (all different types of computer builds of varying componentry and quality but each with the ubiquitous USB port) to the broad range of high quality DACs that exist without any or as good a USB input board. Theoretically, one likely ought to be able to get as good quality from either path. One, however, is easily to implement and, as I said, easier to productise into a solution in a box.

(I presume a bad USB cable, e.g. one collecting additional noise from other system components, might well provide a greater challenge for the USB converter components responsible for isolation and that's why one would generally prefer a better cable as a very general rule.)

FYI these are the oscillators on the ESI Juli@ card which is the backbone of Steve B's formerly preferred solution.

I have seen endless dragged out debates about all things digital, and have yet to find any topic on which there is a consensus (between highly qualified experts and industry pros I might add). So I conclude there is a lot of voodoo going on in digital technology, and to keep my life simple I go with the lemmings (i.e. buy a CAPS 3.0 with battery power if that is the prevailing wisdom). It sounds pretty good, but can I vouch I can tell the difference with my old CAPS 2.0 in a double blind test - nope.

By time the illusive interface that totally isolated digital source from the DAC comes along, I will offload all my digital source gimmicks. Meanwhile, I don't worry about it too much and spend most of my time listening to music.....
edorr is offline  
post #205 of 790 Old 08-06-2013, 03:16 PM
RUR
Innocent Bystander
 
RUR's Avatar
 
Join Date: Feb 2007
Location: California Republic
Posts: 2,216
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 33 Post(s)
Liked: 49
Quote:
Originally Posted by stevekale View Post

......or forget all that and focus on USB galvanic isolation and the power/clocks in the USB converter (and DAC if relevant).

I've yet to read a persuasive argument that anything more than this is required for perfect playback, and it can be had for a few hundred $$.

 

Having said this, I have ready access to the Berkeley converter.  Maybe I'll borrow one and see if any veils are lifted.

RUR is offline  
post #206 of 790 Old 08-06-2013, 04:14 PM - Thread Starter
AVS Addicted Member
 
Steve Bruzonsky's Avatar
 
Join Date: Jan 1999
Location: Gilbert, Arizona
Posts: 17,612
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
I did upgrade my USB cable to a $50-$60 Straight Wire, which was recommended to me by Berkeley Audio. I thought I could afford that. HA!

Bruce Bowen swears by a $1500 plus USB cable from Harmonic Tech. I am not going there, got to put the lid on the $$ at some point!tongue.gif

"Doug Winsor" used to troll at some AV Forums as "Steve Bruzonsky"! My home theater at:
http://www.avsforum.com/avs-vb/showthread.php?t=1158431
Steve Bruzonsky is offline  
post #207 of 790 Old 08-07-2013, 07:04 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Quote:
Originally Posted by Steve Bruzonsky View Post

Bruce Bowen swears by a $1500 plus USB cable from Harmonic Tech. I am not going there, got to put the lid on the $$ at some point!tongue.gif

Yeah that seems a bit nuts - even for you. biggrin.gif

Steve, you are in the enviable position of being able to try a few things out should it ever pique your interest. You can bypass your SOTM card just by plugging the Berkeley into one of your motherboard's USB ports. You an also test the mobo HDMI out versus the video card (admittedly only stereo until we get a work-around from Theta).

The only "weak link" I see with your build is you don't have a high quality (by audiophile standards) power supply in the CAPS/audio server and you are using it for HDMI (a whole different kettle of fish from isolated USB). Sure you're conditioning the 110V going in but the Kingwin is producing the 12V, 5V and 3.3V. It also has a massive amount of amps for what you're running. 6A on each voltage would likely have been more than enough. (12V is mostly used by fans, optical drives etc which of course you don't have). So it's running on a very basic PSU. I wonder how that affects jitter on the HDMI port. Perhaps your video card helps in that regard but it's mostly engineered for video power and not audiophile audio quality. (As noted above, this is something you can test.) You say the quality is pretty good which is all that matters at the end of the day. Perhaps that says a lot about the need for quality mobo power or perhaps it simply is an indication of how good it might be with better power.

Having said that, sorting out a high quality linear ATX PSU is a b%tch. A Paul Hynes ATX setup will cost a bomb (and you'd have to wait until next century to receive it). I sourced a high quality ATX PSU (designed for audiophile use) but the seller wanted $900. I thought I had found a more cost-effective solution but then found the minimum order size for the transformer I needed was 5 at $90 a piece when obviously only one is needed.

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #208 of 790 Old 08-07-2013, 08:39 AM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

Yeah that seems a bit nuts - even for you. biggrin.gif

Steve, you are in the enviable position of being able to try a few things out should it ever pique your interest. You can bypass your SOTM card just by plugging the Berkeley into one of your motherboard's USB ports. You an also test the mobo HDMI out versus the video card (admittedly only stereo until we get a work-around from Theta).

The only "weak link" I see with your build is you don't have a high quality (by audiophile standards) power supply in the CAPS/audio server and you are using it for HDMI (a whole different kettle of fish from isolated USB). Sure you're conditioning the 110V going in but the Kingwin is producing the 12V, 5V and 3.3V. It also has a massive amount of amps for what you're running. 6A on each voltage would likely have been more than enough. (12V is mostly used by fans, optical drives etc which of course you don't have). So it's running on a very basic PSU. I wonder how that affects jitter on the HDMI port. Perhaps your video card helps in that regard but it's mostly engineered for video power and not audiophile audio quality. (As noted above, this is something you can test.) You say the quality is pretty good which is all that matters at the end of the day. Perhaps that says a lot about the need for quality mobo power or perhaps it simply is an indication of how good it might be with better power.

Having said that, sorting out a high quality linear ATX PSU is a b%tch. A Paul Hynes ATX setup will cost a bomb (and you'd have to wait until next century to receive it). I sourced a high quality ATX PSU (designed for audiophile use) but the seller wanted $900. I thought I had found a more cost-effective solution but then found the minimum order size for the transformer I needed was 5 at $90 a piece when obviously only one is needed.

I'm using the $500 wireworld platinum starlight USB cable. The box it comes in is to die for..... Does it sound better than a monoprice cable. Honestly don't know.....

The out of the box PS option for a caps 3 build is the Red Wine Audio battery power with has one feed for the computer and one for the SotM card. The reason Steve cannot use this is the HDMI card draws too much power. I was actually initially planning on getting a chassis with more expansion slots myself and get a state of the art gaming video card for the best video performance, but decided against it because I could then not use the battery PS, and for me audio is always more important than video.
edorr is offline  
post #209 of 790 Old 08-07-2013, 08:44 AM
Advanced Member
 
stevekale's Avatar
 
Join Date: Nov 2006
Location: London
Posts: 969
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 20
Yes but even with that the important 5V and 3.3V are produced by the PicoPSU...

Egglestonworks Andra III, Andra III Centre, Rosa (as surround). Rel Stentor II. Theta CB IV. Krell FPB 200 and two KAV 150a amps. Custom-built audio server. Oppo 103EU. Apple TV. Pioneer PDP-LX608D. Synergistic Research "Element Copper" front speaker cable. Cardas Clear Light bal interconnects.
stevekale is offline  
post #210 of 790 Old 08-07-2013, 08:50 AM
AVS Special Member
 
edorr's Avatar
 
Join Date: Feb 2009
Posts: 2,455
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 35 Post(s)
Liked: 25
Quote:
Originally Posted by stevekale View Post

Yes but even with that the important 5V and 3.3V are produced by the PicoPSU...

There is always a compromise in any approach - this forum would get awfully quiet otherwise. I personally much like the idea of pushing an order button on a website and buy a piece that can be resold for 80%of retail if I move in another direction. I don't shy away from ordering say a PS from Korea and installing it an Oppo myself if this is the only alternative, but much prefer the former approach.
edorr is offline  
Reply Ultra Hi-End HT Gear ($20,000+)

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off