To my ears, the best sounding ripper is Control:Mac, or Control:PC (which is based on EAC, I believe).
I paid for, and have used dBpa a fair bit, but now have gone back to using the Meridian software for this reason, despite it's irritatingly poor metadata/album art handling (I just spend time and do it manually).
I can't back this up objectively, but my ears persistently like those rips better. Your ears might say different (or hear no difference at all), but my advice is to listen for yourself before assuming there can be no possible difference...
How can a rip that has been confirmed identical to xxx other rips sound any different?
Or are all xxx rips wrong?
A number of people have compared the binaries out of different encoders, including dBpoweramp and EAC and found no differences.
And as Sooloos, once decoded in Core to PCM data, subjects all its binary PCM data to exactly the same treatment (high variability network transport, audio endpoint buffering and replay), I cannot see how one version of an accurately encoded lossless file can sound any different to another.
IMHO
Meridian owner since 1992 Prime & PSU, Focal Elear headphones, Roon (ROCK 8Gi5 NUC), Explorer 1 & 2, F80. 200/203, various Sonos (via Roon).
I don't agree with the OP, however if he hears a difference then why not use what ever ripping method suits him, I think there are far more factors to consider that we know do affect sound quality eg: speaker placement, room treatment, decoration and furnishing of the room, quality of audio components and I would rather pursue those, plus they are more enjoyable to do.
Could it be that the difference I'm hearing occurs in processing of the file *after* the checksum stage? A difference in encoding, FLAC compression, etc?
On a slow PC you can see a marginal difference in CPU utilisation when decoding FLACs with different compression ratios - so if you can hear that difference as a change in the noise floor then MAYBE there are audible differences between encoder settings. The decode stage should pass a MD5 checksum whatever the encoder settings.
Mark G68, baby Genelecs, Humax Foxsat, Oppo+Vanity93, Toshiba HD-A35, Pioneer CLD-925 + Meridian 519, various other random bits and bobs Got a black G08.2 you want to sell?
The decode to PCM is done in the core and transferred over the network link/stack to the audio components. Even on a MC200, the connection is via the network stack and subject to all the variables (such as highly variable timing) of that.
With something like ID4x or MSxxx, decode and playout is totally decoupled from hardware and the only bit connecting them is a network and power cables and a series of asynchronous data packets.
The network will not differentiate between lossless PCM packets from dBpoweramp FLAC, EAC generated FLAC, not WAV etc. for that matter. MP3 originated data is slightly different as it's lossy but again, will be converted to exactly same size and format of PCM data just with less discrete data in the signal.
A couple of years back, I posted how to use wireshark to sniff the audio data and confirmed identical packets between different replays on a reliable wired network. This could not be said for a wifi network where packets were intermittently dropped (and retransmitted in batches of 10 * 4xxx ? bytes) as wifi signal quality reduced.
It would be just as easy to test with dBpoweramp FLAC and xxx rip versions.
Meridian owner since 1992 Prime & PSU, Focal Elear headphones, Roon (ROCK 8Gi5 NUC), Explorer 1 & 2, F80. 200/203, various Sonos (via Roon).
Please explain how your explanation of how Sooloos works and my 'theory' about a change in the noise floor are in conflict? I am not referring to any change in the actual signal (hence my comment about passing a hash test) but in the noise level of the system as a whole. Same signal, different SNR (although my opinion is that if properly setup/laid out etc. there should be no audible difference.)
Mark G68, baby Genelecs, Humax Foxsat, Oppo+Vanity93, Toshiba HD-A35, Pioneer CLD-925 + Meridian 519, various other random bits and bobs Got a black G08.2 you want to sell?
Do the two programmes use different FLAC encoders?
FLAC hasn't changed in a major way in several years ... there was a minor release (1.3.0) in 2013 that mainly had to deal with WAV64, 6.1/7.1 files and added some compatibilities, but the last major release was 1.2.1 in Sep 2007.
Keep in mind that Control:Mac and Control:PC are pretty much the same Mono app wrapped in a virtual machine.
Please explain how your explanation of how Sooloos works and my 'theory' about a change in the noise floor are in conflict? I am not referring to any change in the actual signal (hence my comment about passing a hash test) but in the noise level of the system as a whole. Same signal, different SNR (although my opinion is that if properly setup/laid out etc. there should be no audible difference.)
The decode is in the Core not the Sooloos audio endpoint (although some systems such as MCxxx/Ensemble this is not the case but OP and any Sooloos perfectionist will have a distributed system). Thus any theoretical changes in noise in the decode has no relevence to the Sooloos audio endpoint when using a fully distributed Sooloos system like core/store and MSxxx/IDxx. Unless you are saying that a fractional miniscule difference in cpu processing, offset by a miniscule difference in network/disc bandwidth in the core is going to be heard on a piece of equipment typically in a different room over an above all the other environmental and system variables. In addition to these minuscule differences being heard, it has to be constant enough to be reproducable when making a decision that not only A is different to B but A is better than B.
Not only that there will be a miniscule different, but that difference will only be heard for a fraction of the track as decode will happen in a few seconds yet the track will be playing for minutes. I may be wrong about this assumption as I have not monitored traffic between store and core to see if its streamed or not.
As for checksums, pointless in this excercise unless you do not trust what is being read from a reliable closed system, in which case, you have far more to worry about than whether an audio stream is different to another audio stream because the whole program will be behaving differently as its being processed in exactly the same environment.
Meridian owner since 1992 Prime & PSU, Focal Elear headphones, Roon (ROCK 8Gi5 NUC), Explorer 1 & 2, F80. 200/203, various Sonos (via Roon).