Hi
On a whim I offered to record a tape to CD. So now I have 2.5gb of recorded tape on my hard-drive and I am wondering what is best to do with it?
I figure, reduce noise, EQ it to make it sound flatter (anyone know what the freq response of a tape is) and then normalise it and slap it on the CD...
<font size=-1>[ This Message was edited by: cleanbluesky on 2005-02-17 15:57 ]</font>
How besT to put a tape on CD
that's pretty much what I would do.
Cassette tape usually gave a low end boost to the sound, so I would use a gentle (6dB/octave) highpass filter or maybe a low shelving filter, cut 3-6 dB, starting around 60-100 Hertz, just experiment to find a good frequency.
Chances are pretty good that there is no useful info above 16 KHz, so you could cut off somewhere around there in the top end, maybe with a steeper sloped lowpass filter.
R
Cassette tape usually gave a low end boost to the sound, so I would use a gentle (6dB/octave) highpass filter or maybe a low shelving filter, cut 3-6 dB, starting around 60-100 Hertz, just experiment to find a good frequency.
Chances are pretty good that there is no useful info above 16 KHz, so you could cut off somewhere around there in the top end, maybe with a steeper sloped lowpass filter.
R
-
- Posts: 2464
- Joined: Tue Jun 05, 2001 4:00 pm
- Location: Warsaw to Szczecin, Poland
- Contact:
What was the tape type? CrO2 (TypeII) or Fe2O3 (TypeI)? If the first then the frequency response is higher than the second type. There is also TypeIV which offered best frequency responses. but I doubt this tape was of that type. Also you have to check if the recorder head tracking was ok. You can do it by just monoing the signal in your computer. If you hear a flanging effecty in the hiogh area then the tracking was not good. If you don't bother your recorder then you can easily adjust the head degree to the tape with a screwdriver. Usually the aproppriate screws are uncovered or even there is a hole in the recorder cover to manipulate it.
_________________
Sir samplaire scopernicus
<font size=-1>[ This Message was edited by: samplaire on 2005-02-17 16:47 ]</font>
_________________

<font size=-1>[ This Message was edited by: samplaire on 2005-02-17 16:47 ]</font>
I do this kind of stuff all the time. It's a bit silly to suggest a plan of attack until you actually hear the raw audio. You really haven't provided any useful information that would allow someone to suggest "what to do next".
Audio from analog audio cassettes can suffer from any one of a number of problems. Or, it may not be suffering at all. A lot depends on the quality of the tape, how old it is, how well it was stored, the type and quality of the original recording deck, original recording levels, dbx or dolby encoding, etc etc etc.
If you want a better, more accurate response, it might be a good idea to post an MP3 of some of the audio for us to listen to. That will give us a better understanding of what needs to be done.
Anyway, here are some generalizations and observations based on past projects:
* Hiss is usually the biggest problem. As already suggested, hiss can be removed (or at least greatly reduced) by taking a noise findgerprint of just hiss and then applying it across the entire freq spectrum. Any decent noise removal software will do this but you might have to experiment with adjustments to make sure you're not removing too much of the the audio also. There's a lot of entry level software out there for this type of work. One of the programs I love for this is Wave Repair (www.waverepair.com). Doesn't cost much and works brilliantly (as long as you don't mind working at 16 bit). There are lot's of other solutions out there. Do a Google search and you will find whatever you need.
* I would approach any eq treatments with caution. The freq response of most cassette tape recordings doesn't get much past 12kHz unless you used really good tape in a really good deck. There's no point in boosting the high end if there's nothing there to begin with.
* Normalizing may or may not buy you anything. Depends on the source material and how you dumped the audio to your PC in the first place. Would need to see/hear the audio first.
* Spectral enhancements sometimes can work wonders or actually make things worse. Again, proceed with caution and listen carefully at the results.
* Often times I simply slip edit the ends of each clip and then add creative fade curves.
Any other suggestions based on past performance?
Audio from analog audio cassettes can suffer from any one of a number of problems. Or, it may not be suffering at all. A lot depends on the quality of the tape, how old it is, how well it was stored, the type and quality of the original recording deck, original recording levels, dbx or dolby encoding, etc etc etc.
If you want a better, more accurate response, it might be a good idea to post an MP3 of some of the audio for us to listen to. That will give us a better understanding of what needs to be done.
Anyway, here are some generalizations and observations based on past projects:
* Hiss is usually the biggest problem. As already suggested, hiss can be removed (or at least greatly reduced) by taking a noise findgerprint of just hiss and then applying it across the entire freq spectrum. Any decent noise removal software will do this but you might have to experiment with adjustments to make sure you're not removing too much of the the audio also. There's a lot of entry level software out there for this type of work. One of the programs I love for this is Wave Repair (www.waverepair.com). Doesn't cost much and works brilliantly (as long as you don't mind working at 16 bit). There are lot's of other solutions out there. Do a Google search and you will find whatever you need.
* I would approach any eq treatments with caution. The freq response of most cassette tape recordings doesn't get much past 12kHz unless you used really good tape in a really good deck. There's no point in boosting the high end if there's nothing there to begin with.
* Normalizing may or may not buy you anything. Depends on the source material and how you dumped the audio to your PC in the first place. Would need to see/hear the audio first.
* Spectral enhancements sometimes can work wonders or actually make things worse. Again, proceed with caution and listen carefully at the results.
* Often times I simply slip edit the ends of each clip and then add creative fade curves.
Any other suggestions based on past performance?
based on past performance, another problem can be phase problems (i have a fostex R8) due to the hadware ADDA converters, or even the audio soft you use (cubase vst 5 : very bad, sx is better).
I agree with kzirox : if you record from tape, you already have tape sound.
If you do eq stuff etc, in order to creatively make the track sound better, har it again a few days after. the unequalized tape tracks ften sounds better when touching nothing.
Noise : a problem ? it was on tape, why remove it ? i don't understand. plus you might kill higher frequencies. Cutting the tracks exactly where it begins makes noise much more discrete. Of cource excessive noise may be removed if it comes from bad tapes, bad recording, bad converters etc... in this case why not. but if the guys like there material as it is... i wouldn't touch anything (unless you are remastering something).
my 2 cents.
<font size=-1>[ This Message was edited by: Mehdi_T on 2005-02-21 14:10 ]</font>
I agree with kzirox : if you record from tape, you already have tape sound.
If you do eq stuff etc, in order to creatively make the track sound better, har it again a few days after. the unequalized tape tracks ften sounds better when touching nothing.
Noise : a problem ? it was on tape, why remove it ? i don't understand. plus you might kill higher frequencies. Cutting the tracks exactly where it begins makes noise much more discrete. Of cource excessive noise may be removed if it comes from bad tapes, bad recording, bad converters etc... in this case why not. but if the guys like there material as it is... i wouldn't touch anything (unless you are remastering something).
my 2 cents.
<font size=-1>[ This Message was edited by: Mehdi_T on 2005-02-21 14:10 ]</font>
I do a lot of Tape restoration. Most tape recordings can benefit greatly from careful use of a noise reduction app, followed by EQ (Psy-Q does a fantastic job) and some compression. I use Sonic Foundry (used to use Creamware's Osiris in a previous century) - but there's lots of choices. You have to be very careful with how much noise reduction you apply, but with practice you can seriously improve the sound of tape recordings.
/dave
/dave
A very smart suggestion. Many times I leave the noise in there. If you can slip-edit most of it away at the front and back of the song, many times the noise is buried in the music and isn't an annoyance. Depends on the song I guess. Also, I've encountered tapes where the hiss had been previously edited away and I couldn't get a decent fingerprint. What do you do then? Generic broadband reduction will always hurt the music to a certain degree. So I guess you have to ask yourself, is the noise actually an annoyance or can you live with it? If you can live with it, then don't mess with it. Good luck. I find this type of work very interesting. My most succesful jobs were the ones where the music was well recorded in the first place and the tape was high quality and well stored. Sometimes you don't really need to do much of anything.On 2005-02-21 14:05, Mehdi_T wrote:
based on past performance, another problem can be phase problems (i have a fostex R8) due to the hadware ADDA converters, or even the audio soft you use (cubase vst 5 : very bad, sx is better).
I agree with kzirox : if you record from tape, you already have tape sound.
If you do eq stuff etc, in order to creatively make the track sound better, har it again a few days after. the unequalized tape tracks ften sounds better when touching nothing.
Noise : a problem ? it was on tape, why remove it ? i don't understand. plus you might kill higher frequencies. Cutting the tracks exactly where it begins makes noise much more discrete. Of cource excessive noise may be removed if it comes from bad tapes, bad recording, bad converters etc... in this case why not. but if the guys like there material as it is... i wouldn't touch anything (unless you are remastering something).
my 2 cents.
<font size=-1>[ This Message was edited by: Mehdi_T on 2005-02-21 14:10 ]</font>
- ChrisWerner
- Posts: 1738
- Joined: Fri Aug 31, 2001 4:00 pm
- Location: Germany/Bavaria
- Contact:
I´ve done much restauration in the past too and I think I found my way.
Of course every new tape is a new challenge.
First I record the first seconds of the tape, to catch the noise from the deck and the noise from the tape itself before the music begins. I load a denoiser and try to find the best settings (noiseprint) and save the profile. In that step I let the noise recording play in loop mode.
Then I copy the whole tape and reload the denoiser with the profil again.
It gets difficult though, when you have a tape with different recordings situations, so you have to get a profile for each track,
The risk to loose many high frequences while using a denoiser is correct. Anyway, I nearly kill all noise, in fact my denoised version sounds worser than the original.
Now I load a reverb and find a fitting room for the music, gain the reverb EQ to get "new" clean high frequences back. You can reduce the low reverb signals and boost the high frequences, it´s a bit tricky but works for me.
Now I proceed the recording like any other track that I master, better said I remaster the track, eq´s, maybe comps, finalizer etc.
This way is maybe a bit odd, to make a recording worser than the original in the first steps but I like odd things.
Good luck.
<font size=-1>[ This Message was edited by: ChrisWerner on 2005-02-21 18:02 ]</font>
Of course every new tape is a new challenge.
First I record the first seconds of the tape, to catch the noise from the deck and the noise from the tape itself before the music begins. I load a denoiser and try to find the best settings (noiseprint) and save the profile. In that step I let the noise recording play in loop mode.
Then I copy the whole tape and reload the denoiser with the profil again.
It gets difficult though, when you have a tape with different recordings situations, so you have to get a profile for each track,

The risk to loose many high frequences while using a denoiser is correct. Anyway, I nearly kill all noise, in fact my denoised version sounds worser than the original.
Now I load a reverb and find a fitting room for the music, gain the reverb EQ to get "new" clean high frequences back. You can reduce the low reverb signals and boost the high frequences, it´s a bit tricky but works for me.
Now I proceed the recording like any other track that I master, better said I remaster the track, eq´s, maybe comps, finalizer etc.
This way is maybe a bit odd, to make a recording worser than the original in the first steps but I like odd things.
Good luck.
<font size=-1>[ This Message was edited by: ChrisWerner on 2005-02-21 18:02 ]</font>
-
- Posts: 162
- Joined: Sat Jul 31, 2004 4:00 pm
- Location: England
Project turned out to be a little larger than I thought when I agreed to it. I think it is about 1.5-2 hours audio. I am taking samples and applying settings based on the those samples to the entire side.
Have normalised.
Gonna make two versions, one with harsh noise reduction and slight distortion and the other without distortion and more noise.
Might EQ it and give it a little reverb or hamornic boost.
Then will gate and split and burn...
Have normalised.
Gonna make two versions, one with harsh noise reduction and slight distortion and the other without distortion and more noise.
Might EQ it and give it a little reverb or hamornic boost.
Then will gate and split and burn...
Interesting, this riminds me that I had made the utility program as same idea about 3 years ago. Something likes denoise tool as Wave batch processing with the GUI for Windows.
It is standalone and has built-in Reverb, EQ (40 bands, -60dB/-96dB to +12dB/+24dB), Spectrum analyzer and Analog VU meter.
Limitation is that it can takes 16Bit@44.1KHz wave format for Windows ONLY.
Let me know if many of you are interested, I probably will recall this one for beta testing here. But you might need to wait for a while.
LongStudio
<font size=-1>[ This Message was edited by: LHong on 2005-02-23 15:49 ]</font>
It is standalone and has built-in Reverb, EQ (40 bands, -60dB/-96dB to +12dB/+24dB), Spectrum analyzer and Analog VU meter.
Limitation is that it can takes 16Bit@44.1KHz wave format for Windows ONLY.
Let me know if many of you are interested, I probably will recall this one for beta testing here. But you might need to wait for a while.
LongStudio
<font size=-1>[ This Message was edited by: LHong on 2005-02-23 15:49 ]</font>