Solid Snake-Oil Storage: This SSD Is Aimed at Audiophiles

[buzzaro]

I can't believe that people fall for such gimmicky marketing.

Check out this SSD:

Solid Snake-Oil Storage: This SSD Is Aimed at Audiophiles

This SSD purportedly makes your audio sound better.

www.tomshardware.com

One of the comments on Reddit:

"I love that audiophiles still can't figure out the difference between digital and analog data."

https://www.reddit.com/r/hardware/comments/rjfrui

I often feel bamboozled when I see cables and accessories being sold/recommended way over their normal market prices for "audiophile" features which would make no difference in blind tests and yet, they are rampant in the community.

 

[OM_2K19]

Technically an SSD (to be precise, the driver unit) can make a difference in audio fidelity. A cheap SSD can potentially inject a lot of digital noise into the audio chain, thus affecting the overall sound quality. But how much difference a 500$ and a 5000$ SSD will make that's debatable.

 

[buzzaro]

Technically an SSD (to be precise, the driver unit) can make a difference in audio fidelity. A cheap SSD can potentially inject a lot of digital noise into the audio chain, thus affecting the overall sound quality. But how much difference a 500$ and a 5000$ SSD will make that's debatable.

Can you please explain how? I genuinely want to learn and correct my preconceived biases if I have any. As far as I understand, if you store a lossless audio on any SSD, be it $50 or $500, it will only transmit that audio through your computer or other devices in binary data, and only the audio processor, either computer's or AVR's, will convert that binary data to analog and only at that stage you can expect data corruption/loss. How does "digital noise injection" work when data is bit-streamed?

 

[OM_2K19]

Can you please explain how? I genuinely want to learn and correct my preconceived biases if I have any. As far as I understand, if you store a lossless audio on any SSD, be it $50 or $500, it will only transmit that audio through your computer or other devices in binary data, and only the audio processor, either computer's or AVR's, will convert that binary data to analog and only at that stage you can expect data corruption/loss. How does "digital noise injection" work when data is bit-streamed?

In the electronics domain, 1 & 0 are nothing but voltage, usually 0 and 5v. But in practice, it will never be precisely 5v but will vary between 3.9v to 4.9v. This random fluctuation in voltage is generally known as digital noise. A pre-amp sensitive enough will pick up this noise, thus affecting the final audio quality. There are products available to suppress these noises, namely Audioquest Jitterbug.

 

[buzzaro]

In the electronics domain, 1 & 0 are nothing but voltage, usually 0 and 5v. But in practice, it will never be precisely 5v but will vary between 3.9v to 4.9v. This random fluctuation in voltage is generally known as digital noise. A pre-amp sensitive enough will pick up this noise, thus affecting the final audio quality. There are products available to suppress these noises, namely Audioquest Jitterbug.

Makes no sense to me. 1 & 0 are defined by voltages, but whether it's 3.9V or 4.9V, it is still registered and transmitted as 1 only. There's no data loss or corruption. Binary data is math in its purest form. Once you record analog audio and convert it into binary data, no matter which device it uses to transmit or store, the end result is still the same binary data. Or do you think what comes out of some cheaper digital devices and expensive digital devices is different?

 

[raghupb]

It is not the absolute voltage level of 0 or 1. High speed switching can inject noise into the ground plane.
This noise can continue to the next device, affecting SNR. SSD is a high speed switching digital device.
When connected to a streamer/player and the further to a DAC a drop in SNR can reduce the effective bit resolution of the DAC.

Cheers,
Raghu

 

[buzzaro]

It is not the absolute voltage level of 0 or 1. High speed switching can inject noise into the ground plane.
This noise can continue to the next device, affecting SNR. SSD is a high speed switching digital device.
When connected to a streamer/player and the further to a DAC a drop in SNR can reduce the effective bit resolution of the DAC.

Cheers,
Raghu

Again, I am asking in a layman's terms. How is SNR impacted by SSD and what is "injecting noise into ground plane" in binary data? Digital data loss and analog data loss are two different things. A corrupted digital data wouldn't even be processed at all. When we use long HDMI cables to transmit digital signals, the power is not often enough and thus SNR renders the signal unusable. But as long as the HDMI cables are long enough to transmit with enough power, the output digital data is the same as it was at its origin. The digital signature of this signal is not impacted.

For example, if we have two identical systems. A computer (a streamer/player) with SSD bitstreaming digital lossless audio to an audio processor (DAC/AVR etc). Identical analog output system too (same speakers). Only the SSD is different in two systems. How's the output different? The same bits are transmitted to audio processor which it then converts to analog.

 

[OM_2K19]

Makes no sense to me. 1 & 0 are defined by voltages, but whether it's 3.9V or 4.9V, it is still registered and transmitted as 1 only. There's no data loss or corruption. Binary data is math in its purest form. Once you record analog audio and convert it into binary data, no matter which device it uses to transmit or store, the end result is still the same binary data. Or do you think what comes out of some cheaper digital devices and expensive digital devices is different?

It's not about the data loss, nothing is lost but the noise introduced in the chain. The transition between 0&1 happens at a very high switching rate, and at that rate, any slight voltage fluctuation can lead to various problems. The preamp can pick these noises and affect the analog part of the chain, whereas, in the digital domain, they can affect the DAC's reproduction due to a loss in overall S/R. I hope things are clear now.

Edit:-
How s/r affects DAC's performance?

https://www.analog.com/media/en/training-seminars/tutorials/MT-013.pdf

 

[buzzaro]

The transition between 0&1 happens at a very high switching rate, and at that rate, any slight voltage fluctuation can lead to various problems. The preamp can pick these noises

This is where I am stuck. Digital data fed to the preamp (or any audio processor) is the same. It's not 0.00002 1.000002 0.00000099 ... etc. It's always 01001001.... No matter what voltage fluctuations occur, digital software/hardware cannot pick the difference between two signals of the same digital signature. (no matter how sensitive and how expensive)

For example, let's say I put a JPEG on an SSD and send it to my tablet and my OLED TV. The resulting output will be different but not because of the SSD itself, but because of the final digital to analog conversion. The data sent to both output devices is the same to the last bit of 0 or 1. There is no noise/SNR introduced at any stage no matter what different types of cables/wireless instrument i use to transmit that data to those two different output devices.

If I put a lossless audio file on an SSD, a very cheap SSD, send it wirelessly to another cheap SSD, then send it to a cheap USB using a cheap USB cable, and finally put it on an very expensive SSD which already has the same file. There won't be any difference between those two files. When I fed both these files to a preamp, it will only read the binary data as it was in the original file that was first recorded and the converted into the studio.

If I am wrong here, please cite any scientific research into this, or any blind test that can tell two digital signals apart which have the same digital signature.

 

[OM_2K19]

This is where I am stuck. Digital data fed to the preamp (or any audio processor) is the same. It's not 0.00002 1.000002 0.00000099 ... etc. It's always 01001001.... No matter what voltage fluctuations occur, digital software/hardware cannot pick the difference between two signals of the same digital signature. (no matter how sensitive and how expensive)

Digital data is fed into the DAC, and the analog output from DAC is then passed down the line (power amplifier). Any switching circuit generates line noise (we can reduce it by appropriate measures). If the SSD comes with a poorly designed digital circuit, the noise generated due to high-frequency switching has the potential to influence the analog part of the signal. Besides, it may also transmit the data stream with a poor S/R, ultimately taking its toll on DAC performance. Here we are dealing with a two-fold problem, the purity of the analog portion of the signal and the DAC's capability to what degree it can reproduce the original signal.

In the purely digital world, it's all about 0&1, and the rest doesn't matter. There will be enough guards at the protocol level to handle the data corruption due to various line noises. But this is not the case in AV applications; there is generally no error correction to reduce the processing overhead/delay. And that's why it's so sensitive. Things become pretty complicated whenever there is a change in form (A > D or D > A), and many variables come into play.

 

[mbhangui]

Everyone here is right in their own way. SSDs have dc to dc power converters @OM_2K19 and @raghupb have explained it. @buzzaro is also right. This noise shouldn't affect because most audio equipments are designed to shield noise and snake oil industry has the habit of over exagerriating the noise problem. Some chips run on 5v some on 3.3v. Integrated circuits seem to have standard voltages of 5 V, 3.3 V, 2.5 V, 1.8 V, etc. These dc to dc power converters have oscillator to convert DC to ac. AC can be easily converted from one voltage to another. You also have charge pumps, voltage multiplier circuits, etc. All of these will introduce noise typically above the audio band. In effect they will be like the noisy SMPS that computers have. It is because of this noise, high-end audio devices use linear power supplies. Even lithium battery are noisy because they use a dc to dc converter to provide a stable 5v DC.

Audiophiles should never use a hard disk directly on the device which produces the audio anyway. The standard way to do this is have a computer on a very stable power supply. This computer should have the DAC. Let us call this computer 1. The hard disk should be on a totally different computer. Let us call this computer 2.

Now let us have computer 1 and computer 2 on the same network. Now to have the best result, computer 2 should be configured with NFS or Samba. This will allow computer 1 to mount the hard disk of computer 2 and access all the files without connecting the hard disk to computer 1. So if you follow this simple scheme, the noise generating capacity of the hard disk on computer 2 doesn't matter.

Even the traditional hard disks have dc to dc convertors. So by creating an SSD with low noise power supply just to create something that can be sold exhorbitantly is 'real snake oil'. No different than power cords (which you can create yourself by using shielded cable) and audio cables which you yourself can create at home using cables with very good dielectric that gives you the least capacitance per meter.

Noise does introduce jitter which affects sound quality. Jitter is the variation in a signal's timing from its nominal value. Jitter will manifest itself as variations in phase, period, width, or duty cycle. Noise is the variation of a signal's amplitude from nominal. Both noise and jitter can cause transmission errors and increase the bit error rate. It is not just 0s and 1s. It is also about 0s and 1s arriving at the correct time.

But there are ways you can avoid it rather than over engineering a solution to target gullibles with wads of money in their pockets.

 

[mbhangui]

I think if one can't hear the noise it isn't there. Equally if one thinks one can hear noise then it is there.

There is no solution to this argument. Finally one's bank balance and one's perception of noise will drive the decision.

Exactly. Audiophiles are like wine tasters. The costlier the wine, the better it tastes. This is a physchological issue. People who hear the difference actually hear it. Our brain is a marvellous device. It creates perception based on various factors and this is very normal. No one of us are immune from it. The only thing that we should rely on are measurements.

here is a good explanation of noise and jitter. when it comes to music, it is not juse 0s and 1s. It is also about timing where if the 0s and 1s don't come at the right time, the signal waveform gets altered and hence what you get is either a compressed or a stretched waveform of the original.

The difference between noise and jitter

Noise and jitter in electronics are closely related phenomena. Both are unwanted electrical energy that we mitigate where possible and otherwise work

www.testandmeasurementtips.com

 

[superczar]

This is where I am stuck. Digital data fed to the preamp (or any audio processor) is the same. It's not 0.00002 1.000002 0.00000099 ... etc. It's always 01001001.... No matter what voltage fluctuations occur, digital software/hardware cannot pick the difference between two signals of the same digital signature. (no matter how sensitive and how expensive)

For example, let's say I put a JPEG on an SSD and send it to my tablet and my OLED TV. The resulting output will be different but not because of the SSD itself, but because of the final digital to analog conversion. The data sent to both output devices is the same to the last bit of 0 or 1. There is no noise/SNR introduced at any stage no matter what different types of cables/wireless instrument i use to transmit that data to those two different output devices.

If I put a lossless audio file on an SSD, a very cheap SSD, send it wirelessly to another cheap SSD, then send it to a cheap USB using a cheap USB cable, and finally put it on an very expensive SSD which already has the same file. There won't be any difference between those two files. When I fed both these files to a preamp, it will only read the binary data as it was in the original file that was first recorded and the converted into the studio.

If I am wrong here, please cite any scientific research into this, or any blind test that can tell two digital signals apart which have the same digital signature.

You are right - In that digital data you feed into the DAC (not the pre-amp) will get converted exactly the same regardless of whether its a 5000 INR HDD/SSD or a 5000 USD HDD/SSD.

As for how much noise the digital -> analog converter will add during the conversion is a separate debate (and can get quite animated).
But the data source quality or price itself is not of relevance

 

[OM_2K19]

Noise does introduce jitter which affects sound quality. Jitter is the variation in a signal's timing from its nominal value.

Yeah. Jitter/Timing is another crucial factor in digital audio; poor timing can kill PCM stream. Thanks for bringing that up.

 

[superczar]

Yeah. Jitter/Timing is another crucial factor in digital audio; poor timing can kill PCM stream. Thanks for bringing that up.

The catch is that the :
Data blocks (On SSD) -> PCM conversion (on CPU/ASIC) -> Digital to Analog conversion (On DAC) -> Amplification (pre amp/ Amp) -> Speakers

While arguments on Step 2 and step 3 can go on, would we not all agree that Step 1 is out of the noise/jitter/etc etc loop?

 

[square_wave]

@superczar

The data is all good.

Any potential change in noise creation and / or suppression at the digital level of your source affects sound not because the final data is any different. But how the nasties that rides along affects the final analogue sound that goes into the preamp.

This may work differently with amplfiers that work purely in the digital domain. Lyngdorf is an example.

 

[superczar]

@superczar

The data is all good.

Any potential change in noise creation and / or suppression at the digital level of your source affects sound not because the final data is any different. But how the nasties that rides along affects the final analogue sound that goes into the preamp.

This may work differently with amplfiers that work purely in the digital domain. Lyngdorf is an example.

True for amplification for sure..
arguably for Digital to analog conversion..
but is it really true at the decoding stage?

One could argue that if that indeed were the case ..then that could be easily verified by deep pixel peeping into a digital image (say jpeg or heic) decoded by a CPU and rendered on to a monitor.
The same file stored on A low grade HDD should yield a different image (even if very slightly) from a high grade one.

And finding differences here would have very little room for subjectivity as you can zoom almost endlessly down to a pixel level for an image on the spot-a-difference quest

 

[amit11]

This is where I am stuck. Digital data fed to the preamp (or any audio processor) is the same. It's not 0.00002 1.000002 0.00000099 ... etc. It's always 01001001.... No matter what voltage fluctuations occur, digital software/hardware cannot pick the difference between two signals of the same digital signature. (no matter how sensitive and how expensive)

For example, let's say I put a JPEG on an SSD and send it to my tablet and my OLED TV. The resulting output will be different but not because of the SSD itself, but because of the final digital to analog conversion. The data sent to both output devices is the same to the last bit of 0 or 1. There is no noise/SNR introduced at any stage no matter what different types of cables/wireless instrument i use to transmit that data to those two different output devices.

If I put a lossless audio file on an SSD, a very cheap SSD, send it wirelessly to another cheap SSD, then send it to a cheap USB using a cheap USB cable, and finally put it on an very expensive SSD which already has the same file. There won't be any difference between those two files. When I fed both these files to a preamp, it will only read the binary data as it was in the original file that was first recorded and the converted into the studio.

If I am wrong here, please cite any scientific research into this, or any blind test that can tell two digital signals apart which have the same digital signature.

I was also in same confusion couple of years ago. Never believed until i experimented.

The main confusion is created by word "jitter" and we start justifying how can 1s and 0s fallout. That is right.

Jitter is not much of an issue these days.
There is another culprit (the main culprit) which is "electrical noise". This has nothing to do with 1s and 0s or it's timing.

This micro voltage or micro current fluctuations cannot be effectively handled by the receiver.. and it impacts the receiver circuit in one form or the other.

Depending upon the the various permutation And combinations of electrical noise ( we can call as electrical noise profile), the impact is obviously different. The impact is subtle and can be felt in the less enjoyment of the song and many times heard aswell.

 

[square_wave]

True for amplification for sure..
arguably for Digital to analog conversion..
but is it really true at the decoding stage?

One could argue that if that indeed were the case ..then that could be easily verified by deep pixel peeping into a digital image (say jpeg or heic) decoded by a CPU and rendered on to a monitor.
The same file stored on A low grade HDD should yield a different image (even if very slightly) from a high grade one.

And finding differences here would have very little room for subjectivity as you can zoom almost endlessly down to a pixel level for an image on the spot-a-difference quest

The digital image will be the same in both the cases because the image is formed from data only.

 

[reignofchaos]

There is so much misinformation here - it is crazy. This item is pure snake oil. The SSD controller is an arm Cortex CPU and it runs sophisticated code in its firmware to make sure data integrity is maintained. It even detects and isolates any bad blocks due to wear. There is built in error correction. Otherwise what you write and what you read off an SSD will be very different which is never the case.

Timing issues occur in real-time transmission only when clock jitter can cause unrecoverable errors where there is no error correction such as spdif and to some extent USB. That is absolutely not the case here where error correction is in full force and the data pipe is nowhere close to saturation.

Here data is first read into the SSD cache, error corrected and then sent to the memory location pointed to by the host CPU. The total amount of audio data read is miniscule compared to hordes of data read for say something like a video game. If there are no issues there, for hell we can be sure there would be no issues with audio files.

Also anyone who uses SSDs extensively knows that the realtek controller on this SSD is a piece of crap. One is way better off with a WD SN850 or a Samsung 980 Pro.