Digital transport / passthrough

[canard]

Look into the ISP subject...

you'll see the impacts...technical, even on our credibility, and how we might not be more bothered by it, etc...
it's quite terrible...

(Note that it's practically absent in the "classic" version, or only by mistake...it would be too noticeable and regrettable...)

;-)

 

[slartibartfast]

I ought to note that in the 90s I was involved in the implementation of multimedia transmission over TCP/IP, using the Internet, satellites, cable, etc.

We managed to squeeze very high bit rates for audio and video with no processing of the audio payload for audio only.

Do not conflate issues in the transmission of the data with the corruption (or manipulation) of the data itself. Typically the bit rate of the full path transmission is negotiated prior to the start of the transmission. Either you get it or you don't or get dropouts...

Notice your smart phone and Tidal HiFi. Often, out over the wireless network in the car, driving, you get the "low quality" option while at home, using WiFi you can get the "master quality" option - assuming you select that as your maximum rate.

This is all irrelevant to his point.

ISP = Inter Sample Peaks not Internet Service Provider. I've fallen for that one before

You should really look into ISPs... and realize the enormity of their impact ...

 

[canard]

Inter sample peak
(Many "audiophile" considerations are barely, if at all, measurable... whereas this... is perfectly measurable and a very well-known subject... "juste mis sous le tapis" )

So the idea of digital attenuation upstream of the DAC is a good precaution, and... bit-perfect, "a possible problem"
;-)

(Are you aware that the entire production chain has been carried out in the digital and computer field for a very long time, to an overwhelming majority? ;-) )

 

[canard]

To get off this upsetting topic "isp" bitperfect "true peak overload" etc etc...

Have you been interested in "trendy" approaches like Diretta in recent years?
You seem interested in these topics...
(Thread 'Diretta Audio Protocol' https://forum.wiimhome.com/threads/diretta-audio-protocol.9479/)
;-))

 

[TonyE]

This is all irrelevant to his point.

Hmm... so ISP is "Intersample Peaks".. not "Internet Service Provider"

Too many darn acronyms. Gotta be careful when using acronyms... I mean, I'm a networking kind of guy.

Off the cuff, I can see one simple solution. Make the DAC's analog output max output higher than the 0db of the digital mix. Say, +3db of headroom.

My Nitsch DAC outputs a 4vdc max output, which is double what the standard setting for an audio interconnect does.

Done this way you can allow for those peaks.

However, I ought to note that I spent a lot of time with recording digital.. so, yeah, my previous experience matters here as we were encoding analog for digital transmission and then doing the DAC side as well...

I've also spent quite a bit of time with Cubase... so the idea on the recording side it to ensure you always have sufficient headroom for encoding.

I'd guess that with current compressed audio that gets cranked up in volume, ISP is likely. But when recording with high dynamic range, then the problem does not exist.

Seems to be a recording volume problem.

Don't play cr@@p music.

 

[canard]

ISP anticipation must be done upstream of the "DAC" processing in the digital domain...
But there is no shortage of literature on this subject...
Because clipping occurs during this processing... It will depend on the precautions taken, the internal processing, and the relative immunity of the DAC chip to overload, etc. (Tests exist for this.)

The "bit-perfect" "audiophile" stance, when faced with 32-bit and now even 64-bit processing, is "debatable."
;-)

The simple fact of processing the signal the way we do in our Wiim systems imposes an overload indicator, or rather, given that it's an internal DAC... a true peak overload... which explains my insistence...

 

[TonyE]

How is ISP created?

Bad recording techniques?

Somehow, with a 120 db dynamic range, how in the world can you run into this unless the recording gain is set too high?

Don't play Hip Hop...

 

[pjorgenunes]

Inter sample peak
(Many "audiophile" considerations are barely, if at all, measurable... whereas this... is perfectly measurable and a very well-known subject... "juste mis sous le tapis" )

So the idea of digital attenuation upstream of the DAC is a good precaution, and... bit-perfect, "a possible problem"
;-)

(Are you aware that the entire production chain has been carried out in the digital and computer field for a very long time, to an overwhelming majority? ;-) )

I use replaygain "album" and so I have inherent attenuation to my USB DAC. No inter sample peaks above limit.

 

[canard]

Yes
But it's one possible way... but not.. "bit-perfect"
(if we ask him for enough margin)
;-)

 

[slartibartfast]

How is ISP created?

Bad recording techniques?

Somehow, with a 120 db dynamic range, how in the world can you run into this unless the recording gain is set too high?

Don't play Hip Hop...

This article from Benchmark seems to explain it well.

Intersample Overs in CD Recordings

Many oversampled DACs clip when playing CDs. This clipping is caused by a hidden defect in most oversampled DACs. Discover the cause and the cure.

benchmarkmedia.com

 

[pjorgenunes]

Yes
But it's one possible way... but not.. "bit-perfect"
(if we ask him for enough margin)
;-)

Well... bit-perfect is overrated (IMHO)

 

[canard]

(

A look at the music library: How much DAC intersample overhead is enough? True Peak analysis of genres in an eclectic digital collection.

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

)