Ripping CD's in Safemode sounds much better...

Status
Not open for further replies.
I have worked for a long time with systems streaming data in real time to hard drives. There can be large differences in how data is stored on disk - first where on the disk.
-Inner or outer part of platter - disks rotate at a constant speed, so data rates, both read and write, are fastest on the largest diameter sectors.
- Interleave rate - usually set at format, makes a difference in performance.
- File fragmentation - almost always happens, unless special means are used to prevent it.

I believe you can reduce/eliminate fragmentation by copying the audio files from one drive to another, then back again. It may be interesting to try that in this case. Also there are some very sophisticated defrag products out there - for instance Disktrix Inc - The Home Of UltimateDefrag
Who also has a free version UltimateDefrag - The Defragger For Power Users - Trial

These products have an option to let you decide where to put files, i.e. on the outer tracks of your disk.
Not saying that is what's going on here, but it may be.
 
Good information. Thank you.

Do you know more about how data is written to a HDD? Does wow and flutter affect the speed the platter is spinning like an analog tape or LP? Could this alter the space between individual bits when the data is written? Could this affect the playback?
 
Last edited:
Sorry I don't quite understand what you are asking?

Can you elaborate?


I hand you ten files of the same musical selection. Five were ripped with computer in safe mode, five were ripped with computer in "normal" mode. They are numbered, say, 1 to 10. You listen to them using whatever you used to determine "obvious" differences. You sort the files, e.g., you say "3, 5, 7, 9, and 10 are normal, 1, 2, 4, 6, and 8 are safe mode."
 
I hand you ten files of the same musical selection. Five were ripped with computer in safe mode, five were ripped with computer in "normal" mode. They are numbered, say, 1 to 10. You listen to them using whatever you used to determine "obvious" differences. You sort the files, e.g., you say "3, 5, 7, 9, and 10 are normal, 1, 2, 4, 6, and 8 are safe mode."

Cool no worries. download dropbox or something similar, and send me the link in a PM and I'll download and test them.
Actually, if its OK with you, I'd prefer one file as the safemode and one as normal. Ten seems excessive to me. Is there any reason for more than one song to be compared?

But really I'd prefer you to listen for yourself, and report your findings, because I have already reported my findings, now I'd like to see if you or anyone else can hear a difference.
 
Last edited:
If you were to read both files form a shared drive, wouldn't the network protrocol iron out any problem with the rip on the physical media? If the difference is not in the data itself, and it does not seem to be since the two files are bit for bit the same, then the differences in their physical storage would tend to even out by the process of reading the remote drive, packaging the data into tcp/ip packets, sending them, receiving them and having to assemble them to generate the correct order of packets and then feeding the data to the sound streaming software that is requesting the data read. If the file sharing protocol caches data at either end, then this would tend to iron the differences even more. Just wandering if the two files still sound different when both are read by the streaming software from a shared network drive as opposed to a usb stick or local harddisk (which is accessed not by a networking protocol but is read by the local computer's disk controller hardware from physica media).
 
All other things being equal (fragmentation, location on the HDD, playback software/environment), if the checksums match, the files are the same. That's practically irrefutable.

I would have to agree with others here, that if you are indeed hearing a difference, it must be fragmentation, location on the HDD, playback software/environment that are causing the aberrations.
 
Good information. Thank you.

Does wow and flutter affect the speed the platter is spinning like an analog tape or LP? Could this alter the space between individual bits when the data is written? Could this affect the playback?

No. BTW, typical streaming data read and write rates for single hard drives are in the ballpark 15-30MB/s range. Audio wave files need about 1/10 of that. In general, data goes from the hard drive, to memory, to the processor, then out to the sound card / network whatever (I use a squeezebox).

In general, the PC memory buffers any effects from the hard drive. The place for concern is going from the processor to the sound card, but still, in most cases the sound card also memory buffers and does the final clocking onboard to the DAC (exceptions maybe for certain USB cards).
 
Erin,

Sorry not to have heard back from you regarding my suggestion of a binary file comparison. It's really very quick and easy to do, but perhaps I wasn't clear enough on how to perform this. No special tools are required, just a command prompt and the following command (substitute your own filenames and paths, obviously):

E:\Soundfiles>fc /b normalrip.wav safemoderip.wav

If you could please take the time to do this, I think it would be very helpful to this discussion. So, thanking you in advance for your helpful co-operation and eagerly anticipating your results.





 
When reading and writing the data to the disc, its just data to the computer, it dosn't know its music, its just data its got to fetch and transfer on the PCi bus to some other location. It does all data to and from disc with the same way, controlled by low level functions, that form the basis of the operating system (its the DOS in DOS, ok windows is used these days but the acronym gives us the clue).
You cant record jitter during a CD rip nor can you record noise!, and the data transfer system has numerous functions that ensure the data is valid. If this wasn't so a lot more critical things would go down hill fast, that rely on data integrity.
If all the bits match in two files then they are exactly the same, there can be no difference in the data. Any percieved audio differences in a music file would have to be due to the playback system.
Again all this data transfer is done through buffers, so even fragmented data is read into the buffer, the head just has to move a bit more erreticly.
 
The data is ripped not in real time. Even if the PC does something else in that moment, and pauses the ripping process, it will resume the ripping from where it was left.
This it is different from actually playing a CD, when you cannot have "pauses".

There is ONE reason why a copy it COULD sound different. If you verified the bit check-sum on a PC drive, it will be the same. But, if use for comparative listening another dedicated player, the results might vary.
This is when the player that used to listen the discs is an old one not really ment to read CD-R and/or maybe it has with a worn laser. That player cannot read properly the CD-R discs (especially ones burned at maximum speed). Therefore it might loose a few bits on the CD-R disc, bits that get interpolated by the internal circuit and therefore the result WILL sound different. On that POS player.
It doesn't make it "true" for any other player.

If one would just use a modern CD/DVD-player, the audible differences would be zero.
 
Last edited:
Administrator
Joined 2004
Paid Member
But really I'd prefer you to listen for yourself, and report your findings, because I have already reported my findings, now I'd like to see if you or anyone else can hear a difference.
I've spent too many hours listening to these type of files that were ripped differently and thus "sound different." I could never hear any difference.
Various headphones, amps, speakers, DACs, transports, etc. Oh sure, I "thought" there was a difference, sometimes even on the same file, but there wasn't. I could never pass a blind test.

Of course I was told that the resolution of my system is not good enough. Right.
Once I ran a checksum and Diffmaker and found the files to be identical, I was no longer surprised that no difference was heard. A colossal waste of time.

If the files checksum the same, then there can be only two reasons why you hear a difference.
  1. Playback is somehow different
  2. It's your imagination
 
Let's be quite clear about this:

Erin's claim is that he hears a marked difference between playback of rips made in normal vs safe mode. The playback is of the files from hard disk. The two files have the same checksum.

Erin wishes us to try ripping a cd in normal and then in safe mode and compare the results. He is so confident of the difference being real that he is not interested in doing ABX testing, and is dismissive of Pano's suggestion that he can equip himself to do this for free using foobar2000 with the abx plugin.

He does not appear interested in doing a simple binary file comparison, using fc.exe at a command line.

This is not an approach that would cut any ice with any scientist anywhere. Erin's experimental "method" is woefully inadequate and he wants us to spend time repeating it.

Experimenter expectation is an all too common phenomenon and I think this is very clearly just that.

Thank you Pano, for making me aware of the foobar abx, I've downloaded it and am glad to have this facility.

SY, I think your suggestion of a ten track test reasonable and of appropriate statistical sensitivity and specificity.

I'm sorry Erin, but I'm not interested in discussing this any further
 
This is another classic PC thread!
Just as silly as the hand made silver SATA cable thread.

What happens if you rename the files so that you swap the names round? If the files match, surely it's the filename that's causing the difference?
Erin, I think it was your first post where you knew this shouldn't be, but thought you could tell a difference. Shouldn't you be trying to disprove your belief based on what you know?

I'm going to have to think up some obscure PC based discussion to kick off myself. Maybe how fuses affect bass response on low bitrate MP3s. Hmmm...
 
I am quite amazed on the reaction from the viewer of thus thread. Erin thought he found something and reported here. It is a simple thing to do, just rip a cd in safe mode. It is not like he was selling something or asking someone to jump over a cliff. if you don't like his report just skip it. If you are interested then just try it. or, if you don't want to do anything, let it be then. If you do try it and find his suggestion works, let's discuss it. just my 2 cents, peace.
 
My point is (and I believe others share this) is that data on a HDD is data. If two files match (as indicated by a checksum) then they will be identical. If they are different, then it is very unlikely indeed that the second rip has performed some kind of interesting DSP algorithm that makes it sound different. If they are different, it will either be a different file (e.g. a different track - so completely different) or corrupted, so you'd hear pops and clicks. I don't think Erin's hearing differences like that.
The suggestion to do a file compare should resolve this whole argument.
 
Status
Not open for further replies.