Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: mandatory audio processing on imported audio? (Read 3902 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

mandatory audio processing on imported audio?

this link is down from my end http://wiki.hydrogenaudio.org/index.php?ti...cording_(Vinyl)

hi,

as a photographer i know that there are a few mandatory processes that must be applied to scanned film and digitised raw file images, such as levels adjustments and basic sharpening (well, assuming your intent is not to produce an abstract blurry image, etc) because of signal losses via scanning, etc.

does the same apply to digitised audio? Like film and raw files, i always import a wave/raw file to start with and in the past i have only ever applied a slight s-curve adjustment (light amplification) which in a photographic sense increases the contrast/saturation of the image, thereby reducing the amount of visual detail in shadows/highlights

my vinyl importation is simple and does not require 'headroom' (im not going to make large adjustments)

secondly, which ties in to headroom above, i've heard/read varying opinions on which bit rate to capture audio - photographically i could only compare this to editing an image which is 8 or 16 bits in depth - when making moderate/major adjustments, however, the clipping/loss of print information is substantial between the 2 files and can be seen when reviewing image levels (i don't know if it's better to visualise this in audio software or to just listen)

how does this tie into audio and which format is enough to capture/record and/or archive files given my requirements ?

mandatory audio processing on imported audio?

Reply #1
24bit, 48 kHz, set recording level correctly, after recording do peak normalisation, that's it.
Error 404; signature server not available.

mandatory audio processing on imported audio?

Reply #2
I agree.  Nothing's really mandatory.  After digitizing, the digital file should sound exacty like the vinyl (when played through the same system/speakers at the same volume).

This page has lots of helpful information, including software recommendations for "cleaning-up" the digitized copy.


Quote
my vinyl importation is simple and does not require 'headroom' (im not going to make large adjustments)

.... however, the clipping/loss of print information...
You need a bit of headroom so that you don't clip your analog-to-digital converter while recording.    But the levels on vinyl are repeatable and somewhat constrained, so you don't need as much "safety margin" as with live recording.

After digitizing, you can normalize (AKA "maximize" volume so that the peaks hit the "digital maximum" of 0dB).  There is no harm or quality loss in normalizing, although some people worry about "inter-sample overs" and prefer to normalize to around -1dB.  I always normalize to 0dB, because I assume most digital-to-analog converters can handle the inter-sample overs, and I'm pretty sure I wouldn't hear the slight clipping anyway... 
(I normalize the album as a whole, so the volume differences between tracks are not messed-up.)

---------------------------------------
Although it's not "mandatory", I always remove the "snap", "crackle", and "pop", with Wave Repair ($30 USD).  It works great, in the manual mode and and it only "touches" the audio where you identify a defect.  But, it's very time consuming.

I'll usually try some "regular" noise reduction to reduce the background noise, but sometimes this introduces artifacts.

If it's an older "dull sounding" recording, I'll boost the highs a bit.  (Thats optional...  Some people don't like the idea of altering the original sound.)

Then, I normalize as the final step.

mandatory audio processing on imported audio?

Reply #3
Quote
After digitizing, the digital file should sound exacty like the vinyl (when played through the same system/speakers at the same volume).


does this also apply to a burnt CD using the identical file on identical system, but obviously played back via the cd player  ?

Quote
This page has lots of helpful information, including software recommendations for "cleaning-up" the digitized copy.


thank you for the link, very concise resource

Quote
Although it's not "mandatory", I always remove the "snap", "crackle", and "pop", with Wave Repair ($30 USD). It works great, in the manual mode and and it only "touches" the audio where you identify a defect. But, it's very time consuming.


different software obviously employs different ways to remove the rice bubbles - i've heard some subtract the audio and some others copy a small portion of audio surrounding the pops/clicks. Are there advantages to using software that makes use of the latter ?


mandatory audio processing on imported audio?

Reply #4
different software obviously employs different ways to remove the rice bubbles - i've heard some subtract the audio and some others copy a small portion of audio surrounding the pops/clicks. Are there advantages to using software that makes use of the latter ?

Which technique works best will vary depending on the nature of the defect to be repaired. Spectral subtraction is a fairly blunt instrument best suited to broadband noise reduction, although it also forms the basis of the "Younglove" technique for decrackling. Fixing individual clicks is best done by replacing the damaged portion with something that we hope is closer to what it should have been, rather than by applying global processing. Simple or spline curve interpolation can often work quite well. Pasting over the immediately preceding section sometimes works (and is in fact how the old analogue LP declick modules worked). Other times it's better to copy over a similar but undamaged section from nearby (not necessarily immediately preceding - it might even be a section further on).

Bottom line: no automatic mechanism is going to select the best repair method for each click, but it you have thousands of defects to fix, doing it manually isn't really feasible. I believe the pragmatic approach is:

1. Remove any excessive DC offset.
2. Fix big clicks manually, and take the time to select the best repair method for each.
3. If the remaining small clicks are still annoying, run it through your auto-declicker of choice. As it happens, auto-declickers tend to work best on smaller clicks - not because they detect them better, but because the compromised repair methods they use tend to be less damaging on smaller scale defects.
4. If there is a remaining background of crackle that is objectionable, try the Younglove technique (described on the web page referred to by DVDdoug).
5. Perhaps use some light broadband noise reduction on inter-track gaps - but introduce it slowly during the fade-out, otherwise there will be an annoying sudden drop in noise levels as it kicks in. If you can find a noise reduction package that modulates the amount of reduction according to the signal level, that helps.

Once you've dealt with the defects and noise, then you can think about other "tidyup" processes such as trimming, EQ and normalisation.