Clever idea, Scott
I know it will not please everyone, but it's easy to try because the 2 channels should line right up.Before doing the recording, tho, I'd look at both channels of the ripped files to check if they are actually identical.
Clever idea, Scott
Before doing the recording, tho, I'd look at both channels of the ripped files to check if they are actually identical.
I would double check but someone on the iTunes forum posted that they were.
Curious about this, so I checked my rips of Hard Day's Night (Mono Remaster). I do NOT find the left and right tracks to be identical. I tried two different wave editors, nulling the tracks and also extracting each track as a mono file and doing a checksum. They are not identical.
That said, the differences are way, way down. But they exist. Spectrum analysis suggests that the difference is mostly 15Khz, but it makes it impossible to get an identical checksum. Can't see bit perfection that way.
Please let me know if you find something different.
That said, the differences are way, way down. But they exist. Spectrum analysis suggests that the difference is mostly 15Khz, but it makes it impossible to get an identical checksum. Can't see bit perfection that way.
Please let me know if you find something different.
The CD mastering sites have data that shows CDR and CDRW as noticeably worse for C2 errors so this becomes an apples oranges test.
Thanks, the other info might have been wrong or about a different version (?). You find what I would expect, the exact mono analog signal is fed into two A/D's where there would of course be very low level bit differences. You would have to force the digital data to be exactly the same and I am not sure a recording chain would have that option, why would they bother?
Last edited:
I seem to remember reading somewhere about music tracks which had a low level 16kHz signal - from the Rugby VLF transmitter. If the mono signal was simply fed into the L and R channels of the ADC then the output would not be bit-identical, although it ought to be very close.
Understood, but mastering labs can produce master CDRs that have been carefully burned on top quality media and then BER tested. Those can then be send to a stamping facility for duplication. So, CDRs can be very good if they really need to be.
Whether all players will read them as well is another question. In any case, if a CD drive can read a CDR with no non-recoverable errors, then it will probably do the same for a CD.
Also, some people may want to listen to CDRs on their CD players, so having some information about performance for that type of media might be of some use.
Mark, I've done the burn test, and found bit perfection. The big headache was aligning the recorded file with the original on my computer. With test tones and noise I put a 4410Hz maker 3 cycles long after 1 second of silence, as well as another at the tail of the track. That marker is easy to find and not difficult to cut right at the zero crossing. That allows perfect alignment of the track. I find this much, much easier to do in Goldwave than in Audacity, FWIW.
I have also put the marker at the head and tail of prerecorded music tracks to be able to do the alignment. A little trickier, it can work. Without the maker, I find it almost impossible to get bit for bit alignment because the beginning of the recording isn't obvious. Maybe it's worth trying a level zoom to detect the beginning and end.
FWIW, if the Beatles mono remasters where fed into two different DACs (dual mono) then they are superbly matched. Below see the spectrum of the null result between left and right. The difference is boosted 40dB. Any difference is waaay down. But it's enough to make for a different checksum - thus not bit identical between the two channels.
I have also put the marker at the head and tail of prerecorded music tracks to be able to do the alignment. A little trickier, it can work. Without the maker, I find it almost impossible to get bit for bit alignment because the beginning of the recording isn't obvious. Maybe it's worth trying a level zoom to detect the beginning and end.
FWIW, if the Beatles mono remasters where fed into two different DACs (dual mono) then they are superbly matched. Below see the spectrum of the null result between left and right. The difference is boosted 40dB. Any difference is waaay down. But it's enough to make for a different checksum - thus not bit identical between the two channels.
Attachments
For aligning tracks, I think it should be possible in Reaper. You can stretch out the time scale enough to see individual samples. In some cases it may be possible to line up percussion transients, which I have done before for other purposes.
Yes, Reaper or other. In Goldwave I zoom into individual samples. But to see what is going on with leading and trailing silences, you also need to zoom into the amplitude axis. Seeing silence isn't all that easy.
I look for a good transient anywhere in a track and use that. Doesn't have to be at the beginning. I guess I'm not clear why that wouldn't work in your case.
Because I had trouble both finding a transient and then aligning it. I have a hard time finding a certain waveform somewhere in a song. I think I can find it, but then it escapes me. We are talking about finding a 1/44100th of a second in a 3 or 4 minute recoding. For me, it's difficult. Perhaps you have a lot more practice than I do. It just isn't a fast or reliable method for me.
I roughly line tracks up by eye with the time scale zoomed out. Then put a marker on a nice transient on one track, and start zooming in and nudging the other track(s) more precisely into mutual time alignment the first reference track. With the zoom function set to track the timeline placed at the marker and reference track transient, Reaper automatically zooms around that point, so I don't need to keep finding the same spot on the track, it never gets lost. Don't know if you use some method like that, but it worked pretty well when I've tried it.
I'll have to give it another try. I found it very difficult. What do you do about trimming up heads and tails? For example of the re-recording is longer. How do you make sure that you are trimming at the exact sample?
Zoom in until you find the sample you want. At that zoom level, I don't think you can trim at just any arbitrary sample not located on some valid chunk boundary. So, I find the closest time location I can snap the timeline to, select all the tracks, and keeping them in time alignment with each other, move them so the sample to trim at is lined up with the timeline. Put a marker there, zoom out, and trim up to the marker.
You can also perform a mixdown from a time selection region, rather than mixing down from the beginning to the end of the whole project. So you can select a time region for one or more tracks, and then render a new wav file beginning and ending at the selection boundries. Depends on what you are trying to do, but that is one possible option.
There is also a slider function to trim the ends of tracks of tracks or sliced up segments of tracks. You grab the end of the wave file with the mouse and drag it to where you want to trim. You might want to turn off the snap-to-grid function while making fine adjustments.
You can also perform a mixdown from a time selection region, rather than mixing down from the beginning to the end of the whole project. So you can select a time region for one or more tracks, and then render a new wav file beginning and ending at the selection boundries. Depends on what you are trying to do, but that is one possible option.
There is also a slider function to trim the ends of tracks of tracks or sliced up segments of tracks. You grab the end of the wave file with the mouse and drag it to where you want to trim. You might want to turn off the snap-to-grid function while making fine adjustments.
Last edited:
I suppose I can have a try. I found it very difficult to do. I kept thinking that things were lined up, only to find they really were not. 
- Status
- Not open for further replies.