Cambridge DACMagic, anyone? - Page 8  

The front panel clearly shows 96 as the maximum. Had a quick look at the technical blurb just now. It says: Adapted Time Filtering (ATF™) asynchronous upsampling technology converts 16-24 bit audio (at any standard sampling frequency between 32-96kHz) to 24 bit/192kHz.
But isn't that inferior to say the delta sigma oversampling DACs that oversample even the modest 16 bit 44.1KHz by 8X to 252.8KHz? That would make the Dacmagic less detailed with its lower sampling frequency. Why go to all the trouble of the debatable use of upsampling to 192KHz when oversampling would be far simpler? I am also troubled by their statement that any standard 16-24 bit audio and sampling frequency of 32 to 96KHz is automatically upsampled to 24 bit/192KHz. That's 50% more added bits on a CD. Where do these bits come from? Are they divisions of the original 16 bit or software/hardware manipulation of the original data? How trust worthy is then the audio coming from the DacMagic compared to a pure unabridged 16 bit in 16 bit out audio signal? How do you convert a 16 bit signal to a 24 bit signal without compromising the original waveform and what are the benefits? More detail? From where? Can you get more out of a digital procession of 0's and 1's than was in there in the first place? I tried converting 160kps mp3 to 192 or even 256kps, but it didn't sound any better, so how is this different?
Any techy out there with some answers?

How is it you have a knack for masking FUD as innocent questioning? Did you even bother to try and read any of Anagram's white papers? Start here: http://www.anagramtech.com/technolog...ime-filtering/

FWIW, I've heard the 840C and Audio Aero Capitole, which also use Anagram tech, and they're bloody good, but I don't have any presumptions that the DacMagic is going to sound at all similar to either of them just cause of that.

Regardless of the technology here, the bottom line is how this unit sounds, and the ATF is going to just be one factor in it, and not necessarily even a decisive one.

Can't anyone ask questions about the DacMagic that you are not happy with, without you attacking them? Are you a Cambridge salesman or something who is steering away folks from serious hard hitting questions?

Thanks for the ATF link, but they do say in their write up: "...extremely small residual error that is left is designed to be beyond the dynamic range of 24bit audio..". Right. Extremely small... So there is an error as I suspected. They mention 24 bit. I wonder how large the error would then be on a coarser 16 bit signal. Does the wording "extremely" still apply?

By the way, what has the 840C got to do with the DacMagic? Different chipset. No comparison in make or chip model number.

Also, why use ATF upsampling to 24/192 when the Wolfsson WM8740 chip used in the DacMagic can already support 24/192? Is the WM8740 not good enough to do the job by itself or am I missing the jest of the technology? Marketing overkill or genuine musical reasoning to incorporate upsampling?

96khz maximum as input... right but it upsamples everything to 24/192. I am not savvy about the technical design here but listening will be the judge whether this dac is good or not...

So anyone with a dacmagic yet... impressions please

Let's see, StanleyB1, Herandu, even "DC Lee" the happy Beresford zealout, all in the same thread, asking "hard-hitting questions" / knocking a competitor...

Give it up, man...Cambridge salesman? I'm as interested in buying this DAC as I am dignifying your disingenuous FUD.

So now you are resorting to slander of other headfi members instead of giving any answers to the technical questions asked of the DacMagic??? Blame the messenger, not the message is your moto, is it? Your avatar seems most fitting. Palin has the same sort of approach to hard hitting political questions. Poor Obama has been equally mistaken for so many other folks.

Your questions are hardly hard-hitting, yet they're almost identical to StanleyB1's and DC Lee's - to wit, asking irrelevant questions about upsampling / filtering in general, casing, pricing, looks. Slander? What else do you all have in common? Hypersensitivity??

If you're genuinely so interested in upsampling and DAC layouts, go read Horowitz. Otherwise, I suggest we both just sit back and let owners and early adopters post their actual impressions of how the thing sounds...like What Hi Fi did..saying it bettered the Beresford, if you didn't know.

Yeah, I have a feeling I might just get this one too.

do you know any stores in manila that might already have the dacmagic?

Nope. I don't even know which one is the local distributor.

I am just catching up on several weeks of headfi banter and I just picked up on another thread that you poured scorn on the credibility of WHF and its awarding of praise towards any product. You now triumphantly herald WHF as your source of glorious praise upon your own product. What caused you to change your attitude towards WHF?
Why are you so pricky about a few technical questions from anyone? Are flaws being exposed in the design after a concerted marketing drive and ahead of a well planned and carefully timed impending launch?

cambridge dacmagic

Can anyone comment on how the Dacmacic compares to a good reference Dac such as Benchmark Dac 1, that would give a much better idea than the likes of the what hi Fi review which only compared it to low end cheap Dacs which tells us nothing especially as dacmagic is being hyped as a giantkiller- against what?

You missed the point. He is not championing WHF as much as he is using it as an example, hence the use of "like." He, as am I, are waiting for firsthand experiences. I don't care for how it looks, use of poor parts, etc. as long as it performs well. Let's please keep this thread from spiraling further into a slugfest and wait for known Head-Fiers to post their impressions.

Seconded! Please people, let's keep this civil. There's no reason to get so worked up. I think there are some people who are genuinely interested in this DAC's capabilities, and there's no reason they should have to sift through mud to find it in this thread.

I can see you didn't read my earlier post in the thread.

Detail doesn't come from oversampling or upsampling because those processes can't create information that isn't in the original (e.g. 16-bit 44.1kHz) samples. See my earlier comment in this thread.

Upsampling & oversampling techniques are said to help in the analog domain after analog-to-digital conversion because the D/A process creates artifacts in the frequency domain (well above the D/A sampling frequency) that need to be filtered out to reconstruct the original signal. Feeding the DAC chip at a higher sampling rate - even if the upsampled data sequence contains exactly the same information as a non-oversampled data sequence - lets you construct a gentler analog filter which may improve the resulting sound in some ways. There may be other analog circuit benefits as well.

But upsampling does NOT provide any additional information.

If you interpolate or upsample you need to do the underlying arithmetic at a higher resolution than the original signal, otherwise numerical errors creep in that affect (at least) the least significant bit of the result. This is a standard well-known issue in integer (and also in fixed point floating point) arithmetic. So typically for 16-bit samples you'd use 24-bit (or even 32-bit) intermediate computations. When you have your 24-bit result you can do one of two things - keep all 24-bits or truncate back down to 16. Which would you prefer, given that 24 is technically a bit more accurate?

If it's done right the resulting analog audio output is closer to the original analog waveform than would be produced by 16-bit (non-over-/up-sampled) A/D conversion - for the reasons outlined above.

MP3 is a lossy compression technology and you're attempting to decrease the compression level and hoping for some improvement. That won't work because the information that has been thrown away by the heavier compression cannot be reconstructed. However your MP3 changes haven't done anything that will help the analog circuit design of the DAC.

On the other hand, no new information can be created by upsampling either - but the differences are in the accuracy of the D to A process (mainly due to analog circuit design issues).