Hi folks,
Thanks, Tom, for your perspective as it is helpful for people who haven't had a chance to hear the dAck! yet. Viggen, we should be able to get your demo unit out to you shortly. And don't worry, it'll be brand new. I thought I'd clarify a few things for you guys regarding the battery operation because you guys seem to have some questions about this topics.
The 4 hour limit is RECOMMENDED, a cautionary and conservative estimate that is palatable to a wide range of listening styles. This does not mean that you will destroy the batteries by playing it 6 hours between sessions. Or 8 hours. In the 6 hour case, it will give you instead of 800 charges, perhaps 600 or 700 charges. The 4 hours is calculated based on a 30-35% discharge schedule for the main supply cell. You can take it down to 60% and still get quite a lot of charges out of it. It also depends on whether or not you actually play music through it, and what type of music. Some music is more power-hungry than others.
All rechargeable batteries have a slightly diminished charge after every cycle, even Li-Ion, which are designed for multiple cycling - that is chemistry and the laws of physics. Whether or not you will notice this depends on your habits. If you use the unit 10 hours a day, you will get very much diminished battery performance very rapidly. But then again they're pretty cheap to replace. And you get a wonderfully black background that is hard to imagine without hearing it.
Regarding up/oversampling: You cannot make a blanket statement saying which one is better. Up/oversampling is theoretically better if you consider only sampling theory. However, when dealing with very large data transfer rates, signal integrity becomes extremely important - and I'm not only talking about getting that data through your S/PDIF or I2S bus. Every aspect of the digital system becomes extraordinarily sensitive to minor variations in transmission lines, emi/rfi, and piezoelectric effects. By moving to non-oversampling, you can decrease the signal rate by 10-15x in extreme cases. Jitter sensitivity declines, and better yet, you can spend that extra money saved on a less expensive transport and buy more music.
In a related sense, throwing bits at the problem is also superfluous if you look at it from a realistic point of view. In theory it's wonderful - 120dB noise floor! It is a very engineering-type standpoint, but in the realm of esoteric hi-fi, few companies can afford the sort of R&D to properly optimize for these sorts of specs in the real world. Not only does it increase data rate, how many of those noise specs do you really trust? More companies than you would expect simply quote the S/N specs from the digital chips themselves without actually measuring real-world performance. Vinyl has an awful noise floor, yet the concensus is that it has greater apparent dynamics and is able to convey the musical message better than CD.
Apart from the difficulty with jitter and signal integrity, the single most important advance in the non-oversampling approach is complete circumvention of digital filtering in the reproduction. By avoiding digital filtering in the final output stage, one gains enhanced musicality, something quite hard to describe using the typical flowery language that audiophiles like to use. It takes longer-term listening to really begin to appreciate this, but some can spot it within a couple hours of listening. This is what most people buy the very expensive Audio Note and 47Lab for, and is something that is readily available in the dAck! and also the Nixon units.
In a nutshell, it can be done exceedingly well both ways. Compare Accuphase to Zanden, for example. Both are executed beautifully, both have remarkably different approaches, and both are top-notch performers. For the more down-to-earth market, the non-oversampling approach can be done much cheaper with quite stellar results.
Best,
Chris Own
Ack!Industries
