"Hi DC - Thanks for the honest opinion although I do disagree with you on most of it :-) No worries though. This is a very laid back site so I'm sure we'll get some very constructive conversation about this one.
I believe that everything in the chain matters. Some of it may not have a one-to-one effect on the sound. For example switching to an SSD drive v. a standard HDD may not have a direct effect on the sound in some systems, but the little things add up to improved sound in my opinion. I have talked to a bunch of highly respected people who have direct experience with CD-ROM drive differences in quality. In fact some have suggested the very good Plextor drives drastically change around 2000 album rips. As far as applications go, they all seem to have a sound signature and this is one area everyone is researching.
Thanks again for the post DC."
Like I said, nice guy.
CD ripping drive comparison
Software
J. River Media Center 12 (.526)Test Disc
Brand new store bought "Hamiltonia" by Chico Hamilton.Drives
Plextor 716-AL (ATA)
LG GGC-H20L (SATA)
(2) Lite-On SHW-1635S (USB 2.0)
Notes
1. The 716AL is sometimes fussy with some discs as to spin up time (i.e. recognition by Windows), but completes secure rips so I consider it admissible as a test unit and believe the unit adds to the validity of the test by introducing a possible extra variable.
2. EAC may be tested at a future point but most likely filesizes will differ and frequency analysis and listening tests only would be used (this would be due to variance in gap detection by either software).
3. System free of calculation errors due to bad hardware: stress tested with Prime95 (quad-core edition) & memory tested with two simultaneous instances of a separate Windows based tester running on all RAM.
Methodology
J. River ripping setup
Rip to wav iso
Secure rip enabled
Write to log file
Results
Filesize
J. River created three identical ISO files (down to the byte), and reported each as a secure rip in the rip-log.
Listening Tests (ABX)
to follow
Spectrum Analysis
Here's where it gets interesting. I used Global Analysis on the entire wav file in Wavelab; all three wav files show same peaks (0db), but average RMS differed VERY slightly between rips. More intriguingly, was the average pitch of a section from 305ms - 1000ms (the first musical notes of the song). I used Wavelab to set markers at these VERY exact points in each wav file. As you can see, the LG & Plextor are identical, yet the Lite-On reports a lower frequency for the average pitch; it's somewhat of a substantial difference.
Lite-On
.jpg)
LG
Plextor
At look at the rip logs reveals that the Lite-On had issues with this brand new disc whereas the LG & Plextor did not; MC reported a secure rip at finish after retrieval of all data after said number of re-tries. Subsequent re-rip with both Lite-On USB rippers show no problem.
Analysis of Lite-On rips (2nd identical Lite-On unit brought into testing for comparison) with MC12, and LG re-rip with EAC.
MC conducted a re-rip of the Lite-On P: drive and I also ripped the disc with the Lite-On S: drive. Both drives are identical make and model with the same firmware. This time (as you saw in the log pic above), they had no issues ripping the drive i.e. "no re-tries" of any sectors. Analysis reveals that the pitch/frequency matches the initial Plextor and LG rips.
The EAC rip done on the LG gave the same "error" in pitch/frequency that the Lite-On did initially in the first round of ripping. EAC reported 100% accurate rip. Accurate Stream feature disable, secure rip enabled, and direct to wav on disk.
In total, MC ripped 5 "securely ripped" discs and returned 4 with identical pitch/frequency for the period from 305ms-1000ms; this value was LEFT=76.17Hz (D# 1 -36 cents) RIGHT=76.68Hz (D# 1 -25cents).
The 'errant' value was always the same: LEFT=117.11Hz A# 1 +8 cents RIGHT=78.26Hz (D#1 +11 cents). This was found on the LG re-rip (EAC reported 100% accurate) and the initial Lite-On P: rip (MC12 reported re-tries but secure rip.)
LG re-rip with EAC (below) is unlabeled in screenshot
I'm not able to do more testing, or more thourough re-testing right away but this is interesting to say the least. Is it audible? Does it matter? Should the Pitch analysis be identical for every rip? Is this a valid measure of pitch - seems like it should be identical to me but I'm not an audio engineer.
I don't know yet.
I should look at more sample sections; maybe this affects only the very beginning or the rip/wav? I want to do more ripping with EAC and see if the Lite-On & Plextor return the "errant" (could be the more accurate one but I'm an MC fan ;) value consistently with EAC. MC did return this result though as well.....it's murky. The first I need to do is to setup EAC more thoroughly; I haven't used it in a long time. I still think these preliminary results are very interesting. Obviously, a re-rip of the LG with EAC is warranted.
The next step is to setup EAC correctly, and rip the disc again with all drives using both EAC and MC12.
I'm going to test ripping the first track too; as much as I'd dislike changing the test setup. I'm not sure it would make a difference at this point (testing the 305ms-1000ms section). It will save a lot of time.
Thoughts?
Posts
1 comments:
You've already received a lot of good responses to this on HydrogenAudio but I wanted to make some more direct requests.
First, dump the pitch analysis. It's far less accurate and precise than just looking at the raw samples.
Second, look at the very beginning of the samples and ensure that no read offsets in fact exist. It's quite plausible that even though your pitch analysis differs in only one part of the file, it is due to an offset that affects the whole file.
Third, once you're sure the samples are aligned, invert one of them and downmix. Everything should cancel out except the samples that are different, which will be immediately visible across the entire file. This is a far far faster and more accurate method than the one you describe.
Fourth, note that Chris's comments about "everything mattering" are mostly in relation to aspects of digital audio playback that do not affect the bit output. ie, the same bits saved on a SSD sound different compared to saving them on a HDD. Those issues are much different than the ones you describe, and much more controversial.
Fifth... use AccurateRip ;)
Post a Comment