Audacity Forum

Coping LPs via UA25EX sound card to Dell win8.1 laptop.
Trying to get high enough level to avoid noise without clipping.
The last test adjustment in gain gave me a lot of red vertical lines.
However when I play it back I do not hear anything.
When I look at the waveforms up close they do NOT even hit 0dBFS let alone clip.
Some of the clipping seems to allegedly happen when the waveform is at huge negative values down below the noise.

So my question is does anything else besides clipping cause the red lines to appear?

If this is user error then please let me know what I did wrong. thanks.

A high value in either direction will trigger the red bars. In the outside world—outside of Audacity—those are the two places where the digital system "runs out of numbers" and stops following the performance. The damage can be really tiny, fast and high pitched, so if you're old enough to have LPs, you may be too old to hear them.

Audacity uses an internal sound format that is very resistant to overload so you can apply effects or filters that get too high, but leave you the option of resetting the volume later, no harm done. But that's not how the turntable and interface works. Once you run out, it's instant tick, pop, snap and crunch.

It's also possible that the cat hairs on the records are causing pops and they're triggering the red bars. In that case, the pops are just getting worse when they overload.

Are you planning on using any of the vinyl software products?

http://manual.audacityteam.org/man/tuto ... iques.html

Scroll down to Click Repair. It's pay to play, but very well thought of.

Koz

There's an interesting argument with overload sensors (the red bars) You can't really do that.

Any wave that goes right up to the overload point but doesn't go over is perfectly valid and correct.

Anything that goes over is wrong, but there is no "over." The waves will always stop at the overload point, so the best you can do is sense intent. The wave would have gone over had it really wanted to. Some of it is guesswork. I think what Audacity does is ring the alarm bell if the wave gets too close. Another way is to sense the wave has reached overload and then stays there too long. No natural sound does that.

Someone will correct me.

Koz

Coping LPs via UA25EX sound card to Dell win8.1 laptop.
Trying to get high enough level to avoid noise without clipping.

There is virtually no digital noise* so unlike analog tape there's no need for a hot signal. It's OK for your peaks to sit around -6dB. Pros often record at -12 to -18dB (at 24-bits). And, unlike analog tape, there is zero-headroom above 0dB and your analog-to-digital converter will hard-clip at 0dB.

Of course there is analog noise from the record & preamp, but the signal-to-noise ratio is not damaged by a lower recording level.

You can normalize to 0dB after recording. When you do that, you increase the signal and noise together so again, the signal-to-noise ratio is no affected.

The last test adjustment in gain gave me a lot of red vertical lines.

If you're getting red before any "processing" you are hitting 0dB (on the positive or negative half of the waveform) which means you probably "tried" to go over 0dB, and you are clipped.

However when I play it back I do not hear anything.

You don't always hear the distortion, but it's "bad practice" to record that way. Some pro mastering engineers do push the levels into clipping to maximize the loudness (Loudness War) but if you choose to do that, you shouldn't clip during recording, you do it during post-production where you can control it or un-do it.

When I look at the waveforms up close they do NOT even hit 0dBFS let alone clip.
Some of the clipping seems to allegedly happen when the waveform is at huge negative values down below the noise.

As Koz says, the signal swings positive & negative and it can clip either way. Noise should (hopefully) be near the zero-center around the silence, unless you have a big "click" in the record.

The digitized record probably won't sound as loud as the CD version. Older records weren't as dynamically compressed as modern "loudness war" CDs and the process of cutting & playing the record increases the crest factor (peak-to-average ratio) without affecting the sound of the dynamics. But with a (slightly) lower average, it usually won't sound as loud.



BTW - For "snap", "crackle", and "pop" reduction Wave Corrector is now free!



* There is something called quantization noise but it's something like 96dB down at 16-bits and even further down at 24-bits so it's not something to worry about. The analog noise from the record & preamp are far-far worse.

Thanks to all for their insights.

I fixed it by lowering the gain in audacity instead of trying to move the knobs on the preamp which were hard to tweak accurately.

It looks like the LPs are peaking at about -6dBFS now.
No red lines so far. But still several LPs to convert.
When I finish importing them then I will clean up any problems like ticks/pops, hiss, hum etc, IF there is any.
So far on sampling them with brief playback I hear nothing. But the records are pristine, the equipment grounded properly, etc.

One early test did have hiss. I am going to redo that one with the higher gain to see if it is on the record or due to the low level the first setting was at.
The good news is that the hiss filter worked great. But I would have to apply it for each band as doing the whole record at once fails due to lack of memory.

Now I am old, so my ears may not be as sensitive, but these new CDs are for me to listen to instead of playing the LPs again.
I got a CD changer so I can put 5 of them in and play them all in sequence without intervention. Maybe I will end up getting a DJ program someday that can replay all my CDs from digital on a HD. Then I would have to import all the CDs too. More likely I will stick with the changer and put my time into other projects like writing a program to play baroque style music automatically from a melody line.