CD Ripper/Music Server Recommendations

Yes

1 Like

Thanks. Makes sense. I did recall several pieces of info that I picked up in Hydrogen Audio forums over a decade ago including:

  1. Keeping metadata minimal, esp artwork size when embedded into file
  2. Use a less or minimal FLAC compression level - thinking back then seemed to be that if CPU required more decompression effort, it would increase chances of noise / timing issues or something of that sort

I guess this sort of stuff might affect the playback chain and therefore sound quality.

The CPU effort to decode FLAC is the same regardless of compression level.

Oh you have given the game away… it is completely irrelevant.

2 Likes

Guess I am mistaken. But it is good info - esp since I rip at Level 5.

1 Like

MoonDrifter, there once was a pretty good one-box solution to CD playing + ripping + storing + music server-ing and DAC-ing: the Naim HDX. I had found a used unit for $1000 US and loved it. It worked.

Sadly, a power supply component failed and the ‘new’ Naim organization was unable to provide parts to their authorized repair shop here in L.A. They offered a discount on a new Uniti Core that I now use.
Of course now I needed a DAC - found a used DAC-1 - and still no server but I just got a Wiim Ultra.

So 3 boxes to replace that great HDX, but I discovered the Wiim will access the Core stores directly.
The Wiim gives nothing away that I can detect sound-wise and it’s a joy to use compared to the HDX.
I hope this info is of some value to you. Side note: I’m still using the HDX’s 1TB hard drive for back-up.

To clarify, by ‘music server-ing’ I’m referring to streaming. Sorry if I caused any confusion reading this.
It is the streaming that the Wiim Ultra provides that makes it a useful music source for my system.

Clearly you have not read the whole thread. The first post is usually a good place to start.
He already has a world class streamer…the Lumin P1, I know this because I also own one.
Try it (reading the first post), you might have a better idea as to why he started this thread.

Dont overvalue AccurateRip. Look at the source. They use weak checksums not intended for long blocks of data plus the CD-sales of today is very low and my guess is many of those who previously ripped probably stream or download today. To actually update AccurateRip database there is only a few Windows-rippers doing this (maybe this changed the last few years). I still buy CDs but my rips does not update the AccurateRip database.

1 Like

NO-QUARTER, while I did expand on the OP’s enquiry, my first sentence was the response to his.

Actually not so sure CD sales are so very low… in certain markets they are reasonably healthy and even increasing slightly compared to where they were. Physicals have had a resurgence, vinyl, CD and even cassette… they are seen as trendy by the Gen Zs… but yes Vinyl leads the pack on ‘trendiness’… look at the bow wave of cheap CD players coming out of China’s factories.

I rip my new CDs, and even things like monthly mojo magazine CDs… and I notice I am only very rarely the first to upload a rip checksum.

BTW you seem to know one hashing method AccurateRIP use… I couldn’t find it. I assumed it was MD5 which is common in software for none critical / secure functions, do you know?

I have only used EAC which supports AccurateRip. In the logs, the checksum is given as “CRC”, length of the value suggests 32 bit.

For sure, if your cd is not in the database then there is no reliable prediction of the checksum of the original pcm that was used to create the cd. Then we are back to other confidence factors on the rip. If I recall, the database got very large and so there was some throttling of adding new data for a while, focusing on drives with a good offset.

Even a very very weak hash has a high probability of predicting the checksum of the original file once several people have ripped a particular cd.

Then even if the checksum is 32bit, I would invite anyone who is worried that is too weak to think a few minutes about the probability that the rip of a particular track is by some freak of nature a hash clash with the original pcm value. Its probably in the region of 1 in the odd billion or so from memory. Add up the number of tracks you have in your collection and you can calculate the probability that a hash clash will occur mathematically. Then factor in the fact that you will no doubt have looked after your cd collection so the distribution of tracks that have enough scratches and dirt to produce a non accurate rip is skewed. Then if you are anything like me you will not listen to every single track in your cd collection so the probability can be adjusted some more. Oh and then adjust that probability by your estimate of the probability that the error in the ripped track is actually audible in an a/b comparison.

I do think you are more likely to spontaneously combust whilst you are doing this calculation than having to worry that lots of your cd rips in your collection sound poorer than they might if you managed to get a better rip :grinning:

Finally ask your self if, on the balance of probabilities, you would prefer the accurate rip approach to having no clue of the probability that a rip matches the original file. It is up to you though of course. I do think life is too short to worry about such things. Enjoy the music :slightly_smiling_face:.

If I recall spoon said the hash algorithm wasn’t crc32 but I could be wrong.

I looked as well and couldn’t find it, only a reference to an upgraded v2 hashing algorithm.

Thank you, that is helpful. CRC is not a cryptographic hash function and is not a hash function at all as it’s reversible. It is an error detection linear function more suited for detecting noise errors and burst errors rather than affirming authenticity… I am assuming for the limitations of a consumer CD rip these two aspects sufficiently converge to be relevantly equivalent.

2 Likes

Just a final thought.

If you have already ripped your cd collection using a non ‘accurate rip’ based solution but one that has a high quality transport. And you still care enough that the rip is the most accurate it can be from a software engineering perspective. And you are concerned that there might be a sound quality issue on the cds you most enjoy listening to. And you prefer the mastering of that cd over the mastering on qobuz or tidal’s version. Then consider doing the following - just for those cds.

Get a pc, optical drive and dbpoweramp. Only for those cds you care about, re-rip the cd using accurate rip. If you get a positive match back from the database then add that version of the cd into your music database. If it is not an accurate rip then don’t bother. Then play which ever version of the cd rip makes you happier. :slightly_smiling_face:

Edit. I think following this logic I would have about 20 cds I cared about the mastering. Things like early cds of bowie, dire straits and genesis. I don’t like the remasters of those cds as much.

You can check your rips against the AccurateRip database without re-ripping.

Even better!

A question I have put on other Naim threads, but have never received a response, but I rip my CDs to my Qnap NAS, I then copy this to an external hard drive which I connect directly to my Nova. I have used a switch, Cisco, to connect the NAS to the Nova in the past, but this meant additional cables, plus the NAS needed to be on to play music, so for convenience, this works.

When I made the change I did not hear a difference, though any testing was limited, so this may be the answer, but is there anything inherently wrong in having my set-up, and not wishing to go down another rabbit hole, would the connecting cable to the Nova make a difference?

I am assuming that a direct connection would take out other potential issues.

It may be that the Nova is not ‘good enough’ to pick up any issues.

Thanks.

If I get this right, you are concerned about noise from the NAS power supply on the network changing the character of sound due to the ethernet connection?

If so and you can’t hear a difference then I would say don’t worry about it. I think it is highly dependent on the noise profile in your house and network and if you can hear a difference consistently as the noise profile changes. Personally I leave the NAS on 24/7 and don’t care. The backup NAS is setup to come on a 3 in the morning and do its thing.

Edit are you using a music server of some kind to get the file from your nas into the streamer?