SSD Performance Degradation Caused by File Fragmentation Tested

[PeterSt]

12 minutes ago, manueljenkin said:

do the dacs that don't show measurable or audible differences with system tweaks, have actually better fidelity or are they just decorrelating the upstream noise while adding noise of their own (again sine squiggle charts and fft charts are inconclusive in this aspect).

 

Yes, a very good one. I seem to know the answer long gone, but never mind. People will be stirred up unnecessary.

 

[manueljenkin]

1 hour ago, PeterSt said:

 

Yes, a very good one. I seem to know the answer long gone, but never mind. People will be stirred up unnecessary.

 

I too know it, but it's a rigorous task to prove it, not because it's actually hard, but because there's so much aspects of the device properties/characterization/design techniques that is not available to general public, and also even if we happen to measure things, the correlation index of dependent and independent variables to our actual sound perception is quite not clear yet.

 

Trying to aim for 120db SINAD ignoring many of the other factors (most of which we don't have equipment or methods to measure) can be a sub optimal choice for audio signal reproduction. All design choices to overoptimize one area/measurement will compromise somewhere else.

[yamamoto2002]

Created programs for this experiment

 

ReadOneFileWithoutCache : Read one specified file with uncached read
https://sourceforge.net/p/playpcmwin/code/HEAD/tree/PlayPcmWin/00experiments/ReadOneFileWithoutCache/

 

WWFileFragmentationCount2 : Graphical UI to show fragmentation status of specified file on NTFS volume

https://sourceforge.net/p/playpcmwin/wiki/WWFileFragmentCount2/

 

 

[idiot_savant]

@yamamoto2002 - thanks for these, but doesn’t your uncached read only ever read 1Mb of a file?

 

 

 

Your friendly neighbourhood idiot 

[idiot_savant]

Ah, my mistake, you read 1Mb repeatedly into the same buffer. 
 

Although you do use goto, naughty boy!

 

 

your friendly neighbourhood idiot 

[yamamoto2002]

These goto are not absolute necessary and can be replaced with do { } while(false)

 

But some goto is relatively difficult to replace without sacrificing readability of the code.

 

[idiot_savant]

@manueljenkin

6 hours ago, manueljenkin said:

 

Trying to aim for 120db SINAD ignoring many of the other factors (most of which we don't have equipment or methods to measure) can be a sub optimal choice for audio signal reproduction. All design choices to overoptimize one area/measurement will compromise somewhere else.

I actually agree with most of this, apart from the bit in brackets. As to your earlier post, this is a fundamental point. Should a DAC be “sensitive”, or should it be “resolving”?
 

 

your friendly neighbourhood idiot 

[idiot_savant]

@yamamoto2002 I’m teasing you!

 

 

 

your friendly neighbourhood idiot 

[March Audio]

3 hours ago, idiot_savant said:

@manueljenkin

I actually agree with most of this, apart from the bit in brackets. As to your earlier post, this is a fundamental point. Should a DAC be “sensitive”, or should it be “resolving”?
 

 

your friendly neighbourhood idiot 

Indeed.  Whilst there is another forum that seems to focus on a sinad figure, I don't think beyond there many are fixated on such a limited measuring.