ripping in secure mode vs burst mode

dBpoweramp has a similar facility and if you're going to the trouble of ripping and correctly tagging a CD then indeed, why not rip it properly?

No matter how good a CD player is, it will only play back what it has to pass on from it's buffer to ensure uninterrupted playback, missing information and all. The beauty of streaming is that all this error checking is done and checked and rechecked etc. when ripping and the ripper can take as long as it needs to get everything.  Assuming you use a good enough ripper. Subsequent playback involves no interpolation - if ripping was accurate.

It's no wonder to me that streaming at 16/44 knocks spots off CD replay on equipment at similar price levels. 

I too use DBPoweramp and rip using secure mode.

However, I would be very surprised to find any audible difference unless there happens to be a physical problem with the CD that requires additional checking or correction.

I may try a comparison myself when I have time, but (intuitively) don't expect to find any difference. 

I never expected to hear a difference from a brand new CD, but there was, and not a small one.

which now clarifies why products like HDX and Unitiserve exist at all

HDX uses secure mode instead of burst.

Harry posted:

dBpoweramp has a similar facility and if you're going to the trouble of ripping and correctly tagging a CD then indeed, why not rip it properly?

No matter how good a CD player is, it will only play back what it has to pass on from it's buffer to ensure uninterrupted playback, missing information and all. The beauty of streaming is that all this error checking is done and checked and rechecked etc. when ripping and the ripper can take as long as it needs to get everything.  Assuming you use a good enough ripper. Subsequent playback involves no interpolation - if ripping was accurate.

It's no wonder to me that streaming at 16/44 knocks spots off CD replay on equipment at similar price levels. 

I've never understood why high-end CD players don't just buffer a few seconds of material so that they have time to re-read and do the error correction that happens in "secure-mode" ripping. Even a basic CD drive can read a CD many times faster than real-time playback. They could be designed to simultaneously rip and stream instead of simply playing back.

Yup I have said this on many posts and been challenged on this to a point that I wrote some code to compare the output of the two WAV files after the RIFF header per 4 byte block.

It's important to understand that CD red book data is not files and therefore cannot be validated for consistency in the same way as it contains no inherent bit validation. C1 and C2 error correction, while generally good are not fullproof. Although digital, the data on a CD has more in common with vinyl than WAV files stored on your hard drive. It is a series of streams where the TOC provides time offsets for the tracks that translate directly to how far in the laser must track to find the strart of each song. The tracks themselves contain no identifying headers or trailers and, under the right light can be seen with the naked eye. This is why "bit are bits" arguments have no meaning when it comes to ripping CDs and why how you rip is really important.

To me, the difference between secure and burst mode was larger than the difference in SQ between a bit perfect rip of a CD and a 24/192 hi-res download. And considering that your basic software like Media Player and iTunes will use burst mode adds to the argument that if you ripped a collection with substandard software but have spent plenty of money on a quality system, you are not gtting your moneys worth. My findings are buried somewhere in another thread including the percentage difference between the rips (It was in the thousandths of 1% I beleive which is fairly large).

Burst mode is fast and very tolerent of scratches but fairly ugly. For most people with zero patience this is the best option. But it makes no sense for hifi.

As for buffering of CD audio, even older player do. But they do this at 1x speed in CLV so only get a couple passes before the laser must track on. To do this at higher speeds with more re-read like a computer drive, the noise generated by spindle speed changes and laser tracking to move back and forth would probably be more detrimental to SQ in a CD player than just making your best pass first time round.

I too have analyzed using file debugging tools data chunks within WAV RIFF files produced by different rippers from the same CD (Unitiserve, EAC, dBPoweramp, iTunes).. They all apart from iTunes produced identical bit perfect identical copies that additionally sounded identical... The only difference in iTunes was that it had a few milliseconds offset .. where there was samples missing from the very start of each track .. not normally audible unless the tracks ran into each other) .. Other than that the data produced was identical. It turned out this was a PC running iTunesand there was some mismatch between iTunes and the CD-ROM (iTunes wasn't able to calibrate the CD-ROM)

Now if a CD is damaged or marked, and/or the CD drive is worn or badly calibrated the burst mode may introduce errors . However these errors are in the data domain, not the analogue-time domain, and so they appear as clicks or skips. It is conceivable in burst mode the CD ROM drivers are changing the data in a particular set up..though it feels strange. However why not use AccurateRip which checksums the ripped sample data and that way any driver errors or read errors can be identified.

My method is to burst rip and compare with AccurateRip .. if match fine.. If not in the database or not matching then automatically rip using two different methods (burst and secure) .. If these match then fine.. If not flag.. Out of the 30 to 40,000 tracks I have ripped I have probably had no more than 20 or so track failures, and these have always been through damaged or marked CDs. Further I can then decide to accept the failure and the worst side effect will be an audible skip.

Simon

Found it: https://forums.naimaudio.com/to...70#47746327389651270

Findings not dissimilar to SImon's. I could't be bothered to copy out the post again but there is a comment in there about repetative testing I did back in 2014.

Simon the 3 rippers you mentioned Unitiserve, EAC, dBPoweramp, are very good rippers and are expected to have identifical ripping capabilities if secure more is used.

i did my fair tests with EAC burst vs secure mode... and honestly, no matter how you put it, if the drive for ripping is good (i have an LG model that seems to be on the good side), and if the disc is in pristine condition (mine are all like this), there is no perceivable difference (at least to me).

also, no real difference found between the EAC rip and the same album's rip with HDX or unitiserve...

if differences are found, i'm suspecting a rather not so great disc or a rather not so great drive. FWIW, i already "finished" 2 cd drives with a few (30 or 40) rips in secure mode... it is very challenging for a drive to match the exact patterns of the highest security settings in EAC. and is more likely you start with a good drive, and after a while you and up with a not so good one...

burst mode is way faster, and as i said, if drive is good and disc is good, i can see no issues in transferring way faster the tracks read into WAV (or flac) files, any reasonable computer of these days does not have a problem with this kind of processing in a fast mode.

i have a collection of around 5-600 cds, and i haven't ripped them all so far, starting this summer... if i had used all the time secure mode, it would probably take me a hundred years to rip them all, and added cds if i will further buy any...

IF there is a difference, it is so small that to me it does not justify the cost.

for a more accurate comparison, and thus eliminating the placebo effect, i suggest let the remote or iphone or whatever, in other hands, and play randomly from both secure and burst mode rips, and/or HDX rip... if you still can separate with a good percentage all 3 rips then yes, there is a difference.

If the rip verifies against the accuraterip database then secure mode ripping will produce an absolutely identical file to burst mode ripping. This is easy to verify using the comparison tools in dBpoweramp and foobar.  Only in the case of a cd with errors will secure mode ripping possibly produce a different rip. My suggestion would therefore be to use burst mode as your default because it is quicker, but if a cd fails to verify against the accuraterip database, attempt to re-rip it with secure mode (and be prepared to wait).  

btw, if you haven't used a ripper which checks against the accuraterip database but would like to be reassured your rips are good, there is a product called PerfectTunes (from the guys that wrote dBpoweramp) that will check your entire library against the accuraterip database and report any errors, and also (optionally) suggest tag and artwork corrections. Theres a free version with most of the useful functionality.

Analogmusic you had me worried for a second there. I rip using EAC so pulled the laptop out to check that I had been ripping in accurate mode and luckily it seems to be a default setting. Didn't fancy having to re-rip an entire collection due to not checking something simple like that!

I use XLF for ripping.

It gives : CD paranoia

               xld secure ripper 

               cd burst

I use xld secure ripper , is it the best choice ?

XLD secure ripper is fine. as long as it gives an accurate rip match from the database, nothing to worry about.

ok, thanks

For optimum speed and 100% or near 100% reliability  (for slightly damaged or out of spec discs) my suggestion is try and use a ripper that rips in burst mode, and then validates rip in the accuraterip database. If then there is mis match or the CD is not in the database .. often the case for recent Mojo magazine CDs for example, then the ripper should validate the rip by re reading in secure mode.. If these two match then great, if not, let the ripper do brute force rips, where the CD is read multiple times in secure mode and the most common match is then used... 

Dbpoweramp and I believe EAC can be set to work this way or very similar. They key thing is to try and use the accuraterip database whenever you can.. for your confidence and assurance.

Simon

Simon,

I agree that your suggestion is the optimum approach. However, despite my (untested) conviction that there should be absolutely no difference between a non faulty CD ripped in burst and secure mode, I must admit that I do rip my CDs in DBPoweramp in secure mode, just for ease of mind. But then, my CD collection has already been ripped, I don't buy a huge number of new CDs regularly so ripping isn't that much of a pain, and I do have a very fast PC which makes ripping pretty rapid anyway.