mastering a noisy speech recording for car audio

[cordura21]

heya guys.

I have to make a cd for listening on a long trip on a car, that has those Pioneer stereos with that V eq curve.
The source is a noisy speech, transferred from a cassette.

Since I am mastering it just for car audio maybe there was something I can use to improve the intelligibility. As it is from a cassette, it has plenty of hiss. But when I remove I also understand less.

Is it ok to use hard filters (as I am doing) on the trebles, or would you do something more subtle, like using a multiband eq and taking the mid highs more upfront?

I am using a band reject for the lows cause it's speech and it's a car stereo. I guess this is the right thing to do. Any other idea?

Cheers, Andrés

 

[TexRoadkill]

If the noise is a constant pitch you might have some luck using noise reduction in a DAW.

 

[Sangram]

I've had a lot of success using the cassette tape NR filters on a program called Diamond Cut audio. It's got a nice 40 day trial without limitations. I'm converting a lot of my cassette tapes into CDs and MP3s over the last few months. It's an ongoing project.

I lose very little of the top end this way (I actually mod the filter response heavily, I first boost (!) the treble and expand the dynamic range a bit, you may not want to do that if it's speech only.

I have a tape I did, which was recorded 25 years ago on a flatbed Japanese tape recorder, on some pretty good German tape, I don't remember the brand. It was baby talk (OK, mine!) which my parents decided to keep for posterity. About 10 years back we noticed the tape was detoriating significantly and I transfered it onto a new tape, thus pulling the recording down one generation, blissfully unaware.

On returning home after those 10 years, I decided this would be a great way to learn some mastering. Let me see if I can send over some sound samples once I get home (I'm at work now). I managed to get the recording into some shape. It was terrible before that, lots of pops clicks and ambient from the condensor mic, and a huge amount of motor hum, stop/start clicks, etc.

Sang

 

[JR#97]

ahhh... one of my specialties.... Like the others have said, the DAW makes it a lot easier. But if you dont' have a DAW, to take care of the tape noise, cut with a narrow bandwidth at around 3.5k and also at 7k. Parametric EQ is what you want to use. Sweepable mids on most mixing boards are too broad and most don't have adjustable bandwidth. If there is other hiss other than tape noise, than try and notch it out as well. Keep in mind that the goal here isn't to eliminate the hiss, but to pull out back so that it isn't dominating. For voice only stuff, cut everything above and below the spectrum of the voice in question. Definitely everything below 60hz and maybe even upwards of 80hz. On the high end, you need to be a little more careful because you don't want to lose intelligibility on T's and S's and some of the hard consanants. Also, maybe boost in the mids a bit to compensate for the tape noise notches.

CEP is one of my personal faves for noise reduction. Sound Forge is good as well.