[watchnerd]

As I said, all these abstract discussions lead nowhere - just try galvanic isolation between PC & DAC to check it out for yourself in a practical way.

 
Errr...what does that have to do with software processes?

 

[watchnerd]

  I'm sure not everyone can because they don't own a few highend CD/SACD transports like me. I used to own a few Esoteric/Emm Labs highend CD/SACD transports and sound quality from computer audio was really sad comparing to those reference equipment. I wrote Fidelizer as a tool to correct some software implementation problems related to audio playback so I can enjoy the music better at quality that I can accept when playing neck to neck with those reference transports.

 
I find the comparison to CD/SACD transports odd.  That's a real apples vs oranges comparison to a computer.
 
A more logical comparison to me would be...
 
Computer HW + (Windows / OSX / Linux) + DAC
 
(you could do this on a Mac set up for multiboot)
 
...and then comparing how the sound changed with different OS and playback software, keeping the rest of the HW the same.

 

[watchnerd]

 
The best I can think of right now is I found Fidelizer solves audio playback issues like stuttering/clicks/pops for some users so it really did improve audio performance.

 
Fixing dropouts makes sense if the machine is underpowered.  But that's very different from saying it makes a machine with enough horsepower sound better.

 

[watchnerd]

   
Second, I did write technical explanation about Fidelizer in here.
 
http://www.fidelizer-audio.com/about-fidelizer/
 
and ask me out if you have any questions about my explanation. I wrote Fidelizer exactly does so I'm not really sure at what grounds I can convince you that my project will really improve audio performance in your sense. I already told you I raised the priority of audio task and provide better clock resolution, core affinity tweaks and stuff inside Windows' multimedia platform directly but you don't seem to believe me. Even foobar2000 got bombed hard as you can see here when touch subjects like this.
 

 
Okay, the 'Extremist' option doesn't make a lot of sense from a computer science point of view in the context of modern hardware.
 
It doesn't take 90% of a modern, multi-core >1 GHz CPU's time to process audio.
 
And, in fact, if you did assign 90% of resources to those processes, they're going to spend most of their time idling.
 
The 'Extremist' option really only makes logical sense if the computer is woefully underpowered by modern specs.

 

[WindowsX]

I mentioned CD/SACD transports as a reference level of sound quality. Fidelizer helps me getting closer to that level. If you have real highend transports as refernce and find computer audio lacking a lot for acceptable performance, Fidelizer may help you for Windows platform.
 
When I wrote bits are bits list, i keep the same machine requirements in mind so those who is curious and try and findout if "bits are bits" won't really affect perceivable audio performance. From my experience, I have yet to see any "bits are bits" believer owns a real highend cd/sacd transport. Well, that's given considering they believe upgrading cd/sacd transport won't change anything because bits are bits.
 
As you can see that Fidelizer can fix dropouts and stuff, that's concrete proof for Fidelizer to be effective in improving audio performance. Perceivable or not, it' seems to do good more than harm as you can see.
 
The word modern hardware sounds really boring to me personally. I've been developing system optimizer software since I was 14 and I'm turning 30 this year. I've been hearing this word as counter argument that optimizing software won't be needed for modern hardware for over a decade already. When will we get over this? Even i7 processor can get dropouts and stuff if implementation isn't done right.
 
As for Extremist optimization level and stuff, I once removed it from Fidelizer to avoid the risks and problems but a few users requested me to bring them back again because they really needed it.
 
Why don't you just give Fidelizer a try and see for yourself? If you find nothing, it'll be only few minutes of your life spent on this and I'm sure it'll be shorter than time you're going to spend in debating with me. It doesn't cause any permanent effect in free version so it's safe to try without worries.
 
Regards,
Windows X

 

[watchnerd]

 
 
The word modern hardware sounds really boring to me personally. I've been developing system optimizer software since I was 14 and I'm turning 30 this year. I've been hearing this word as counter argument that optimizing software won't be needed for modern hardware for over a decade already. When will we get over this?
 
 

 
LOL..."boring?"
 
I work in software.  When the hardware becomes powerful enough relative to the size of the workload you don't need to micro-manage processes to ensure they're completed on time unless the machine is overloaded.
 
If you need to micro-manage processes to properly handle audio, it's really time for a HW upgrade.
 

   
As you can see that Fidelizer can fix dropouts and stuff, that's concrete proof for Fidelizer to be effective in improving audio performance. 

 
It's proof that it can fix a broken stream.  It's not proof that it can make an intact stream sound better.

 
Why don't you just give Fidelizer a try and see for yourself? 

 
I don't use Windows.

 

[WindowsX]

It's boring because software developer thinks upgrading hardware will solve all performance problems instead of improving the implementation and establishment of software implementation side. We all know Windows isn't configured to work best on multimedia alone by default. Well, they did try on Windows Vista such as 4ms latency for default USB Audio driver, guarantee 100ns time slice for multimedia resource scheduling and received a lot of complaints like battery drainage, freeze and hangs for some cases so they lowered audio standards in later updates to be more compromise.
 
I don't see why you shouldn't optimize software for the better as long as it doesn't cause trouble and easy enough to use. If you don't believe in Fidelizer's principles, that's fine by me. People find different products with different principles in mind and that's the charm of this hobby after all.
 
Regards,
Windows X

 

[mmerrill99]

As I said, all these abstract discussions lead nowhere - just try galvanic isolation between PC

Errr...what does that have to do with software processes?

As I said at the start - the mechanism of operation of Fidelizer may be that it changes the spectrum of noise within the ground plane of the computer & therefore could affect a connected audio device through this ground noise entering & affecting the sensitive circuitry in the D to A stage.

You could object to this with arguments:
- you don't believe there is any noise on the ground plane of computers
- yes there is ground noise but the changing processing in the computer does not cause any change in this, it is a constant
- this changing ground noise cannot be transmitted to an audio device connected via wired connection

Using galvanic isolation between computer & DAC will show if noise from a computer has any role in sound quality. Are you really asking what it has to do with the discussion?

 

[watchnerd]

Using galvanic isolation between computer & DAC will show if noise from a computer has any role in sound quality. Are you really asking what it has to do with the discussion?

 
Yes, I'm asking, because it's a roundabout way to assess RFI.
 
If you really think RFI is a problem, you don't need Fidelizer or galvanic isolation DAC to test that.  You crank up the CPU to max with a simulated work load and use a RFI detector, which will give you a reading in dB.

 

[watchnerd]

  If you don't believe in Fidelizer's principles, that's fine by me. People find different products with different principles in mind and that's the charm of this hobby after all.

 
It's not a matter of principles, it's a matter of data.
 
I haven't seen any data other than:
 
1. It stopped drop-outs in some cases.  
 
Sure, this seems reasonable, prima facie, because the explanation is obvious, logical, and the phenomenon is unambiguous.
 
2. It makes audio sound better
 
I haven't seen any data on this, other than anecdotal user testimony, nor measurements.

 

[mmerrill99]

Using galvanic isolation between computer

Yes, I'm asking, because it's a roundabout way to assess RFI.

If you really think RFI is a problem, you don't need Fidelizer or galvanic isolation DAC to test that.  You crank up the CPU to max with a simulated work load and use a RFI detector, which will give you a reading in dB.

Archimago has already measured the noise coming over USB cables & even using his mediocre measuring equipment, he measures a 5 to 10dB drop in all noise (it's not just RF)
http://archimago.blogspot.ie/2015/05/measurements-corning-usb-3-optical.html

 

[bfreedma]

Using galvanic isolation between computer

Yes, I'm asking, because it's a roundabout way to assess RFI.

If you really think RFI is a problem, you don't need Fidelizer or galvanic isolation DAC to test that. You crank up the CPU to max with a simulated work load and use a RFI detector, which will give you a reading in dB.

Archimago has already measured the noise coming over USB cables & even using his mediocre measuring equipment, he measures a 5 to 10dB drop in all noise (it's not just RF)
http://archimago.blogspot.ie/2015/05/measurements-corning-usb-3-optical.html

First you discredit Archimego, then you use his data.

You may have missed it, but the USB noise prior to reduction was already at or below -120db. That's way below the audible noise floor.

Not that any of that is relevant to the discussion of Fidilizer changing process priority and processor affinity.

 

[watchnerd]

Archimago has already measured the noise coming over USB cables & even using his mediocre measuring equipment, he measures a 5 to 10dB drop in all noise (it's not just RF)
http://archimago.blogspot.ie/2015/05/measurements-corning-usb-3-optical.html

 
Okay, that all makes sense.
 
But if you do have RFI issues, using Fidelizer (or any other software) to address it is very roundabout, expensive, and not universally applicable.  If I have an RFI issue, I should tackle it electrically (the optical USB is pretty clever) so the problem is reduced, regardless of what software I use.

 

[watchnerd]

You may have missed it, but the USB noise prior to reduction was already at or below -120db. That's way below the audible noise floor.
 

 
I was thinking the same thing, wondering "why does this guy care about stuff that is at -120 db".
 
My assumption was that his graph scale must be miscalibrated, because he claimed to be able to hear 60 Hz hum when it was at -110 dB.
 
If the graph isn't miscalibrated, why bother?

 

[mmerrill99]

Using galvanic isolation between computer

Yes, I'm asking, because it's a roundabout way to assess RFI.

If you really think RFI is a problem, you don't need Fidelizer or galvanic isolation DAC to test that. You crank up the CPU to max with a simulated work load and use a RFI detector, which will give you a reading in dB.

Archimago has already measured the noise coming over USB cables & even using his mediocre measuring equipment, he measures a 5 to 10dB drop in all noise (it's not just RF)
http://archimago.blogspot.ie/2015/05/measurements-corning-usb-3-optical.html

First you discredit Archimego, then you use his data.

As I said if his amateur pseudo-measuremnts show a reduction in the noise floor of 5-10dB, who knows how much more a real measurement would show?

You may have missed it, but the USB noise prior to reduction was already at or below -120db. That's way below the audible noise floor.

Who said anything about hearing the noise floor directly? It's about the effect a fluctuation ground noise spectrum can have on the sensitive analogue circuits inside D to A converters - the clock & the voltage or current references - it's the secondary effect of noise fluctuation on these processes

Not that any of that is relevant to the discussion of Fidilizer changing process priority and processor affinity.

As I said, a possible mechanism for its effect is changing the noise spectrum of the computer