Announcement

Collapse
No announcement yet.

78rpm questions...

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • 78rpm questions...

    Hi all,

    I read with great interest the 78rpm restoration - DC7 steps thread on this forum. I too am starting to do my own 78rpm transfers and have some questions of my own...

    1)
    I have learnt that i'll need a flat preamp as opposed to one which performs RIAA correction, which is understandable. I have seen the CTP-1000 on the DV website - would this be suitable for use in the UK? Does anybody know of a UK supplier or something that is suitable for UK power?

    2)
    With acoustic recordings, is there anything I can do to the equalisation once it's in Diamond Cut to enhance the sound, or do most people just apply the impulse remover etc and leave the EQ alone?

    3)
    For electrics, a lot of my 78s use the Westrex recording system. I have found an EQ guide here to http://www.rfwilmut.clara.net/repro78/computer.html recreate the curve. I was wondering if I was best to create this on the equaliser in DC, or if there are better presets available inside DC itself? OR, am I just best using the VPP, selecting flat pre-amp under Preamp Hardware and European 78's under Record Type? Does this 'perform' a Westrex EQ?

    4)
    What do most people record the source in? I have been recording at 44.1khz, but at 32 bit due to the quietness of some of the acoustics. Is this overkill?

    Many thanks for any suggestions - and well done to Diamond Cut for the great software...

    Thanks,
    Chris
    Last edited by Craig Maier; 05-04-2019, 10:57 PM.

  • #2
    Hi Chris -

    I think most people are using higher sampling rates to help with de-clicking I have used 96 but now use 88.2 so that the fraction works better when you downsample to 44.1. For the bit rate, I use 24-bit, but my hunch is that 16-bit would work just about as well with 78 transfers. Craig will be the expert on all this.

    Dan
    Dan McDonald

    Comment


    • #3
      To add to what Dan stated, I find myself always trying to kill a myth. The 78 EQ chart that you pointed to is not correct. 78 rpm records were not recorded with a Rolloff (or pre-emphasis) curve. They used only low frequency turnover.

      The Diamond Cut VPA re-creates several common turnover curves perfectly. I recommend that you use those.

      If you want to tweek the sound after you have completed the restoration process, that is a matter of your own taste.

      As for 32 bit resolution, that would imply that the system has 192 dB of Dynamic range (I never want to hear a signal with that amount of dynamic range as it would literally kill me). 78s have around 40 dB of Dynamic Range so 16 bits which are capable of 96 dB of Dynamic range is fine. Of course, more will not hurt, but will not help. But, the higher sample rate as recommended by Dan makes sense.
      "Who put orange juice in my orange juice?" - - - William Claude Dukenfield

      Comment


      • #4
        Acoustic EQ

        Originally posted by ChrisSavage
        2)
        With acoustic recordings, is there anything I can do to the equalisation once it's in Diamond Cut to enhance the sound, or do most people just apply the impulse remover etc and leave the EQ alone?
        Thanks,
        Chris
        There are some acoustics which can be enhanced with a little judicious EQ. The "horn" sound or honkiness typical of acoustics is caused by a midrange bump that can be tamed. You can use parametric EQ or the built in 20 band equalizer. Hunt around the 1-2k range (if memory serves).

        Good luck!
        Doug

        Comment


        • #5
          Good point - - - sometimes you can actually measure the acoustical resonances by playing the entire record into the Diamond Cut Spectrum Analyzer with the Analyzer operating in Averaging mode.
          "Who put orange juice in my orange juice?" - - - William Claude Dukenfield

          Comment


          • #6
            Originally posted by Craig Maier
            Good point - - - sometimes you can actually measure the acoustical resonances by playing the entire record into the Diamond Cut Spectrum Analyzer with the Analyzer operating in Averaging mode.
            Wow! I am going to have to try this. I may find myself convinced that a little EQ is a good thing for acoustics!

            When it comes to sampling rate, I have been going so far as to record at 96 KHz at 33 rpm. I find that the impulse filters really do a better job of separating the noise from the signal when the original is recorded at slow speed. Of course, you get a really huge file and your mental health might deteriorate as you listen to 78s played at 33!

            After applying an initial round of impulse filtration, I then use the pitch control to bump the speed up to 45 then to 78.26 and often apply another round of impulse filtration. (This can take out the "thuddier" impulse noises missed at the slower speed; but it is also much easier to distort you signal. So be careful!)

            Hope this helps.

            Norm Kleve

            Comment


            • #7
              Norm,

              I have gotten to the point where I always start out using the EZ Impulse filter rather than the Expert Impulse filter. I have learned that 90 percent of the time, I get the desired result with the EZ Impulse filter. I like that approach because it is much easier to operate. But, now and then, I do end up in the Expert Impulse filter (for unusual material with strange types of impulsive noise).
              Last edited by Craig Maier; 05-31-2008, 09:22 PM.
              "Who put orange juice in my orange juice?" - - - William Claude Dukenfield

              Comment


              • #8
                Craig,

                I usually use the EZ Clean and find that, in most cases, it does the most complete job while keeping the signal intact. I will have to experiment more with the EZ Impulse feature.

                Norm

                Comment


                • #9
                  78 Pre-Emphasis ????

                  Craig and Other's

                  When I read the old books on audio like the Radiotron Handbook and the "Tin Foil To Stereo" book, I read about various Pre-Emphasis added to 78 records . While each company seemed to have different amounts, it seems that it was done to some degree.

                  The diamond cut tool that I use, the conversion from RIAA to 78 uses various turnover values but seems to not have a compensation for the pre-emphasis.

                  Is there some insight into the confusing world of pre-emphasis ?

                  Sure was a good thing when all the companies used RIAA compensation !

                  Regards,
                  Marc

                  Comment


                  • #10
                    Quoting Marc,

                    "The diamond cut tool that I use, the conversion from RIAA to 78 uses various turnover values but seems to not have a compensation for the pre-emphasis.

                    Is there some insight into the confusing world of pre-emphasis?"

                    -------------------------------------------------------------------------

                    Again, I think that there is a lot of "urban myth" out there about 78 RPM pre-emphasis. The standards that we have encountered (including the reverse - engineering of old recording and playback equipment) do not have pre-emphasis, just low frequency turnover. Thus, that is what we created in the VPA.

                    Now, it is possible that some mastering engineers "rolled their own" pre-emphasis but simply turning up some sort of treble control inbetween the mixer and the cutting head power amplifier. But, if they did that, it was not to a standard, but to their taste and thusly, no way to be precisely reversed.

                    If you want to improve the sound of a 78, set it up appropriately with the VPA hardware and EQ checkboxes. Then, enable the tone controls (which are of the shelving variety) and you can create some very nice variations to the treble, mids and bass response of your 78 RPM transfers.

                    Craig

                    ps - standardized pre-emphasis really came about by Columbia when they introduced the 33.33 RPM LP - - - the Columbia LP Curve. You can find that in the Diamond Cut Virtual Phono Preamplifer when you are working with some early LPs.
                    "Who put orange juice in my orange juice?" - - - William Claude Dukenfield

                    Comment


                    • #11
                      Pre-Emphasis

                      Craig,

                      Indeed, the concept of Pre-Emphasis could have been continued via the Urban Myth idea. Wouldn't it be interesting to find some of the engineer's from that period and talk with them ?

                      One section in the "Tin Foil to Stereo" book talked about how with pre-emphasis applied to records, the users would use the treble tone control to turn down the extra high frequencies.....

                      I find that with all the adjustments with the Diamond Cut Software I can easily add or subtract highs to suit my taste. It's a fine software tool.

                      Marc

                      Comment


                      • #12
                        Acoustic EQ

                        Returning to the question of helping with the sound of acoustic 78's, Rick and Craig already have you covered.

                        If you go to the 20 band equalizer, there are presets for Baritone, Tenor, and Soprano emphasis. Select the most appropriate and use it as a starting point. I was restoring a 78 of a tenor and it really helped with the intelligability of the song. Usually in these songs the background band is so soft any effect on the instruments is unnoticeable.

                        Doug

                        Comment


                        • #13
                          Acoustic EQ

                          Doug,

                          I too have used the various voice boosts in the 20 Band EQ. The Tenor one seems to be the best for the Edison 2M Cyclinder ones.

                          I'm currently using the EZ Enhancer and finding that it has a number of extra features over a straight EQ. This tool seems to add various harmonics and maybe even "reverb".
                          As such, the recording seems to take on an almost 3-D performance.

                          It would be easy to overuse this effect though.....

                          Regards,
                          Marc

                          Comment


                          • #14
                            Continuous Noise Philosophy and Filter Use

                            I've decided audio restoration is as much art as science, and the treatment of continuous noise is one of the primary areas where differences in restoration philosophy is most apparent.

                            Continuous noise, by definition, contains elements at all frequencies of the audio spectrum. The "Sample Noise" feature of the Continous Noise Filter is great for identifying the noise inherent in a recording and the filter does a great job of using the sample to eliminate noise. There's a delema, though.

                            As anyone who has checked "Keep Residue" knows, a significant amount of content is also removed along with the noise. Toggling back and forth between the filter and bypass demonstrates that along with noise being removed, a recording can be rendered muted and lifeless. Some restorers are willing to trade off "presence" for the most quiet possible result. OTOH, there are those, including me, who will tolerate some noise in exchange for a "livelier" sounding restoration. I have worked out a method of use of the filter that gives what I consider the best comprimise. Here's my method:

                            I search both the run in and run out areas for noise. I've found that there's usually more noise in the run in, which may actually be a lot noisier than the rest of the recording. If so, I sample the run out.

                            After sampling a non-musical area for noise, I select the entire file and click preview. Sometimes this leads to a resampling. I don't know if I can adequately describe what I listen for in the treatment of the recording. Experience will help get you to the point of knowing your sample is representative of the CN in the recording.

                            After I've got a good sample, I listen to the recording with "Keep Residue" clicked. I use the "Threshold" arrow buttons to move the blue line down to a point where I hear no or very little musical content. Since the upper frequency range of a 78 is around 8k at best, I click on the two rightmost data points on the blue line and move them above the red noiseprint. I also often move the 10k datapoint to the left, lowering the frequency of the datapoint. As I do this, I toggle between residue and the actual content. I also use the threshold toggles to hunt for the sweet spot; the most noise removal with the least effect on content.

                            The result of this is file that is not noise free, but has the least possible noise. I've done a casual survey of family and friends who'll tolerate my mad experiments. They prefer a little noise in recordings that are 60 to 100 years old. They find a file that is "too quiet", where all noise has been eliminated, as not "sounding right".

                            I'd be interested to hear how others use the continuous noise filter.

                            Doug

                            Comment


                            • #15
                              CNF Useage

                              Doug,

                              Good note on the Continous Noise Filter.

                              At first, I just ran the level up to 70-80 (artifact Suppression ON) and left the threshold where it was after sampling.

                              As time went on, I too have found that some (just a little bit) broad spectrum noise seems to be the right mix between reducing the sound to dull vs lively.
                              For most records I have been using 60 and the highest setting and around 4-5 mouse clicks up on the threshold.

                              I have found that the highest setting on the FFT sample size really helps and at times it even clears up some distortion on the original recording.

                              The CNF has to be one of, if not the hardest filters to use correctly. Often, I will have to play the whole song using preview and tweek the settings.

                              So far, the most sensitive sounds to have distortion due to the CNF settings seem to be quiet voice passages and muted trumpet/cornet . When a record has these sounds and I set up the filter to have them "sound good" the rest of the record comes out O.K.

                              I guess that the toughest one so far is "Rhapsody in Blue" by Oscar Levent. Getting the piano to sound good was a tough one.

                              Regards,
                              Marc

                              Comment

                              Working...
                              X