I don’t think any DAC can accept compressed files, because in that case the decoder would be a separate module in front of the actual DAC, which necessarily needs raw samples.
But many receive the FLAC file by some protocol like UPnP, Tidal Connect or whatever and decode it right on the machine. And there the EMI noise debate starts, which we don’t need anymore at all but in the Roon case is simply moot because this does not happen at all on the endpoint.
Which has been my point from the very beginning of this thread…
I know, just agreeing. And I thought your rhetorical question was a neatly recursive one that evoked the puppets in my head.
I’m almost certain that the default of LMS (for a FLAC file) is to send the FLAC file to a squeezebox player (e.g., TOUCH or TRANSPORTER*) and the FLAC is decoded by the TOUCH before passing on to either the internal TOUCH DAC or an external DAC. One can change this behavior in LMS (so that LMS decodes the file to PCM before sending to the Squeezebox player).
*for example, in the distant past, a firmware update was required for the TRANSPORTER to play certain FLAC files encoded at -8, because the TRANSPORTER stumbled in decoding these files. So obviously, these files weren’t being decoded first by LMS.
Yes, and I would argue that in this situation, the audio player is built into the receiving device. So, strictly, LMS as server isn’t an audio player.
The same holds true for receivers or streamers enabled to read files from a mass storage device… But yes, as already stated, in Roon’s case this is moot as it is always done on the Core server.
I agree. LMS Is not the “player”, only the server. And as you and others have noted, this is not relevant in the case of Roon/RAAT, as all decoding is done on the Roon core, never by the endpoint.
I also prefer WAV over FLAC. It is a personal preference for me, but Marcus on this forum the consensus is definitely towards FLAC. You are wasting your time if you try and convince them otherwise. Do like me, ignore it and go with WAV.
–MD
… and that’s totally ok, no disputing personal preferences. What’s not ok, though, is to spread false claims… out of ignorance, misunderstanding the inner workings of tech, or whatever. This should and will be disputed.
Ripping Wav is 100% safe and effective 
9 out of 10 doctors recommended FLAC to their audiophile mental patients 
But on a more serious note, the resulting WAV files do not include a checksum that allows a user in the future to run a simple batch check to ensure that the audio of the file is not corrupted. With a FLAC file, one can do this in batch, with FLAC file automatically decoded and compared to embedded audio checksum from the time of creation. If they don’t match an error is reported. This is not possible with WAV or ALAC files (unless one manually created such checksums and stored them in some way). This aspect of FLAC files is the main reason I use FLAC.
edit: I’ve seen horror stories in the past where a drive going bad resulted in some corrupted files. With WAV, that would mean the user would have to open and play each track to confirm that it is not corrupted. With FLAC, one could copy all these files to a new hard drive, then run a simple batch command and identify easily any files that are corrupted.
I love Paul and his enthusiasm.
He’s not always right and, sadly, has been known to change his narrative over the years based on a PS Audio product release. It’s healthy to learn and change your views on these things but that should happen organically. Not because the thing you’re selling requires a change in opinion for that thing you’re selling to make sense.
I use the same software to rip to AIFF. My understanding is that this is essentially a .wav file but with metadata embedded.
A quick google search would settle it but it’s almost 2:30 am in Toronto now so it’s way past my bedtime. Perhaps tomorrow I’ll double check. 
I haven’t viewed Paul’s videos here but I’ve probably seen them before anyway. I think what Paul is trying to say is that DSD is more analog-like than PCM. Given the considerably higher sample rate of DSD that’s not entirely incorrect. If you’re going to diss someone, it would help if you know what you’re talking about.
As someone who studied electronics engineering 30+ years ago I am certainly no expert on current standards, but I do have a very good understanding of how communication protocols and sample rates work.
I’m going to side with Paul on this one, at least in that he’s expressing a reasonably accurate over generalization.
Yes, AIFF is derived from IFF, which is also the base for RIFF, the format used by WAV. It seems AIFF standardizes how metadata is embedded, so from that point of view it is better than WAV, but you still get unnecessarily humongous files compared to FLAC.
The original question was around ripping CDs to WAV or DSD, though.
Ripping a CD to DSD isn’t bit perfect and why sample all the way to a large DSD file when all that’s there in the first place is 44.1/16?
I don’t consider high sample rate an indicator of an analog-like experience. High sample rate are mostly so that filters can have more to work with. In the beginning, DSD was touted as a space saving strategy, not as a sonic improvement.
Very true.
Then there’s this appropriate observation.
This is absolutely true. Paul said many times at first that DSD 64 was more than enough as, purely coincidentally I’m sure, PS Audio products couldn’t handle more than that. Now that the product can handle up to DSD 256, that is suddenly the new sweet spot.
I preordered the PS Audio Air Lens hoping to push some DSD 1024 files I have and see how they sound. I’m disappointed to see in the Air Lens spec that it will support only, you guessed it, DSD 256. So Paul is building the PS Audio ecosystem around DSD 256.
Right now I’m using ASIO USB drivers on a Windows notebook to push DSD 512 files to my DAC. Although it is a subtle difference, I do indeed notice a difference and I prefer the DSD 512 files. They seem ever so slightly richer.
My Macs, of course, can only push up to DSD 256.
Someone mentioned file size as an issue. I just read where Seagate should release 20TB drives this year and 50TB drives by 2026. Most everyone has 1Gb ethernet and most new PCs come with either 2.5Gb or 10Gb ports. I copy my DSD 512 files and lower resolution DSD and PCM files to my iPhone 13 and play them directly with the free version of VOX. I have approximately 1TB of music files so not all of it fits on the iPhone. But most of it does and for a portable device that’s not bad. And now there is ROON ARC for the rest.
I agree with everyone here that thinks ripping a CD to a DSD doesn’t make much sense. I’m happy with AIFF.
I wish people would be more clear when they talk of streaming. Yes, most commercial streaming services over the internet use either lower resolution or compressed file formats. There was a great YouTube video a couple of years ago where someone uploaded master files to Tidal and when he downloaded them again they had been altered. When he approached Tidal tech support to make sure he was doing everything right at his end, Tidal’s response was to cancel his account. I stream. But I stream from my own files on my own fileserver. So please be careful how you throw the term streaming around. When you use ROON, you’re streaming, even if the files are stored locally.
I hope that doesn’t sound too harsh. Black type on a white background can do that. I’ll beg my apologies as I seem to have caught a nasty tummy bug that kept me up running to the bathroom all night. I don’t mean to be harsh, I’m just mostly brain-dumping. I hope everyone will understand.
One last thing. With respect to digital versus analog. If you believe in quantum physics then there is no such thing as analog at all. The shortest distance isn’t zero length but the Planck length. Similarly the highest sample rate is based on the Planck time. At what point does the quantized (digital) world become analog? We, as humans, don’t notice Planck-scale things. To us it’s all analog. But similarly there is a case to be made that at some point long before we reach Planck-scale values, the digital approximation is so close that humans can’t tell the difference. So when Paul argues that digital can be better than analog, you have to ask, what is he really saying? It seems to me that he is arguing that current technology finally allows us to reproduce sound in ways that create a better analog experience. While that may sound counter-intuitive, it doesn’t mean it’s not true.
To help put my thoughts in perspective, I turned 60 last year. So I’m not some technical whiz kid that disrespects more traditional ways of thinking. Ultimately what I think doesn’t matter. It’s what you think your ears tell you that’s important and it is very likely that your experience is different than mine. And that’s ok.
Thanks,
David.
As the question here is if redbook CDs should be ripped to DSD, this is however all irrelevant. The answer is FLAC.
Yep, no problem. I had the same problem about a month ago. It put me in a mood. Have you fainted, yet? Mickey Spillane pens a great line, “He went to brush something from his cheek and found it was the floor.”
Peace.
That is anyone’s guess. Only Paul really knows.
The part I referred to is when he opines that PCM isn’t analog-like, but DSD is.