Robbie Busting Audio Myths at CanJam NYC 2024

[IAtaman]

https://physics.stackexchange.com/questions/110540/how-loud-is-the-thermal-motion-of-air-molecules
I don't know enough to check it, but sources are cited putting it at -20 to -30 dB SPL depending on assumptions, so the Microsoft anechoic chamber isn't far off. Between that and the point where the atmosphere starts clipping we have the dynamic range of the atmosphere, around 220dB.

... atmosphere starts clipping ...

Oh wow. Never thought of that. So obvious when someone points it out - there is indeed a maximum pressure air can transfer.

From Wiki:
While 1 atm (194 dB peak or 191 dB SPL)[11][12] is the largest pressure variation an undistorted sound wave can have in Earth's atmosphere (i. e., if the thermodynamic properties of the air are disregarded; in reality, the sound waves become progressively non-linear starting over 150 dB), larger sound waves can be present in other atmospheres or other media, such as underwater or through the Earth.

 

[somebodyelse]

It's the low pressure side of the wave that has a hard limit - it can't drop below zero. The high side can go higher. For comparison, analysis of the SLS launch compared to the Saturn V:
https://pubs.aip.org/asa/jel/articl...nd-power-of-NASA-s-lunar-rockets-Space-Launch

 

[MaxwellsEq]

I give

It's the low pressure side of the wave that has a hard limit - it can't drop below zero. The high side can go higher. For comparison, analysis of the SLS launch compared to the Saturn V:
https://pubs.aip.org/asa/jel/articl...nd-power-of-NASA-s-lunar-rockets-Space-Launch

Wow, fascinating paper, thanks! A new rabbit hole beckons.

 

[AdrianusG]

Sure, In any case if you have the option like at HDTracks, don't pay more for 24/192 files. You'll never hear a difference unless
one of the files has been purposely manipulated to sound different and milk your wallet.

Hi Sal,

In how far can we trust this web side (HDTracks), as in, are their High Res files really High Res, and not up-scaled CD'quality, is there a way to check this?
I once bought an album through them (Lee Morgan - The Sidewinder 24-96) and sure enough when played with JRiver it says it's correct.

Here's the thing , if for instance i Rip a CD with DB Poweramp i could Rip it to 24-96, you would of course not get any extra information or better fidelity, just a much larger file, but JRiver would still display it as 24-96 as well, so....

 

[IAtaman]

Hi Sal,

In how far can we trust this web side (HDTracks), as in, are their High Res files really High Res, and not up-scaled CD'quality, is there a way to check this?
I once bought an album through them (Lee Morgan - The Sidewinder 24-96) and sure enough when played with JRiver it says it's correct.

Here's the thing , if for instance i Rip a CD with DB Poweramp i could Rip it to 24-96, you would of course not get any extra information or better fidelity, just a much larger file, but JRiver would still display it as 24-96 as well, so....

Well, if you can not tell by hearing, why pay more to 24-96?

 

[AdrianusG]

Well, if you can not tell by hearing, why pay more to 24-96?

I know I know, but still would like to know how to check whether High Res is really High Res.

 

[IAtaman]

I know I know, but still would like to know how to check whether High Res is really High Res.

One way might be to check for ultrasonic content in a simple audio software like Audacity I suppose. If the file has ultrasonic content, then it might be likely due to the fact that its sampled at > 44.1K.

 

[MaxwellsEq]

I know I know, but still would like to know how to check whether High Res is really High Res.

Use SOX.

 

[SharpEars]

Yep, with real optical fibers there are no limits and kilometers of multimode fiber is no problem.
There are, however, many different types of cable connectors requiring factory made cables and proper cleaning when handling the cables.
The fun part of POF cable is that you can basically DIY those cables.
Plastic optical 'fiber' (we call those light sewers in the fiber-optic world) is severely limited.
It worked fine for CD when it was first used but with multichannel and unnecessary high bit rates/depths it falls short.

If it's distances in kilometers that are being covered, it isn't being done using multimode fiber, I can assure you.

Let me quote an article on the topic, one among a plethora of similar instructional articles, that describes the differences between multimode and single mode fiber, their differences, use cases, and the costs associated with each type of fiber (and associated connectors):

Choosing Between Single Mode vs Multimode Fibers (emphasis, mine)

For these reasons, multimode optical fiber systems continue to be the most cost-effective fiber choice for enterprise and data center applications up to the 500 – 600 meter range.

See also 1000BASE-SX (upto 550 meters) and 10GBase-SR (upto 300 meters) multimode cable maximum distances. The latest bleeding edge OM5 fiber can carry 40G-SWDM4 for runs as long as 440 meters (Wow!) (See: OM5 Fiber FAQs: Must Know for High-Speed Transmission)

With regard to TOSLINK, here is a Wikipedia link and a quoted summary describing its shortcomings:

Toslink / Properties and Problems (Wikipedia)

Due to their high attenuation of light, the effective range of plastic optical cables is limited to 5–10 m.

Or the more nuanced explanation from the article being paraphrased in Wikipedia:
ePanorama.net - Audio Documents / SPDIF / Cabling Details / Third Paragraph (RE: Toslink) (emphasis, mine)

There also an optical version of S/PDIF interface which is usually called Toslink, because uses Toslink optical components. The transmission media is 1 mm plastic fiber and the signals are trasmitted using visible light (red transmitting LED). The optical signals have exactly the same format as the electrical S/PDIF signals, they are just converted to light signals (light on/off). Because high light signal attenuation in the Toslink fiberoptic cable, the transmission distance available using this technique is less than 10 meters (with some equipments only few meters).

 

[solderdude]

yep, kilometers of data kabel are usually single mode. This is because of the data bandwidth and color multiplexing.
MM cable still can be used with km lenghts but is limited in bandwidth in this case. Still more than enough for digital audio bandwidths.
Usually in MM the lenghts are short (patch cables) and the used light sources are cheap 850nm or 1300nm LEDs where longer SM uses lasers or VECSELS as lightsources.

 

[Hayabusa]

So what I was able to get from the transcript:

1. Jitter
2. RF noise flipping bits
3. RF noise correlated with your music somehow, flipping bits

I guess I won't argue that those can be audible in principle, but I am not convinced these are present or audible or make it into the output in typical use cases, since you know, I've rarely if ever seen evidence of it happening. His assertions about hearing it aren't super compelling...

flipping bits: if so, each bit has equal chance of flipping.
If it is the MSB you will know it!
So, highly unlikely I would say.

 

[SSS]

Flipped into this thread today by chance. Listened to Rob in the video. In my knowledge echnically it is all right what he states. The phenomena exist. The only question is whether it really influences audio sound quality such that it is perceivable. And there I döubt it when no big design and component flaws exist. Todays time to market is crucial so may be that a design is not worked out properly.
DC linear power supplies can also have RF-filters like switchmode ones. Resistors can be voltage dependent but normally at high voltages and not really at the low ones in a DAC, and there it must be in the circuit where the audio signal flows. Jitter in an USB connection will be cancelled by the DAC internal reference clock. RF-noise thru the USB can be a issue but when proper designed and filtered input it should be removed.

 

[BrooklynNick]

If bits are being flipped in USB Audio 2, there will be checksum errors. So this is very easy to prove and has not been demonstrated.In other words, this is BS. Show us a working USB cable that is getting a high rate of checksum errors.

 

[ahofer]

Same old same old - something is theoretically possible (and inaudible tiny) but we need gigabuck components (expensive tinfoil hats) to rid ourselves of our self-imposed anxieties.

He really should find a more honest line of work.

 

[Palladium]

If bits are being flipped in USB Audio 2, there will be checksum errors. So this is very easy to prove and has not been demonstrated.In other words, this is BS. Show us a working USB cable that is getting a high rate of checksum errors.

You see, the most common sense approach when meeting extreme pathological usage scenarios not covered by the design spec like exceeding USB cable max length is definitely throw lots of money running it out of spec.