A CD is a CD is a CD?

[Dan McDonald]

Here is a puzzler I have. A friend in Ireland and I were talking about HDCD formats and what was different about them, and whether there were any noticeable differences. He had an HDCD, so he ripped it with software that supposedly converts the CD to 24-bit, 44.1 kHz files. He ripped the cd using that software and also additional software that is supposed to reproduce the HDCD information using 16-bit files. He sent those files to me. I compared them using the spectral filter (spectral difference) and saw very little difference, except for slightly higher treble end in the 24-bit file (on the range of 2dB or so increases, up to about 4dB increase at 20kHz). So far so good.

Then I got curious as to whether there was any difference between that and a standard CD ripper that doesn't take account of the HDCD information. It turned out he didn't have a standard ripper handy, so I started poking around and found I had an HDCD cd, so I ripped it in Diamond Cut and then another piece of software, and saw similar differences between the two files. I didn't think the other software was supposed to take account of the HDCD tags, so that surprised me. Then I thought that it might be a normalizing difference, so I gain normalized each file to 0 using DC8.
The differences were still there.

So... I took a standard commercial CD and ripped both with the DC8 and the other software. When I compared them, I saw similar differences. I gain-normalized both, and still saw the differences. In this case, there was a bump of around 4dB around 125 Hz, and also some differences in the higher frequencies, as well as just minor differences throughout.

I always thought that ripping a CD gave exactly the same information, no matter what ripping software was used, but the spectral difference filter is saying that these aren't the same.

Is my logic wrong or is there some easy explanation? Do different CD rippers produce different results?

Dan

[Craig Maier]

Hi Dan,

I am under the impression that 16 bit, 44.1 kHz CDs are exactly the same as 16 bit, 44.1 kHz .wav files. Ripping them just re-arranges the file headers so that they look appropriate for the format being converted to.

So, I am confused by this.

Craig

[Dan McDonald]

That was what I thought too - I thought that if you took a wav file, burned it to cd, extracted the wav from the cd, the two wave files would be identical, or that if you ripped the cd with other rippers they would be identical as well. This has me very puzzled.

[Dan McDonald]

I think I have this figured out. If I burn a cd with DC8 and rip that CD with DC8, I get a file that produces only one line in the spectral filter (the green one we start with; the blue one is behind the green one). But, if you look at the red "difference" line, it is very jagged for some reason. I expected a straight line. But even when I zoom in, there appears to be no difference between the actual spectral profiles. The same thing happens with a commercial CD ripped by DC8 and another software product. The red line is jagged, but there appears to be no difference between the green and blue lines.

When I rip the HDCD using the two different rippers DC8 and free Brand X, I get the same thing - no difference in the green and blue lines, but it looks like as much as a 3dB difference in the red line.

Dan