Spoken Word: Tweaking for Maximum Quality

Special Interest Groups Audio Processing
1

Hello Folks,

I’ve been using Audacity for awhile, and I have a few questions for the guri.

I am editing lectures that were given 20 some odd years ago. Some were recorded professionally, some with a shoebox recorder; they all are now on cassette, which of course gets played into my computer, etc… Generally, I can use Noise Reduction to eliminate the ambient noise in the tapes that were professionally recorded. I can even get the cruddy ones sounding OK, but I want to know if there are any plugins that y’all know of that I might use to make the files sound even better.

I have to say, the one impediment I get with Audacity is the included EQ. Well, that and the 5 second preview. On my machine (Vista, AMD Turion 64x2, Aud. 1.3.5), the EQ will only let me preview one time with new settings. If I adjust the settings again, the preview will only give me the adjustments of the first “change”. So I have to click Cancel, then open up the EQ again ad nauseam. I can really tweak the voice with a good EQ, but this one is annoying. Maybe it is just a 1.3.5 glitch!?

I am using Mag’s Notch Filter to weed out feedback and other anomalies. Often, I use the EQ with everything below 140± turned all the way down prior to running Noise Reduction to avoid too much of the speech being taken out. I have found that the Noise Reduction plugin included with Audacity is better than Peak’s Sound Soap for my particular application. Do I correctly understand that it is a combination Noise Gate and Filter? In other words, not only will it cut audio below the sample I give it’s volume, but it will also filter some of those frequencies out of the retained audio as well?

After losing my MAC and having to go with a PC for Mastering, I was using Nero Wave Editor (6), but this won’t run with Vista, so I switched to Audacity, which I soon learned is much more powerful, though the GUI is not necessarily as friendly.The EQ was much more to my liking. I tried, without success, to import the Nero plugins. I would like to know if there are similar plugins available, in the nonexistent budget range.

I am also on the lookout for some good literature on the subject of Digital Audio. I have been mastering for 4 years, and I have a good knowledge of computers, digital audio and mastering, but there are still some holes. I would like to study up on the subject, particularly in relevance to editing spoken word and other types of old recordings. I appreciate any help on the subject.

2

Forgive me for editing your post to put a little air in there… I just couldn’t read one enormous block of text like that.

You obviously want the “Professional Audio Filter.” Whatever you shot comes out perfect after application of the filter. This is an old video joke and any time the programmers get a little uppity, we ask for that filter.

The problem is you appear to be head and shoulders more advanced than all but the very top of the pile of experienced users. And I’m not in that pile, although I do OK.

<<<it [noise reduction] is a combination Noise Gate and Filter?>>>

Yes, but everything’s dynamic. You can change and tune the effect of the reduction not only in absolute amount, but in attack and decay, and in both frequency and amplitude domains. I had to read the documentation several times and I’m still not sure I have an instinctive feel for all the adjustments.

http://audacityteam.org/wiki/index.php?title=Noise_Removal#Improved_Noise_Removal_in_1.3.3

I, too, can do a much better job with this one than in almost anything else. Not only that, it does look-ahead which helps a lot. No effect trails. You do have to be careful not to overdo it. A clear voice on a dead velvet “black” background isn’t natural. If you do that by accident, you end up putting a little bit of echo or noise back in in post.

Two of the special effects producers in our shop once owned an actual perfect echo-free room and tried to record in it. They had to go out and buy flats of plywood to put an echo back in here and there.

Koz

3

<<>>

http://www.soundonsound.com/

I buy the smeared ink on smashed trees version when I can, but it’s on my HTML speed-dial.

Koz

4

Thanks for your reply. Well, I know better than to expect the “Professional Audio Filter”. What I am looking for is information indicating that either I am doing as much filtering as possible or I am possibly missing something that might give a slightly better result. Maybe I’m not providing enough information for anybody to respond accurately.

I didn’t intend to make myself appear as an advanced user. I am relatively new to Audacity, though I have been editing for about 4 years. All of my experience has been “on the job”; I have made many mistakes, and gone down wrong roads. I would like to think I’ve gotten better since I began, but who knows. My customers might not think so. :ugeek:

I downloaded somebody’s free packet of VST plugins, most of which I am sure I won’t need, though there isn’t much documentation to tell me how to use them in the first place anyways. So I have to weed through them.

Thanks again for your help.

5

There’s a lot of LADSPA and Nyquist plug-ins available from the main Audacity site. For Audacity I would recommend these over VST plug-ins as they are generally less trouble (some VST plug-ins can cause problems with Audacity, and many are hard to use because of the limited GUI of the VST bridge).

Personally I like to do as little processing as possible, restricting myself to what is necessary. Filtering (Eq) and noise removal are the main effects - on Linux there is a program called “Gnome Wave Cleaner” that has a very effective noise reduction effect with lots of tweakable parameters for optimising the process.

Noise gates can benefit some recordings if used in a subtle way. As with noise removal, gating can often sound worse than the original noise if used too harshly.

With old tapes that have been recorded with Dolby B, the digital copy can sometimes be better if played without Dolby, then using the noise removal effect instead of having Dolby handle it. Whether this is better or worse than playing back with Dolby will depend on the specific tape, so you will need to try it both ways. Dolby C, Dolby S, and DBX noise reduction should always be used on playback if they have been used when recording as these technologies use a lot of dynamic processing as well as pre-emphasis/de-emphasis and so cannot be corrected with EQ alone.

The length of the preview can be adjusted in “Edit > Preferences > Audio I/O”
The “first change only” preview is a glitch - the preview should respond to changes that you make. You could try using 1.3.4 instead of 1.3.5 - this version did not have any preview problems on my machine.

6

Right…how do you get 1.3.4??? I don’t see it anywhere on the site…never mind I got it. Yes, the preview does work in this version. Do you know what the major differences are between 1.3.4 and 1.3.5? What will I be missing, in other words?

7

For the English, Windows version - not much. The main difference that you may notice is that if you need to set up latency correction (for multi-track recording), in 1.3.4 it has to be set up manually.
The change log is here: Redirecting to: https://www.audacityteam.org/blog

8

Thanks for the link. 1.3.4 should do me just fine. We’ll see how it holds up.

9

Ok, just an update:

The audio I am stating with are lectures recorded in the late 80’s. They have a moderate amount of background hiss, though it varies from file to file. Some were recorded with a lapel mic and some with a shoebox recorder on a table with a bunch of people sitting around. You can guess which is better.

Anyway, I have formulated a process that seems to work well. I always run Noise Removal last. I found that if I ran it prior to running Hi/Low Pass filters, it took out too much of the vocals.

I first run the Hi-Pass Filter, setting the rolloff to 24db and the frequency to 165Hz.

Next the Low-Pass filter, setting the rolloff to 3db and the freq. to 3250. This seems to have been the best improvement to the process yet, as it removes alot of higher frequncy hiss. Any lower, and the lecturer begins to sound muffled. I leave the q value alone for both filters. I do not quite understand the function of this parameter.

I then run Noise Reduction. This produces a clear product for me. If you have any further tips, I’d gladly appreciate them.

Thanks,
DS

10

That is how I would do it, although (depending on the original recording) I may prefer to set the lowpass filter a little higher.

As a point of information, the 3dB setting is actually 6dB per octave - they have been incorrectly labelled, but I believe that this has now been corrected in the current alpha version.

Perhaps you would like to add support to my “feature request”: http://audacityteam.org/forum/viewtopic.php?f=20&t=7584

11

Steve, I know this was a number of years ago. I just started using Audacity 1 week ago. What you and ogelquaeler talk about here has helped me.

I am using Audacity 2.0 on a Mac with OS X

Just wondering, when you say you would “set the Lowpass filter a little higher”, do you mean that you would set the cutoff frequency higher or that you would set the Rolloff higher? If it is the Rolloff, how much higher would you set it? I used cutoff Frequency =3250, and Rolloff = 36 and it sounded ok… but wonder if my ears might be missing something. So just wanted to clarify what you meant by “Set the lowpass filter a little higher”.
Thanks in advance for your help.

12

Wow, it certainly was :smiley:

I mean set the Low-pass filter frequency a little higher.

3kHz is about the highest frequency on a traditional telephone land line, but you will probably be aware that sometimes it can be difficult to distinguish the differences between an “f” and an “s” over the telephone. The “main” frequency range in speech is around 300 Hz to 3000 Hz but some sounds (“s”, “f”, “t”, “th”) have frequencies that go quite a bit higher and these higher frequencies can be important for recognising these types of sound.

One study I read suggested that for maximum intelligibility the frequency bandwidth should go up to 7 kHz, though if there is a lot of hiss in the recording it may be better to limit the frequencies to less than that. Try experimenting with different frequency settings for the low-pass filter and see which works best. On moderately good recordings you should find that a 7 kHz corner frequency gives noticeably clearer sound but if the recording quality is poor it may be more intelligible set closer to 3.5 kHz.

13

Hi Steve,
Way more than a number of years ago now but I have been searching for exactly this information for quite a while now. Thank you very, very much.
Amanda