Scratchie Static Record Filter?

Comments

johnmeyer wrote on 12/6/2009, 9:10 AM
I have transferred quite a few of the "Recordio" live cut discs for various clients. The sound quality is AWFUL, even if they haven't been touched. The first ten seconds and last twenty seconds are often almost unintelligible. So, your family may not be entirely to blame. Having said that, the discs I have seen actually have a substrate that feels more like a cardboard or wax paper base, and definitely would not take kindly to moisture.

It's been several years since I transferred one of these, but my memory is that they required a much heavier than normal tracking force (5 g. is "normal" for most 78s from the 30s, 40s, & 50s). I seem to remember using the maximum I could crank in. As always, if you have multiple styli, it would be worthwhile trying various truncated versions to see if you can get better sound higher up in the groove.
farss wrote on 12/6/2009, 12:34 PM
This disk feels like it has a metal core. It's music, client's mother singing with a piano. Part of the noise seems to come from the acetate layer having crazed over the years but one side is way better than the other.
I've tried a regular and a 78 stylus, 78 seems best, had hoped the normal stylus might have found a cleaner groove but not so.
The good news is a lot of the noise is HF surface noise that's amenable to being removed with a HF shelf. The actual plops SF can tame a bit. I can get something better out of it than the previous attempt, I hope the client realises there's limits to what can be done.

Bob.
johnmeyer wrote on 12/6/2009, 12:57 PM
I've tried a regular and a 78 stylus, 78 seems best, had hoped the normal stylus might have found a cleaner groove but not so.A non-78 styli will always do a terrible job. The LP needle is so much smaller than those used for 78 that it will go all the way to the bottom of the groove and hit all the gunk that has built up down there. Also, since it is so small, it rattles around in the groove rather than actually moving with the groove.

Most "big-time" 78 rpm restoration outfits use truncated needles which are designed to track higher up in the groove and obtain their vibrations from the upper part of the groove. This not only keeps the needle away from the microscopic debris at the bottom of the groove, but also puts the needle in the upper part of the groove which is usually less worn. Thus, a bigger fatter stylus is often preferred.


The good news is a lot of the noise is HF surface noise that's amenable to being removed with a HF shelf ... I hope the client realises there's limits to what can be doneThe other thing that is so important about 78 rpm restoration is getting the turnover and rolloff settings correct. As you know, all recordings (at least those done after the early acoustic recordings) reduce the bass so that the cutting needle doesn’t swerve into the adjacent groove during a loud bass note, and it increases the high frequency sounds so that during playback these can be decreased which in turn will decrease the amount of high frequency noise. When 33 1/3 LP records were introduced, this equalization was finally standardized, but during the reign of 78 records, every manufacturer did it differently.

So, if you don’t do proper equalization that matches how the 78 was recorded, some 78 rpm recordings may sound very "hissy and scratchy."

The proper way to do 78 equalization is with an old amplifier that has these two knobs. The second best way (and this is what I do) is to use a standard amplifier that has a phono input. This input is compensated using the standard RIAA curve. The trick is to then create a spreadsheet which computes what additional EQ needs to be applied in order to get from the RIAA equalization to the equalization used by the particular record.

I think that many of the record-as-you-play technologies didn't use equalization at all. Thus while you don’t have to apply any turnover and rolloff settings you WILL have to un-do the RIAA equalization.

If you are interested, I have an Excel spreadsheet that took me many years of trial and error to perfect which lets you input the turnover and rolloff settings for the record you are transferring, and it then spits out the EQ settings you need to apply in Sound Forge in order to get the proper "net" EQ. The spreadsheet assumes you are capturing the audio through an RIAA-compensated preamp.



farss wrote on 12/6/2009, 1:16 PM
Thank you, I completely forgot about the EQ problem. My preamp has standard RIAA EQ built in. I'll try undoing that after the capture later today.

Bob.
johnmeyer wrote on 12/6/2009, 4:21 PM
Here are the settings (in dB), directly from my spreadsheet, needed to undo the RIAA equalization. The frequencies shown exactly match the 20 band equalizer settings in Sound Forge:

Frequency         20 	 28 	 40 	 56 	 80 	 113 	 160 	 225 	 320 	 450 	 640 	 900 	 1,300 	 1,800 	 2,500 	 3,600 	 5,100 	 7,200 	 10,000 	 15,000 
RIAA 19.3 18.8 17.8 16.5 14.6 12.3 9.9 7.5 5.2 3.2 1.7 0.4 -0.8 -2.0 -3.6 -5.8 -8.3 -11.0 -13.6 -17.1


If you think the recording was made without any equalization, this will give you the correct result. However, if it was made with equalization, you will then have to add those numbers to these in order to get the correct net result.

[Edit] Actually, what I gave you above is the playback equalization for RIAA. To undo this once you have recorded the audio through an RIAA amp, you actually have to do the negative of what is shown above (i.e., cut the bass and boost the treble).