Update as of 1/1/2014:
What an embarrassing revelation. After more listening tests, I have discovered that I was wrong.
The differences I thought I heard were merely my overactive imagination. You could say that I was analyzing little details so closely that I began hearing things that weren't there. Ironically, it was only when I listened more casually that I saw the error of my ways. I should have known this all along, since the objective data already told me the truth.
I hereby rescind my mistaken assertion and formally apologize for any confusion caused by this thread.
My personal choice will now be dBpoweramp, due to its higher ripping speed and ease of use. Exact Audio Copy will be used as a backup when need be. Both are excellent programs nonetheless.
On a more positive note, at least now I (and indeed, all of us) can move on to better things...like gear that actually will drastically improve the sound.
(original post is below)
Head-Fi has been invaluable to me for info about audio equipment. While I have never posted here myself, I thought a little experiment of mine would serve as a worthy introduction.
I have ripped thousands of CDs with dBpoweramp and Exact Audio Copy and can declare with 100% certainty that each program has its own unique sonic signature.
These are obvious (not subtle) audible differences that are often heard immediately on just about any song.
As far as subjective listening is concerned, I have tested this extensively, over a period of many months, across a spectrum of styles, including pop, rock, metal, dance, rap, jazz, ambient, orchestral, and everything in-between.
Here's where it gets confusing.
To analyze the data, I used foobar2000's Binary Comparator component, Exact Audio Copy's WAV comparison feature, and Liberty Instruments Audio DiffMaker signal difference extraction software.
All three indicate that the files are bit-identical; yet, they sound drastically different!
Objectively, this should not be happening at all...but it undoubtedly is.
I also went to great lengths to ensure I had the proper settings configured for both programs to generate flawless rips, so I know the problem doesn't lie there.
The following parameters were constant throughout the experiment.
--- Rip settings ---
*File format: Wave (.WAV)
*Read sample offset correction: +6
*C2 error pointers for error detection
--- Ripping equipment ---
*Computer: Alienware M11x laptop
*Operating system: Microsoft Windows 7 Home Premium 64-bit
*CD drive: external LG Portable Super Multi Drive model GP10NB20
--- Listening equipment ---
*Digital audio player: Apple iPod Classic 160 GB (on Electronic EQ setting)
*Headphones: Bose QuietComfort 15 (with Duracell PreCharged Rechargable AAA battery)
Since I do not have a dedicated DAC, listening directly from the laptop with headphones (even with JPlay Mini software) doesn't sound as good as the iPod.
Admittedly, my setup is not what you would call high-end, but I can clearly hear innumerable differences nonetheless. I suspect that those of you with better gear will be able to hear even more.
EAC rips were typically done in secure mode. (I have experimented with paranoid and burst modes, if this is relevant.) dB rips were done in ultra secure mode, which ultimately rips in burst mode for each pass, even if there is a problem with AccurateRip verification.
I initially ripped everything to a 12 TB Buffalo DriveStation Quad external hard drive array, but recently found that, for whatever reason, ripping directly to the laptop hard drive produces noticeably superior sound quality.
(A similar phenomenon applies here. All the files are bit-identical, but the old rips sound somewhat distorted compared to the new ones. The sonic differences between the two ripping programs have become even more apparent, now that the music has greater clarity and detail.)
As a result, I am now re-ripping my entire collection. I'm inclined to do it with just one program this time around, to reduce the workload, if I can figure out which one has objectively better sound quality. (Then again, I may end up using both, to have aural options to choose from, depending on my mood.)
Here's where you come in! I would like qualified individuals to participate in this experiment with me and share their results.
If you have ripped a significant selection of CDs with both dBpoweramp and Exact Audio Copy and listened closely to the files, or are willing to do so in the near future, you qualify.
I cordially invite you to join me in my quest for musical excellence.
For your convenience, feel free to refer to albums that you are most familiar with. Naturally, some songs will be easier to detect anomalies on than others, so please be thorough and strive to include a broad range of genres to evaluate.
To replicate the experiment and remove as many variables as possible, it would be best if all participants used Wave files ripped directly to their computer's internal hard drive. (While compressed lossless formats are bit-identical to uncompressed Wave, a growing number of audiophiles have reported hearing differences on their system.)
Remember that you need the correct read offset setting for your drive. This is done automatically when you configure AccurateRip, but if you run into any problems, you may be able to find the number you need on their website. (Resource links are listed at the bottom of this post.) Just look for the model of your CD drive.
As an optional precaution, you can analyze the ripped files like I did, using any of the aforementioned free resources.
In addition, the Binary Comparator in foobar2000 lets you compare files to the CD itself. Unsurprisingly, all rips came up identical, unless the wrong offset was used.
(You must set the right offset setting in foobar2000 for this to work. Click File, Open audio CD..., Drive settings, and either manually input the number in the field next to Read offset correction:, or click auto, then click OK. To run the bit comparison, right-click on at least one of your ripped files (or two if you are comparing digital files only) and click Play in foobar2000, or Open with... and select foobar2000. With the files open in the playlist, click File, Open audio CD..., Add to playlist. Click on one of your files in the playlist, hold Ctrl and click on the corresponding track on the CD, then right-click that track, hover over Utilities and click Bit-Compare Tracks... If all goes well, a box will pop up with the message "All tracks decoded fine, no differences found." and "No differences in decoded data found.")
This is beside the point, since the purpose of this exercise is to subjectively examine the qualities of sound files during playback. First and foremost, this is a listening experiment.
Whenever I pinpointed a section of interest where deviations were heard in a song, I would then implement a repeat comparison and listen to that section over and over on both files, to make sure it wasn't merely my imagination. I recommend that you do the same.
You can take the sweeping approach and listen to full albums back to back, to get a general feel, but I find that targeted listening is more effective.
If you are able to listen to CDs directly with the same hardware setup as the digital files (the higher quality the system, the better), you can also share your thoughts on which ripper sounds closer (or identical) to the original CD.
Once you make an informed decision, I encourage you to vote in the poll at the top of this page.
There are four possibilities for your results:
1.) You do not hear any difference at all.
2.) You hear differences and believe that dBpoweramp sounds better.
3.) You hear differences and believe that Exact Audio Copy sounds better.
4.) You hear differences, but have no preference and can't tell which sounds better to you.
After listening to a LOT of music ripped by both programs, I can say that I enjoy the sound of dBpoweramp, but at the same time get the feeling that EAC may sound more accurate to the CD source. There is tighter bass and a greater sense of realism, for example.
Again, I understand that these files are technically bit-identical. (After all, I conducted objective tests to confirm as much.)
If you are convinced that they *must* therefore sound identical (which they indeed do not) and refuse to attempt the ripping/listening experiment altogether, I respectfully request that you refrain from posting in this thread.
I hold nothing but esteem for music enthusiasts - particularly those who are passionate about computer audio - so please do not interpret my proposal as intending to be provocative or create controversy for its own sake. I simply want to uncover the truth.
I am equally aware that this has been debated elsewhere, but it is rather difficult to find sufficient information on the subject.
Given the stellar reputation of these digital audio extractors in the audiophile community, I am astonished at the lack of deliberation regarding this matter.
From the rare occasions that I discovered descriptions of discernible differences (sorry, I delight in alliteration!) between dBpoweramp and Exact Audio Copy rips, I estimate the public's preference to be 50/50.
I need to experiment further, as well as gather more opinions. There is definitely something going on here.
I would love to get a productive discussion going. If you have any questions, don't hesitate to reply. I will be happy to help.
I especially look forward to hearing about specific details of variations in sound, be they real or perceived.
(For the moment, I will decline to disclose my own experiences, to prevent any bias on your part...not to mention that it would take eons to document the fine points.)
--- Resource links ---
Edited by Music Alchemist - 1/1/14 at 3:44pm