Jump to content

Apparently different storage sounds different


Recommended Posts



  • 1 year later...

My read of the article was it is not so much the NAS selection, as the drive selection.

 

The "worst sounding" drive was a HDD running at 7200 RPM with max read throughput 85 MBps and seek latency 22 mSec. The middle sounding drive had advertised read throughput of 112 MBps and seek latency of 22 msec ( I did not chase IOPS but they are usually in the 70 to 90 IOPS range). The "best sounding" drive was a SSD with read throughput of 300 Mbps, basically no seek latency and can sustain 4,000 IOPS.

 

It is not so much how your storage sounds, it is as a system is it up to the task you have asked it to do or not.

Edited by Cruncher
Link to comment
Share on other sites

From what I read there there seems to be differences coming from a range of different factors, the drives themselves but also disk controllers and so on.

 

Not as clear cut with drive tech' either reading what they didn't like about the Corsair SSD when compared to the Kingson SSD. So just because it is SSD doesn't seem to imply SQ will be equally better.

Link to comment
Share on other sites



I've read that some people think different brands of micro SD card sound different in Volumio on Raspberry Pi use.

For most Volumio setups on a Raspberry Pi, the OS is loaded from the micro SD card, it isn't a music storage location/media. I've also read that the OS loads into memory at boot time and the micro SD isn't accessed during normal operation.

 

It all seems a bit crazy to me, but apparently people are setting mobile phone tower equipment on fire in the belief that 5G wireless is causing/transmitting the COVID-19 outbreak.

Edited by pwstereo
Spelling & capitalisation
  • Like 4
Link to comment
Share on other sites

@muon* We test a lot of SSD at work and then it occurred to me I was not familiar with this manufacture and had not heard of "DuraClass" technology. On Wikipedia ( https://en.wikipedia.org/wiki/SandForce ) it says ... " SandForce controllers did not use DRAM for caching[2] which reduces cost and complexity compared to other SSD controllers.". Reading on this series had a bunch of problems. Then I realized the drive was released around 2010 !!

 

I am now questioning how old this article is ( circa 2010). SInce the NAS and the drives are now all 10 years old I wonder how relevant this article is.

 

I would use a current generation SSD from a market leading brand e.g Samsung or good brand of HDD in RAID and I would not expect to see any problems.

 

 

  • Like 3
Link to comment
Share on other sites

  • 1 month later...
On 06/04/2020 at 12:25 PM, pwstereo said:

I've read that some people think different brands of micro SD card sound different in Volumio on Raspberry Pi use.

For most Volumio setups on a Raspberry Pi, the OS is loaded from the micro SD card, it isn't a music storage location/media. I've also read that the OS loads into memory at boot time and the micro SD isn't accessed during normal operation.

 

It all seems a bit crazy to me, but apparently people are setting mobile phone tower equipment on fire in the belief that 5G wireless is causing/transmitting the COVID-19 outbreak.

Peter,

I have to say that I have experienced something very similar to this in using different USB sticks as a source for playing music.

I had a 16g stick and the system was sounding very good with that config but was limited in music choice.

I bought the exact same model but in a 128 gig and the sq took a noticeable dive.

After. much head scratching went back to the 16 and all was good.

The difference was quite repeatable.

We attribute it  possibly to the controller chip accessing multiple memory chips on the larger capacity device and this process potentially injecting noise/jitter into the data stream.

Pretty sure the 16gig is a single chip device.

But this is just an hypothesis with no certainty that this is the cause.

All ideas are welcome.

Please don’t come back with ‘placebo’ anybody since both I and the gent that I was listening with at the time were quite perplexed as to what was going on and it took us some time before we even thought of going back to the original 16 gig stick ----- upon which we both went Aha!

The Usb stick was way down the list on the possibilities that we discussed.

 

We then went on to experiment with usb connected SSD vs usb connected spinning drive vs small usb stick vs large usb stick and the differences are there.

Frustrating?

You betcha

Edited by rawl99
Link to comment
Share on other sites

Goodaye all

 

l really struggle with this and cant see whats going on.

In the old days and l am sure its still applicable, files get loaded into Ram before they are played.

If you dont have enough Ram l could see a difference, people seem to think there is  wholesale loss of bits between the HD and processor.

If that was the case the there would be huge computing problems everywhere.

 

What l have struck in the past is interference from other electrical source's interferring with data cable's and computers.

So maybe its a case of some of the eqiupment tested is sensitive and poorly shielded.

 

l used to build Ultra high frequency plastic welders when we had a car industry, we had to shield them well.

One customer we had l used to speak to on the phone l could always hear the interference form a high frequency welder 50m away.

 

regards Bruce

Link to comment
Share on other sites

On 06/04/2020 at 12:25 PM, pwstereo said:

I've read that some people think different brands of micro SD card sound different in Volumio on Raspberry Pi use.

In the "enjoy the music" article their drive had known problems. The advantage of that article was I had enough data to chase down a reason. These others I am just speculating why this may be the case.

 

The thing I struggle with is the amount of variation in Computer Audio.

 

For example : I bought a Digital Camera many years ago it had a SANDisk memory card and the speed was 1x. I replaced it recently with a SanDisk that speed was 1,000 x  (cheaper, bigger, faster by multiples) the camera takes photos faster (shoot, save , shoot) and scrolls faster and I would never have thought. I have two versions of Raspberry Pi. I used Volumio for a while but I moved away from it because towards the end I felt each time I upgraded the software the software quality got worse. So in one house if I look at the permutations and combinations  : I have 2 memory cards from the same manufacture who's performance varies by 100,000 %, two versions of the same microcomputer and multiple versions of Volumio (N x M x O ).  So it does not surprise me that uses on are forum are reporting different results, I don't think it should be that way, but it is possible.

 

On 29/05/2020 at 6:18 PM, sloper said:

In the old days and l am sure its still applicable, files get loaded into Ram before they are played.

I am with you that I would have thought there should not be much if any disc activity / storage dependency .

 

I while back I bought Amarra music player because it loaded the music file into RAM and played it from RAM. When I added a track to the playlist it would stop playing the music while adding the track. My takeaway was just because it plays from RAM don't mean that there was other things going on and / or it was not well written or did not have bugs etc.

On 29/05/2020 at 3:42 PM, rawl99 said:

We then went on to experiment with usb connected SSD vs usb connected spinning drive vs small usb stick vs large usb stick and the differences are there.

Curious and frustrating. Be interesting to run some diagnostic and tracing tools and see what is going on.

 

I wonder if you copied a large file from these devices to the computer (if using one ) and see if there is any performance difference and if this married up with your listening experience ?

 

 

 

 

 

 

 

Link to comment
Share on other sites



7 hours ago, Cruncher said:

For example : I bought a Digital Camera many years ago it had a SANDisk memory card and the speed was 1x. I replaced it recently with a SanDisk that speed was 1,000 x  (cheaper, bigger, faster by multiples) the camera takes photos faster (shoot, save , shoot) and scrolls faster and I would never have thought.

 

Fully expected by any IT person.  Comes down to the time it takes to read and write the data from the card between taking and viewing photos.  Nothing magic here. 

 

However it has nothing to do with how sound might change on a raspberry pi with different cards though, as ALL cards are MUCH faster than any required reading speed for audio, and anyway, it is buffered in RAM, which is the same on each rPi.

  • Like 2
Link to comment
Share on other sites

 

3 hours ago, aussievintage said:

Fully expected by any IT person.

I brought up the camera because I thought it may be a good analogy / proxy. I deliberately did not name the brand or the age of the camera so we did not go down any rat holes. What I was somewhat clumsy trying to point out was a chain is only as strong as it's weakest link, or what I usually come across system scalability / performance / reliability is determined by the weakest sub-system. Without knowing the system chain in my camera I was not expecting the result.

 

Since you are an IT person I could say that when I bought my camera, 16bit applications were still being shipped  and there may have been some 16 bit Operating Systems around. If I ran them up today it would not matter if I threw 1TByte of RAM at them there is only so much RAM they can address and take advantage of.

 

My camera is size, power and cost constrained and from what I can gather reading the Raspberry PI site one of it's key objective was low cost, power and size. I am not familiar with the hardware and software architecture of the Pi so I was hoping someone else would add their expertise.

 

@rawl99 experience with the 16 G vs 128 G USB comparison is something I feel is constrained enough that we could probably get to the bottom of it. I know members here have the expertise, but I wonder if it would look too much like their day job ?

 

I am always happy to learn from others experience.

Edited by Cruncher
replaced, 'was' to 'was not'
Link to comment
Share on other sites

20 hours ago, Cruncher said:

I know members here have the expertise, but I wonder if it would look too much like their day job ?

I doubt it .....  while it's possible this is a "IT" cause/solution (eg. condifiguration, driver, prformance, etc.) ......  at it's core, this is more of an "audio engineering" issue.... not an "IT person" issue.

Link to comment
Share on other sites



16 minutes ago, davewantsmoore said:

I doubt it .....  while it's possible this is a "IT" cause/solution (eg. condifiguration, driver, prformance, etc.) ......  at it's core, this is more of an "audio engineering" issue.... not an "IT person" issue.

I feel you give it too much credence to label it "engineering".  I have 3  rPis and numerous other devices that can play digital music files, and on NONE of them does the storage medium make ANY difference to the quality of the sound.    I am so disappointed in the quality of my ears :(  

Link to comment
Share on other sites

1 minute ago, aussievintage said:

I feel you give it too much credence to label it "engineering". 

If it did make a difference.... it would be "engineering".    Jitter, power supplies, noise, etc....  affecting the DA conversion stage, or later.

Link to comment
Share on other sites

21 minutes ago, aussievintage said:

Needs to be conclusively demonstrated first.  Why spend time speculating on causes for something based on opinions only?

That is about how I make most of my choices trying/buying stuff. Based on opinions.

Most of the gear I use is well designed but love to hear personal opinions in forums and make my judgments from there.

 

eg. If there is a lot of feedback on a Samsung memory stick sounding the best, then I may buy that one, Toshiba Hard drive, same I might look for that one.

 

Edited by rocky500
Link to comment
Share on other sites

19 minutes ago, rocky500 said:

That is about how I make most of my choices trying/buying stuff. Based on opinions.

Most of the gear I use is well designed but love to hear personal opinions in forums and make my judgments from there.

 

eg. If there is a lot of feedback on a Samsung memory stick sounding the best, then I may buy that one, Toshiba Hard drive, same I might look for that one.

 

 

Fair enough, it is, in the end up to you.  I think I made my opinion clear :) I agree with the tag applied to this thread.

  • Like 1
Link to comment
Share on other sites



Who's up for the triple blind test after 5 or 7 red wines..............................yeah/nah...............me neither.

 

Sticking with my 2tb Kingston SSD inside my Caps.   60 yr old ears are probably not the best for the assessment.

 

Happy camper here.    I'm thinking there are better ways to improve sound rather than worrying about Raid 1, 2, 3 or 4 in my NAS.

 

Regards Cazzesman

 

 

  • Like 1
Link to comment
Share on other sites

On 03/06/2020 at 10:21 AM, aussievintage said:

Needs to be conclusively demonstrated first.  Why spend time speculating on causes for something based on opinions only?

And there we start with the crap about double blind testing or measurements to prove that there is a difference.

Had to happen soon enough.

You really ought to take a road-trip around the major symphony orchestras and tell the conductors that they are all idiots.

I mean, since audio memory is only a few seconds and objective opinions don’t mean squat, how could a conductor decide when he or she is happy with the performance of the orchestra.

Maybe you should invent a ‘quality’ meter that orchestral conductors can have sitting on their podium to know when the musicians in front of them have got it right.......

 

I find it amusingly arrogant that because you don’t experience it, it cannot be so.

 

Ever driven a race car?

Ever driven a high performance go-kart.

2 cars/karts that may time the same in respect to lap time can have a very different feel in how they drive.

Some just fit in the groove and others’ one has to work really hard at.

Objectively they are the same.

But I will bet the opinions of a driver will vary a wee bit.

Tweaking suspension geometry can make a car ‘nicer’ and less fatiguing to drive without changing lap times

But it would only be an opinion so therefore obviously not worth anything.

Edited by rawl99
  • Like 1
Link to comment
Share on other sites

On 07/06/2020 at 3:57 PM, rawl99 said:

And there we start with the crap about double blind testing or measurements to prove that there is a difference.

Had to happen soon enough.

You really ought to take a road-trip around the major symphony orchestras and tell the conductors that they are all idiots.

I mean, since audio memory is only a few seconds and objective opinions don’t mean squat, how could a conductor decide when he or she is happy with the performance of the orchestra.

Maybe you should invent a ‘quality’ meter that orchestral conductors can have sitting on their podium to know when the musicians in front of them have got it right.......

Art vs science.

 

Quote

I find it amusingly arrogant that because you don’t experience it, it cannot be so.

Ever driven a race car?

Ever driven a high performance go-kart.

2 cars/karts that may time the same in respect to lap time can have a very different feel in how they drive.

Some just fit in the groove and others’ one has to work really hard at.

Objectively they are the same.

Objectively in JUST laptime.    As you have said, evrything else about the cars is different.   Objectivly different.  Quantitatively different.

Edited by davewantsmoore
Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top