POSITIVE FEEDBACK ONLINE - ISSUE 1
An Interview with Tony Taddeo
Dave Clark: Is the Digital Antidote Two compatible with the new formats?
Tony Taddeo: Yes. The Digital Antidote Two will upgrade either narrow band SACD or 24/96 to be comparable to wide band SACD or 24/192.
Dave: What is upsampling and is it compatible with the Digital Antidote Two?
Tony: Yes, the Digital Antidote Two is completely compatible with upsampling. Upsampling is a new term for an old technology. It is a variation of (subset of) oversampling.
In oversampling the filter frequency is raised to a much higher frequency that is a multiple of 44.1 kilohertz. The information remains bandwidth limited to 44.1 kilohertz, thus causing phase distortion.
In upsampling the filter frequency is raised to a higher frequency that is not a multiple of 44.1 kilohertz, but an arbitrary frequency that may be divided by 44.1, 48, or 96 kilohertz. This gives the upsampler the ability to mate dissimilar frequencies. Thus a 24/96 DAC may be used with a normal 16/44.1 CD. This cuts the cost of having to have two separate DACs. Because of this cost cutting most 24/96 DACs will be upsampling. Like an oversampler, the information is bandwidth limited to 44.1 kilohertz, and is phase distorted.
This is a point of great confusion to consumers, because they wrongly believe an upsampler somehow converts the information to 24/96. This absolutely is not true! An upsampler is just another form of oversampler. Some have suggested that an upsampler results in slightly lower jitter, but even if it does, this is not the best way to reduce jitter. Because 44.1 kilohertz is not an integer of 96 kilohertz, you do have to lose some information by upsampling. This is why the industry went to oversampling rather than upsampling over a decade ago. You could legitimately call your present oversampling CD player a synchronous upsampler.
Dave: What about interpolation? Is the Digital Antidote Two compatible with interpolators?
Tony: Many of the new 24/96 DACs employ interpolation. Yes the Digital Antidote Two is completely compatible with interpolators. The jury is still out as to the benefit of this technique. The interpolators do not add real information but "smooth out" the existing information. It is effectively a mild form of synthesis. It is marginally audible depending on the software. The reason I say "the jury is still out" is any type of synthesizer can at first sound pleasant, but in time, the "ear" begins to recognize its signature and it becomes a negative. I am not trying to knock this technology but time will tell its real benefit. My initial opinion is mildly positive, but I am still a bit apprehensive about the long term. This is my subjective opinion. I believe that a good sounding DAC will be a good sounding DAC with or without upsampling or interpolation.
Dave: Are there any negative aspects to the Digital Antidote Two?
Tony: Sure. The first and greatest negative is the price. All good things in life come at a price. With the Digital Antidote Two, it happens to be $995.00. There is also very slight shelving of frequency. It is down about one decibel at 10 kilohertz and approximately another decibel and a half at 20 kilohertz. A bat might be able to hear this, but it is unlikely an adult human being could. As the vast majority of recordings are overmiked, I am sure the bat would prefer the modest roll off.
Dave: How high a frequency response is really necessary?
Tony: This is a complex question as the frequency response of a digital audio system is impacted in two separate ways. First is the direct high frequency response of the human ear, which is a maximum of 20 kilohertz at best. This has been well documented and many millions of people have had their hearing tested. 24/96 and SACD have indirectly also provided more evidence for this. Metal dome tweeters have a nasty fifteen-decibel spike at 25 kilohertz. If people could hear this high a frequency, metal domes when played on 24/96 or SACD would be unlistenable, bordering on painful. Yet the first round of reviewing of the new formats have revealed nary a complaint about this.
The second benefit is the improved phase performance indirectly due to the extended frequency response. This is the reason the new formats sound so much better than C.D. This is also the problem the Digital Antidote Two fixes by another means.
Dave: How many Bits do I need?
Tony: Theoretically it has been shown that 108 dB at approximately 3-kilohertz frequency is the limit of human hearing. 108 dB is 18 bits of resolution. A sixteen bit CD with noise shaping at the recording can achieve this. Noise shaping at the recording level can add real information. The Digital Antidote Two does add almost one bit of real information (unlike the interpolating filters in DACs), because the Digital Antidote Two interpolates in the time domain. The information added is real, not synthesized.
Well this is theory but how about reality? You have to answer that one for yourself. HDCD can get a legitimate 20-bit resolution. The Digital Antidote Two is completely compatible with HDCD as well as interpolating DACs. The current noise floor limit of the best electronics is about 20 bits. Unfortunately so much compression is used in 99% of recordings, that this whole question is moot.
Dave: The Digital Antidote Two is another box in the chain; is this a disadvantage?
Tony: Quite the opposite. While it is true that simpler is better (and for that reason there are no switches on the Digital Antidote Two), a digital source is an exception to the rule. A digital source is an extremely noise polluted environment and an additional stage of isolation is very beneficial.
Dave: How much of a difference will I hear with the Digital Antidote Two?
Tony: This is a real tough one to answer. Typically, people who listen to analog or live music a lot hear a dramatic difference immediately. People who listen to CD exclusively find it subtle at first, then it becomes progressively more dramatic and after time they cant live without it. What really happens is that the poor phase performance of CD has caused them to be a bit tone deaf. This is temporary and once exposed to a low distortion source their hearing quickly improves. To put it on a somewhat objective scale: loudspeakers, Digital Antidote Two, power amplifier, pre-amplifier, and cables. That means loudspeakers will make the biggest difference on sound, cables the least. This is based solely on measurable differences. This is as close as one can get to objective. Clearly there is a very strong subjective component, but there is no way to honestly comment on that meaningfully.
Dave: Do you ever plan to license the technology?
Tony: Eventually, yes. We have recently seen a large growth in demand. Initially, we were somewhat disappointed in the industrys acceptance of our technology. Before the introduction of the new formats, there was no standard to compare to, and while most people liked its effect, they were not sure enough of their own judgement to wholeheartedly endorse it. The new formats have changed all that. Now one can compare a Digital Antidote Two equipped CD to identical software on the new formats. The two sound virtually identical. Obviously the technology has a very bright future. We will face a lot of opposition from various vested interests in the audio industry, but because the technology works, long term we will win out. Short term, it is probably prudent not to license, but put out the best version of the technology we can to establish the technology. Then we can license it later.
Dave: Does any other manufacturer attempt phase correction?
Tony: Yes, there are only two other companies, namely Wadia and Pioneer. Both Wadia and Pioneer Legato Link implement phase correction in the digital domain by using synthesis techniques. We believe our technology is superior to theirs. Quite simply, our technology will make CD sound like the new formats, theirs wont. This is not to say theirs wont improve CD, but not to the extent ours will.
Dave: Why does the Digital Antidote Two operate in the analog domain?
Tony: When we first looked at the problem (almost a decade ago), we considered working in the digital domain. We then realized that there are great many unwanted artifacts caused by digital processing. Working in the digital domain would only add to these artifacts.
Working in the analog domain we actually reduce these artifacts. A good example is zero crossing noise (pointed out by Ed Meitner in his interview in Positive Feedback magazine). The Digital Antidote Two reduces this noise by about 50%.