I removed all the metadata (not just the titles).
I also removed the leading/trailing silence. But I’m happy to leave that in if you want to try the test again.
You must be joking! I have spent all of today hearing my music better than ever before, as I keep re-ripping, (and that’s more than 60 years), and I know that those who will visit me here in the next few months will hear it as well.
You are cordially invited, if you are ever in Dublin, to experience what induced me to embark on a complete re-rip, which is the last thing I expected to have to do.
Thank you for the generous invitation. Dublin is a lovely city and it is with some regret that I say that, at least in the near future, my travel plans don’t include it.
There is nothing mysterious about the leading silences in David’s rips. It’s common for different drives to have them due to the drive read offsets. No drive can serkl to the exact location of the audio it can be within 1/75th so depentant on the drive the silence will be different with some adding more at the end and some more at the front so they get all the samples.
Tried several players, including roon, JRiver, foobar, VLC with a Sennheiser 25-1. Also through an Eximus DP1->Mark Levinson-27.5->B&W 802. Couldn’t hear any difference.
An interesting list here from Spoon (dBpoweramp). It’s a compilation of CD ripper accuracy, reverse engineered from the AccurateRip database:
Absolutely no difference.
And ultimately, if one gets an AccurateRip match, that’s all that matters. That’s the beauty of the AR database matching system. The key breakthrough in bitperfect ripping technology (created by Spoon, the person behind dBpoweramp).
Full Disclosure: I’m a long time user of dBpoweramp for ripping, PerfectTunes for post ripping AR checking among other things, and TuneFusion for converting and transferring files to my iThings.
Thanks for that! Since audio CD data is addressable down to individual frames (588 stereo samples = 1/75 sec), I naively figured that drives could seek with that level of accuracy.
I still have the residual puzzlement that a ripper typically isn’t seeking. It’s just reading the whole disk, from the first frame to the last. In that scenario, there seems to be less scope for being off by a frame.
In response to
@David_O_Higgins said
Actually, I wasn’t.
The only difference between your two files was the metadata and the amount of leading/trailing silence. Removing both of those, you clearly couldn’t tell the difference between them. The logical thing to do would be to add back in the leading/trailing silence and see if that allowed you to distinguish between them.
I kinda doubt it, but hey …
1 Like
Did David’s files get uploaded. Would be interested to have a listen.
Indeed. What I found interesting though is just how vanishingly small the chances of an inaccurate rip actually are with modern mass produced, standards based, and inexpensive consumer devices. There are actually several Pioneers at the top of a very long list of 1,120 devices (but not the BDR-206MBK, I guess it is a relatively recent model.) The best performing drives are hitting 99.5% accuracy with a few dozen inaccurate rips being reported over samples of thousands. But even the worse performing drives are hitting about 91% accuracy and I assume all the reported inaccurate rips can be corrected by the AccurateRip database.
1 Like
Excellent point. The AR database doesn’t provide for correction possibilities. But CueTools has the ability to correct small errors based on its database. I’ve found this very useful in the past.
I’m afraid this really just isn’t good enough Jacques.
How can we possibly trust someone who doesn’t have a visit to Dublin on their travel itinerary?
.sjb
2 Likes
Well, I got my priorities wrong here. I started with the D100 ripper, because it gave me a relatively slight, but worthwhile, improvement, at half the price of the S100 switch, albeit with a huge time investment.
On Friday last I replaced a Netgear GS108 with a Melco S100 and in an instant I got a transformational improvement. The difference is astonishing. ‘The Science’ will eventually explain it, but meanwhile I have a huge grin on my face, despite the ongoing state of The World.
2 Likes
This is a great thread, but clearly too many engineers who only hear what they measure and have their brains shut to audible difference that other people can clearly hear. I have heard so many unexplained improvements in audio that are seemingly ‘magic’ until someone finally comes up with an explanation. Digital is not just digital!
THANKS for having good ears and not being afraid to say what you hear! @David_O_Higgins, saving up for my Melco ripper
It also works the other way round that those that hear the differences do so justify the expenditure as it must be better as its more expensive.
I am no engineer but as its a very expensive drive and if there are claims it sounds better, when all logic and understanding off the ripping process dictates otherwise I would want some hard facts to back it up and justify this increased price.
I personally heard no difference and I could play them in sync and seamlessly, switching from one version of the file to another with no interruptions it played as it was one recording no change in anything But if your happy to give Melco money for what seems like a different drive offset be my guest.
Subjective opinion. The evidence is presented in the thread. Both rips were identical. Perhaps you should look you into why people hear differences that can’t exist?
2 Likes
very interesting:
they rip with one d100 drive and different usb cables and did not get bit perfect results.
the files are all different!
strange 
https://www.hifistatement.nethttps//www.hifistatement.net/tests/item/2825-wie-wird-das-digitale-zum-file-usb-kabel?showall=1
its in german!
The guts of this very expensive drive is a standard Pioneer BRD transport, so you’re paying a grand for the case and power supply.
1 Like