Talos2000

Is it just me, or does anybody else have a problem with the way the search engine works on this site? If I search for something - here is an example: "Saracon" - I get a list of topics in which the word "Saracon" has appeared. These topics can contain hundreds of individual posts and there is no way I can see to home in on which one(s) contain the actual search term. Is there something I am doing wrong?

Is this still their position? Whether it clicks or not is kind of irrelevant. These are lossless, uncompressed, native file formats. Either sacd_extract produces identical audio content in the DFF and DSF extractions, or it doesn't. If the two are different, then at least one of 'em must be wrong! Miska (and others) has pointed out precisely where to look for the differences. It is 5 minutes of work with a Hex Editor. An easier way is to look at the very end of the DSD audio content in the DSF file. The DSF spec calls for the audio data to be stored in blocks of 4096 bytes. If the last block is incomplete, the spec calls for it to be padded with zeros. So unless the DSD audio content just happens to fit exactly into an integer number of 4096-byte blocks, then the last 4096-sample block will be padded with a bunch of zeros that are easy to spot with a Hex Editor. If the zeros aren't there, then either you've hit the one-in-a-thousand chance that the audio data just happens to exactly fill the last block, or else the data must have been truncated. Oh, and thanks for pointing out the DFF2DSF converter from the Yate guys...

I'm still not totally clear on this. Is the SHM-SACD version a direct Analog-to-DSD remastering, or just another PCM-contaminated half-breed? Can anybody categorically answer this one way or the other? Over the years I have accumulated several versions of this album - including several LP pressings - in the vain hope of coming across one clean recording. To my ears, it sounds like the original recording was done with every single microphone in DG's possession disported about the recording venue. Unless the analog master tape is the original 48-track studio tape (rather than the original mastered stereo mix), and a significant job of remastering can be done, I suspect even a true SHM-SACD analog-to-DSD conversion will still sound (sonically) mehh. I say that as someone who owns a few stunningly magnificent SHM-SACDs, and the occasional mehh.

Your point is well made. But I just want to observe that a SDM can also output a multi-bit signal, including "N-and-a half-bit" (where zero is an explicitly coded value. Many (most?) DACs already do this, since 1-bit SDMs inherently develop progressively serious stability issues when the noise-shaping filter becomes aggressive enough to generate a desirable SNR. In fact, most SACD DACs will remodulate the incoming 1-bit data to a multi-bit, higher-sample-rate SDM format which is then used for the actual ADC.

Thanks for pointing this out. I did not know that. Is there any easy way to establish which DAC chips do this and which ones upsample properly? Maybe you could write a brief post on the subject

I'm hoping Larry Ho of Light Harmonic will chip in here. He told me that the i5 is "quieter" than the i7 and thinks it is better for music playback, but I didn't get the chance to quiz him on the details.

Yes, I do. And just to be clear, I most certainly, verifiably and definitively do have album art embedded into each and every one of my 27,384 (and counting) tracks This is the same process I use, except that I have not tried repairing disk permissions. Not really sure why this would work, though, but stranger things happen at sea... Yes, I do have a Synology NAS, but again, I am at a loss as to how that could affect this particular problem. The album art that goes missing is cached in the iTunes directory on the local HD. Recall that if I select "Get Info" on the first track in the album, the album art magically re-appears. Can anybody confirm whether or not executing "Get Info" causes the metadata embedded in the file to be re-read? I have always assumed that it does not (it seems to me it would not really make sense to do so), but you know what they say about assumptions...

Yes, it can split the files, but can it use the CUE sheet? I haven't found a way to make it do that.

iTunes has an annoying habit of 'losing' the Album Art on some albums. Not losing entirely (in the sense that all I have to do is "Get Info..." on the first track in the album and the Album Art mysteriously re-appears), but the Album Art is just not displayed - only the 'grey quavers' box. Usually, it only happens on a small selection of albums that are related in some way. The most regular culprits are albums tagged with "Opera", or that are physically located in a folder where a lot of opera resides. I am beginning to suspect that some opera-hating programmer at Apple might have put something in there intentionally ... Yesterday, I opened iTunes and the whole music collection is showing no artwork. That's over 2,300 albums. As usual, if I do "Get Info" on the first track of any one album, the artwork re-appears. I have tried various schemes to achieve the same result with one operation on multiple albums, but so far without success. I suppose I could delete and re-import the entire collection, but for various reasons I am very reluctant to go that route. Has anyone experienced this problem? How about solved it ?!?

So here is a provocative assertion. Analog is actually DSD. An analog signal is normally viewed as a voltage waveform. But the transmission of an analog voltage from one place to another is done by a current waveform. Electrical current comprises a flow of electrons. The more electrons flow, the higher the current. Electrons carry a very specific electrical charge determined by the laws of Physics. For a certain current (expressed in Amperes) to flow, it requires a very precise number of electrons to flow through the current sensor. We can therefore represent the electrical waveform with total and absolute precision as a digital bitstream - albeit one with a phenomenally high sample rate. This bears a strong similarity to DSD. Both effectively represent the analog waveform in a manner that we recognize as Pulse Density Modulation. Aside from the sampling rate (which differs by 19 orders of magnitude!!) the only real difference is that DSD is synchronous, whereas Analog is not (actually, there is a school of Physics which suggests that time itself passes in a sequence of discrete chunks - which would mean that Analog is actually synchronous as well). So is it any surprise - I ask with my tongue only partially in my cheek - that many listeners suggest that DSD is more "analog sounding"?

Realistically, there is no such thing as "hard evidence", because nobody can measure unambiguously what a person hears. Yes, you can generate a sound which has measurable property X and ask if a listener can hear it. But that's not entirely the same thing. There are a couple of points I have made previously, but want to set out again, because I feel they are really important: (Also it is Jan 1st, and I am still a little hung over) First, I accept that there is a body of information theory that purports to set forth a reasonable basis for supposing that 16/44.1 PCM is capable of representing everything that we can hear. But any real-world implementation of PCM requires either (i) serious pre-sampling (ADC) and post conversion (DAC) filtering in the analog domain, which creates known, quantifiable, measurable, and irreversible distortions; or (ii) expressing the PCM in an ultra-high sample rate format that reduces the filtration requirements to benign levels. Second, all PCM data is derived from a Sigma-Delta Modulation bitstream, due to the nature of practical ADC design. The algorithms which convert SDM to PCM (usually referred to as decimation filters), analogous to analog filters, create known, quantifiable, measurable, and irreversible distortions. Again, these distortions can be reduced to benign levels if one is willing to accept PCM coding formats at ultra-high resolution, with the heavy baggage of data density that this implies. So, in summary, while it may be possible to argue that PCM can be theoretically transparent to the source, it is meaningless to do so without the context of a real-world implementation.

There's the mass market, sure, but what does the industry need to do to satisfy an audiophile minority who might be looking for DSD support? Right now, you might argue that there is a reasonably well-supported move towards the release of popular albums in high-resolution PCM formats. Yet nobody would suggest that high-res has gone mainstream. Something similar in DSD would be counted as reasonably successful in the audiophile realm. But I don't think anybody here is seriously advocating DSD on iTunes/iPod as an evidentiary yardstick of success.

First of all - I jumped the gun and installed iTunes 11 before getting BitPerfect's official blessing, but it seems to be working fine for me. No problems at all. However, iTunes 11 itself seems to be a bit of a let-down after Apple's bally-hoo. Its nothing more than a slight freshening up of the interface. Now, to be fair I am just an audio guy, and I don't have any use for any of its other capabilities, so maybe I'm missing the bits where all the fanfare is directed. I had more or less got iTunes 10's user interface to work for me from a perspective of presenting my music in a way such that I can navigate reasonably efficiently through it. So I am looking for iTunes 11 to at least leave that experience none the worse, and hopefully to improve it. No such luck. The "standard" view in my iTunes 10.x set-up is now the "Songs" option. Unfortunately, although it is possible to pretty much re-create what iTunes 10 looked like, they have removed the option to "Always view Album Art". Album Art now appears only with all of the other views, but in such a way as to limit its usefulness. I really don't want to get into a detailed review here, but I am quite disappointed. The much touted "Play Next" feature is there - but I don't expect I am going to use it. It does not seem to offer anything unless you want to make like a part-time DJ. Anyway, the important thing is that nothing crucial was broken, and BitPerfect still works. Phew! BTW, there is a new iPad/iPhone Remote App that goes with it. I haven't had a chance to download it yet, but I have been told that it is an improvement on the old one. We'll see...

I just saw Vintage Trouble live, supporting The Who on their Quadrophenia tour. They were such a high-octane stage act that I had to go out and get their album. Sadly, the CD is a typically dynamically compressed affair. I note that there is a Vinyl version available, and I am wondering if anybody has heard it? If so, is it a separately mastered release?

I enjoyed reading your lengthy post, but there was one aspect of your analysis that I have not been able to get my head around.... You began by describing how quantization errors in multi-bit PCM encoding are correlated, and therefore comprise distortion rather than noise. You went on the suggest that this no longer becomes a problem when the noise within the signal itself is of a larger magnitude than the quantization error distortion components. You go on to point out for this to be the case with 1-bit encoding, then in effect the noise would have to wash out not only the distortion, but the signal itself. Therefore, quantization distortion is unavoidable in this case. Here is where I am having trouble. With PCM the multi-bit encoding attempts to encode the instantaneous value of the signal at the sampling instant. But with single-bit (single-level) SDM, the one bit signifies something different. It signifies the fact that the output value of an integration circuit has exceeded a certain threshold at the sampling instant. What I cannot get my head around is whether or not the argument you make for PCM amounts to the same thing in the SDM case. I would appreciate it if you could clarify this. (Sorry it took me 3 weeks to respond - I only just saw the post)