the blitter is an actual device in the GPU of your computer; it is used to accelerate primarily 2D graphics and emits HF interference. move windows around while listening and if it is blitter, you will hear obvious correlations with the frequency of the noise as you do so
it has nothing to do with the power supply. i'm not saying its definitely blitter, but its a simple test to rule it out. everyone seems keen to banter about power supply noise, jitter etc, but has it been confirmed as solved?? not from what i'm reading
it has nothing to do with the power supply. i'm not saying its definitely blitter, but its a simple test to rule it out. everyone seems keen to banter about power supply noise, jitter etc, but has it been confirmed as solved?? not from what i'm reading
Last edited:
There are two discussions here, both intertwined. The first is about if it's beneficial to provide external "clean" power to the HiFace. There's a position that it should be unnecessary, but based on the numbers of users reporting that performing the mod improved matters we're now in somewhat speculative mode as to what causes the reported improvement.
The main working theory is that there is a lack of suppression on the power input of the interface. That in turn appears to be allowing noise into places where it should't be and causing degraded sound.
The other question is related to a more general case: is it worthwhile providing stable power to all audio devices linked to a pc because there is a belief that power rails are noisy. As evidence we have a Creative sound card that appears to be picking up noise correlated with the operation of the host PC.
This noise is what you are referring to as "blitter noise", although I'll admit the last time I heard of a blitter it was in reference to my Amiga 500 many moons ago.
Blitter - Wikipedia, the free encyclopedia
You say that the noise is due to the "Blitter" in the GPU and not the power supply. While the source of the noise may be the gpu, it could easily be something else too. How does the sound manifest itself on the audio output? RF? Is the GPU olluting the 12 volt rail and the noise picked up by the sound card? Does it come in via the grounding or data lines to the sound card? Something else? That's something I'd love to know a clear answer to.
Anyway, back to the question about clean power, I have a lot of experience dealing with many different audio cards and interface solutions in a studio and apart from an old crusty pc I had I've never seen anything like has been described. That points (at least in my mind) to a hardware fault in the computer, or poor design on the creative sound card.
Until we get some firmer test results or more data we've appeared to hit a wall with how much further this discussion can go.
No, that's not what the manufacturer is saying here, you have misread it. I'm not blaming you for this btw as their English expression does have considerable room for improvementThis 2.5ppm is the longer term accuracy (they say precision) of the frequency setting, not the jitter.
That could be! I did find their report a little difficult to wade through. Still, I would argue that a jitter event, which would momentarily alter frequency, can not fall out side of the specified precision.
I agree. I think in this case they were making reference to a sampling frequency error of brand x, that is as much as 10% supposedly.
I do not. However, the manufacturer reports 78 dB below fundamental and better, I highly doubt that is audible. Reasonably, the onus of proof is on those claiming the mod's performance gains.
Impressed!
It seems you are more interested in debating the mechanics of debate.
Indeed. So to short circuit that process and return the discussion to the technical rather than the philosophical, how about someone posting some actual experimental data? If these grave effects are happening, they will certainly appear in spectra of the analog outputs.
FWIW, there's no evidence of power supply issues for the PCI card I use (no untoward noise, no sidebands, no distortion components above -110dB). I do, however, see odd order harmonics of 60Hz in an external Firewire card that I've been using with the same computer, but though visible on the spectrum, the worst one is the third (180 Hz) at -90dB, and frankly, I can't hear it unless the volume is turned all the way up, nothing is playing, and I have my ear within an inch of the speaker. The spdif square wave outputs of both the PCI and Firewire cards look totally clean on my scope.
I do not have the particular interface under discussion so I won't claim that my own observations are general.
jkeny has posted up his listening impressions somewhere - not sure if here or another discussion board. Descriptions of what people hear constitute experimental data and they can be correlated between various listeners listening in different conditions.
Yet to return to philosophy for just one moment, I have the impression that you'll not be accepting of descriptions of things heard as experimental data. Or have I got the wrong end of the stick?
Who said anything about 'grave' effects? No doubt at all that there will appear some differences at the outputs, just how to acquire those spectra so as to correlate with what is heard is the tricky bit. What block size would you suggest for the FFT?
Yes, you do.
If by "block size," you mean acquisition data length, that will depend on the sample rate. You'd want sufficient frequency resolution to separate a sideband 50 Hz from the fundamental for a worst-case jitter modulation or power supply modulation. Higher harmonics are easier to resolve.
Did you notice if they were quoting the manufacturer of the crystal oscillator's spec or had they actually measured this in the circuit? (Can't recall myself). Your doubts are based on ignorance of the audibility of phase noise then?
A really weird meaning of 'reasonably'.
In jkeny's case, he's not merely claiming, he's describing the changes to the sound he perceives on the modified version. No onus on him whatsoever to 'prove' anything - rather the onus is on you to provide evidence (proof is for mathematicians) for your various claims - for example the one that opined that the mod had 'no value'. So far, none forthcoming but plenty of diversions...
At 44k1, even an acquisition length of 2048 points is more than enough to resolve sidebands. Make it safe and go for 8192. That won't keep you waiting much longer. Window won't be too critical, so just for poops and grins, use Blackman-Harris (since most cheap audio software doesn't have my favorites...).
Whoever does this should do it with and without an external supply.
Window won't be too critical, so just for poops and grins, use Blackman-Harris (since most cheap audio software doesn't have my favorites...).Whoever does this should do it with and without an external supply.
an aside...
to this discussion: the fact that most recordings these days are processed with internal sound cards/ProTools seems to me to be a good explanation for a lot of the sound quality problems with many of todays recordings, and not good evidence for why their is no bad electrical interference with digital audio inside a computer.
to this discussion: the fact that most recordings these days are processed with internal sound cards/ProTools seems to me to be a good explanation for a lot of the sound quality problems with many of todays recordings, and not good evidence for why their is no bad electrical interference with digital audio inside a computer.
Listen guys - there's so much bandwith, energy & hot air wasted on discussing this & that, looking for proof of this & that - if you are really interested in getting to the bottom of this & learning something why not try it & then start to theorise?
The O/P asked a simple question & two of you guys jumped all over the answer saying that it was pointless - & that it had to be proven that the mod would be of benefit - no proof has to be given! - you are deluding yourselves to even ask for one. The rest of the thread has been about endless to-ing & fro-ing most of which proves that you don't understand what you are talking about or are making all sorts of suppositions plucked out of the air - for instance what is phase noise & how does it relate to ppm? Does it even relate to ppm? What does noise on the PS do to a clock? You really should go away & read up before jumping on a thread with some ill conceived notion of how the digital world works.
I call all this thread-jacking. I'm gone again as I don't & won't be drawn into the inevitable round-robin, mindless responses that are going to ensue.
BTW, there is a thread on this forum in which modifying the Hiface is discussed where various people have improved the PS & reported their results - it might bear looking at?
The O/P asked a simple question & two of you guys jumped all over the answer saying that it was pointless - & that it had to be proven that the mod would be of benefit - no proof has to be given! - you are deluding yourselves to even ask for one. The rest of the thread has been about endless to-ing & fro-ing most of which proves that you don't understand what you are talking about or are making all sorts of suppositions plucked out of the air - for instance what is phase noise & how does it relate to ppm? Does it even relate to ppm? What does noise on the PS do to a clock? You really should go away & read up before jumping on a thread with some ill conceived notion of how the digital world works.
I call all this thread-jacking. I'm gone again as I don't & won't be drawn into the inevitable round-robin, mindless responses that are going to ensue.
BTW, there is a thread on this forum in which modifying the Hiface is discussed where various people have improved the PS & reported their results - it might bear looking at?
Last edited:
I did and I reported my results with PCI and Firewire. I didn't with the USB because I just don't see the reason to spend a chunk of money to buy something redundant when the folks promoting this idea can do these measurements to validate the hypothesis simply and quickly, then report their results.
Come now, there is no need for hostilities. On the subject of audio signal distortion content I have some personal experience, and have learned from others along the way. I do not doubt that phase noise can be audible, but I reasonably doubt that it is when it is 78 dB below the fundamental (and on the low end of the audio spectrum no less, where our ability to hear distortion is reduced when compared to higher frequencies).
A really weird meaning of 'reasonably'.
Such subjective data is of no value, especially considering the commercially biased source. This same logic is used by proponents of Totem Beaks, to prove their worth.
Of course there is. If you tell me you've invented a time machine, is it on you to prove it exists, or is it on the world to prove that it doesn't?
JF
I said I have scope shots of the SPDIF signal with/without battery but can't publish them yet. They clearly show a difference in the signal waveform.
When these shots are released I bet you are going to argue over whether the test set-up was rigorous enough, what the sonic effects of that difference is anyway, & why didn't such & such get measured instead.
BTW there are also scope shots showing a clearly demonstrated effect on the SPDIF of using attenuators. All these same arguments were used in an attenuator thread I started where other, different scope shots produced - stating "show that these actually reduce jitter, where is the jitter, etc - prove it to me".
It's the same old story, people with a fixed immovable view of the world obviously being discomforted by evidence that they don't like & putting up the usual set of false arguments supposedly based in science (yes they are usually Ee's or ex-Ee's) to justify their view of the world!
I even got accused of demeaning their profession in that thread when, in fact, they did a great job of demeaning it themselves 
Edit: so, lets get it out now - what do you want to see in scope shots to prove that there is a significant effect from this mod or would any gross change to the waveform show that the PS has a significant effect on the output?
When these shots are released I bet you are going to argue over whether the test set-up was rigorous enough, what the sonic effects of that difference is anyway, & why didn't such & such get measured instead.
BTW there are also scope shots showing a clearly demonstrated effect on the SPDIF of using attenuators. All these same arguments were used in an attenuator thread I started where other, different scope shots produced - stating "show that these actually reduce jitter, where is the jitter, etc - prove it to me".
It's the same old story, people with a fixed immovable view of the world obviously being discomforted by evidence that they don't like & putting up the usual set of false arguments supposedly based in science (yes they are usually Ee's or ex-Ee's) to justify their view of the world!
I even got accused of demeaning their profession in that thread
when, in fact, they did a great job of demeaning it themselves Edit: so, lets get it out now - what do you want to see in scope shots to prove that there is a significant effect from this mod or would any gross change to the waveform show that the PS has a significant effect on the output?
Last edited:
Certainly a valid question.
Changing a system in any way will have an effect, but is it a significant and positive effect?
On this, we agree.
I knew this rubbish would be spouted - it's very predictable - but one thing, I believe you said at the beginning that the mod would make absolutely no difference, now you are saying that any change will make a difference & your argument has now changed to prove to you that this is a positive difference. When all else fails you will play the commerce argument as a reason to distrust all/any of what is presented.
Edit: now you are going to try to prove that this is not what you said - it's very predictable, really
Let me quote your post before you dig yourself too deep a hole:
Edit: now you are going to try to prove that this is not what you said - it's very predictable, really
Let me quote your post before you dig yourself too deep a hole:
Last edited:
I agree.
Perhaps I did not express my position as well as I could have, but what I meant was that I suspect this modification is pointless at best. Of course anything can have an effect, but I highly doubt it can have any real, positive effect on sound quality. There is a chance the DIY supply might even make things worse.
- Status
- Not open for further replies.