Audacity Forum

Re.: Basic Recording, Editing and Exporting Tutorial
The tutorial advises saving a project as a WAV file for future use, e.g., re-editing. Why can I not just save my unedited .aup project files and use those for the same purpose. Those get created when I make my recording and surely have all the data that a .wav file exported from them would have, right? I save the .aup projects without any editing (I export to .flac after editing) So, should I be unhappy with one of my end-products ( a .flac file) could I not just reload the project and re-edit and export? IOW, why have both .aup and .wav raw master backups?

I am seeing two different suggestions for Noise Removal settings: Freq. Smoothing of 500 Hz & 300 Hz and Attack/decay Time of 0.5 sec and 0.25 sec. Is this because of the changes in the beta's version of Noise Remover? What is the safest setting to use for the beta version of Audacity?

And, if I use Chris' Compression would I then not need to use either Normalize or Amplify effects?
Thanks,

Richard,

I can only really comment on your first question: the reason for exporting a WAV is that it is a single robust fairly transportable music file - and Audacity project is a very complex beast with lots of little audio clips strung together by the "recipe" in the .aup file - and sometimes dependent on external files (depending on your preferences settings). This makes them very prone to damage if moved or similar, unless you really know what you are doing - and thus not very suitable for archive.

So for simple single stereo track projects, then exporting as a WAV is almost certainly the best option - however if you have a multi-track project that you wish to preserve, then saving the project is the easiest way to back it it. Easiest but still not the safest - cautious people with such projects will export each track as a separate WAV file (IIRC there is a feature request on the Wiki for an export of a multi-track project into separate WAV diles - and for the reverse import).

WC

Thanks WC,
That makes sense. All those multiple bits in the Audacity Project certainly do make the data more vulnerable. It is an easy thing to save the .wav, so I will do so. Thank you,
R

rorser wrote:
I am seeing two different suggestions for Noise Removal settings: Freq. Smoothing of 500 Hz & 300 Hz and Attack/decay Time of 0.5 sec and 0.25 sec. Is this because of the changes in the beta's version of Noise Remover? What is the safest setting to use for the beta version of Audacity?

And, if I use Chris' Compression would I then not need to use either Normalize or Amplify effects?
Thanks,

I'm not sure what the defaults are for NR. I think you can restore all defaults by exiting Audacity, deleting the audacity.cfg file, and restarting Audacity.

I've seen suggestions of 18, 500, and zero. I've been using 12, 150, and .15.

Frankly, I've never had a reason to experiment with smoothing and decay. The NR slider alone does what I want--if you move it too high, you will begin to hear obvious artifacts.

Move it intentionally way up so you can hear the artifacts and are therefore aware of what to look for and avoid in the future.

I have seen a situation in which I could detect no artifacts, even at 48 on the NR slider. On other material, I hear no artifacts at 12 and do have artifacts at 18, so it's a moving target depending on the source material.

Re: Chris' compressor:

I have experimented with it quite a bit lately.

It is NOT a substitute for normalize or amplify.

It can create audible artifacts, depending on the setting.

My general advice would be to use it sparingly if at all, UNLESS you actually prefer a compressed sound with little variation in the level as a song plays. I don't like that, but it is common on most retail CDs and radio listening in the last 20 years.

The only reason I personally would use it is to attempt to get equal volume across a bunch of unrelated tracks on a CD burn. It helps for that, but it is not perfect. You can do the same job, probably with less fiddling when all is said and done, by listening to the first 20 seconds of each prospective track in the burn before committing to the burner--and adjusting volume by ear then and there.

I'm sure the source material matters a lot as well. You probably couldn't even detect compression on the average yammering high volume modern rock record, but would pick it up easily on a finger picked acoustic guitar instrumental, where there is space between notes and the natural attack and decay of the strings is the attraction.

So the question is----why are you considering compression? You like the sound? You see it as a replacement for amplify? Other?

Well, I just checked the NR defaults by deleting Audacity.cfg and restarting, as mentioned above.

They are: 24, 150, and .15----so I have been using the defaults for smoothing and attack/decay all along.

That file is found at the following location in Windows 7 and probably in a very similar location in Vista:

C users yourusername appdata roaming audacity

You have to "show hidden folders" to see that location. Otherwise, appdata and its subfolders won't appear.

Thanks, Ignatz,
I think I am getting it now. I really see no reason for me to use compression for what I plan to use these audio files for.
As for the effect Amplify, might it cause any artifacts or does is merely raise the amplitude to the max. I set, e.g., -1 dB. IOW, it would not distort or change the music content itself???

Thanks for all your help.

I can't speak as a software or audio engineer, but Amplify is not supposed to cause distortion (clipping) when it increases volume to 0 db.

Whether it does that strictly and at all times, I can't say.

The rule of thumb I have heard is to leave some headroom as there is no point to going for 0 DB.

You are leaving 1 DB headroom, which I think is common.

I leave 2 DB.

Amplify does not relieve you of the responsibility to set a decent level when recording. Again, the rule of thumb seems to be circa .5 if you are using the waveform scale on the left or peaking at around - 6 DB if you are reading the red meters. Setting a level was very critical in the days of tape, but is less critical on digital recordings so long as you never go above 1 (zero DB) on peaks.

rorser wrote:As for the effect Amplify, might it cause any artifacts or does is merely raise the amplitude to the max.

ignatz has covered most of the points.
The Amplify effect will "merely" raise the amplitude to whatever you have set it to. If you set it too high the OK button is (by default) greyed out to prevent you from accidentally clipping the waveform. If your track is in 16 bit format there will be an extremely small amount of distortion due to "rounding" errors. If your track is 32 bit (which is the Audacity default) then scaling (Amplifying) will be virtually perfect.

Thanks Guys, Just what I needed to know.

stevethefiddle wrote:

rorser wrote:As for the effect Amplify, might it cause any artifacts or does is merely raise the amplitude to the max.

ignatz has covered most of the points.
The Amplify effect will "merely" raise the amplitude to whatever you have set it to. If you set it too high the OK button is (by default) greyed out to prevent you from accidentally clipping the waveform. If your track is in 16 bit format there will be an extremely small amount of distortion due to "rounding" errors. If your track is 32 bit (which is the Audacity default) then scaling (Amplifying) will be virtually perfect.

Since these are reel-to-reel recordings of LP records, I should, according to the Tutorial, use click removal before using Amplify. I do not see much discussion about the batch Click Remover (by Craig DeForest, but more about removing individual clicks by hand--something that I am unwilling to do because of the time involved. Is there much of down side to running all my recordings through that before Amplify?