Quoting Maluvia <terakuma(a)imbris.net>et>: This is what happens: the data is read from the file via the filesystem. The IDE/SCSI/Firewire/USB controller will do error detection & correction when data is passed from the disk to the controller. Through DMA, this data will be stored in memory. This buffer is the buffer which your soundcard uses as a memory mapped output buffer. (There might be DSP phases between reading the signal and storing it in the output buffers, but let's forget about that for now). Now, this output buffer is the same place in memory where your DA unit reads data to convert into an analog signal. The source of the data inside the memory has no impact on where the data is. The jitter of the data source has no impact on how the data is stored in the buffer. This is pure digital copying, all done with error correction. I can write a piece of software which will read a signal from file(s) to this buffer, and compare them to eachother. If the files are copies of the same data, the signal in the output buffer will not differ. At all. After this phase, the DA unit will convert the data to an audible signal. What the DA unit does with the data is totally independent of how the data in the output buffer has been gathered. At this point, jitter or other differences between data sources have no effect on how the DA consumes the signal. The hard drives do _not_ produce any reading or writing errors, because otherwise your system would be totally unusable, as the programs you are running to read this email would be broken and incapable of performing their duties. The whole of computerland depends on one thing: predictability. Without this, no computer system would work at all. The same error detection & correction which makes your software work, will keep the audio data (copies or not) correctly on disk. Let go of your anxieties. You can make perfect digital copies of your sessions, and that will not make any difference in how they sound. Why do you think the RIAA is after us all? :) Sampo