So I've finally come across an album where there is an actual visible - and audible! - difference in dynamics between the CD and vinyl versions: Radiohead's "King of Limbs."
The CD is heavily brickwalled (the worst I've seen on a Radiohead album by a pretty good margin), while the double 10" (clear!) vinyl has some actual life to it, so I'd like to digitize the vinyl as accurately as possible and use that as my main copy of the album.
The thing is, the vinyl is 45 RPM, and while my eBay-special Pioneer PL-4 is rock-solid at 33 RPM, its speed varies quite a bit at 45.
So, I was thinking about digitizing the vinyl at 33 RPM, changing the playback sample rate by the appropriate percentage in Cool Edit 2000 to speed it back up, and then resampling to 44.1 kHz to make my FLACs (and MP3s).
I haven't tried it yet, it's just a thought at this point, so before I headed down that road, I wanted to ask our esteemed community if there are any caveats/gotchas/pitfalls to this method.
Thanks!
One thing that won't be quite right is the RIAA eq applied by whatever preamp your turntable is connected to - it applies a frequency dependent treble cut and bass boost in simplistic terms. However, whether or not the result is unpleasant - or even noticeable - will depend on how you like you music to sound. I'd give it a go and see what you think of the result.
If you find it a bit bright and perhaps a bit muddy in the upper bass you could try to eq it to you liking via Cooledit. Or if you want to do a more accurate job you can always use Google to find the precise details of the RIAA curve and figure out where the offending frequencies are likely to be (f/45X33) and try to be a bit more specific with your eq
So, playing a 45 at 33rpm (and vice versa, presumably) messes up RIAA playback eq? I did not know that.
Yes (thanks, botface!), as the frequencies coming off the record will be shifted down (in the case of 45 vinyl -> 33 playback), but the RIAA curve will remain the same.
So, for instance, if you have a 100 Hz tone on a 45 RPM record, it will receive a roughly 13 dB boost on playback from the RIAA EQ when played at its original speed of 45 RPM.
However, if you play that record back at 33 RPM, that tone will now be at 74 Hz, where the curve is applying 15 dB of boost.
When sped up via a sample rate change to its original, correct playback speed, that 74 Hz tone will be back at 100 Hz, but 2 dB higher in level.
Heh, hadn't even occurred to me, either, but as soon as I read botface's post, a little light bulb flickered on.
One thing that won't be quite right is the RIAA eq applied by whatever preamp your turntable is connected to -
The
Audacity website[/u] (http://manual.audacityteam.org/index.php?title=Recording_78rpm_records) has the solution:
1. Record the 78 at 45 or 33 1/3
2. Apply the Inverse RIAA EQ (to make it like a "flat" recording)
3. Change the speed to 78rpm
4. Apply appropriate EQ for what the 78 was recorded with
I don't know if CoolEdit has an RIAA filter. If not, the Audacity website also has an XML file that can be used to set it's equalizer to approximate RIAA.
So, playing a 45 at 33rpm (and vice versa, presumably) messes up RIAA playback eq? I did not know that.
Yeah... Since it's a "curve", and not a straight line, and since frequency/pitch changes with speed, the relative levels at various frequencies get fouled-up. . (But, I’d
guess it’s only a couple of dB…)
You could try the clickrepair equalizer (freeware, java), which comes with various (also RIAA) inverse curves included.
http://clickrepair.net/equalization.html (http://clickrepair.net/equalization.html)
The maximum equalization that you would need to apply to correct the frequency curve for the different rotation speed is only about 2 dB. You may not even notice it.
Yeah, I ran a few more calculations at random frequencies and as you and DVDdoug noted, it's pretty mild.
Looks like I've got a couple freely available options for RIAA EQ processing, so I'll give those a shot.
Thanks all for the responses!