- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,822
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
More Rob Wattage
- Thread starter Purité Audio
- Start date
solderdude
Grand Contributor
Some things Rob Watts says are shocking but he does know a thing or two about designing good electronics.
This is why i find some of his more asinine comments particularly disappointing, he is not the usual golden eared half whit snake oil peddler but is a genuinely competent and capable engineer who does know his stuff. I expect a bit more from such an individual than from idiots who just regurgitate nonsense they're spoon fed by someone else.
OP
- Thread Starter
- #24
Today ‘Back story’ is everything !
Keith
Keith
The TI C6678 supports 64-bit floating-point and has 8 cores at 1.25 GHz. I should think it's fast enough to implement a volume control.
- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,822
As I said I don't follow the digital end of things much anymore. But in any case this numerical resolution is internal to the DSP, what DAC preserves this?
Just tried to resample a 16/44.1k flac to 32 float/2822.4k, with volume control (Amplify plugin) after upsampling, with my super outdated i3-4160. Still have 15.38x real time speed using a single thread. Though I saved to a 8-bit wav file to reduce harddrive I/O bottleneck as it is not related to processor speed, otherwise bitrate would be as high as 180.6Mbps, good enough even for 8k video LOL.
xr100
Addicted to Fun and Learning
Using Reaper's "JS: Bit/Dither w/ Noise Shaping" plug-in...
1kHz Sine at -120dBFS: (64-bit float/96kHz sample rate)
Truncated to 18-bits:
Truncated to 20-bits:
18-bits output, Noise Shaping ON, Dither OFF: (c.f. NO output when truncated to 18-bits.)
18-bits output, Noise Shaping ON, Dither (TPDF) ON:
I offer the above "as is" without implication either way as to audibility. Now, what happens with a multi-order noise-shaped feedback loop...
Oh, but your puny software volume control isn't as "sophisticated" as the mighty Watts attenuator.
Then use Miska's
xr100
Addicted to Fun and Learning
Then use Miska's
?
Someone who sells software player with 80-bit float processing. Forget about it, don't want to promote his software.
xr100
Addicted to Fun and Learning
Ah, n/m, I've found who you mean anyway. Other than the price, I actually like his software (though not used in "DSD" mode) but I'll avoid an unfortunate quagmire by leaving it there.
OP
- Thread Starter
- #34
Apologies in advance for even more Rob Wattage,
does any of this make any sense,
Quote
,What can we say about timing errors?
If you would have asked me this a few years ago, I would have said uS accuracy was needed. Now I make no such assumption - there is perhaps no limit to how good the timing of transients need to be. So how can I substantiate that bold statement? Unlike noise shapers, it's rather difficult to put a number to timing accuracy. I guess I ought to state what I mean by transient timing accuracy. I do not mean - unlike the rest of the audio business - ringing performance; this is absolutely not what I am thinking about when I talk about the time domain or timing accuracy. Ringing uses an illegal signal from sampling theory POV as it is not bandwidth limited, so you would not actually get a perfect impulse from a perfect legal bandwidth limited ADC. So why worry about a signal you will never get? So it is actually pointless talking about it. What I mean is the accuracy of the timing of transients. Imagine a bandwidth limited analogue signal that is being sampled in the ADC - it is fully negative, goes positive and at some time crosses through zero. Let us say it is sampled at 44.1 kHz, so every 22,676 nS it's sampled. Let us imagine that the signal is sampled, and then crosses through zero at exactly 20,155 nS after sampling. Of course, when it gets sampled again at +22,676 nS it will now be a positive value. The question is, when the DAC reconstructs the sampled data - converting sampled data back to a continuous analogue signal - when will the signal cross thru zero? Theory is completely clear and undeniable - if we use an infinite oversampling FIR filter with a sinc response at 22,676 nS and a perfect DAC we will reconstruct the time it crosses thru zero absolutely perfectly at 20,155 nS. But with a finite non sinc function reconstruction filter, it will not cross thru at exactly 20,155 - maybe at 19,000 nS or 21,000 nS. And it is these differences in the timing of transients, are what I am talking about. Now in the past I would have said that getting it right to a uS was perhaps OK (timing errors can be as big as 100uS in conventional filters) - now I know that instead of worrying about uS we need to worry about getting it correct to nS's.
What is the evidence for that view? In designing Dave, I wanted to discover what I had done in the Hugo design (it was a happy accident) to give me the timing performance that I so enjoyed with it. By this I mean the ability to hear the stopping and starting of notes. After trying different things, I chased down this quality to the interpolation filters after the WTA filter. Now with Hugo, I used a 16FS WTA filter, followed by a linear interpolator and a two stage IIR filter filtering up to 2048 FS. Changing this to a 256 FS WTA filter followed by my usual 3 stage filtering gave a massive change in sound quality - at this point Dave was sounding impossibly rich and smooth and almost soft sounding. By changing it to 256FS WTA gave a substantial change in character - it was still smooth, but very fast and you could hear the starting and stopping much more easily. It went in character from soft and smooth to fast and sharp - when the occasion demanded.
Now replacing the WTA from 16FS (data every 1,417 nS) to 256 FS (data every 89 nS) is technically very small in the sense that transient accuracy using a WTA against an IIR filter at this speed is not a vast change in the time domain - it is a very subtle difference, but was nonetheless extremely audible. What it tells me is that very small - impossibly small - timing errors are very significant for the brain's ability to process the ear data."
Again, I admit I cannot comment on the technical aspect of what Rob Watts says and that is why I need your help to understand your claim that "There's no point providing timing accuracy at greater than 2x our max timing resolution ability and that currently stands at 0.05ms".
does any of this make any sense,
Quote
,What can we say about timing errors?
If you would have asked me this a few years ago, I would have said uS accuracy was needed. Now I make no such assumption - there is perhaps no limit to how good the timing of transients need to be. So how can I substantiate that bold statement? Unlike noise shapers, it's rather difficult to put a number to timing accuracy. I guess I ought to state what I mean by transient timing accuracy. I do not mean - unlike the rest of the audio business - ringing performance; this is absolutely not what I am thinking about when I talk about the time domain or timing accuracy. Ringing uses an illegal signal from sampling theory POV as it is not bandwidth limited, so you would not actually get a perfect impulse from a perfect legal bandwidth limited ADC. So why worry about a signal you will never get? So it is actually pointless talking about it. What I mean is the accuracy of the timing of transients. Imagine a bandwidth limited analogue signal that is being sampled in the ADC - it is fully negative, goes positive and at some time crosses through zero. Let us say it is sampled at 44.1 kHz, so every 22,676 nS it's sampled. Let us imagine that the signal is sampled, and then crosses through zero at exactly 20,155 nS after sampling. Of course, when it gets sampled again at +22,676 nS it will now be a positive value. The question is, when the DAC reconstructs the sampled data - converting sampled data back to a continuous analogue signal - when will the signal cross thru zero? Theory is completely clear and undeniable - if we use an infinite oversampling FIR filter with a sinc response at 22,676 nS and a perfect DAC we will reconstruct the time it crosses thru zero absolutely perfectly at 20,155 nS. But with a finite non sinc function reconstruction filter, it will not cross thru at exactly 20,155 - maybe at 19,000 nS or 21,000 nS. And it is these differences in the timing of transients, are what I am talking about. Now in the past I would have said that getting it right to a uS was perhaps OK (timing errors can be as big as 100uS in conventional filters) - now I know that instead of worrying about uS we need to worry about getting it correct to nS's.
What is the evidence for that view? In designing Dave, I wanted to discover what I had done in the Hugo design (it was a happy accident) to give me the timing performance that I so enjoyed with it. By this I mean the ability to hear the stopping and starting of notes. After trying different things, I chased down this quality to the interpolation filters after the WTA filter. Now with Hugo, I used a 16FS WTA filter, followed by a linear interpolator and a two stage IIR filter filtering up to 2048 FS. Changing this to a 256 FS WTA filter followed by my usual 3 stage filtering gave a massive change in sound quality - at this point Dave was sounding impossibly rich and smooth and almost soft sounding. By changing it to 256FS WTA gave a substantial change in character - it was still smooth, but very fast and you could hear the starting and stopping much more easily. It went in character from soft and smooth to fast and sharp - when the occasion demanded.
Now replacing the WTA from 16FS (data every 1,417 nS) to 256 FS (data every 89 nS) is technically very small in the sense that transient accuracy using a WTA against an IIR filter at this speed is not a vast change in the time domain - it is a very subtle difference, but was nonetheless extremely audible. What it tells me is that very small - impossibly small - timing errors are very significant for the brain's ability to process the ear data."
Again, I admit I cannot comment on the technical aspect of what Rob Watts says and that is why I need your help to understand your claim that "There's no point providing timing accuracy at greater than 2x our max timing resolution ability and that currently stands at 0.05ms".
Why care what he says when he presents no evidence that his hypothesis is audible?
OP
- Thread Starter
- #36
Just wondered if there was any shred of technical merit , ‘transient timing accuracy’ for example,I realise he has never substantiated his ’longer tap length = better sq’ argument.
Keith
Keith
Never mind Rob , Almost turned up at your gaff looking for a bed for the night today .
Trains and wind causing issues , fortunately it's all ok now apparently... .. .
OP
- Thread Starter
- #38
Do you have wind down there?
Keith
Keith
Next thing I expect from him is to make his own rubidium atomic clock in conjunction with his bazillion taps (the clock makes the taps in sync to the timing he's babbling about) and now he's on the league with Auralic or dCS and some other companies that use a master clock snake oil box
D
Deleted member 9286
Guest
I do find it a shame that there are a lot of people here who are very closed minded to some of his ideas. I’m not saying anyone should just believe everything he says and he needs to prove stuff just like the rest of us, but it seems much more popular to try and belittle him here rather than ask him a question and debate what he is claiming. I’m not claiming he is right either, I have no idea.
He is clearly very knowledgeable though and is one of the only designers I’m aware of that also made a very good living designing DAC chips for the major players, before designing his own DACs, so I suspect he does know a thing or two. Also, if his own APX555 measurements were verified, his latest DACs would be the best measured here so far. We don’t know why there are differences between Rob’s measurements and Amir’s but I find it hard to believe he would go to the effort of measuring only to lie about the performance afterwards.
It might turn out he is talking absolute rubbish and is just like a lot of others, or we might all learn something new. It can happen. Just continually trying to bash him here seems a bit pointless.
Similar threads
