Difference between revisions of "Digitising cassette tapes"

From TMB Wiki
Jump to: navigation, search
(Capturing/Recording)
(Capturing/Recording)
Line 18: Line 18:
 
#:'''tip:''' fast forward and rewind the cassette tape before playing it. This will tighten up the tape inside and help prevent it being 'chewed-up' by the tape deck  
 
#:'''tip:''' fast forward and rewind the cassette tape before playing it. This will tighten up the tape inside and help prevent it being 'chewed-up' by the tape deck  
 
#Sample rate should be set to 44100Hz.  Sample format should be set to 16-bit.  Given the relative low quality of the analogue audio in the cassette tape, you'll be very unlikely to notice any improvement in sound quality with higher settings; CD is 44100Hz!   
 
#Sample rate should be set to 44100Hz.  Sample format should be set to 16-bit.  Given the relative low quality of the analogue audio in the cassette tape, you'll be very unlikely to notice any improvement in sound quality with higher settings; CD is 44100Hz!   
The only time I'd recommend higher settings, such as 48000Hz and 24-bit/36-bit, is when you have a very good soundcard and a good quality recorded tape and/or you will be doing lots of manipulating of the recording. Each transform, such as Noise Reduction, will produce quantization errors. Ideally the editing program dithers the transform (if working in 16-bit), which means more noise added. However, due to the high intrinsic noise of recordings from cassettes, this is unlikely to be distinguishable for the 'normal' amount of editing (see 7. & 8. below)
+
#:The only time I'd recommend higher settings, such as 48000Hz and 24-bit/36-bit, is when you have a very good soundcard and a good quality recorded tape and/or you will be doing lots of manipulating of the recording. Each transform, such as Noise Reduction, will produce quantization errors. Ideally the editing program dithers the transform (if working in 16-bit), which means more noise added. However, due to the high intrinsic noise of recordings from cassettes, this is unlikely to be distinguishable for the 'normal' amount of editing (see 7. & 8. below)
 
#Play the tape and start recording: bthis is a test!/b though, because you'll need to fast forward to random parts of the tape to try and find the loudest section: use that as your basis to adjust the line in recording volume so that there is no clipping.  Ideally, you want to capture the audio at -3-5db.  In [[Audacity]] you can view the sound wave in db format, so that will show you what db level you are recording at.  It's always better to record at a lower volume and then increase it after you've captured the audio
 
#Play the tape and start recording: bthis is a test!/b though, because you'll need to fast forward to random parts of the tape to try and find the loudest section: use that as your basis to adjust the line in recording volume so that there is no clipping.  Ideally, you want to capture the audio at -3-5db.  In [[Audacity]] you can view the sound wave in db format, so that will show you what db level you are recording at.  It's always better to record at a lower volume and then increase it after you've captured the audio
 
#After you've sorted out the line-in recording volume level, you can start to record the audio... press record basically
 
#After you've sorted out the line-in recording volume level, you can start to record the audio... press record basically

Revision as of 08:38, 6 April 2006

jaybeee's guide on recording cassette tapes onto the computer:


What You Need

  1. A Hi-Fi separates cassette tape deck (a portable cassette player walkman will work, but with reduced audio quality)
  2. 3.5mm stereo to two phono plugs cable lead: approx £5 in the UK (if you have a really good soundcard it will have two phono line-in sockets: so just get a standard interconnect cable)
  3. A soundcard that accepts line-in: most motherboards these days have this on board
  4. A cassette tape cleaner to clean tape heads or use pure isopropyl alcohol and lint-free swabs (cotton wool buds will be ok also). This will help to capture a better sound since the cassette tapes will likely be old and dirty
  5. An audio capturing/editing software application. I personally use Audacity as it's very easy to use, is free and open source, and has many plug-ins available (LADSPA / Nyquist / VST Plug-Ins). There are other apps that will do the same job
  6. A lossless audio encoder like FLAC or WavPack: this is so that once you've got the audio in the state you're happy with, you can encode it to a lossless format for archival purposes. You can then transcode it to mp3 etc for upping
  7. A lossy audio encoder like LAME for mp3

What To Do

Capturing/Recording

  1. Connect the left & right plugs of the 3.5mm stereo to two phono plugs cable lead to the cassette tape deck and the single 3.5mm stereo plug to the line-in on the soundcard
  2. Ensure the sound recording capabilities on your computer are enabled: (for windows) Control Panel > Sounds and Audio Devices > Audio tab > Sound recording > Volume > tick 'Line In'
    tip: fast forward and rewind the cassette tape before playing it. This will tighten up the tape inside and help prevent it being 'chewed-up' by the tape deck
  3. Sample rate should be set to 44100Hz. Sample format should be set to 16-bit. Given the relative low quality of the analogue audio in the cassette tape, you'll be very unlikely to notice any improvement in sound quality with higher settings; CD is 44100Hz!
    The only time I'd recommend higher settings, such as 48000Hz and 24-bit/36-bit, is when you have a very good soundcard and a good quality recorded tape and/or you will be doing lots of manipulating of the recording. Each transform, such as Noise Reduction, will produce quantization errors. Ideally the editing program dithers the transform (if working in 16-bit), which means more noise added. However, due to the high intrinsic noise of recordings from cassettes, this is unlikely to be distinguishable for the 'normal' amount of editing (see 7. & 8. below)
  4. Play the tape and start recording: bthis is a test!/b though, because you'll need to fast forward to random parts of the tape to try and find the loudest section: use that as your basis to adjust the line in recording volume so that there is no clipping. Ideally, you want to capture the audio at -3-5db. In Audacity you can view the sound wave in db format, so that will show you what db level you are recording at. It's always better to record at a lower volume and then increase it after you've captured the audio
  5. After you've sorted out the line-in recording volume level, you can start to record the audio... press record basically
  6. Once the recording process has finished, you might want to go through the audio to edit out any parts that you do not require. Namely the silence at the start and the end of the tape. If you want to, it would also be wise to fade in and fade out now whilst you have the audio in this 'raw' format

8. You might also like to remove any 'hiss' or reduce certain noises (tape hum). http://audacity.sourceforge.net/Audacity has a 'Noise Removal' tool that is quite good. Be warned though: it will often result in a slight loss of high-end audio that might make it sound worse than the original. I don't normally perform this action unless there is a lot of hiss. http://www.adobe.com/products/audition/main.htmlAdobe Audition has a very good Noise Removal tool, but the program is also quite expensive

9. Once you've completed manipulating the audio, I'd recommend running the file through WaveGain (available at http://www.rarewares.org/others.htmlhttp://www.rarewares.org/others.html). This is based upon the http://www.replaygain.org/Replay Gain standard and will effectively apply gain (volume) adjustments directly by adjusting the scaling of the samples. Unlike Replay Gain, this is not a lossless process an cannot be reversed. It basically means the files are adjusted to -89db, which normally results in a smaller file size too, as most people tend to record at too high a level. If all music was the same relative volume, we'd never need to change the volume control on our equipment if we always wanted the same absolute volume /list

What To Do: Encoding list 1. Given the effort and time taken to capture and manipulate the audio, I'd highly recommend that the final file is encoded to a lossless format such as http://flac.sourceforge.net/FLAC or http://www.wavpack.com/WavPack. This means you are left with a perfect archival backup of the audio captured, it's smaller in file size than a wav file, and you are free to convert (transcode) to any lossy audio format you like whenever you like, such as mp3 or Ogg Vorbis or AAC etc etc

2. For upping at themixingbowl.org I'd recommend the file is encoded to mp3 using the http://www.mp3dev.org/LAME encoder. See my post http://themixingbowl.org/topics.php?action=viewtopic&topicid=9060here for more info about that

3. And don't forget to tag the file. If you can add as much relevant info into the tags and in a logical way, then the downloaders will not need to change the tagging and thus can keep seeding for longer (ever!)


If you breally/b value your cassette tapes recordings then I'd recommend you buy an audio editor as it will give better results than free audio applications. You can also search the internet for probably better, and certainly more detailed technical information about cassette tape audio capturing. This is just how I do it and I get pleasing results from fairly cheap equipment and free software.


This page is a stub. You can help us expand it by logging in with your TMB username/password and editing it.